Rules and standards discussions/Archive/Archive06

This is an archive page for the Rules and standards discussions page. Please do not edit the contents. To start a new discussion, please click here.
This archive includes discussions from November 2008 - July 2009‎.

Archive Quick Links
Archives of old Rules and standards discussions.

---

1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 · 9 · 10 · 11 · 12 · 13 · 14 · 15 · 16 · 17 · 18 · 19 · 20 · 21

---

Expanded archive listing

ARC first printing, in or out?

I know ISFDB standards don't allow advance reading copies, but I've come across a situation that might lead to a discussion of exceptions. I have two copies of Kathe Koja's Bad Brains. The first is an ARC with a complete number line. The second was the first publicly available copy (she signed my copy at a local bookstore signing shortly after publication), but it's number line drops the "1", indicating a second printing. Is this an unusual situation or was it a common practice for Dell or other paperback publishers? I have a copy of her first novel The Cipher and it's also a second printing. There's a verified copy that matches mine but the printing isn't stated. MALloyd hasn't responded to his talk page in quite awhile, so if anyone else has a copy can you check its printing. Thanks. MHHutchins 02:01, 22 May 2008 (UTC)

I have never seen such a case, at least that I know of, but who knows what some publishers may have done. It does sound like a fair case for treating an ARC as the true first printing. -DES ^Talk 02:40, 22 May 2008 (UTC)

On what grounds would you assume that the book-signing would definitely be for a first printing? I know I have a 3rd printing of Douglas Adams and John Lloyd's "Meaning of Liff" from what I believe to be the first signing-tour - I presumed it was so much more popular than expected that it needed reprinting multiple times during the tour. I can't imagine a tour being canceled for being too successful if they can reprint fast enough... BLongley 19:08, 22 May 2008 (UTC)

True, some books have been reprinted 2-6 times within a few weeks of publication. Much depends on the pre-orders from the major retailers and distributors, so if there is a big spike in pre-orders right before publication I suppose it is even possible to run into a second printing on the first day of the book's availability. We are probably better off documenting the details of this case in the Notes field of the second printing's Publication record. Ahasuerus 22:02, 22 May 2008 (UTC)

Yes, it's an assumption on my part that the book signing copies would be first printings, but I can't imagine that the second book by a relatively unknown author would have already gone into a second printing within weeks of publication (again, I admit, an assumption). All but one of the abebooks.com dealer that mentions first printings of this title also admit that it's the ARC. I've sent that bookseller a message to verify whether their first printing has the full color die-cut cover of the regular edition. There are more than a few listings that explicitly state 2nd printing. In the meantime, I've deleted the ARC and modified the note in the 2nd verified printing. MHHutchins 00:12, 23 May 2008 (UTC)

And after doing all that I checked Locus1:

Bad Brains (Dell Abyss 0-440-21114-X, Apr ’92 [Mar ’92], $4.99, 367pp, pb) Horror novel of an artist whose head injury exposes him to a strange and terrifying world lurking at the edge of human consciousness. Highly recommended (SW). Despite the copyright page denoting this as a second printing, this is actually the first after a large advance proof edition.

Guess I should have stuck with my instincts. MHHutchins 00:17, 23 May 2008 (UTC)

Good catch! I don't think I have ever seen an ARC counted as an official first printing, but the publishing world is nothing if not weird :) Ahasuerus 01:19, 23 May 2008 (UTC)

There are times when I suspect that there's a gnome in Zurich whose job it is to create bibliographic nightmares like this, just to keep people like us on our toes. At any rate, this is a new one on me, too. The closest I've ever come is noting from one source or another that a given book may be found as a first printing or a third printing, but not a second printing, as the second printing was done for, and distributed in, the Canadian market. ARCs are, to me, interesting collectible items, but have no bibliographic value as separate "editions". The only reason I'd question this approach is if I found one that had substantially different artwork than the finished product, but I've yet to come across one such. Jayembee 06:17, 2 November 2008 (UTC) Jayembee

The situation that I described above was highly unusual. Even so, we chose not to include a complete number-line printing, because it went against ISFDB policy concerning ARCs. The note in the first publicly available edition was enough to warn bibliophiles and bibliographers of the strange situation. MHHutchins 06:41, 2 November 2008 (UTC)

Commenting on the first comment in this thread - I don't have a problem with someone adding/updating publication records based on an ARC as long as the notes clearly state it's an ARC. The ARC books clearly exist as "verifiable publications." I could see one reason for stating ARCs are not allowed is support the position that ARCs should not be sold/redistributed as the neither the publisher nor author gets paid for them meaning we don't want to encourage or support ARC redistribution. Marc Kupper (talk) 18:57, 15 October 2008 (UTC)

As long as we are on the topic, Baen books has fairly recently created something that they call an "e-arc". This is an ebook version of an unproofed, late but not final draft of a book, which they put on sale usually 1-2 months before the release date of the final version of the book. Since they sell this (typically for more than the cost of a pb, but less than half the cost of a paper hc, and more than the cost of a released individual ebook), and it is downloadable like their other ebooks, it surely exists in a verifiable state. Should such "e-arcs" (their term) be listed here? They will generally differ at least in detail, because the final copyedit pass has not been done on them, and in a few cases significant late rewrites have, i understand, occurred after the release of the e-arc. -DES ^Talk 19:54, 15 October 2008 (UTC)

[unindent] I believe all ARCs should be out, even electronic ones. I can't see why an "e-arc" should be treated differently. MHHutchins 06:41, 2 November 2008 (UTC)

The difference, in my mind, is that these "e-arcs" are actually being advertised and sold. The traditional arc was only distributed to reviewers and buyers fro bookstores and distributors, and indeed generally carried a large "NOT FOR SALE" notice. Fans only got these if they knew someone on the distribution list, or if someone on that list sold a copy in violation of agreement. But with Baen marketing these, arguably they aren't "really" ARCs at all, but separate editions. -DES ^Talk 12:31, 2 November 2008 (UTC)

Since this is a bit separate from the ARC questions above, and has gone active qagain, I'm goint to copy this to a new thread below. -DES ^Talk 12:31, 2 November 2008 (UTC)

Partial entries for non-genre magazines

I had somehow thought the policy was not to enter general fiction mags such as The Strand, The Saturday Evening post, Argosy, and All-story, and I have advised others to indicate first pubs in such places via the note field. But a discussion on User talk:Kpulliam has called this policy into question. I want to start a general discussion on this issue. To simplify things, i am copying here several comments originally posted on User talk:Kpulliam. -DES ^Talk 17:23, 14 August 2008 (UTC)

Start comments copied from User talk:Kpulliam

Is there a particular reason that the early pulps are not included in the database? I am particularly interested in The Argosy, The Argosy Weekly, and The Argosy All-Story Weekly, but there are several other listed in the timeline here that do not appear in the database yet. Magazine:1920-1929

Thanks - Kpulliam 03:57, 13 August 2008 (UTC)

Good question. Here's a link to the ISFDB:Policy#Contents.2FProject_Scope_Policy database content policy. Also check out the rules of acquisition. To be more specific in answering your question, if a magazine had a majority of speculative fiction content, then it can (and probably should) be included in the database. The magazines you list appear to lean more toward the adventure genre, although that may include specific works of speculative fiction. Those stories might have been reprinted later either in collections or anthologies, thus being included in the database, with a citation in the notes about their first publication. This is the ideal situation, but it would vary because we depend on individuals to enter contents. If you have specific issues of those titles and feel strong enough that they should be included in the database (and are willing to enter their contents), I think most of the moderators would allow the non-spec-fic content. Looking at my copy of Ashley and Tymn's Science Fiction, Fantasy, and Weird Fiction Magazines, I see that those titles are listed, but with a caveat: "Neither The Argosy nor All-Story was a science fiction or fantasy magazine, and yet they probably influenced the course of science fiction more than any other magazine either within or outside of the field..." Pretty strong endorsement, I might add. MHHutchins 04:43, 13 August 2008 (UTC)

I don't personally have any copies of Argosy or it's follow-on publications, but there are some issues available online, notably 5 issues at the Internet Archive from 1922 (With Ray Cummings 'Fire People' which at the moment exists in the database but with zero publications listed which is what prompted the question to me in the first place) and a few more from the 1890s.

As above, it was the Ray Cummings entry Fire People and the magazine images at IA which started me wondering why the magazines didn't exist as entries. I've read the contents policy and I understand that the line must be drawn somewhere. But this line is very blurry. I've been using ISFDB as a look-up source for years, and I truly love the ability to provide updates when I find omissions, etc. This is one of those cases where because the magazine has been excluded (even in partial form) from the list of magazine indexes ISFDB Magazines that I was unsure how to get the info into the database. Kpulliam 05:14, 13 August 2008 (UTC)

There is a bit of history here. The original Rules of Acquisition (which are no longer accessible since most older versions of Wiki pages were dropped a few months ago to reclaim space) called for the inclusion of all non-genre magazines and anthologies as long as they contained any speculative fiction contents. We then quickly realized that if we entered all non-SF stories in Argosy, The Norton Book of American Literature, etc, we would end up with thousands of extraneous records -- and besides we don't support non-genre short fiction particularly well -- so we updated the Policy page to state that we should create abbreviated publication records for these books which would only include SF titles. That way we could still Verify them and perform other bibliographic operations.

Unfortunately, the Policy text was not very clear (my fault) and in many cases the original publication data was recorded in the Notes field as opposed to a separate Publication record. I know that I used Notes numerous times simply because it was faster and I figured I would come back to it and create a new record... some day. Over time, this approach has become more popular and I assume that some editors took it to be the standard since the Policy page could be read that way and the prevailing practice suggested this interpretation. We definitely need to clarify the Policy one way or the other, and I still think that abbreviated Publication records are a better way to capture this data to facilitate Verification, but that's something to discuss on the Standards and/or Policy page. Ahasuerus 12:31, 13 August 2008 (UTC)

I'm quite happy with abbreviated publication records - they may even grow over time as we discover there was more than one SF story in it. One of the things that put ME off was the minimum requirements for Magazine entries: e.g. there's a couple of dozen notes saying that a story was first published in "The Strand" magazine of a certain date, but if you try and enter just the information we have you get a big "*** ERROR: No authors were specified" message. And I for one don't know who the Editor of a magazine that old was... still, if we can figure out what to do in these cases ('uncredited'? 'anonymous'? 'unknown'? 'dunno'?) and make searching for magazines a little easier (as basically you cannot find them via anything other than advanced search, or a link from the wiki which we probably won't have in these cases), it shouldn't be too difficult to convert notes to abbreviated publications (or 'stubs' as some people have been calling them). Whether we want to start 'proper' magazine wiki pages for such is a matter we should probably address sooner rather than later: maybe separate them out into another section like we did with Fanzines? I can search notes offline if people tell me what to look for - e.g. we have almost 600 "First Published" notes, but those won't always point to a non-genre magazine, they may have been written before we got the genre magazine, or may refer to a book - or even a newspaper. BLongley 20:22, 13 August 2008 (UTC)

If partial publication records are to be the way to go, fine, but at a minimum the db demands a magazine title (and I would hope it would be standardized), an editor (perhaps "unknown" will do when a name is not available), and a date, which really ought to be as accurate as possible. On sources, Project Gutenberg has a few issues of The Strand up, and more coming I think. Wikipedia lists the founding editors for both The Strand and Argosy, but not a complete list of editors for either.

If we are going to encourage partial entries for issues of general fiction (non-genre) mags, we really ought to agree on a few standards, and write up a "how to" for such entries. I am OK with either decision. -DES ^Talk 17:23, 14 August 2008 (UTC)

End comments copied from User talk:Kpulliam

(unindent) I don't find anything about partial magazine entries anywhere in Help:Screen:NewPub or Help:Screen:EditPub. RoA #1 clearly makes SF content published in English IN, "including works published within and outside the genre.". Nothing is said about non-SF works published alongside SF works. We need to make a decision, and once it is made, clarify the RoA and the Help one way or the other IMO. I don't feel strongly about the decision either way, but it would be better to be clear, i think. -DES ^Talk 17:23, 14 August 2008 (UTC)

The history of dedicated sf magazines begins with the April 1926 issue of Amazing Stories. With a couple of very minor exceptions the only other legitimate title that predates that is Weird Tales. Partial entries for magazines that are predominately non-genre should include no more than the minimum amount of date required by the system and only stories that meet our existing standards. There is no reason we should use any other than the current standards for non-genre material - only material by authors whose data is primarily fantastic should have any non-genre entries. Finding them - use either the generic magazine title as the editor name or something like "Argosy editors", "Blue Book editors", etc.--swfritter 18:47, 14 August 2008 (UTC)

I like that idea for finding them, and we seem to have done that for Playboy. We've a few Saturday Evening Posts to adjust already. Strand, Pearson, Pall Mall, London Magazine etc are still to create. For Collier's, we'd have to lose attributions for Walter Davenport and Louis Ruppel though - would anyone miss those? BLongley 19:14, 14 August 2008 (UTC)

If I understand Swfritter's comments above correctly, he is saying that even when we know who the editor of a non-genre mag was at a particular time, we should not enter it, but instead use "The editors of X" or some such? This seems perverse to me, although i could support using such a term in any case when the editor is not known. I can agree with his suggestions that when creating entries for issues of non-genre mags that include some SF, we include only the SF content (and clearly related art, I would add) and otherwise minimal meta-data. -DES ^Talk 19:34, 14 August 2008 (UTC)

Well, neither Davenport or Ruppel are particularly noteworthy in an ISFDB sense, and I only found the names from Wikipedia and some other source I forget - it's not 100% certainly 'known' as these aren't primary verified. If you want to keep the names though, what do you suggest for making finding the magazines easier? A Non-Genre Magazine/Newspaper Wiki section would work but seems a lot of trouble for what may be one-off issues. BLongley 22:10, 14 August 2008 (UTC)

I think that we will find, that due to the inclusion of some popular authors (Doyle, London, Burroughs, etc, etc) that quite alot of the back issues of the primary magazines are going to pop up with at least one genre story (or genre related author). (I'm personally estimating 60-80% of the issues)Kpulliam 23:53, 14 August 2008 (UTC)

Agreed, but for every magazine that published SF often, we're going to find one that only did it once or twice. And "genre related author" is probably a separate can of worms - we could end up with a lot of magazines added just to keep the Asimov and Clarke essays complete. BLongley 01:36, 15 August 2008 (UTC)

As to keeping only SF content, that is typically what we've done, e.g. here. So far the issue of artwork hasn't come up as these are all from secondary sources. BLongley 22:10, 14 August 2008 (UTC)

OK, after only a couple of hours of chasing down examples caused by ONE book, I've found examples of artwork: Moderators, see the submission for "Rogue, August 1962". That also shows the problems with finding new titles by accident, the Wilson Tucker story is not one we had already. And a new author: presumably it's SF as that whole issue was. BLongley 01:36, 15 August 2008 (UTC)

I too like that idea for finding them. Personally... I was more interested in finding them from the authors bibliography. Starting with an author, and then attempting to find early publication records. (And in this case, my purpose was for searching out public domain items, for eventual inclusion into Gutenberg).Kpulliam 23:48, 14 August 2008 (UTC)

FYI: Robert H. Davis was editor for All-Story (January 1905-July 17, 1920, through several title changes until it merged with Argosy). Frank A. Munsey edited Argosy from December 9, 1882-November 27, 1886, succeeded by Matthew White (December 4, 1886-June 2, 1928, the "Golden Age" of Argosy). Other editors followed but the title's importance diminished. They're all listed at Phil Stephensen-Payne's Galactic Central. Add me to the list of supporters for actually crediting the editors by name. MHHutchins 00:16, 15 August 2008 (UTC)

I see we have a Robert H. Davis here in his own right, so presumably that's someone notable enough to keep recording. However, I don't want people to NOT enter the magazines just because they don't know the Editor, so perhaps the significant non-genre magazines can get their own wiki page with such information on, where people can find such data? The rest can be left as swfritter suggested, IMO: or do you feel as strongly about editors for 'Continental Airlines Magazine', 'Cosmopolitan', 'Ladies Home Journal', 'Games Magazine', the 'Philadelphia Sunday Bulletin', the London 'Evening Standard', to mention just a few that we currently have notes for? 01:50, 15 August 2008 (UTC)

No, the line should be drawn somewhere. Perhaps we search for editors only if the periodical is a majority fiction magazine. That would rule out Playboy, Penthouse (and other men's magazines), Redbook, Good Housekeeping (and other women's magazines) (am I being delightfully sexist here?), and even Saturday Evening Post and Boys' Life. MHHutchins 02:11, 15 August 2008 (UTC)

I suspect most of the "abbreviated publications" have been/will be created from secondary sources: so won't be verified, and therefore can be improved without too much questioning if somebody wants to add the Real Editor, price, publisher, binding, etc. So I'm happy if we start with a title, Dummy editor, date, and SF content: if some people want Real Editor then I think they can just update the dummies, and start a Wiki page so that it doesn't take hours to find that sort of detail for future entries. (Converting Real Editor to Dummy should be a big No-No, hence my asking about Davenport and Ruppel in advance even though it was me that found them.) The Wiki pages for such needn't be anywhere near as complex as the ones for Genre magazines - definitely no grids for all known issues, for instance. But they should cover the details that some people want recorded (which should be reasonable SF-based requests, not "which playmate appeared on the cover of that issue of Playboy?") and they should be easily findable. (I'm still thinking on the lines of a (small!) Wiki Section for Non-Genre magazines where people want a few rules for that title.) BLongley 03:43, 15 August 2008 (UTC)

As a 'line' may I suggest that you stick to what I already thought was the line Magazine:1900-1919. If it's worthy of being included in the magazines timeline ... then it is probably worthy of additional correct information. If it's truly a 'transient' publication of SF, then it isn't worthy of inclusion in the timeline. This way, if someone comes along and suggest an additional publication be 'upgraded' with researched editors, etc, if it is accepted, the timeline update can be part of that initial research? Kpulliam 03:11, 15 August 2008 (UTC)

Magazine:1900-1919 and its siblings were an attempt to organize early magazines in a readable/helpful fashion, but their periodization was fairly arbitrary -- note that the 1909-1919 page discusses pre-1909 magazines as well.

Substantively, I had assumed that we would want to enter all editors when known, but I can see how adding another few thousand effectively non-SF Author records to the database could be problematic. Also, keep in mind that many early SF stories were serialized in newspapers, which may make it an even bigger headache. Ahasuerus 03:25, 15 August 2008 (UTC)

(Unindent) I've been trying to do some research along these lines to see what we end up with. The results for new publications like this can currently mostly be seen on the Submission queue: where I couldn't find an exact magazine I left notes on the title instead. I just picked the first Anthology to hand that mentioned original Magazine publications: unfortunately it was The Best of Sci-Fi No. 4 and so after a couple of hours of fairly easy research I got into the "Honorable Mentions". Which leads to new titles we don't currently have, e.g.

Jules Feiffer, The Lonely Machine, Playboy, December 1962
Robert S. Malcolm, A Free Fall, Free for All, Gent, August 1962
David Newman, Three Fables for the Atomic Age, Esquire, April 1962
John Novotny, Snug as a Bug, The Dude, November 1962
Leonard Wibberley, The Man Who Lived on Water, Saturday Evening Post, 28th March 1962
Gahan Wilson, Horror Trio, Playboy, June 1962 - which also contains an Arthur C. Clarke essay "From Lilliput to Brobdingnag" and "The Tryst" by Jules Feiffer.

I believe these should all be IN, assuming Judith Merril got the credits right: but it's now half past four in the morning and I've been looking at just one book's worth of data (and discussing it here, of course) and realise that this is not going to be a quick task, whatever we decide. BLongley 03:43, 15 August 2008 (UTC)

(A bit off-topic) I just took a look at "The Best of Sci-Fi No. 5" (aka "The 5th Annual of the Year's Best S-F") to see how much trouble the NEXT book will give me tomorrow (or sometime - I acquired 60 other books today, I suppose I don't HAVE to do this next), and see that the UK edition explains the abbreviations used - and then omits the section where the abbreviations ARE used entirely. Anyone with a book like that with the same sort of credits can see how far this topic can take us though. I could read the entire book faster than I can track down the original publications satisfactorily. (And may well do so.) BLongley 04:12, 15 August 2008 (UTC)

If we know the editor and want them to be listed we can enter their name and also the magazine name as editors. We need a simple mechanism to link to and find the magazines. Expecting people to add to a magazine wiki, or even create one, for a single story is too much work to require. I think we are mostly in agreement about non-genre being out and remember also that there is no way to designate a story as both shortfiction and non-genre.--swfritter 12:42, 15 August 2008 (UTC)

Bill said above "So I'm happy if we start with a title, Dummy editor, date, and SF content...". I would also be happy if we start with that. If the "real" editor isn't known by the person doing the entry, fine. But while we would, in effect, have a rule against entering the non-sf content of such magazines, there would be no rule against anyone who had a reliable source entering the correct magazine editor, and such metadata as price and page count. As for searches, i would think that searches by the author of the SF content would be the most common, and those should work with no special work. If people really want to find all examples of, say, Argosy and see what is entered for them, we could always create a publication series for all such magazine issues. Whether a wiki page would be worth creating and maintaining I'm not sure, but creating one would be easy enough. Are we approaching a consensus here? The points which people seem to agree on are:

1. It is acceptable to make partial entries of issues on non-genre magazines (or collections?) where speculative fiction was published, particularly ones where such fiction was first published.
2. Such entries should not include content entries for non-genre content in the same publication.

   Perhaps a note indicating that the entry is known to be incomplete would be a good idea.

3. It is acceptable to make such entries with minimal metadata: Publication title, Publication date, and a dummy entry for the publication editor such as "Argosy Editors".
4. When entered from a primary source, or determined from a reliable secondary source, it is acceptable but not required to enter additional metadata, including the actual editor's name.
5. The issue of how best to search for such publications should be considered, but need not hold up data entry.

Are we in fact in agreement on the above? Does anyone have other items to add? Are we ready to consider modifying the RoA and the help in line with the above? -DES ^Talk 15:16, 15 August 2008 (UTC)

The magazine editor series method will work. Might be a good project to find all the magazine editor records that are not in a series. Of course, we still have a lot of magazines, mostly the minor ones, that don't have editor records.--swfritter 16:31, 15 August 2008 (UTC)

Should we also mention that it is still acceptable to enter the data in the notes for the title? I would rather have the data there then not entered at all. There are also cases where it would require a fair amount of research in order to determine the actual issue in which a story appeared and the notes are the only logical place for partial information.--swfritter 16:31, 15 August 2008 (UTC)

Guinea Pig #1 in this editor series.--swfritter 17:19, 15 August 2008 (UTC)

I think notes are always acceptable: I'm not deleting them when I add the stub publications either, using both is fine by me. There's at least one of my example submissions where I could only find Magazine name and year. However, can I point people at The Fiction Mags Index? It's a very useful resource for this sort of thing, although it does have the problem of unstable URLs. BLongley 18:16, 15 August 2008 (UTC)

In the meantime I'm quite happy to continue with the guidelines above and see what problems turn up. I'm sure there's a few guidelines we will want to add: e.g. "put year in full, '90 doesn't give enough details when we have magazines from 1890 and 1990". BLongley 18:16, 15 August 2008 (UTC)

Will anyone object if I write up a new help page based on the above, and add links to it from some of our existing help pages? -DES ^Talk 18:26, 15 August 2008 (UTC)

Not me. I might note one other issue that occurs a few times in the current database but will become more common if sf content from older mags is entered. Many of the stories in the old magazines were published as serials and it was not unusual that these were of novelette and novella length. If they were later reprinted they were likely to be printed whole. These cannot be merged so I guess the only way they can be documented is in the notes.--swfritter 15:53, 16 August 2008 (UTC)

Go ahead. I'm eager to see how many guidelines my submissions have broken already. (Nobody's given me any feedback on the ones on the queue.) BLongley 17:47, 17 August 2008 (UTC)

Response to David: yes, please rewrite the help pages, as I believe we've come to a consensus (hurrah!) And a response to Bill: Stop boo-hooing :-) I really like what you've done (something that I'm not able/willing to take on myself.) Nothing I've seen has broken any rules (even the "Editors of..." credit isn't that bad.) AFAIC, let 'em through to join their fellow records. MHHutchins 18:09, 17 August 2008 (UTC)

If anyone wants more entries to play with or to bump rules against to see how they work, a lot (but not all) of Bleilers "Science Fiction: The Early Years", Short story and book index up to 1930 is available at Google Books, I haven't talked myself into buying it yet.... kpulliam 18:30, 17 August 2008 (UTC)

Looks good to me. I know I did at least entry with a note about first pub in Blue Book so I'll dig that one out and add it. Dana Carson 22:48, 17 August 2008 (UTC)

Ok, I've approved the "Playboy" and "Saturday Evening Post" entries as those are the most common. Go check, and add comments to the talk page for the Help that DES is going to add. I still feel we're going to encounter problems, so I haven't done any merges yet. If this is a mistake, I want to take some of you down with me! ;-) BLongley 00:49, 18 August 2008 (UTC)

The couple of playboys I looked at looked fine. And now I have an excuse to look for old playboys on ebay... Got to verify those pubs and input page numbers. Hehe kpulliam 01:12, 18 August 2008 (UTC)

Just spotted one thing... Do we want the 'Editors of' marked as a pseudonym of 'Unknown'? Isn't that going to cause problems if we ever want to 'upgrade' one of these magazines series by adding the known editors? kpulliam 01:18, 18 August 2008 (UTC)

Is the consensus to go with "PUBLICATION EDITORS" or "EDITORS of PUBLICATION", and is the appropriate lastname "PUBLICATION", "EDITORS", or "LAST WORD OF PUBLICATION NAME"? I was going to enter a couple of Boy's Life and I ran into a previous entry. Boy's Life that went with "ED of PUB" with lastname of "Life". What do you folks suggest? kpulliam 00:31, 19 August 2008 (UTC)

[unindent] I think we came to a consensus of "Editors of ...". You won't need to create an author record, so you don't have to worry about which word is the lastname. The system will set up to author data record and "Life" would be this "author's" lastname. (Look at this author record.) Keep in mind that the publication is titled Boys' Life (note the placement of the apostrophe). MHHutchins 00:44, 19 August 2008 (UTC)

Someone with more experience than me might want to run a search for "Editor" in authors and change a few things around. I count 12 of 23 as not conforming to this standard and one of them Argosy Editors was one of the test subjects above done as PUBNAME EDITORS with lastname of EDITORS. That's why I wasn't sure we had reached consensus yet. Thanks!kpulliam 01:07, 19 August 2008 (UTC)

I went with "Editors of" rather than "The Editors of" in the end as it saved my fingers a bit. I've adjusted the ones I entered with a "The" now. I haven't messed with existing entries. The lastname is a minor issue, only really important if you're searching for an author via the author directory rather than simple name search. If someone cares, I'd suggest "Editors" for them all so we can find all such easily, or Magazine name if that's the way people think of them. But there's no rush to fix those, as Mike says, the system will assign something anyway. BLongley 02:16, 19 August 2008 (UTC)

A couple of oddities I did note: The Saturday Evening Post Reader of Fantasy and Science Fiction is currently by "uncredited". There WAS an entry by SEP editors, but it was a duplicate with no contents. It might be a good idea to make this by "Editors of The Saturday Evening Post" to keep it with the magazine entries, but then we need to clarify the rules for unknown editors of Anthologies. (We've only been dealing with magazines here, I think.) There are similar anthologies by Playboy editors which are credited to them as such, so that should probably be addressed anyway as we're already breaking a rule. BLongley 02:16, 19 August 2008 (UTC)

The other is a bit trickier: I DID find a genuine SEP (I think) Editor name, and left it in notes so I can't find it now. :-/ However, using such instead of an "Editors of" entry makes searching harder: even using it AS WELL AS an "Editors of" means we split up the entries and anyone searching for SEPs in general would have to search all the "Editors of The Saturday Evening Post/Real Editor Name" versions. When it's such a forgettable person as this one was, I'd prefer to leave it in notes: however, we could make such a person a variant of the "Editors of The Saturday Evening Post" name if that's all they're notable for. (Maybe a certain editor would indicate an increased likelihood of SF entries, for instance.) If they're more notable in ISFDB terms, then the magazine probably needs promoting to a more prominent status - not Magazine Wiki, but maybe Fanzine Wiki. As to making Editors variants of "Unknown" - I've never understood why that's been done before, so can't recommend doing it now. Can anyone explain the history of such? BLongley 02:16, 19 August 2008 (UTC)

I like the idea of making the few real names, variants of "Editors of" unless / until they get promoted (and if promoted, some of the work is already done). It seems like an all inclusive idea. kpulliam 02:42, 19 August 2008 (UTC)

I tried it with one example and it looks OK so far. As we can't undo pseudonyms at all easily though, I'm loath to try it with any more prolific editors just yet. BLongley 13:33, 19 August 2008 (UTC)

I would not make this more complicated than it needs to be and would definitely avoid pseudonyms. The key data in these pubs are the stories and not the editors.--swfritter 16:01, 19 August 2008 (UTC)

The consensus I spoke of was based on the current discussion about the future creation of stubs for non-genre magazines. Your count is somewhat misleading and only accurate if you discredit the pronoun "the" before "editors", otherwise only 6 of the 23 fall out of the format. And one of those 6 is a legitimate author name used as a byline in many publications ("The Editors"). In any case, those credits were created before the current discussion, and a search easily brought up the records, even for "someone with more experience". :) MHHutchins 02:12, 19 August 2008 (UTC)

I knew that the discussion was about the future... but with only a dozen records... now is probably the best time to try and clean them up, before people start looking for examples and follow the wrong example. (I almost used the Argosy entry as my template, as an example). No Biggie. kpulliam 02:42, 19 August 2008 (UTC)

Good eye. Have might my entry consistent. Even more important is to start putting the pubs in editor series so we can easily find them.--swfritter 16:01, 19 August 2008 (UTC)

I've just tried this. It seems an awful lot of edits just to be able to use "http://www.isfdb.org/cgi-bin/pe.cgi?24776" instead of "http://www.isfdb.org/cgi-bin/ea.cgi?Editors_of_The_Saturday_Evening_Post". What advantages have I missed? BLongley 16:44, 19 August 2008 (UTC)

Also note that I have started a section on the Magazine Page to link to the non-genre series. Strictly a trial run - open for suggestions.--swfritter 16:01, 19 August 2008 (UTC)

Well, looks like Bill likes the idea. The other thing I would suggest we think about doing is to merge the editor records by year.--swfritter 16:47, 19 August 2008 (UTC)

I'm still ambivalent about the idea of editor series for non-genre magazines, don't count that as "support" or "liking the idea" yet. I just tried them and stopped when I thought "what has all that work given me?". Which is "nothing", so far. Your suggestion of "merge the editor records by year" looks useful in the same way that it is for normal magazines, which is for when author pages get too big, I think? I'll count that as an advantage, but if a non-genre magazine gets that big then we probably shouldn't be treating it as non-genre, IMO. I note your comment "I would not make this more complicated than it needs to be" - well, variants are a bit complex for our average editor, and I'm happy to drop "Louis Ruppel" entirely. (I'm not sure Mike or DES are though.) EDITOR records ARE complicated - I think if you did a poll of current moderators half would admit they don't fully understand them, and if you polled active editors 90% or more would admit no clue. There IS an advantage, yes - thanks for pointing it out - but only in extreme circumstances. If there's other benefits I've missed, please state them: but at the moment I can't support the idea for small groups of publications. Just telling people what to do when there's a magazine worth recording, but the submission checks prevent such, is a step towards more data. Add too many rules and we just won't get people adding it. BLongley 22:17, 19 August 2008 (UTC)

(unindent) To answer Bill's question about the history behind creating vts for "unknown", I believe the primary reason was to get "house names" to display correctly. For example, take a look at Alexander Blade's bibliography. At this time we have two essays by unknown pseudonymous authors which are credited to Alexander Blade. Parenthetically, one of them is in a verified 1989 publication, which makes me wonder if we may be able to identify the real author? In any event, the way the bibliography is currently displayed, it looks like "Alexander Blade" was a real person whose name was also used by a dozen writers. Once we set up vts for the two outstanding essays, the page will change to say "Pseudonym: See [list of authors here]", which is presumably what we want to happen. Does this make sense? Ahasuerus 16:55, 19 August 2008 (UTC)

Not really. The thing that makes him look like a real person to me is "Used These Alternate Names: Henry S. Lewis". I suppose if we hadn't identified any of the real writers or had that alternate name then he might look real too: but just setting the Legal Name to "Unknown" or "House Name" would indicate the nature of the name just as well as a variant does, as far as I can see. I thought the real reason might be something along the lines of "let's make it easy to find all unknowns in case anybody wants to work on those in particular". BLongley 17:05, 19 August 2008 (UTC)

"Used These Alternate Names: Henry S. Lewis" was just someone's attempt to set up a pseudonym except that it was done backwards and now we are stuck with it until Al adds the ability to delete pseudonyms. As far as setting the Legal Name to "House Name" goes, well, it would be functionally equivalent, but as far as I can tell our users don't always notice such things, which is why we added "Pseudonym: See [...]" in the first place.

As an aside, there have been requests to display "all Titles published as by a certain pseudonym". This information is currently available under "Titles", but it's not easy to find and perhaps a new highlighted link next to the "Pseudonym: ..." line would be useful. Ahasuerus 18:15, 19 August 2008 (UTC)

True: it took me months to find, Mike admitted such too, and we've only just explained it to Don. BLongley 18:54, 19 August 2008 (UTC)

(inindent) Help:Entering non-genre magazines is now in somethinng like a complete state, although far from graven in stone. Please comment at Help talk:Entering non-genre magazines (as several people already have, and thanks much). I think this represnets something close to a consensus, but i would like cofirmation and suggestions for improvement. -DES ^Talk 05:53, 20 August 2008 (UTC)

Please see Help talk:Entering non-genre magazines‎#Editor names, where I have tried to lay out the options on whether and how to use actual names as opposed to "Editors of Mag Name". -DES ^Talk 16:00, 22 August 2008 (UTC)

Doubles again

I have moved this high traffic thread to Rules and standards discussions/Doubles and shortfiction --Marc Kupper|talk 00:39, 13 November 2008 (UTC)

Doubles again, redux

I am also one person who favors a fairly strict definition of "omnibus" and since doubles are being recorded as omnibuses, IMO the definition of omnibus must be considered in dealing with doubles.

IMO an omnibus should be limited to publications that: a) include two or more novel length (over-40,000 words) works, previously separately published, and remaining reasonably distinct in the combined publication; or b) include one or more novel-length works plus one or more anthologies or collections that were previously published in book form, and that retain their contents as a unit, and retain a separate title and identity; or c) two or more novel-length works that are presented with separate titles and identities, and at most one work of short fiction (a bonus story), where one or more of the novels has not been previously published, or has been published only in serial form.

These standards would mean that any omnibus would include at least one novel-length work of fiction; any publication that did not would be an anthology or collection. These standards would mean that any omnibus would contain multiple works of short fiction only as part of a previously published anthology or collection; any publication that did not would be an anthology or collection. These standards would mean that some ACE and Tor doubles would be omnibuses, and some would not. If that last point is too unwelcome to others, i could support a special exception for the historical ACE doubles only: they could contain two novellas or one novel and one novella, and still be considered omnibuses. The Binary Star publications, as described above, seem more like anthologies to me than omnibuses anyway. The Tor Doubles I'm less sure about, it would be desirable to treat them alike, but not IMO as important as the ACE doubles: there are far fewer of them, and thy don't play nearly the same role in the history of the field. -DES ^Talk 20:02, 30 September 2008 (UTC)

I'm against strict standards but rather use guidelines and try to find appropriate fit between what the books appear to be, what ISFDB can do, and personal opinion on the part of the editors/moderators involved. The guidelines give guidance so that we tend to operate in a consistent form.

I don't like the 40,000 word definition for novels as we don't know the word length for nearly all stories and it's difficult to get an exact count. Thus, I use a more easily verifiable standard that if a story is published as a book with it's title/author "headlined" on the cover and title page then it's a novel. If a story never appears as the "headline" then it's not a novel. The length of the work is not a factor and it also means that works that never appear in print (mostly electronic works) are not novels. If someone's willing to vanity press an empty book they have created a novel. If a work is usually short then this can be mentioned in the title notes and also the storylen field can be set.

While we have long had a storylen field the value input here has always been entirely subjective and as far as I know, has never been based on an official source. For example, I don't think anyone has ever gone down one of the awards lists and noted citations that this story was entered in the nv, na, nt, ss, or vi category for a judged event. We have some support for those stories that win awards but I'm not aware of effort to make sure the storylen values for those titles match the awarded category. The only time I pay the slightest attention to an editor's changing the storylen is if then go from something to sf or whatever the default is when a storylen is not defined. When entering and verifying books I copy/paste the contents into a spreadsheet that then parses and tells me if any storylens should be adjusted. The spreadsheet estimates the storylen values based on the page count.

I believe we are in agreement on what an omnibus is with one exception which is that as both the double and dos-a-dos formats "headline" story titles and their authors that automatically tends to make the stories "novels." It's not an absolute thing for doubles and but dos-a-dos publication has a stronger case for classifying a story as a novel. Marc Kupper (talk) 21:03, 30 September 2008 (UTC)

Actually i have adjusted storylen fields based on actual wordcount. Any time I enter an ebook publication, i download the texts and check actual wordcounts on all contents, and adjust the storylengths according to our announced standards. Similarly, i have used the published word counts from the Vance Integral Edition to adjust the story lengths of many of Jack Vance's works. I hadn't thought of comparing with award nomination lists, but it now seems like a very good idea to me, and I'll probably do at least some of that. And when i am unsure, but have the actual publication in front of me, i frequently do physical counts on a sample page, and multiply by pages printed to get a pretty accurate word count. I don't buy the "headline" argument at all. If we followed that rule, than anything listed as a "complete novel" in its serial publication should probably be a novel. The length rules in Help:Screen:EditPub are not worded as "guidelines" but as clear-cut hard and fast "definitions" and i treat them as such. Granted we can't always get exact wordcounts, but when we can, we should follow them, IMO. And we can get counts in far more cases than you might think. -DES ^Talk 21:23, 30 September 2008 (UTC)

I have been sitting on the fence re: this issue for a long time. As I wrote back in July:

• Currently, we use a very simple and inflexible categorization system, which is based on an objective criterion, i. e. the number of words in the text. Of course, counting the words is only possible when you have access to the publication, but once you have done it, there can be no disagreement whether the text in question is a Novel or a Novella, which at first glance is a Good Thing.

• Unfortunately, this lack of flexibility sometimes leads to situations where we violate the Principle of Least Astonishment. For example, a faithful application of our rules would change many YA books to Novellas (or even Novelets!), which is presumably not what our users expect. Similarly, many older novels like The Trouble with Tycho -- which were first serialized in the 1940s-1960s and appeared as standalones in the 1950s-1970s -- would become Novellas which is again probably not what our users expect.

I am still torn on this issue, but I definitely don't think that it would be useful to call 6 page stories published as standalone chapbook "novels", which I recall Marc experimenting with at one point. That would just absolutely slaughter the Principle of Least Astonishment :)

On the plus side, we can always find works of short fiction published as standalones with a simple script and adjusts them later on, perhaps even automagically. I am usually more worried about proposed band-aid solutions that do permanent violence to the underlying data.

As far as comparing our short fiction categorization data with the categories used by the Hugo and Nebula crowds goes, it's a useful project, but we have to be careful. Their current guidelines match ours, but I seem to recall that their standards have changed over the years and, besides, they have been known to move borderline works from one category to another on a case by case basis. Other awards use very different criteria, e.g. World Fantasy Awards have categories for Novellas and Short Stories, but not for Novelets, so everything over roughly 25 pages is a novella in their world. Ahasuerus 22:25, 30 September 2008 (UTC)

Ahasuerus wrote a faithful application of our rules would change many YA books to Novellas (or even Novelets!), which is presumably not what our users expect. - The rule writers are smarter than us as the rules already have at Help:Screen:EditPub#EntryType under NOVEL "If you know something has been independently published as a novel, it is safe to mark it as a novel." :-)

I tend to give more weight to this latter criteria over the 40,000 word criteria as it's easy to verify. At present there are no easy methods for getting an accurate word count for most of the 321,809 titles in the database though we could do use the page count method to narrow that down to a list of shortfiction titles that are 100 or more pages and also a list of novel titles not published as novels of the same name that also are less than 100 pages.

DES wrote: anything listed as a "complete novel" in its serial publication should probably be a novel - The headline in that case is the magazine title and not the "complete novel" it's featuring. The help already discusses that "complete novel" on a magazine cover is not evidence that the story over 40,000 words. Marc Kupper (talk) 22:58, 30 September 2008 (UTC)

(unindent) Someone PLEASE split this thread, it was started as a nice simple discussion about Ace Doubles, and how we actually achieved consensus for Months (if not quite a full Year), and could we now update the help without any further disagreement? Note that I wasn't proposing any CHANGE to what we actually HAVE. Just what other Doubles might be IN. BLongley 23:25, 30 September 2008 (UTC)

"Binary Star series" are obviously OUT as too controversial. "Omnibus" is a separate issue - you can argue about the content categories as long as you like, fixing an OMNIBUS status for DOUBLES will not affect the content lengths at all. Two "Books" might define an Omnibus, two "Novellas" might, a "Collection" as one half might - that's irrelevant though. We have the opportunity to agree on DOUBLES. If we can agree on that (as we seem to have done last year) then people can go back to arguing about whether the CONTENTS are right. BLongley 23:25, 30 September 2008 (UTC)

What's the point/question? You said we have consensus about something on Ace doubles and for some reason you say "Binary Star series" are obviously OUT". What are these in and out of? Marc Kupper (talk) 23:47, 30 September 2008 (UTC)

I linked to the original article at the beginning of this one, and that points to the Help pages that disagree. The one that was suggested for update was The FAQ for dos-a-dos books such as Ace Doubles Binary Star have been established as NOT being dos-a-dos, I understand? That would make them out. BLongley 18:20, 1 October 2008 (UTC)

But we don't really have consensus. I agree that we should change the help to better match what we are doing, and/or what we now agree we ought to do in future, and that the various help pages should be consistant. We agree on much of what should be done on doubles. But the help pages are going to contain instruction on how to enter a double, adn that is either going to say that all doubles will be omnibuses, or that soem will be soemthign else. It will also say either that all content items in doubles will be recorded as novels, or that some will be recorded as novellas. Unless we agree on which of those things the help will say, we don't have consensus to change the help. Those are the points at issue above, so they are all relevant to the question of changign the help on doubles, and so this thread can't usefulf be split into doubels and otherstuff. It could be split into Novels (what are they?) and Omnibuses (what are they?) but both would need to be settled before the help could be changed. -DES ^Talk 00:49, 1 October 2008 (UTC)

I should add that I don't understand what Bill means by some kinds of doubles being IN, and others being OUT. In and Out of what? -DES ^Talk 00:51, 1 October 2008 (UTC)

The help under discussion was for "How does the ISFDB deal with dos-à-dos books such as Ace Doubles, and Tor Doubles?" and it already does say that they're all Omnibuses, and that they will contain Novels or Novellas as appropriate. It doesn't contradict the word-length definitions of such, and in most regards reflects the current state of Ace doubles: it's titles and authors that need fixing to match the current practice. Given the complications with Binary Star, and the fact that they're not dos-a-dos, it seems wise to leave what's covered as is. BLongley 18:38, 1 October 2008 (UTC)

Actually i have a suggestion. Can we change the help to cover my points 1 & 3-6 above, and state that ACE doubles can be entered as omnibuses no matter what the length of their contents, but that doing so is still under discussion for any other doubles. It would also state that listign items in a double as "novels" if under the 40,000 word limit is still under discussion, and does not have uniform approval. Would doing this handle the ares where we do have consensus, and allow further diuscussion of the points where we differ? Will that work for everyone? -DES ^Talk 01:48, 1 October 2008 (UTC)

I'm fine with 1,3-6 except

• #4 - the bit about "or of the anthology when two ...".
• #6 - When I do doubles I add a note to documents both the authors and artists as ISFDB does not maintain the order reliably.

A couple of my doubles are either anthologies or collections and so if someone's constructing a help page then you might as well add notes about these. I wished someone had linked to the help articles that are under debate rather than "'as according to help' (if you read a certain section of help)" as I already thought this sort of stuff was in the help. Marc Kupper (talk) 09:45, 1 October 2008 (UTC)

Since I am the 'Tor' Transgressor, My thinking came from pb to the page length criteria. That caused me to 'double/triple' think the issues. Sorry but the 'backtracking' extends to other areas such as the Ace Perry Rhodan series. Word length should never trump common db user perceptions. Title to novel is his most common usage point.

1) First admit that 'Doubles' are a publisher grouping of stories, basically to get the smallest novels and larger novellas into a marketable state. 2)State that all publisher doubles (not author(s) duology sets) be stated as dos + omnibus at all times irregardless of actual word length. Adding page counts and cover artist separate crediting to make it easier on the user. 3) Limit the page determinations for category to the contents area. Parse this so the 'db user' knows that novel is per publisher criteria at the category selection area, while novel, novella, short story are parsed at the title entry (in notes) and content area in the classical way and even then not by 'Awards' commissions, etc. Note the variances in the notes field (title & specific entry) as much as possible, but do not let them 'reshape' the classification as the publishers have done for their publishing purposes.

3) is the important factor of this discussion. If I am right this method generates the least confusion for the db user, not always necessarily for the moderators and editors. The db user cannot be expected to parse the finer points as he is looking for general data and the researcher db user should be expected to understand the rules easily enough, though he will undoubtedly have even more ways to subdivide each type of novel, novella, etc I am sure ISFDB does not want to go there.

Sorry that this is muddled, but this seems to be basically what is needed. Thanks, Harry. --Dragoondelight 12:50, 1 October 2008 (UTC)

Oh, by the way, I use a spreadsheet that allows the user to enter a fairly small amount of data from relatively random pages to get an estimate. It did not get a great deal of applause from the people I showed it to but it seems to be relatively accurate. And back to this discussion, Ace Doubles should probably have their own type but the current solution is quite acceptable.--swfritter 22:34, 10 November 2008 (UTC)

It looks like this. Keep it simple - allow the editor some discretion when determining whether a book with a word count less than 40k should be classified as a novel. There are just too many special cases to be able develop any concise and acceptable definitions. For shorter pieces of fiction those awards that use the same word count standards as we currently do (Hugo, Nebula) are significant.--swfritter 23:02, 10 November 2008 (UTC)

Techie-bit alert

My original thought was that it's a single-edit fix per pub, and such an ANTHOLOGY is picked up easily with SQL like:

select p.pub_id, p.pub_tag, p.pub_title
from pubs p, pub_content pc, titles t
where p.pub_id = pc.pub_id
and pc.title_id = t.title_id
and p.pub_title = t.title_title
and p.pub_ctype= 'ANTHOLOGY'
and t.title_ttype = 'SHORTFICTION'

and we don't have to go check whether there's any other publication contents. For a COLLECTION type, there's probably hundreds or thousands of collections named after one of the constituent short-stories. Not sure about OMNIBUS type, I'll run some queries and see how often we get those collisions. BLongley 19:02, 7 October 2008 (UTC)

OK, it seems that as of the last backup I loaded we have 137 Anthologies with a same-name shortfiction content record already. We have 1479 Collections with a same-name shortfiction title/author content (as I suspected). Only 3 Omnibuses with the same problem, so maybe that would be easiest for the workaround if people like it. Of course, I could spend more time writing more comprehensive SQL (my quick solution doesn't check authors, count other contents, etc.) but a standard publication note for anything we use this workaround for wouldn't add much complexity. BLongley 19:32, 7 October 2008 (UTC)

Bill - this thread got lost and I'm not sure what you are proposing other than you are reporting on anthologies and collections that have the same title as one of their contained stories. -Marc Kupper|talk 19:22, 10 November 2008 (UTC)

It relates directly to the suggested workaround(s) in the "Chapter Books yet again" thread immediately above (at the moment). That seemed to end with an agreement that as Chapterbooks are fundamentally broken at present we'd try keeping the Shortfiction as Shortfiction, but put them in a temporary "container" title type that's a bit more stable. However, as it's planned to be only a temporary fix till we get something else that pleases everybody, it's best to use a workaround that is going to be easily found and converted later. The statistics show that although a single-content "Collection" might look most reasonable in the mean-time it'll be difficult to find ones used for Chapbook-evasion purposes alone. "Anthology" is better, Omnibus better still. Of course, in this particular case people went with the multi-step "Magazine" solution instead so the discussion died out. BLongley 20:16, 10 November 2008 (UTC)

Musicals

A new editor just entered Weird Romance: Two One-Act Musicals of Speculative Fiction. Seeking to confirm it, I found that, according to OCLC at least, the same publisher also issued another SF musical libretto under the same ISBN, and I have duly entered it: Gorey Stories: A Musical Entertainment. I don't think there is any question that these are IN, both are written and published works of SF, or appear to be from secondary sources. But there are some questions about how to record them. User:Jayembee recorded the first as a collection, and since there are separate entries for the two acts, which are based on different printed stories, this seems reasonable. I recorded the second as a novel, although perhaps it ought to be a chapbook or a collection: It is said to be based on 18 short stories by Edward Gorey, but I don't know if those stories have been treated as separate mini-plays (which would probably make this a novel) or if this is a sort of musical fix-up; I rather suspect the latter. There is also the question of author listings: Should the authors of the stories on which these musicals are based be listed as authors here, or not? I could make arguments either way. OCLC lists them as authors, and so did User:Jayembee (who may well have a physical copy of Weird Romance), so i followed suit. Still, a general policy may be a good idea in case this situation comes up again. Of course Implementing Feature:90164 (based on) would help a lot with this, but I'm not holding my breath. -DES ^Talk 17:30, 29 October 2008 (UTC)

Jayembee here. I found the wiki. I'm slow, but I CAN be taught. First off, yes, I have a physical copy of the book I was working off. I see that I haven't verified the entry yet. I shall do so. As for crediting the authors of the original work, in the case of the second act ("Her Pilgrim Soul"), Brennert is the original author, so the question's moot. But for the record, the title page of the book reads:

Weird Romance

Two One-Act Musicals of Speculative Fiction

Music by Alan Menken Lyrics by David Spencer

"The Girl Who Was Plugged In" Book by Alan Brennert and David Spencer based on the story by James Tiptree, Jr.

"Her Pilgrim Soul" Book by Alan Brennert based on his original story

It really seemed wrong to not list Tiptree as a co-author of the first, since at the very least, she was responsible for the plot (I've never gone back and compared the script with Tiptree's story to see how much dialogue is taken directly from Tiptree; I should do that, if only for my own curiosity). For the record, while the script refers to "Girl" as "Act I" and "Soul" as "Act II" - which suggests that the musical is a single work - the subtitle of the work refers to them as two one-act musicals, which suggests that they be treated as separate works. That and the fact that they come from separately previously-published stories was the reason why I felt it should be entered into the ISFDb as a collection. Jayembee 08:56, 2 November 2008 (UTC) Jayembee

Well, we don't seem to usually credit the original Script-Writers for novelised TV shows except in notes, and even the book of the film of a book by someone else doesn't lead to a co-author credit. We seem to have Once More with Feeling here already though so maybe musicals are an exception. Or maybe they're not - I see no Rocky Horror of any kind here, and Little Shop of Horrors looks a bit too short to be a valid entry. I can see a slippery slope here: doe anyone want me to enter "Spock in Manacles"? I'd be quite happy to enter it, or my favourite musical. BLongley 20:20, 29 October 2008 (UTC)

If the script or book of a Musical, or for that matter of a TV show, is published in print (or as an ebook i suppose) It seems to me pretty clearly in. We do have "Phoenix Without Ashes" which is the script for a TV episode. (As for Return to the Forbidden Planet, it seems to be in print in pb form. Assuming it is in fact SF, I would enter it if I had a copy.) -DES ^Talk 21:09, 29 October 2008 (UTC)

It is SF - Laurence Olivier Award winning SF even. (And that is a significant award. I've held it, and could easily use it as a lethal blunt instrument.) That book is actually more tp than pb but at least is more script-like than some other books based on the show, which are just the songs with keyboard and guitar arrangements. The songs are an integral part of the plot, but so are William Shakespeare's The Tempest and the original Forbidden Planet film. As I say, slippery slope... still, we're sliding down it. I see that under current practices Belaboring the Obvious probably now gets in for its audiobook excerpt. We never had problems like this when I was an ISFDB trainee... :-/ BLongley 22:19, 29 October 2008 (UTC)

Return to the Forbidden Planet is most definitely SF. I saw it during it's original West End production. Anyway, I was going to say that maybe the original script-writers of novelized TV show episodes or films (or original authors of stories they're based on -- like "Arena" in Star Trek 2 being a 3-way collaboration of Blish, Coon, and Brown :-)) SHOULD be given a co-author credit. But then, as the Trek example shows, that could open a can of worms. At least we don't have the Thunderball mess to deal with (aside: I'm rather surprised, though, that neither Live and Let Die nor Moonraker - or the novelization of the film version of the latter - are in the database)

As a counter-example, perhaps, there's the case of I, Robot: The Illustrated Screenplay, in which Asimov is presented as co-author, even though (as far as I can see) he is only responsible for the original stories that Ellison adapted, and didn't have any hand in writing the screenplay itself. Jayembee 08:56, 2 November 2008 (UTC) Jayembee

Whether to grant the authors of the works adapted a co-author credit is another matter. I am working from secondary sources here, or from the the submission of an editor who has not yet found the wiki, so i can't say with assurance if such credits were given on the publications themselves. All I can say is that the secondary source (OCLC) gave such credit. -DES ^Talk 21:09, 29 October 2008 (UTC)

We've got a lot of rework to do if we start giving co-author credits for authors of an otherwise non-enterable version of a work, that get credited in the later entered versions. (Does that make sense? I basically mean that we could go back and massively increase the number of books co-written by Joss Whedon, Gene Roddenberry, etc, and even introduce some like "Constance M. Burge".) I'm not against such - it might keep me busy this winter. It does make us look a bit more visual-media orientated than I think we usually are though. BLongley 22:19, 29 October 2008 (UTC)

Anyway i wanted to call attention to what I had just done, and get opinions on whether it was a good idea or not, in the matter of the author credits. I also wanted to draw some attention to the somewhat odd (though not unheard of) situation of two quite different books with the same ISBN, at least according to OCLC. -DES ^Talk 21:09, 29 October 2008 (UTC)

Different books with same ISBN shouldn't happen, but I have seen it elsewhere. All we can do is note the collision and cross-reference - it's rare to be able to say which one is "right". BLongley 22:19, 29 October 2008 (UTC)

(Unindent) I think we've got a few issues here, it might be worth separating them out fast. 1) SF musicals printed in book form - I suspect OUT if they're just songs, IN if they have the plot? 2) Co-credits for the authors of media versions, or only if they wrote it before it got a media version before a book version? (Obviously co-credited if the book co-credits them as authors.) 3) Duplicate ISBNs - I think all we can do is as I stated above, but I'm open to suggestions if people think there's a way of distinguishing the "correct" one. BLongley 22:19, 29 October 2008 (UTC)

I agree that these are separate issues. In fact, i'll create sub-threads for them. -DES ^Talk 22:58, 29 October 2008 (UTC)

SF musicals printed in book form

If these are nothing but song books, I agree that they should be OUT. But if they have significant text beyond song lyrics (novelizations, sort of), or dialog as a printed script would or some combo, then i would say IN. Actually I'm rethinking the first. We already have SF poetry collections, so a book of song lyrics, provided that the lyrics are SF, should probably be IN, but I'm less sure of it than a true script or libretto or "book". -DES ^Talk 22:58, 29 October 2008 (UTC)

OK, let's confuse the issue even more. I'd be inclined to say that song books from SF musicals should be in if they they are simply published lyrics. If they are actual music books, that is contain sheet music, then no. Then again, I see that Swann's The Road Goes Ever On is included, so... Jayembee 08:56, 2 November 2008 (UTC) Jayembee

Ok, submitted but immediately held this and cover image is here. Other background here. The next slippery step is to allow Filk songbooks in ("Yay! Bill gets four more credits in ISFDB! And then I can push for Award entries for such, and get one of those too!"). We've already got people slightly misusing "POEM" for filks in certain C. J. Cherryh, Leslie Fish and Mercedes Lackey books. (Less important standards discussion - capitalisation of BECCON.) BLongley 23:27, 29 October 2008 (UTC)

Well, what is printed in the Cherryh and Lackey books are the lyrics to filks, which are poems, or at least verse, by any definition i can think of, and many of Kipling's poems were apparently written to fit popular music hall tunes of the day ("The Liner She's a lady" I've heard sung to a period tune) and we have some of them on record. Verse, whether constituting lyrics or not, included in a professionally published SF work, seem to me clearly IN, and "Poem" fits them better than "Shortfiction", IMO. Filks, and filk songbooks, are a step down towards amateur levels, but then we already include fanzines and self-published books that perhaps have sold 2 copies. If we include them, I'm in, as I was the editor and publisher of the MSU Tolkien Society Song Book, vols 1-3 (Singer's and Musician's editions). I still have copies, too. On the rest of this, my opinions are up-thread, i won't repeat or expand on them, for once. -DES ^Talk 23:47, 29 October 2008 (UTC)

I think we have discussed POEM or ESSAY use/misuse for "SONG" entries before (probably numerous times in fact) and I find I use POEM more often. SONG might be nice if we have the musical arrangements to go with the lyric. (But then we'd get complaints about whether a piano or guitar arrangement was enough, and why don't we credit the arranger(s)...). So long as there's words, and they're SF, I'm inclusionist. I have no idea how my published filks would be categorised really: one I'd count as OUT as computer-related only, another OUT as only related to another filker (although a filker that is a fictional character in some major fantasy books), the award-winner might be out as it's a filk of a filk of a Comic-book story, and the last is so unmemorable I can't even remember what it was about. But the overall song-book might be in, despite my contributions. I sing in the Key of Z and it sounds like L - but I apparently write nice words at times. BLongley 00:13, 30 October 2008 (UTC)

FWIW, I have been using POEM as a substitute for SONG as well since songs are generally works of fiction and making them Essays seems rather counter-intuitive. I always enter filk pieces as POEMs when I encounter them in magazines and anthologies (not that it happens very often), but sheet music books seem to be a step too far. Ahasuerus 01:22, 3 November 2008 (UTC)

Co-credits

The case I originally raised was a co-credit for the author of the conventional story. For example, if the script for the Peter Jackson LOTR movies was to be printed, should J. R. R. Tolkien get a Co-credit? All of the plot and much of the dialog is his, after all. In any case, if the printed book co-credits a "based on" author, the answer is easy. If it merely has a credit "This musical based on the stories of Joe Doaks 1905-1972" for a 1997 work (so Doaks clearly did not contribute to the musical itself) then probably only a credit in the notes? -DES ^Talk 22:58, 29 October 2008 (UTC)

Co-credits to the creators of movies or TV shows in the novelizations thereof is a different matter, although related. Probably just in notes unless credited as a co-author on the title page? -DES ^Talk 22:58, 29 October 2008 (UTC)

Again on the slippery slope, and with the example above: we don't have the "Spock in Manacles" fanzine by Kate Davies. The Fanzine is probably in if we could find it, after recent changes, the entry in "The BECCON Plays" is certainly credited. Date issues might ensue. I assume "Musical Arrangement" and "Producer" or "Director" credits and cast-lists are out, however famous the names are. (Yes, Lionel Fanthorpe did appear in his own musical, and you might have heard of some bloke called Geoff_Ryman appearing as Spock. Presumably in manacles of some theatrical sort.) BLongley 23:43, 29 October 2008 (UTC)

Duplicate ISBNs

split off from discussion in the prevuous section: "Musicals"

Not much to do except note them, that I can see.

The most extreme case that I know of: the 44 volumes of the "Vance Integral Edition" (we have the first 13 or so listed in the db so far) use exactly 2 ISBNs, one for the volumes issued in 2002, one for the ones issued in 2005. They were only sold as a complete set, you see, so only one ISBN was needed, they thought. -DES ^Talk 22:58, 29 October 2008 (UTC)

I'm in favour of claiming all SF-related ISBNs. Not so keen on applying an overall ISBN to each individual book. BLongley 00:19, 30 October 2008 (UTC)

The OCLC entries list an ISBN for each book in the VIE. It just happens to be the same one, over and over. Since it was never for sale at retail through any of the usual outlets, none of them needed a unique code. Makes life harder for us, and for other bibliographers, of course. -DES ^Talk 03:30, 30 October 2008 (UTC)

Quite often specialty publishers, e.g. Phantasia Press will place the same ISBN on both the trade edition and the deluxe edition. Is the deluxe edition a truly separate edition, or is it simply a boxed and signed version of the trade edition? I chose to create separate records for each. Sometime I need to check and see which edition our ISBN links lead to. MHHutchins 06:49, 2 November 2008 (UTC)

I have been using Michael's approach as well since it makes physical Verification easier. Also, keep in mind that a publisher may claim on the copyright page that other editions exist, but for various reasons they may never appear. Ahasuerus 01:16, 3 November 2008 (UTC)

Baen's "e-arcs": IN or OUT

First three entries copied from the ARC thread well above, as the topic has driftend and goe live after a lapse. -DES ^Talk 12:33, 2 November 2008 (UTC)

Baen books has fairly recently created something that they call an "e-arc". This is an ebook version of an unproofed, late but not final draft of a book, which they put on sale usually 1-2 months before the release date of the final version of the book. Since they sell this (typically for more than the cost of a pb, but less than half the cost of a paper hc, and more than the cost of a released individual ebook), and it is downloadable like their other ebooks, it surely exists in a verifiable state. Should such "e-arcs" (their term) be listed here? They will generally differ at least in detail, because the final copyedit pass has not been done on them, and in a few cases significant late rewrites have, i understand, occurred after the release of the e-arc. -DES ^Talk 19:54, 15 October 2008 (UTC)

I believe all ARCs should be out, even electronic ones. I can't see why an "e-arc" should be treated differently. MHHutchins 06:41, 2 November 2008 (UTC)

The difference, in my mind, is that these "e-arcs" are actually being advertised and sold. The traditional ARC was only distributed to reviewers and buyers for bookstores and distributors, and indeed generally carried a large "NOT FOR SALE" notice. Fans only got these if they knew someone on the distribution list, or if someone on that list sold a copy in violation of agreement. But with Baen marketing these, arguably they aren't "really" ARCs at all, but separate editions. -DES ^Talk 12:33, 2 November 2008 (UTC)

I guess the key issue here is that these "ARCs" are not exactly "advanced reader copies", at least not in the sense that the term is usually understood. As long as they are sold to the public, you could argue that they represent just another (albeit raw) Publication of the Title. Hm, maybe we need a new term for them? Ahasuerus 01:12, 3 November 2008 (UTC)

That was more or less my thought. no new term springs to mind, however. -DES ^Talk 02:18, 3 November 2008 (UTC)

This begs an even deeper question. Why would an author allow an unedited version of a novel to be sold at all? Has all integrity flew out the window in order to get a few extra bucks? Jeez, what's next? Hawking your wares out of the trunk of your car on the corner of an abandoned strip mall's parking lot? MHHutchins 04:37, 3 November 2008 (UTC)

$15.00 and apparently you don't get a copy of the final product. The process does allow input to the author from the readers. They are only available for purchase for a limited time - although they can still be downloaded if already purchased. As long as there are not multiple versions of the "e-arcs" I think they are valid.--swfritter 15:11, 3 November 2008 (UTC)

Before e-arcs Baen was known to (and perhaps still does) release an arc or two on ebay, and gauge the print run order (or at least order another print run, or perhaps just cackle with glee) before release based on the bidding or demand for the arc. I have no doubt that Toni & Co. continue to use e-arc sales to justify or re-adjust expectations before some books even hit the shelves. They have been known to squeeze $15.00 (e-arc) + $3.75 (1/4 $15.00 for month of e-books) + $26 (HC) = $44.75 out of my wallet before they are done on some (apparently addictive) works. Personally I think all ARCs should be in. The binding on print editions is usually TP when the first official printing is HC, and I have at least one arc with a different (non-artistic) cover in mass market. On the electronic editions, they are available to own, pay royalties, and justifiably catalog-able in a personal or institutional collection. What is a Bibliography if not a checklist of total published works and editions for the completist collector.Kevin 04:39, 6 February 2009 (UTC)

Non-Roman characters in titles

This pub of Autumn Angels includes the introductory essay "The Phi Psi Omega of A B C" by Ellison. It also includes the note "The title of Ellison's introduction uses the actual Greek letters for "phi", "psi", and "omega" rather than spelling them out in English". Now we could enter this as "The Ψ Φ Ω of A B C". But should we? It would be rather hard to search for. That probably wouldn't matter much with an introductory essay, but might matter if such characters were used in the title of a work of fiction. Opinions? -DES ^Talk 14:36, 2 November 2008 (UTC)

Somehow use variant title? one of which would not be in a pub.--swfritter 14:59, 2 November 2008 (UTC)

That would work. In line with our usual rule, I suggest that the version with the actual characters should be the parent, and should be in the actual pub, at least in this case? -DES ^Talk 15:20, 2 November 2008 (UTC)

That's the way it is most likely to be listed elsewhere so that makes sense.--swfritter 16:05, 2 November 2008 (UTC)

Roglo and I played with non-Latin-1 character sets back when he was cleaning up Stanislaw Lem and I was trying to add foreign language sequels to well known English language series. It didn't work too well in the context of the Author field, but Titles were mostly OK. There were some strange ramifications, though, e.g. do a regular Title search on "42" and review the first three records. Also, if you go the Variant Title route, pseudonymous Titles become even more convoluted than they currently are, but not insurmountably so.

My suggestion would be to pick a few sample Titles, change them in accordance with this proposal and see how it works. Ahasuerus 16:55, 2 November 2008 (UTC)

Done for this case, see the links above. -DES ^Talk 17:22, 2 November 2008 (UTC)

Novelizations

Based on the discussion above, do we want to replace the highlighted words in Help:Screen:EditTitle:

• nvz - Novelization - Indicates that this story is a novelization of an earlier shortfiction. Note that this code is not available when editing the Contents section of a publication but you will see it when editing title records.

with a movie, TV show, game or other media [work?] to reflect the current usage? Ahasuerus 01:28, 3 November 2008 (UTC)

I could go along with that. Do we want to include the case of expansion of a short fiction work to novel length as an alternate use for this code, or use it only for the media->novel case? -DES ^Talk 02:16, 3 November 2008 (UTC)

Yes, the help pages should be changed as Ahasuerus suggests, and no, it should not include the expansion of a shorter work into a novel. IMHO. MHHutchins 04:30, 3 November 2008 (UTC)

I've never used it for expansion of short fiction, only for books based on other media. And not often then - e.g. despite re-doing over 50 Doctor Who books yesterday (for foreign prices and covers) I didn't feel up to adding Wikipedia entries at title level for them all (although I suspect they're all there), and even the few I did I forgot to put 'nvz' on. I totally agree with Michael on help changes. BLongley 19:23, 3 November 2008 (UTC)

Help changed. -DES ^Talk 19:54, 3 November 2008 (UTC)

Seems a bit quick, but that may be the only way we'll ever change things nowadays. Talk too long and we go nowhere. BLongley 23:21, 3 November 2008 (UTC)

Anyway, while the iron is hot I've adjusted Starfleet Academy. I suspect the next question is whether the "Wikipedia Entry" at title level should be used for links to a Wikipedia article that doesn't include the NOVEL that's here. But if that's worth talking about, please split it out as a separate topic or we'll just get the current help change reverted. (Same if you want "excerpts" recorded better than I did there, or any other issues. Small, bite-size topics please!) BLongley 23:21, 3 November 2008 (UTC)

Actually, looking at that, where does "nvz" help anyone anyway? I don't recall seeing it any time apart from title edits. BLongley 23:25, 3 November 2008 (UTC)

"nvz" and "jvn" were implemented way back in 1995 because John Wenn was using them in his bibliographies and we just followed his lead. I think we displayed this information in Summary Bibliographies for a while, but the current code doesn't display them just like it doesn't display our ancient Author biographies which are still buried in the guts of the database. Shouldn't be too hard to make the abbreviations reappear, though. Ahasuerus 01:39, 4 November 2008 (UTC)

New Infinities and Byron Preiss (Concept copyright publishing?)

This. [1] . When I verified this publication the notes tell it all. New Infinities is not a publisher, but holds copyright material used in the book. They may have even hired Ace to print and distribute their books. New Infinites left logos on covers and spines, but again seems to be a 'jobber' for Trigee Enterprises Corporation (A Gygax after TSR event). I do no wish to list New Infinities as a publisher and wondered if New Infinities should be tagged as not being a publisher. Byron Preiss fits the same mold. Do you want it to say ACE / New Infinities or just be noted. Do we give publisher credit for 'Concept' copyrights? I also did Cyborg Commando #1, before I realized the strange relationship of New Infinities. #2 is at hand and it matches #3 my example. Thanks, Harry. --Dragoondelight 21:39, 3 November 2008 (UTC)

I would document what is stated on the front cover, spine, back cover, title page, copyright page, etc. I personally consider the "publisher" to be what's stated on the bottom of the title page and so if I had a publication that used New Infinities there then that's what I use. FWIW, here's a New Infinities / Berkley Publishing Group book. I also would consider them a "Publisher" as it seems they performed many of the roles associated with publishers. Book manufacturing is only a small part of it. For example, I don't think DAW Books has ever manufactured books and yet most people would consider them to be a publisher.

As for the concept copyright - You can document it in the notes. We use the author/edit credit that's stated on the title page as the author/editor of a publication. When I see that the copyright is in someone else' name I document that. For example, I recently verified this. -Marc Kupper|talk 03:56, 4 November 2008 (UTC)

Thanks. I usually put the copyrights in for someone may find them useful. I enjoy thinking the why. I always put it in if it is not the author or even if the author uses a variant name. Your publisher example would fit into the same scenario and when I do my Gygax's I'll cross check for New Infinities, etc.

In the Cyborg Commando series, the title page is no help. Nobody is there. The copyright page lists Ace and New Infinites for their identity & logo copyrights, but none are specifically to the book. New Infinities sprinkles it's logos and gets them on the front and spine. Trigee, owner of New Infinities, gets a copyright credit but it's placement could very well be for the Cyborg Commando series itself or just maybe the book. Ace is on the spine.

The problem then is of definition. Ace does not claim the book. New Infinities owns it's trademark, but does not specify a claim to the book. Berkley group distributed the product to the book trade. Trigee owns the copyright to 'Cyborg Commando'.

Top of copyright page reads thus. Chase Into Space (over) (spacing) A Cyborg Commando(tm) Book (over) (spacing) Copyright 1988 Trigee Enterprises corporation. All Rights reserved.

Problem then is who? It is like a group of lawyers were able to take and not take all creit for everything and nothing. How should it be displayed. Trigee, New Infinities, Ace. With the imprint publishers how? Trigee / Ace, Ace / Trigee, Ace / New Infinities, New Infinities / Ace, Trigee / New Infinities / Ace. Trigee + Ace + New Infinities since none are parts of the other? Thanks, Harry. --Dragoondelight 16:58, 4 November 2008 (UTC)

Since they are using an Ace ISBN Ace should definitely be in the publisher name meaning it's probably "New Infinities and Ace Books" (or Ace Publishing if it's a book done before Charter bought them). -Marc Kupper|talk 06:22, 5 November 2008 (UTC)

I did not think of the ISBN publisher angle. Since Ace Books has the history and ISBN, and New Infinities may well be (probably is) already in the wind. I think Ace Books and New Infinities. If someone had to get records on the book, Ace would be the place I would start. Thanks, Harry. --Dragoondelight 13:18, 5 November 2008 (UTC)

Standards discussions

(Moved to Talk:Rules and standards discussions#Recurring topics.) -DES ^Talk 00:07, 13 November 2008 (UTC)

Thanks, folks, now I can go back to working on User:Fixer and not feel guilty. Or at least less guilty :) Ahasuerus 01:40, 13 November 2008 (UTC)

Anthologies with no stated or known editor

Should an anthology which combines several novellas by different authors, with no stated editor, and no secondary source info indicating an editor, be credited as a joint work by the authors, like this one or attributed to "uncredited", like this one. I just entered the latter, and I thought i was following our practices as described in the help. I think i was, but arguably the first example, which i found because it shares several of the authors, gives a better result. Should we change our policy and document the change in the help to credit such an anthology as a joint work of the authors involved, or what? -DES ^Talk 18:14, 18 November 2008 (UTC)

I have to admit that several Nora Roberts (a.k.a. J. D. Robb ) co-authored anthologies are down to me converting NOVELs with four titles by four different authors in back to something more workable - they're not finalized by any means. They might well be OMNIBUS for all I know. But as Dissembler (presumably) keeps adding them, I find the Nora Roberts bibliography a good example of where to experiment with some big changes that usually lead to some cover-artists getting demoted from co-author, and improve my language skills in the meantime. We don't seem to have many Vampire-Shagger editing fans around so that's always a good place to look for some messy data you can practice on. BLongley 21:52, 18 November 2008 (UTC)

Back on topic - I think three or four co-authors of an otherwise uncredited Anthology is fine. Too many and you've probably found an RPG sourcebook instead, and if it really IS SF then a big number of co-authors is probably unworkable. In these examples I'd guess someone at the publisher chose a "headline" name but I don't think the headline name deserves more consideration than the others, unless the publication says otherwise. BLongley 21:52, 18 November 2008 (UTC)

In the case of the two I linked to above, I think we have pretty clearcut anthologies, and i suspect any Dissembler items with multiple co-authors by Robb/Roberts (and Roberts isn't her legal name either, she was born Robertson) are also anthologies. On Suite 606 the front cover shows all four names, although Robb is in larger type, and the title page shows all four names with none given precedence over the others. Silent Night is, i belive, also a true anthology. -DES ^Talk 23:28, 18 November 2008 (UTC)

Uncredited makes sense to me. Crediting the authors assigns them a role they did not have. In fact, just got through un-crediting some authors. If you are going to credit any of them then they all should be credited.--swfritter 22:04, 18 November 2008 (UTC)

Yes, they are listed as "editors" which they really arent. OTOH, this is the only way to have the book listed on their summery bibliography page, or indeed on anyone's, adn where an anthology combines the work of three or four authors without a listed editor, it is a fair bet that some or all of the authors played at least part of the usual role of the "editor" of an anthology. Still, I'm not sure. -DES ^Talk 23:28, 18 November 2008 (UTC)

Oh I should add that I fully agree that there is no reason in this case to treat the authors differently, either all should be credited or none should, unless ther is actual evidence that one (or more than one but less than all) did actually take on the role of "editor". Being a "headline" author is IMO not relevant here. -DES ^Talk 23:30, 18 November 2008 (UTC)

A "senior policy" in my mind would be to go by what's stated on the title page. Sometimes that is ambiguous, for example, with the infamous Binary Star books I've attempted a reproduction of a title page. A case could be made that the title is Binary Star No. 5 with an uncredited editor. Or that it's Binary Star No. 5: Nightflyers / True Names with an uncredited editor and two credited authors. DES just entered Suite 606 as edited by uncredited. To me, that means the title page has Suite 606 and either does not credit the story authors, or does so in such a way that DES felt it was decoration and not part of the title. He later wrote that the title page is much like the cover other than the author names are all the same size. In that case I'd enter the book exactly as DES did which is an anthology Suite 606 edited by uncredited though I'd also add a pub-note that explains what's on the title page, that the anthology editor is not credited, and that the copyright is "Copyright © 2008 by Penguin Group (USA) Inc." implying it's an in-house editor.

A literal reading of the title page would be Suite 606 by the four authors. In a sense that is true in that that that all contributed to a package called Suite 606 though someone's bound to be astonished. --Marc Kupper|talk 01:47, 19 November 2008 (UTC)

Your description is pretty accurate. The title page has "Suite 606", and under it, the names of the four authors of the four constituant novellas. None is indicated as editor, and none is give any particular promeninece over any of the others. They are listed in the same order as that in which the novellas are placed in the book. (By the way, the novellas are linked, in that each of the first three shares a scene, generally a fairly minor scene, with the fourth.)

J. D. Robb has also participated in four similar anthologies. In each, there are four novellas by four different authors. In each, no author is listed as editor, and all authors are given equal billing on the title page, at least in the ones i have been able to check. The other four are Silent Night, Out of this World, Bump in the Night, and Dead of Night. Each of these credits the four authors as being also editors, so that the books are listed on the author's bibliographies under Anthologies. It seems clear to me that we should treat all 5 in the same way. If I got Suite 606 right, then we ought to change the other four to match. If not, we ought to change Suite 606. In some ways these are like Ace or Tor Doubles: call them "Quads". Some of the arguments for special treatment of doubles also apply, and some of the arguments for treating them as anthologies edited by "uncredited" would also apply to doubles, or to the Binary Star books. But I think we need at least a rule for "Quads", if not a general rule, agreed to and then spelled out in the help. -DES ^Talk 16:36, 19 November 2008 (UTC)

Essentially a hybrid? Somewhere in between anthology and collaboration? It kind of depends in how the user is going to search - and in this case it may well be the anthology title rather than the individual stories. Should only be done if stories are connected by at least minimal action (not just theme) and they were specifically written to appear together in the anthology.--swfritter 20:40, 19 November 2008 (UTC)

I guess these could be called hybrids, yes. In all five books the novellas were, i think, written for the joint publication, or at least had their first publication therein. In Suite 606 a set of time-traveling characters in one story walk into each of the other stories, although they make no real difference in any of them. But in some of the other four books, there is only a shared theme, no shared action I think. Yet it seems to me that all five of these really should be treated in the same way, one way or the other. -DES ^Talk 21:06, 19 November 2008 (UTC)

Uncredited anthologies with 3-4 novellas/novels are common in the "paranormal romance" niche, e.g. see Maggie Shayne. Sometimes the constituent stories are related, sometimes they aren't. I have been entering them either as anthologies "edited" by all 3-4 contributors or as omnibuses (depending on the length of the pieces), but I can see the arguments for using "uncredited" as well. I don't have a strong preference either way, except that consistency would be nice, of course. Ahasuerus 22:02, 19 November 2008 (UTC)

I think the final decision should be based on what the user is likely to search for - and the series should be done the same even if some of the anthologies have only theme connected stories. There seems to be an intuitive feel that the user will expect to find the anthology listing.--swfritter 22:10, 19 November 2008 (UTC)

"uncredited" adds no value, almost by definition. I find it easier to go search for other co-authors if they're co-credited, but as I say above that's often to find that they aren't actually authors at all, or are non-genre at best. But the end result should be on "least confusion" basis - I think the user will find the anthology on title search fine, but on a name search might be slightly confused if they find somebody credited as an Editor when they were just a Contributor. But there's plenty of examples of people Editing an anthology just so they can get one of their own works included, or a headline Editor being credited when they obviously weren't (the Silverberg/Clarke US/UK example comes to mind - they can't BOTH have done it) - and Editors of collections are demoted to notes. So I don't expect much confusion, and don't really have a strong preference either way, particularly in this sub-genre. BLongley 23:12, 19 November 2008 (UTC)

Which workaround?

We have a new Kurt Vonnegut CHAPTERBOOK that we can't create a Kurt Vonnegut, Jr variant for. Is everyone OK if I make this a single shortfiction content COLLECTION for now, or shall we leave it till chapterbooks are fixed and/or shortnovels are introduced? BLongley 19:46, 18 November 2008 (UTC)

Single item collection works for me. Waiting would also be ok, but probably a poorer idea. -DES ^Talk 19:50, 18 November 2008 (UTC)

It looks to me as if the real problem is that the story has also been made as type chapterbook, rather than shortfiction, with only the publication as type chapterbook. -DES ^Talk 19:54, 18 November 2008 (UTC)

Yes, changing to the workaround isn't a simple edit. It's fine as it is (perfect example of how chapterbooks should be displayed) but as chapterbooks are unstable I thought I'd volunteer to change it rather than end up with a chapterbook variant of an anthology, which seems to be the default the software gives people when they mess with things they don't understand (and shouldn't really HAVE to understand). BLongley 21:31, 18 November 2008 (UTC)

Ah, I see Mike's created a variant collection already. So we get the "Type Mismatch (Pub=CHAPTERBOOK, should be COLLECTION)" message and we can't see the chapterbook as Al originally coded. I can't say I like it that way, does anyone else? BLongley 22:10, 19 November 2008 (UTC)

I was clearing up the "Kurt Vonnegut" page by creating the "Jr." variants. This title appeared under "Anthology" on the "Jr." summary page, which is obviously wrong. When I tried to change the "Jr." title to "Chapterbook" it was not one of options on the dropdown menu. The closest I could find was "Collection" which is more descriptive of the title, according to Wikipedia is "a collection of short fictional interviews". MHHutchins 22:47, 19 November 2008 (UTC)

Yes, Anthology is the default that ISFDB gives you when you try any sort of edit on a Chapterbook Pub or Title. (Hence my "unstable" comment - it's an insidious change that people usually won't even notice is happening. It'll occur if you try to add any more contents like interiorart, or try and put the title in a series, or create a variant - basically anything will destroy a chapterbook entry. You might be able to change a review of such but I wouldn't even bet on that.) I believe the current desired workaround is a Shortfiction "God Bless You, Dr. Kevorkian" entry inside a "God Bless You, Dr. Kevorkian" Collection - and in this case, make the collection and shortfiction entries "No Jr" variants of the "Jr" canonical named versions. As it's been broken now, there's no way back to the Chapterbook status as originally entered. BLongley 23:29, 19 November 2008 (UTC)

Sub-titles that are series indicators

I have run across quite a few books where the title page will include a sub-title of sorts that indicates the story is part of a series. More recently, I was verifying a publication where one of the stories had one of these sub-titles.

Often times the series sub-titles are in a much smaller or different font than the main title and author name implying they are not part of the story title. For example, waiting in a to be-verified stack is Exile's Honor: A Novel of Valdemar where I'll be leaving the sub-title in simply because it's already there and all of publications use it. The publication's title page has Exile's Honor in a 38pt font and underneath this a shield/sword illustration with "A Novel of" and "Valdemar" to the left/right of the illustration in a 9pt font. The author name follows and is in a 32pt font though the same face as the main title.

Template:PublicationFields:Title and Template:TitleFields:Title are silent on if these should be considered as part of the story title though the latter gives a hint that even something as trivial as "A Novel" would be something you'd enter into ISFDB as part of a publication or story's title.

I've seen a mix of formats in use on ISFDB including

• Series: Title - for example Manifold: Origin.
• Title: Series - for example The Beasts of Barakhai: Volume One of The Books of Barakhai.
• Title: (Series) - for example A Cavern of Black Ice (Sword of Shadows)

As far as addressing this in the help - I'd vote for making the series sub-title optional and it's the editor's discretion on if he/she wants to include it when entering or verifying a publication. --Marc Kupper|talk 23:35, 30 November 2008 (UTC)

I tend to drop series-specific subtitles from Title-titles except when the connection seems particularly strong. However, I often leave them in Publication-titles, especially if the subtitle looks like it's a part of the title proper as opposed to thinly veiled advertising along the lines of "Dragon Errant: Volume 4 in the [exciting] new series Erroneous Dragon!". As far as non-series subtitles go, I always drop minor ones like "A Novel" and try to remember to record borderline cases in Notes.

My reasoning -- such as it is -- is that Summary Bibliography pages would be overrun with extraneous information if every other series entry contained the series title as part of the Title record. Publication records, on the other hand, are closer to the raw data found in the book, so I tend to be more forgiving there. Unfortunately, it makes it harder to search for Title-Publication discrepancies programmatically, so there are trade-offs.

I would be against making it a policy to add series titles to Title-titles for the reasons outlined above, but I don't have strong preferences re: Publication titles one way or the other. Leaving it up to the editor's discretion Ahasuerus 03:02, 1 December 2008 (UTC)

Story titles in anthologies

If you do "Diff publications" on the Robinson and the Carroll & Graf editions of The Chronicles of the Round Table, you'll notice that the story titles are essentially identical except that in the Carroll & Graf edition each one starts with "Sir [Whatshisname] in [...]". This is the way Harry entered them earlier today and I am sure he used what the title pages said, but I wonder if we really want to list the first three words in the title. I would be inclined to follow The Locus Index's lead and drop the "Sir"s, but I thought I'd mention it here first. Ahasuerus 02:25, 2 December 2008 (UTC)

Sounds similar to the Date prefixes for Martian Chronicles entries discussion. As it's a little bit of extra data I'm happy to keep the prefixes, I can never remember which Knight did what. But I'm happy to leave it to Editor's choice per pub so long as the variant links are in place. BLongley 19:27, 2 December 2008 (UTC)

Thanks. I will leave them as I entered then, unless told not to. This leaves the option for future users or verifiers to go either way as the will and be aware of the differences. I think this allows more versatility than making everything the same, even if there are differences. Thanks, Harry. --Dragoondelight 14:37, 5 December 2008 (UTC)

What consitutes a "series", and when to change series information?

I was wondering what constitutes a series which should be created. Are two connected stories enough? And how to name a series? And in which cases series name should be changed? The last question arose when I was creating a series for a couple stories written by Robert Chase. I did some background checking, and found that in Locus magazine’s database those stories are combined to a series “Skylandia”. So, I linked one story (which had no series information in ISFDB) to “Skylandia” series - only to notice that the other stories connected to it are already in a series named as “Changeling Hunt”. For some reason that series isn’t listed as a series on the Robert R. Chase’s summary page, so I didn’t notice it before trying the edit information on the stories themselves. So which way to go – to correct my submission, and change the series to “Changeling Hunt” - or follow the lead set by Locus, and change the series info in all of the stories to “Skylandia”? Tpi 11:46, 6 December 2008 (UTC)

I don't think there's any hard rules on when a series should or should not be created though there is one about changing a series around - mainly to help slow down possible edit wars between two or more conflicting versions of a particular series. Two connected stories may be reason enough for one person and not for another to create a series. It's up to you. I know I've created shortfiction stories that only have two titles though in anticipation that more titles will be discovered.

As for “Skylandia” vs. “Changeling Hunt”, that's also up to you. Google finds:

• Skylandia is only used by locus and has
  • The Changeling Hunt (nv) Analog Jul 1987
  • The Wellness Plague (nv) Analog Oct 1995
• Changeling Hunt is used by several sites but it turns out all of them are based on ISFDB data and at the moment they are using the same two stories.

I personally would change the series name to Skylandia on ISFDB and would add the third story. That way someone who happens across the name Skylandia, I assume it's a character or place in the stories, and does a Google will find both the Locus and ISFDB references. The series has a story named The Changeling Hunt meaning a search for that name will find the series. If you have a copy of one of the stories then you could look up who or what Skylandia is and add that to the series notes.

My general thinking is if there are two or more possible names for something then I at least try to document them. For example, I may see that a story or author name name tends to be misspelled frequently - usually because my own fingers fumbled into the misspelling. In that case I'll add a title or publication note explaining the misspelling. The intent is if someone's using Google that they will hit the page and will learn the correct spelling or wording. Likewise, at times I've run across an alternative series order. I'll usually document that on the ISFDB series' wiki page and again the intent is someone using Google will find it. If possible, I try to locate the source of a particular series (the publisher, author's web site, fan site, Fantastic Fiction, etc.) and will cite that. --Marc Kupper|talk 16:32, 6 December 2008 (UTC)

I read the story just a few days ago. The series is named by the planet where the action happens. But there is a slight fly in the ointment. The problem is that it seems Locus got the spelling wrong. :-) In the story itself it is spelled "Skiland" - at least in two separate places. I wonder I should go with that name, and make third series name for same stories... Tpi 16:54, 6 December 2008 (UTC)

It's sounding complicated enough that we should leave the series name at Changeling Hunt and instead to document on Series:Changeling Hunt the Skylandia name used by Locus and to document, and that whichever story you read is based on the planet "Skiland". It sounds like someone will need to read all three stories as I can't imagine someone would accidentally change Skiland into Skylandia meaning there should be some basis in the stories for the name used by Locus. I'm also confused as to how many stories there are. The Skylandia series has The Meeting of Pilgrims and the Changeling Hunt series has The Changeling Hunt and The Wellness Plague. --Marc Kupper|talk 07:05, 7 December 2008 (UTC)

The problem arose when I made the "Skylandia" series here (following the lead from Locus). There are (at least) three stories, as in the end The Meeting of Pilgrims is editors note: "This story is sequel to "Changeling Hunt" (July 1987) and "Wellness Plague" (October 1995)" . And all three (The Meeting of Pilgrims included) of them are in series "Skylandia" on Locus's database. Tpi 10:01, 7 December 2008 (UTC)

I moved the other two titles over to the Skylandia series and added Series:Skylandia. The Changeling Hunt series is gone. --Marc Kupper|talk 04:53, 8 December 2008 (UTC)

Currey "n/a" designation??

I am unsure of when and when not to use the "N/A" designation when verifying a pub using L.W. Currey's book. I have a first edition, but he has apparently revised it. I have no idea what that has entailed. Whether he has extended it past 1979; included more authors up to 1979 or just fixed some of the errors from the first edition. Technically any edition/pub that isn't a first edition would qualify, but it seems that without knowing what's in his revisions it should only be applied to pre-1979 pubs. Any suggestions? ~Bill, --Bluesman 04:24, 12 January 2009 (UTC)

My personal take: don't "N/A" anything unless you're positively certain that the dates fall outside the scope of the verification source. For instance, any book published after 1968 would be "N/A" for Tuck; any book published before 1984 would be "N/A" for Locus1. I don't believe Currey's revisions added anything published after 1977, but because I'm not certain, I wouldn't mark anything "N/A". In fact, I don't recall ever marking any pub "N/A", as I don't believe it does much to help when using the verification system. BTW, I just checked at Currey's website, and it appears that the CD-ROM version of his book (published in 2002) only goes through 1977. MHHutchins 21:55, 12 January 2009 (UTC)

That's kind of what I figured. And I have no idea why there even is an "N/A" category except to note that what you may be looking for "ain't here". Thanks. ~Bill, --Bluesman 23:06, 12 January 2009 (UTC)

I've never used "N/A" either. If we trusted the sources implicitly, then Al could just do a mass-update to mark anything out of the date-range anyway, and automagically do that for new pubs. No point us doing them one by one. And I don't think we can rely on the dates that the respected sources claim to cover: I'm sure I've had comments from Ahasuerus that as Tuck Vol 1 didn't appear until 1974 (by our own verification!), he had plenty of potential to get confused by some post-1968 pubs. (I think I was posting 1970/1 pubs.) But Tuck is probably definitely irrelevant by 1975 and later. There COULD be some use in mass-updating "N/A" verifications, but not as clearly as you might think. BLongley 23:21, 12 January 2009 (UTC)

Variants: Digits & Dastards

In the pub Digits & Dastards there is a story With Redfern on Capella XII that is currently listed in the pub record as "by Charles Satterfield". The original Galaxy publication was under this Pohl pseudonym and that is mentioned in the copyright acknowledgements. But the contents and story title page don't have anything about Satterfield (indeed it is unusual in any collection to find individual authorship on the story title page). Since the collection is as by Pohl, should this story credit be changed to Pohl, [NO! NO! NO! see below. MHHutchins 19:05, 24 January 2009 (UTC)] merged and then it would read With... by Frederik Pohl; (aka With... as by Charles Satterfield) ? And if so, at what point should such changes be done? I had one today (another Pohl/Slattery) that made no mention of the pseudonym at all, so that was an easy one. ~Bill, --Bluesman 15:34, 24 January 2009 (UTC)

I believed this was caused by an import of content from Bill Contento's Index to Science Fiction Anthologies and Collections, where he will list the name of the author credits of the story's first appearance. The only way to correct this is the old add/drop method: choose to edit the pub, add the record showing Pohl as the author, drop the record showing Satterfield, merge the new Pohl title record with the existing Pohl title record. It takes several submissions to fix it, but I can't think of any other way of doing it. At one time we had a data consistency project which included this task. (Look and you'll see three pubs of Digits and Dastards on the list.) I worked on it for awhile, but became bored soon afterward, and have never returned. Because it's a multi-step process I don't suggest non-mods take it on. But if you come across individual pub records with the problem, don't hesitate to use the old add/drop method. But PLEASE DO NOT CHANGE THE AUTHOR in the content title record. It would cause almost irreparable damage to the data. Hope this helps. MHHutchins 19:05, 24 January 2009 (UTC)

Take a few moments to read this help page about The Method (my own terminology, but I had to come up with something impressive.) MHHutchins 19:10, 24 January 2009 (UTC)

Last night a new editor overwrote the first line of the contents that used to say COLLECTION with the introduction essay. As they had added a ton of stories what I did was to pull up the the collection's title record in edit mode, approved the editor's update, edited the publication to add-title the essay as a new record, and then hit [Submit] on the edit of the title I'd left up to punch the original values back into the title record. The ISFDB song is Dancing With Editors. :-) --Marc Kupper|talk 07:14, 31 January 2009 (UTC)

Vonnegut (Jr.)

I just finished importing contents into Wampeters Foma & Granfalloons. This pub is listed correctly with no "Jr.". The one I imported from still had the "Jr." designation. Now it gets interesting. Every essay in this 'collection' has a copyright under Vonnegut Jr. They all existed before he dropped it. Since they are appearing in a pub under his name without the Jr. should all of these titles now get variants?... or just leave them the way they are?? I don't think they should get VTs since they are not stories. This same issue is going to come up with any reprints of his earlier collections as they too are being released without the Jr. Anyone besides me getting a headache?? ~Bill, --Bluesman 01:21, 30 January 2009 (UTC)

Personally, I don't think we should create variants when this situation arises. When even one reprint of a story appears under a different name, and a variant is created, the author's summary page will display the story as "STORY TITLE [as by Kurt Vonnegut]", even if 500 previous publications had the story as written by "Jr.". A display problem that I can see no way of getting around, but then I'm not a programmer. And yes, I get headaches when I run headlong into the database's limitations. MHHutchins 01:36, 30 January 2009 (UTC)

Have an easier one, I think. All the pubs of "Galapagos" are missing the "...´..." over the second "a". They all should be "Galápagos". In this case, instead of going through all the unmerge/merge dirge, should just one be changed so all follow suit? ~Bill, --Bluesman 01:50, 30 January 2009 (UTC)

No, they should not all be changed. Mine doesn't have it on title page. BLongley 19:07, 30 January 2009 (UTC)

That is an easy one, as far as I'm concerned. I don't think the database recognizes those accent marks, because whether I enter "Philip Jose Farmer" or "Philip José Farmer", everything is recorded under "José". Maybe it's been programmed just for specific names. As for titles, I did a search for "Galápagos" and all titles with that spelling came up, even the ones without an accent. So if you're certain (and I am) that all pubs had the accent in their title, submit a change to the TITLE RECORD (here) and I'll accept the submission. Then if you want, change the PUB RECORDs for those that you can verify. I'm going to change the pub that I verified. MHHutchins 02:06, 30 January 2009 (UTC)

To simplify somewhat, the database effectively stores two versions of each author's name. The first version has the correct capitalization, all and any diacritics, etc. It's the version that is displayed everywhere in the application. The second version is a "dumbed down" version of the name, with the letters converted to the same case and all non-standard letters changed to their English equivalents, e.g. "é" is changed to "e" and "á" is changed to "a". This version of the name is never displayed, but it is stored internally and points to the first (i.e. full) name. When a Name search is performed or a new record is entered, the software first "dumbs down" the entered name and compares it against its list of "dumbed down" names already on file. If there is a match, it pulls up the related full version of the name and uses it for search/data entry purposes.

As a side effect of this approach, you can't create a new version of a name by simply varying capitalization. Whether you enter "E. E. Cummings" or "e. e. cummings", the software will find and use the same Author record. However, if you edit the name in Author Data, the software will let you change the capitalization/spelling of the author's name. At the moment, the software doesn't check for duplicates, so it will let you change "Ray Bradbury" to "Harlan Ellison", so we would have two records for "Harlan Ellison" on file, one of them inaccessible. That would be, naturally, a Bad Thing, but it can be reversed by creatively using Author Data. Still, not an experiment that I would recommend :) Ahasuerus 02:27, 30 January 2009 (UTC)

Ebook page numbers

I was wondering, do we have a general consensus yet on handling pages in ebooks yet? Some previous discussions [[Publisher_talk:Project_Gutenberg#Pages_fields, and a new discussion User_talk:Mhhutchins#unpaginated_Rejection_of_an_ebook reminded me that it is still kind of 'up in the air'. Since the old discussion was on the Gutenberg talk page I thought I would bring the question to the center stage so to speak. Currently I'm thinking of the following choices.

• Ebook has defined pages: Use the defined pages just as in a printed book.
• Ebook has fluid or undefined pages (but references to pages as in Gutenberg html editions): Use the referenced pages just as in a printed book.
• Ebook has fluid or undefined pages: Use 'SOMETHING'

The 'SOMETHING' is what I am thinking about. The choices are "Leave it blank"; 0 or zero; and 'unpaginated' from previous discussion. My thoughts are: Blank is wrong, and invites second guessing by later editors; 0 or zero is also wrong, the book definitely has at least 1 page even if it is 2000 lines long and this also invites doubt (though less than blank); unpaginated is both clear (that the editor input something) and understandable (there are no page numbers). I'm open to a fourth option if anyone has one, and one objection to the third has been noted that items with 'unpaginated' entered look bad because the current ISFDB interface diplays this with the appended 'pp' so it is displayed as "unpaginatedpp". See A Just Determination for an example of this behavior. While not the prettiest result and I understand that updates / improvements to the underlying system are slow to materialize but eventually a request to change this behavior could be implemented if we have a standard that we want displayed. What are your folks' thoughts? Kevin 06:25, 31 January 2009 (UTC)

Template:PublicationFields:Pages says to leave the "pages" field blank for unpaginated works and to add a note. Putting "unpaginated" in the field is a good idea if we added code to recognize this and to not show the "pp". As you noted, features appear slowly.

Something that's related is last night when I saw a publication where the stories were on page 1, 2, 3, ..., 15. I read the notes and learned it was an unpaginated eBook with the editor adding the numbering to show the order of the stories. The main disconcerting aspect was I was looking to see if I should approve someone's proposed merge of a type NOVEL with type SHORTFICTION meaning I was looking at the story length using page numbers. The novel was 204 pages and at first I thought the shortfiction was one page until I saw the 1, 2, 3, ... While it seemed odd at first once I understood what the editor was trying to show then I'm fine with numbering the stories like this. --Marc Kupper|talk 07:06, 31 January 2009 (UTC)

By their very nature most ebooks don't have a valid page count. Since that is the default state I see no reason to enter a value or even to document the fact. In the case where there is a discernible page-count of one sort or another it should be used and documented. Bogus numbers are an old issue - until there is some sort of a background method of defining order I think they are valid as long as they are documented.--swfritter 22:15, 31 January 2009 (UTC)

I think the "Missing page count:" warnings make our data look dodgy (far more so than an unwanted "pp") so would put something in there. I do put a "0" in for audio-books for instance. When Al sorts out the display problems I might change my practice. I don't usually handle ebook submissions though, if you want a separate standard for those. (Although I think there are some similarities in the two issues.) Does an audiobook have at least 1 page? I can't think of an example audiobook with multiple contents with faked page numbers on though. BLongley 22:52, 31 January 2009 (UTC)

Warnings are strictly a display issue and can be resolved sometime in the future. It would actually make sense to me that the warnings should only appear when the user is an editor logged into the system. The information is of little value to the transient user. It would be nice if we could have play length for audio books. I most commonly used bogus page numbers in cases where I want to group artwork with a story - in the case of Baen's Universe there is actually an index number in the HTML version so in that case I give both story and artwork the same page number. Not a lot of artwork in audio books.--swfritter 23:29, 31 January 2009 (UTC)

We're drifting across a few formats here, but generally "unpaginated" for page-count just tells me that each page isn't numbered (for printed books where somebody could count them anyway), or that it's irrelevant for this sort of publication (in which case we should eventually tie the error messages in with the "binding type" and prevent errors for formats that don't HAVE pages). What we seem to want is a "length" definition of some sort. Page counts work for some formats of books, but then you need to know page size and whether it's Large Print to be even in the right range. Play length or running time could be good for audio books. (In lieu of such, I at least record "Unabridged" or "Abridged" in notes where I can determine it.) "Number of cassettes" or "Number of CDs" or "Megabytes of MP3 data" might work. For Ebooks, character count of the ASCII version might work. We're unlikely to get such from secondary sources, and even if I listened to audio books I'd never do a word-count on such, which might the ideal. :-) BLongley 00:11, 1 February 2009 (UTC)

Page Numbering within a book is a separate issue, and I'm OK with fake page numbers to maintain order so long as they're explained in notes. I still recommend that people do such with big gaps to make it easier for other editors: e.g. if you add fake page numbers for 3 Novels in an Omnibus, don't make them pages 1, 2 and 3. When somebody wants to go into more detail and add the introductory essays, or afterwords, they'll have to redo all your page numbers. Make them 10, 20 and 30 and someone can just add introductions on 9, 19 and 29 and/or afterwords on 11, 21 and 31 without having to redo all the other pages. Think of us poor Mods that have to look at such changes! BLongley 00:11, 1 February 2009 (UTC)

Since it was my rejection of a submission that started this topic, here's my POV:

From the Dictionary for Library and Information Science - unpaginated: A book or other publication in which the pages of the text are not numbered or sequentially marked. The total number of pages in an unpaginated work is noted by the cataloger as an interpolation inside square brackets in the physical description area of the bibliographic record (example: [118] p.)

Let's recap: "Unpaginated" does not mean "no pages", it means "pages are not numbered". So if an e-book has no pages then "0" should be placed in the page count field. It will be displayed as "0pp". How can that be considered ambiguous? 0 pages, no pages, zero pages. If an e-book does have pages, but they are not numbered, then they should be counted and placed in brackets, or the field should be left blank. If they can't be counted, then leave the field blank. My ha'penny. MHHutchins 00:41, 1 February 2009 (UTC)

A blank field invites 'correction' later. A blank field does not indicate that an editor made 'any' determination (and some ebooks have pages defined). Nature abhors a vacuum. 0 is incorrect and in an electronic world, 0 is also often considered the same as 'blank' or nothing. I agree that an ebook has no 'folios' or physical papers, but anything that is readable cannot have a page length equal to zero. My computer has a 'page down' key; My ebook reader software (mobipocket) refers to 'First Page', 'Last Page', 'Next Page', and even 'go to page'; My reader software on my cellphone has a setting "Clicking on page turns page". The definition of 'page' has changed (hence the Dictionary you refer to is out of date). It no longer is limited to the verso or recto of a folio. Maybe we need a new term, or we can continue to co-opt unpaginated... I'm completely open to finding a better 'SOMETHING', but my original point is that both 'blank' and '0' have more compelling limitations. Like I suggested in our discussion... go try out some ebooks for the experience of it (It's free of monetary cost to you, and only costs some time). Kevin 02:01, 1 February 2009 (UTC)

Mike, I'm with you on the meaning of "unpaginated". It's a cop-out to put it in a page-count field. And I support "inside square brackets" for page count entries for unpaginated books that do actually have pages that can be counted. I don't like "leave the field blank" in the meantime, due to the number of error messages it generates. We've let in publication types we never used to deal with before, and the display software hasn't kept up. How about "N/A" or "Not Applicable" for the page count field, if people don't like to use "0"? (I'm actually quite happy with "0" myself in the meantime.) But "1" is arguably just as valid for formats that show as one big HTML page, for instance. But this is just working around a display format problem - if "total pages" for a book doesn't really mean anything in the format it's delivered in, it shouldn't be a field that causes errors if left blank. BLongley 02:29, 1 February 2009 (UTC)

Until I've educated myself about e-books, I will recuse myself from involvement with any further submissions containing them. Or should I just imagine the time when a database field designed for numbers will reject any letters placed in them? MHHutchins 03:50, 1 February 2009 (UTC)

"n/a" sounds good to me. I might note that we are way behind in ebook editions. There are many thousands of ebook editions that we do not have - unless Fixer is adding them.--swfritter 20:05, 1 February 2009 (UTC)

I don't think Fixer is finding them. He's finding a lot of Audio books though - not surprising as Amazon apparently own "Brilliance Audio" and "Audible.com". Still, they apparently own "Mobipocket.com" too, so I'm not sure why we would be lacking those. And they seem to own "AbeBooks", which owns "Shelfari" and has a big stake in "LibraryThing", so eventually Amazon will no doubt integrate all this data and make us redundant. Or just "assimilate" it and make us more useful in recording what the books actually state. BLongley 21:25, 1 February 2009 (UTC)

Fixer has found a lot of Kindle editions and such -- 11,711 at last count -- but most of them do not have ISBNs and Fixer doesn't submit books without an ISBN. I couldn't think of a straightforward way to determine whether we already had ISBN-free books on file, so I asked Fixer to leave them alone for now. When an e-book has an ISBN, Fixer does submit it, although Amazon isn't always very clear about the book's format (Adobe vs. Mobipocket etc), but then what else is new?.. Ahasuerus 00:51, 2 February 2009 (UTC)

The Kindle format is mobipocket but modified so that DRM pubs cannot be read with any other device and vice-versa. Kindle may also be able to process other formats but I think the Amazon titles are all in mobipocket format. Ebooks often have the isbn of the print edition as part of their contents but it is entirely possible that it is not an official isbn for the ebook.--swfritter 01:27, 2 February 2009 (UTC)

UNINDENT So is there consensus yet? (This is not a question about TOC numbers, but a question of what to put in the PAGES field of an ebook publication)

• Leave it Blank (as the guidelines for Publisher:Project_Gutenberg#Pages_fields Gutenberg now states)
• 'unpaginated' (as the consensus of Publisher_talk:Project_Gutenberg#Pages_fields September 2008 seemed to reach)
• 0 (as has been proposed)
• n/a (as has been proposed)
• some other proposal? - I'm welcome to hear them.

Thanks - Kevin 01:22, 5 February 2009 (UTC)

As far as approving an editor's entries go - all of them are fine with me. I have used all of the above at times plus "(none)" and "None". One problem I have with n/a is I don't know if it means "not available" or "not applicable" though in ISFDB we generally leave a field blank if the data is not available. Using "unpaginated" seems clearest as it's assumed the bulk of the people seeing a record will be visitors and not ISFDB editors that would be more versed in the details of how a particular field is coded. --Marc Kupper|talk 08:08, 5 February 2009 (UTC)

I might also note that when you are looking at pub contents like this an empty page count does not show up so the user is not even aware that there is any potential data there to miss. Putting data in the field merely allows for the display of information that may not be of much value to the user. Page count shows up in other contexts as a Bibliographic Warning - a warning is not exactly the same as an error and is primarily a cue to editors rather than information the casual user cares about.--swfritter 01:00, 6 February 2009 (UTC)

Brits solve the apostrophe problem!!!

This solution might be applicable for isfdb. SQL works so much better without them.--swfritter 20:08, 1 February 2009 (UTC)

Aaargh! And only as far back as 2001 Britain won the Ig Nobel Prize for Literature for preserving them! Time to revive the Apostrophe Preservation Society and march on Birmingham. BLongley 20:56, 1 February 2009 (UTC)

Actually, forget the marching, it's cold and it's far simpler to Nuke Birmingham from orbit, it's the only way to be sure. I'll just stay down here in the South and rebut the "SQL works so much better without them" - I don't think I've got a single useful SQL query against the ISFDB database that doesn't use them. And even on the Wiki side, they're rather useful - I used SIX in my previous comment for instance. And TWELVE in this one. BLongley 21:09, 1 February 2009 (UTC)

I do not even enter apostrophes in my own personal Access book database. But I do put 'The' at the end of a title like most civilized people do. When I was a paid programmer we used to have a subroutine called "twiddle_the_double_quotes" that was used to pre-process SQL statements.--swfritter 22:14, 1 February 2009 (UTC)

Definite article at end of title makes sense not. Civilised people evidence for, wrong also. Microsoft Access usage - clear sign Dark Side belonging. Jedi programmer you are not. (Not that I'd prefer Yoda to determine the rules of English anyway. That was a long time ago, in a Galaxy far far away, and England only came about in fairly recent times - but we have got it sorted out now. Just ignore our kids that tell you to "Ph3ar R l33t-speak!" or suchlike. ) BLongley 23:04, 1 February 2009 (UTC)

I guess the libraries in Britain do their index cards differently. If any library still has index cards.--swfritter 01:18, 2 February 2009 (UTC)

I remember my father (a professional librarian) coming home with a large supply of blank index cards that the University of Essex didn't need any more. That would be 25 years or more ago, as I used such as separator cards for my 7" singles collection. Still, he was apparently one of the first British librarians to have to deal with those new-fangled "computerised catalogues" and was grumbling about such before I even got my hands on a ZX81. But it's quite possible that such records got converted exactly and Ahasuerus is dealing with the consequences even now. ;-) (And no, my Dad isn't interested in SF and won't come out of retirement to explain Z39:50 or such.) BLongley 22:41, 2 February 2009 (UTC)

I use Bookpedia for my local book db. When it sorts on titles it ignores The, A and An with an option in prefs to ignore foreign articles also. Dana Carson 02:10, 2 February 2009 (UTC)

It would require an extra field in isfdb for the sortable title - either that or some kind of smart indexing that automatically lops off the articles. Guess I just have to get used to living in the 21st Century - and it's already almost 1/10th over.--swfritter 21:40, 2 February 2009 (UTC)

I've seen databases with title-prefix, title, and title-suffix fields to allow for sorting and reconstruction of the "exactly as stated" title. And other databases with "smart" algorithms to ignore leading articles. (They're "deterministic" functions apparently, so you can still index efficiently.) And others that just assume they're so efficient you can do a "like '%title%'" query on any title anyway. (Oh wait - ISFDB would be an example of that.) There's more than one way to skin a cat, I'm told, but there's probably even more book-database designs. And I could probably break every single one of them. BLongley 22:41, 2 February 2009 (UTC)

Bookpedia does have a hidden sort_title field so you can force it to sort in the order you want. I've never had to use that. Dana Carson 23:17, 2 February 2009 (UTC)

Link Reviews to Canonical Title or Actual Title?

I just approved some that linked them to the totally different canonical title. Has there been a discussion about whether they should be linked to canonical or actual title?--swfritter 22:32, 1 February 2009 (UTC)

The Hugh Walters titles from Tpi? I'd considered holding them, but that would have made them my problem and I've been very busy today anyway. I don't know how the display software works, so would experiment on whether this means they're lost or will display under the canonical title. Not that we have any pubs under either the canonical title or the variant - I should have kept them when I had them. I think the 2d per day [1] "late return penalties" from my library at the time might not seem over-punitive given the costs of acquiring them now. BLongley 22:51, 1 February 2009 (UTC)

[1] Yes, I did borrow those books in pre-decimalisation times. Unfortunately I returned them in those times too. My mummy wouldn't let me steal books. BLongley 22:51, 1 February 2009 (UTC)

I thought about it. I decided that the content (not the name of the content) is what is reviewed, so it is best to link to canonical title. And I didn't even realize you could link reviews to variant title. I didn't try, but it seems to be possible. Tpi 20:25, 2 February 2009 (UTC)

Actually reviews are linked to title records, not "titles" (a subtle, but important, difference). So anyone can choose, after the review record has been created, which title record to link it to, no matter what the "title" of the pub is. The system will link them automatically if there's an exact match when first submitted, but be careful. Sometimes that "exact match" will be the first "title" it comes to. For instance, a short story or novella by the same author that was later expanded into a novel of the same name. I've had to change many review links for this reason. MHHutchins 21:17, 2 February 2009 (UTC)

As I way typing when Mike was saving - When you add a new review it automatically links to the title which matches lexically, even if it is a variant title. The same is also true if the review credits a pseudonym as the author - it does not link to the title record which credits the canonical author. From the standpoint of the user who wants to find all the reviews for a certain work it does seem to make sense to link to the canonical title. Ideally, you should be able to link to the variant title but have the review also show up under the canonical title.--swfritter 21:28, 2 February 2009 (UTC)

I've done a lot of work with reviews, having entered many review zines (with hundreds more waiting in the wings, alas), and the point that Stephen brings up is quite valid. Sometimes reviews are "hidden" away from the canonical title because they're only listed under the actual title and author credit. Perhaps only a change in the software for display of reviews would do the trick (or is my naivete about programming showing?) MHHutchins 22:27, 2 February 2009 (UTC)

I guess the desired outcome is to have all reviews show up as reviews of the canonical title, same as all other variants of the canonical title show up. And maybe that if the review was of a variant, it would show up under that variant when you select that subset. But if we're now allowing linking of anything that the review says was reviewed to anything else in the database, I have no idea how it should be displayed. If there's a review of "Mohole Mystery, The" by "Walters, Hugh" do we demand regularisation of title and author first or let it through? I know there's already author pseudonyms set up only because they exist due to reviews. I'm not particularly happy with those, but they get past my main objection ('authors' that don't link anywhere useful, and you can't see the reason for their existence at all) but titles that only exist in reviews are a bit different - again, they could be created as variants but the direct linking should avoid any need for that. (And of course, what I want is for reviews to link to the edition(s) being reviewed, which is not necessarily a separate title.) BLongley 23:07, 2 February 2009 (UTC)

The display of variant title reviews from the canonical title can be put on the long list of things that would probably be easy to implement. If we could plan on the change being made sometime in the future I think it would be better to link to the variant title when the title matches the review. That would also follow the default system functionality for auto-linked reviews.--swfritter 23:39, 2 February 2009 (UTC)

Linking to the title actually reviewed is closer to my ideal, so I'll support that in the meantime. I never work the other way: a review may show me missing books, I never look at a book and try and find reviews. Even if ISFDB told me a book was reviewed in another publication I actually have, I'd still just Google for reviews rather than go find where my cleaner put the magazine or book with the review in. BLongley 23:57, 2 February 2009 (UTC)

FWIW, I generally link reviews to canonical titles due to the "burial" problem referred to above. Ahasuerus 17:32, 3 February 2009 (UTC)

Just what we need, another schism. :-( BLongley 18:15, 3 February 2009 (UTC)

This one can be addressed via a trivial software fix, though. If/when Marc manages to get his development system up and running, these types of fixes shouldn't take too long. Ahasuerus 03:07, 4 February 2009 (UTC)

Oh well, I can live with either way. In the meantime, having just obliterated "Albert Nuetzel", "Doris M. Paine", "E. J. McFall", "Gray Morrow", "Jack Bechdolt", "Maxine McArthur" and "Rachel Maddux" (and still want to obliterate "Robert W. Lowndes" but can't find him), can I remind people that linking reviews shouldn't be a substitute for getting the Reviewee right? (Don't worry, all the above had extra spaces in their names, I haven't obliterated the correct ones.) BLongley 18:15, 3 February 2009 (UTC)

I keep seeing more and more reasons to drop everything ISFDB related except to get coding of features and bug fixes up to speed. I'd rather link a review to the title that the reviewer inspected and ideally to the publication record and that it'd be visible in the canonical author bibliography. --Marc Kupper|talk 19:44, 3 February 2009 (UTC)

We could definitely do with some bug-fixing sooner rather than later. Features are going to be a bit harder when the requirements are not only unclear, but sometimes pulling in completely different directions. :-( Given some recent cases, I think one of the things we could do with is just plain diagrams to go with the help - some ideas just aren't getting across at all when we express them in words alone. Maybe it's time to boot up my old work laptop, there might be a copy of Visio or something like that on it. BLongley 21:25, 3 February 2009 (UTC)

There are more than 1500 reviews linked to variant titles - about 800 of those reviewed books have the same title which indicates the only difference in the variant/canonical title data is a result of pseudonym use. If we want to make the band-aid approach of linking to the canonical title the standard then there is a project for someone. Otherwise we can wait and if a different display method is used there will still need to be some work done.--swfritter 22:47, 4 February 2009 (UTC)

I can't seen any way that the "trivial software fix" (which I think is for display issues only) will address "bad" links created in the meantime. As I've no use for links created to Canonical title/author when they are actually of a variant, I'm not going to create such in the meantime, or approve such. I can only see it leading to more work in the long term. BLongley 23:28, 4 February 2009 (UTC)

I believe we should continue to link to the actual title (even if's a variant by virtue of title or author), and that the software should be changed for all reviews to be displayed under the canonical title (just as pub records are now). We can't change horses in the middle of this stream. I probably did more than my share of those 1500 reviews, and I'm not taking the time to change them now when a future software change will have made the effort moot. MHHutchins 01:32, 5 February 2009 (UTC)

I don't think we want to spend any time on relinking reviews if the issue can be addressed in the software. On the plus side, once the software has been updated, it should be easy to find all reviews that link to a wrong title. Ahasuerus 03:51, 5 February 2009 (UTC)

One pub, two ISBNs, two pub records

I handled a submission a few days ago that created a new pub record, simply because the pub printed both US & Canadian ISBNs and prices. When questioned, the submitter stated that there had been an earlier discussion about having two pub records so that both will show up when googled. I let it slide, but today there's another one in the submission queue, so I think the issue needs to discussed before it gets so far out of hand that it can't be easily fixed. I know that this practice has been going on for more than two decades, with some publishers printing both ISBNs and prices in the same publication (Tor and Baen come to mind as the ones who seem to have done it the most.) So have I misunderstood ISFDB standard that one pub=one record? Doesn't recording the Canadian ISBN and price in the notes suffice? Or do we make an exception for Google? MHHutchins 20:46, 2 February 2009 (UTC)

Also, as I noted elsewhere many ebooks also have the same ISBN as the print edition. Whether this is technically the ISBN of the ebook I am not sure. It may be that they just don't bother to take it out of the text when they do the conversion from the files used to create the print version. I am not so sure that we should let Google display issues drive our decisions one way or the other.--swfritter 21:35, 2 February 2009 (UTC)

Same ISBN is less of a problem I think. Certainly not a problem if it's just a different format from the same publisher. The UK seems to be fragmenting titles across ISBNs far more often now - e.g. a hardcover and paperback from the same publisher will now get different ISBNs, whereas they used to get different ISBNs only because one company would do the hardcover and a different one the paperback. Nowadays there's been so much consolidation of publishers that this differentiation is often irrelevant, they're all "Hachette Livre" or something in the end. In the old days, where the ISBN belonged to a particular separate publisher, the differentiation was important, but each publisher would often keep the same ISBN over all editions and printings as long as possible, as new ISBNs cost money. It's not necessarily an error to have the same ISBN over multiple publishers, particularly if there's a history of ownership changes and rebranding. Within the same overall publisher I wouldn't notice anything wrong until the imprint appears for the wrong years, or for a completely wrong publisher. (For instance, the Fixer "Roc" additions today were completely wrong at least four times, but pointed me at other previous (Dissembler?) entries that needed cleaning.) BLongley 22:07, 2 February 2009 (UTC)

Hmmm. That's not how I recall the discussion. I support two records here because ISBN searches are common, and if we don't have a separate pub record for the Canadian ISBN we won't find it with a search here as we can't search notes. (And wouldn't even expect to have to.) Googling would be a fallback position to find such, but I do not recommend relying on Google to do our work for us. However, I do support recording the alternative ISBNs in notes as well, it is information recorded in the publication. I've got no readily-available examples to compare, but the impression I got was that a US edition could mention a Canadian ISBN, and a Canadian edition might mention the US ISBN? Are people not clear which one they're entering? (I suggested only a stub publication for the other.) BLongley 21:44, 2 February 2009 (UTC)

If they're actually exactly the same book I can see that it would be considered a problem by some, which is why I think the cross-references would help. But I see fragmentation occurring anyway as one book with one ISBN gets entered several times, one with the US price and one with the UK price. (This is why I'm now adding ALL prices shown on my publications, so that Australian and Irish and Canadian Editors don't enter the same book again - and if we had had more editors from New Zealand and Trinidad and Gibraltar and South Africa and East Africa and Cyprus and Spain we might already have a bigger mess.) BLongley 21:44, 2 February 2009 (UTC)

I know I've had several books where the paperback contained enough information to create the differently ISBNed hardback stub and the format difference made it clear which was which. I also know that an ISBN and date and printing number and price and even a country of printing doesn't necessarily distinguish two publications. Even a City doesn't help at times. I can show you two "Buffy the Vampire Slayer" books apparently identical in all but name of printer. And both printers are in Glasgow, Scotland. :-( BLongley 21:44, 2 February 2009 (UTC)

Anyway, given all that background - I still support two publication records for the two ISBNs. Even if they're really identical books, cross-reference away and don't break the searches. I don't pretend to be an expert on US and Canadian editions of the same book, or how one book with both prices and ISBNs on can be determined to be either Canadian or US, but I'm sure we don't want a civil war over the matter. And demoting Canada to notes automatically does sound rather unfriendly. BLongley 21:44, 2 February 2009 (UTC)

Perhaps I'm being ethnocentric, but I think these editions are printed in the US for both markets. So placing the Canada info in the notes doesn't appear to be no more prejudicial than what Bill does when he puts the New Zealand price in the notes and the UK price in the price field. Is Bill a Kiwi-basher? I don't think so. :) MHHutchins 22:33, 2 February 2009 (UTC)

As it was my submission(s) that started this... so far all (maybe 15) of the ones I've done the duplicate entries for are wholly printed in the US. Indeed if they were printed separately, there would be no mention of the other's ISBN. So far they are all limited to Tor & Baen editions and from a narrow span of years (88-91). Have yet to see one later than that. I always enter them as 'mirror images', mentioning the other ISBN and price in the notes. TOR Doubles come to mind. TORDOB09 has the double entry. Can't think of a Baen off hand. --Bluesman 23:37, 2 February 2009 (UTC)

I go by first price listed. This isn't always the UK price. (Although I can only recall recording one definite Australian book, and no New Zealand books. There are definitely some net-exporting book countries, and some net-importing book countries, based on the data we have here so far.) I don't recall any book where I've listed the Canadian price over the US one, but based on my collection, Canada is somewhere we Brits usually send books to, not where we get them from. But the price issue is different from the ISBN one. I'm seeing more and more records clarifying "Cover printed in USA" from the book being printed in Canada. If that's going to be important, we all need to do a lot more rework. ISBN differences I think are a big bibliographical difference that needs recording. I'm rather hoping that we don't all need to go back and record exact printer on everything we've already verified. If I can delay that by a US/Canada dispute I can stay out of, I'm fine for a while! :-) (Although if there's any sort of "US is the default" attitude expressed I will eventually have to reclaim all the BBC books for a start, and that would take months... :-( ) BLongley 23:46, 2 February 2009 (UTC)

I've entered a number of these in the past and from my point of view there should only be one record. The US ISBN# is the primary number and I found searches of the Canadian ISBN returned limited or no results. This was a short lived experiment by Tor & Baen. If we do decide to have two records then we better make it VERY clear that it refers to only one publication.Kraang 01:39, 3 February 2009 (UTC)

When I've run across books with a second ISBN on the copyright page I add notes rather than assuming that a book was manufactured with that ISBN barcoded on the back cover. My notes always use both unhyphenated and hy-phen-ated versions of the ISBN so that someone using Google can spot the note. The only times I will create a second publication record is those books that note the ISBN of the trade, hardcover, paperback edition(s) on the copyright page. Even then I look around to see if publications with those bindings are available for sale.

One option for a book with both USA and Canadian ISBNs would be to check on a site like AbeBooks. If I have the USA edition then I'd check the Canadian ISBN on Abebooks to see if a bunch of Canadian dealers have it. That'll give me confidence that the Canadian edition was manufactured and I can add the AbeBooks findings to the notes when I create the publication record.

Related to this is a guiding philosophy I try to keep in mind which is that if a publication has more than one possible interpretation of how someone would describe it then I'll document them. The thinking is someone doing secondary source only research on a subject will find the record and that it will also help them consolidate records they have found on other sites that all described the same publication in different ways. Another reason is that it'll help someone with a publication accurately determine if their publication is the same as the one I verified or if they should use or add another publication record.

re: "Cover printed in the USA." I used to document this but no longer do as I decided the odds are really low that the exact same book was manufactured with both USA and Canadian covers. I'd have to think one of the other key points we use in describing a publication will change and that'll be enough for a new publication record. If someone ever does get evidence of nearly identical USA and Canadian editions then sure, we create separate publication records and explain in the notes what the differences are.

re: which country is "primary." I go by country of manufacture for the publication (not the cover) and also which price is either first or most prominent. The DAW books are curious in this aspect in that for a long time they listed the Canadian price first inside parentheses followed by the USA price in bold. I always went with the USA price and these days also document the Canadian price in the notes. --Marc Kupper|talk 19:35, 3 February 2009 (UTC)

Can you clarify your last point please? Your example suggests that you found the DAW US price more prominent, but you don't say where the books were printed. BLongley 21:19, 3 February 2009 (UTC)

Sorry - all DAW books have been printed in the USA since the early 1980s and have been dual priced though in March and April 2008 DAW published some single-priced books before reverting to dual pricing. I had assumed then the intent was the stated price was to be used in both the USA and Canada but now I'm wondering if there are separate Canadian editions.

To back up a little. From 1972 to ~1983 DAW single priced the books. Both stated $ without the country and you need to look inside for the Printed in the USA or Canada to see if the price is in U.S. or Canadian dollars. In 1983 DAW switched from the yellow box in the upper-right corner of the front cover to a pair of lines with the first having the genre and ISBN and the second having the price as "(CANADA $4.50) • U.S. $3.50". I believe all of these are printed in the USA though it's possible DAW did not stop printing in Canada until later as I've seen some reprints of 1984 books that say "Printed in Canada". --Marc Kupper|talk 06:17, 5 February 2009 (UTC)

The last book "Printed in Canada" that I have is dated Sept 1988 Daw# 757, after this there all printed in the US.Kraang 03:21, 6 February 2009 (UTC)

(Unindent) I think I don't really know what people have been doing any more. For instance, this was added today - lots of notes, amongst which it's stated that it was printed in the UK but also that it only has US and Canadian prices. No indication of order of prices. It looks pretty similar to this apart from the price. It may be the same publication - I don't know if the British publication exists, or if it's just the export edition over-priced with a UK price. (That publisher is too close to RPG for my taste.) Dissembler and Fixer are never going to go into that much detail, so we'll always have to cope with such differences. Where do we fragment suspected different publications and where do we combine them? BLongley 21:19, 3 February 2009 (UTC)

I did your example above, as far as I can tell that book was printed in Great Britain for RPGers and action bashem readers in Great Britain with two separate issues, one £ amd one U.S./Canada. I did a 2007 at the same time that implies the publisher is capable of U.S. publishing. [On real close inspection the person who inputed the data assumed a U.S. branch. I will have to correct.] TSR does the same. There is often a note saying the distributor's name. Thus the implication is that they can/do produce multiple price books and can/do produce market specific editions by price. I have not enough of this publisher to be dead sure on all specifics. They were thrift store buys. So the answer is the printer/printings are done in Great Britain. So I think, the pricing determines the entry. Thanks, Harry. --Dragoondelight 00:57, 8 February 2009 (UTC)

That's a good point on the order of the prices. The implication is USA is first as that's in the main price field. "Where do we fragment suspected different publications and where do we combine them?" I don't think a hard and fast rule is possible as you may be dealing with a mix of verified, data imported, etc. records plus at times a person may include their knowledge of the publisher's practices, etc. If we are doing a "ISFDB editing for dummies" page then the only distinguishing factors are the ones in the metadata. Everything else is optional. You may end up with two records for a publication if one person used the catalog # and the other the ISBN and in that case a decision needs to be made on if they are in fact referring to the same publication. Ideally, both records are verified and you ask the two verifiers to compare notes. --Marc Kupper|talk 06:40, 5 February 2009 (UTC)

If Tor & Baen are known to put multiple ISBNs on a definite single publication, I can live with hiding the less-used one away in notes if that's what people prefer. (I've never searched for a Canadian ISBN in my life to buy it, only for ISFDB purposes.) But I for one will need to know which publishers are dual-ISBNing single publications and which are indicating that other editions truly exist. Wasn't it one of those two that had near-identical covers for US and Canadian editions apart from a little maple-leaf logo? Or was that DAW? Seriously, if such is becoming a big issue and there's no sharing of knowledge of such practices I may have to recuse myself from moderating such publishers. (Sorry, Fixer - if that's the case then you'd better learn to deal with Amazon UK sooner rather than later if you want my help.) BLongley 21:19, 3 February 2009 (UTC)

Don't foget the ebook issue lurking in the background of this conversation. Some (Pengiun/Ace, Penguin/ROC, Penguin/DAW, HarperCollins, Macmillan (but they only use mobipocket)) publishers are issuing a different ISBN to ebooks sold in multiple (Mobipocket, eReader, Microsoft, Adobe) encrypted formats (I find that the unencrypted publishers (Baen+Webscription, Fictionwise, NightShade+Webscription, Tor+Webscription)just re-use the last print edition and don't differentiate between formats with ISBNS). I assume it's for book-keeping and comparison of different encrypted markets since the encryption is an added cost that varies by encryption standard. So not only is this a problem for some of these experimental items from way back when.... it's a 'new' problem all over again, and the 'Less Used' one will not be obvious. Kevin 00:08, 4 February 2009 (UTC)

It looks like there are two competing priorities here. The first one is to make it easy to search by ISBN. The second one is to avoid creating two Publication records for one physical book. Both are worthy goals and it's unfortunate that they collide in this case since our ISBN field doesn't allow multiple ISBNs.

Given the limited scope of the Tor/Baen experiment and the existence of Google, which generally indexes our Notes, I think that burying Canadian ISBNs in Notes is the lesser of two evils.

I agree with this. I just document the anomalies as best as I can with the assumption that someone using Google will find the record and then understand why in their searches they have found references to what seem like two separate publications. While my personal database supports multiple codes and ISBNs per publication I've never felt inclined to push that ISFDB also support this as I've found doing it as notes is better as I can describe where on a publication the various codes can be found. --Marc Kupper|talk 07:05, 5 February 2009 (UTC)

Re: Kevin's point, as far as I can tell it's a somewhat different but related problem. I don't think there is anything preventing us from creating one Pub record per ISBN for e-books, although library catalogs typically dump all ISBNs (hs, tp, Adobe, etc) in the same field (field 20 in the MARC family of standards) in the same record. It makes Fixer's life more difficult since not only does he need to figure our which ISBN represents which format, but he also has to decide whether some of the ISBNs are duplicates because of the ISBN-10/14 transition. No wonder the poor dear stays up all night... Ahasuerus 03:02, 4 February 2009 (UTC)

I believe it's long established that the ISBN is largely meaningless but can be used as a hint. The first part hints at the country or language, the middle part hints at a publisher, and the last part is subject to the publisher's whims but usually hints at a title. --Marc Kupper|talk 07:05, 5 February 2009 (UTC)

I believe pretty much the opposite. ISBN is largely meaningful. Overall, it is a VERY good sign of a title. The beginning is pretty good at defining a language, especially for English, the middle part defining a publisher is the only bit that causes confusion. OK, I'm probably biased as I come from the country that invented them, and am used to several generations of publisher inheriting (I)SBNs. BLongley 22:57, 7 February 2009 (UTC)

As far as robots and fields go - as long as the robots leave evidence that the records are from a robot then I'm happy. If a robot gets smart and is evaluating multiple fields, or deriving information from a field, to decide which should be the ISBN, etc. then ideally it documents the raw field values so that the humans can see if they agree with the robot's logic. --Marc Kupper|talk 07:05, 5 February 2009 (UTC)

Fixer will be dumping a lot of stuff in the Notes field when the time comes to submit library records. You may yet regret asking for more "supporting documentation" :) Ahasuerus 05:53, 8 February 2009 (UTC)

Enough chit-chat. A decision must be made. Yes, duplicate records. No, a single record. Every current editor, whether mod or not, can make their choice known here:

• No - MHHutchins 00:27, 7 February 2009 (UTC)
• No - Kraang 03:56, 7 February 2009 (UTC)
• No - Ahasuerus 04:12, 7 February 2009 (UTC)
• No - --Marc Kupper|talk 05:35, 9 February 2009 (UTC) (this vote from me is only about those publications that state a USA and Canadian ISBN on the copyright page)

• Yes. BLongley 22:57, 7 February 2009 (UTC) Because you're all just talking about US / Canada differences from two or three publishers and I don't think you're looking at the bigger picture. If the Canadians want to roll over now and let their ISBNs go to notes, fine. I am not going to vote for such for all other discrepancies. If we haven't got the other ISBN recorded as a separate pub, record it, I say, and note why it is there. BLongley 22:57, 7 February 2009 (UTC)

I believe that the current practice for dual-ISBN'd publications is to record the other ISBN in Notes if there is only one pub record. I don't think we record all related ISBNs for books that share the copyright page (e.g. one ISBN for the hardcover edition and another ISBN for the trade paperback edition), but that's a different issue since they are not true "dual-ISBN" pubs. Ahasuerus 05:53, 8 February 2009 (UTC)

My current practice is to create the other ISBN'd edition if we don't have it already, either explaining the record came from a verified pub, or more usually finding it from another source. The problem is that (apart from these few example publishers) nobody has explained how to spot a dual-ISBN'd pub from a pub that references other pub's ISBNs that deserve their own entries. BLongley 20:54, 8 February 2009 (UTC)

Ah, I see! Well, if a book has 2+ ISBNs printed on the copyright page, they usually have disambiguating information printed right next to them, e.g.:

Merril, Judith, 1923-
Survival ship and other stories
0919588085 (cloth)
0919588077 (pbk.)

so in this case the copyright page claims that there are two editions of this book, one a hardcover and the other one a paperback. If your copy is a hardcover, then the first ISBN applies to it and if your copy is a paperback, then the second one applies. If, on the other hand, the copyright page has a separate ISBN which says "Can." or "Canadian" next to it, then both ISBNs refer to the same physical book. Ahasuerus 02:17, 9 February 2009 (UTC)

Is that an absolute rule, or just for the publishers identified so far? BLongley 21:38, 9 February 2009 (UTC)

Sometimes things get more complicated, e.g. "library binding" versions may looks a lot like regular hardcover versions to people who are not familiar with library binding books, but overall it should work reasonably well.Ahasuerus 02:17, 9 February 2009 (UTC)

I'm extremely unfamiliar with Library Binding books, and the only probable example I have is ISBN'd differently on back cover and copyright page - as it seems to be the paperback rebound with harder covers. But that's a separate issue. BLongley 21:38, 9 February 2009 (UTC)

Of course, any claims made on copyright pages need to be taken with a grain of salt. As our record for this Merril title explains, "The copyright page of the paperback edition also lists ISBN 0-919588-08-5 for a "cloth" edition, but according to Lloyd Currey's "Science Fiction and Fantasy Authors", it was never published". But then again, any secondary source can be wrong and, besides, we may want to create 8888-00-00 pubs for such. Ahasuerus 02:17, 9 February 2009 (UTC)

"8888-00-00" might be a solution, but might also be misleading if the reference is "Published simultaneously in Canada". BLongley 21:38, 9 February 2009 (UTC)

And I must admit I'm slightly disappointed in Ahasuerus voting "No" before he's seen what Fixer will do with Amazon Canada or Amazon UK data. This is a question that doesn't need to be addressed this fast. Unless it's only a question about Baen and Daw - Tor titles (from recent Fixer submission evidence) need a lot of work, as we're getting a lot of false entries for US versions that never existed as far as I can see. And I don't appreciate being rushed on such decisions. BLongley 22:57, 7 February 2009 (UTC)

I am not entirely sure I see how Fixer's misadventures with "Tor UK" books are related to the issue of dual-ISBN books. Do Tor/Tor UK print two ISBNs in the same book? Ahasuerus 05:53, 8 February 2009 (UTC)

Not as far as I know, they tend to have exactly the same Title, Author, ISBN and publication date. Only the price is wrong, and that's down to Amazon.com being the default source, and Fixer won't submit a publication if we've already got that ISBN. Dissembler seems to have done so too. BLongley 20:54, 8 February 2009 (UTC)

OK, so they are not "dual ISBN" books, just two separate books that share the same ISBN, a common (if irritating) practice. Unfortunately, as long as they have the same title/author/ISBN/publication date and the only differentiating field is price, I don't think that Fixer can be realistically expected to create two separate records based on Amazon US/UK's data -- even if his IQ suddenly doubled. After all, many "Tor US" books are also sold by Amazon UK using UK prices, so how can an honest hard working bot tell whether a Tor book available from Amazon UK is a duplicate of a "Tor US" one or a separate "Tor UK" book? We would have the exact same problem if we processed Amazon UK's data first, so I don't think there is a silver bullet here. On the plus side, regardless of which store is processed first, Fixer will at least create a valid pub record for a real book and the other book can be added manually later. Ahasuerus 02:17, 9 February 2009 (UTC)

That's fine for the future, what about the publications we have now? I suspect we already have duplications. If the goal is to end up with "one pub, one ISBN, other ISBN in notes" then we need to figure out how to get there from the current chaos. BLongley 21:38, 9 February 2009 (UTC)

The practices of our major bots are introducing a "US First!" policy I dislike intensely. BLongley 20:54, 8 February 2009 (UTC)

As long as Fixer can create only one pub record per ISBN, we can try to make sure that we create a record for the publisher's home country version by identifying "primarily US" and "primarily UK" publishers and submitting their books from the respective country's Amazon store first. This will also address a different (but related) issue with prices for books that were never officially published in a particular country, but are sold by that country's Amazon store. This is especially desirable since I have finally found the Amazon.com flag that indicates that a book is an import (clearly, I need new glasses) and then discovered that it is set for only about 50% of all Gollancz books :( Ahasuerus 02:17, 9 February 2009 (UTC)

Glad to hear of the find. I suspected it would be about that accurate and/or useful. :-/ I don't think "primarily US" and "primarily UK" publishers is going to help much though. For instance, "obscure British Author" published by "primarily US publisher Tor" was a warning sign to me that all his books were wrongly entered. I don't think we can program for such (or you could replace me with a bot too) but we can share such knowledge. I just don't want the knowledge you apparently already have about a few US/Canada publisher practices to become canonical for ALL books that mention another country's edition. If the vote was just on Baen and Tor and maybe DAW (I've lost track of the ones where everyone agrees) US/Canada differences I can live with that, as I've already stated. Someone needs to update Help and such, and there should be a plan to fix existing dual entries. BLongley 21:38, 9 February 2009 (UTC)

I believe that all Amazon stores list just one ISBN-10 (and one ISBN-13) per book, so dual-ISBN books will appear as regular single-ISBN books on Amazon. Granted, if Amazon.com uses the US ISBN for a dual-ISBN book and Amazon.ca uses the Canadian ISBN, then, given the "relegate Canadian ISBN to Notes" solution, Fixer, which operates based on ISBNs, will not find the Canadian ISBN on file and merrily create a new publication record for the Canadian ISBN. I am not sure what we could do about that short of asking Fixer to add a special warning about any 1980s Tor and Baen records imported from Amazon.ca. Ahasuerus 05:53, 8 February 2009 (UTC)

This is exactly the sort of experiment I think we should wait for. Let Fixer loose and see how the "one-pub, put the other ISBN in notes" people deal with it. There's a lot of examples here already I think, and I've not seen any attempt to fix such. BLongley 20:54, 8 February 2009 (UTC)

If the true "dual ISBN" books are limited to Tor and Baen in the 1980s -- and if there are any other known suspects, please speak up -- then we should be in pretty good shape. There are only 9 seemingly legitimate Baen ISBNs and ~20 Tor ISBNs from the 1980s that Amazon knows about and that we don't have on file, so we should be able to handle them easily. Ahasuerus 02:17, 9 February 2009 (UTC)

Go for it. If this is just Tor and Baen for a limited period, you can document the exceptions. BLongley 21:38, 9 February 2009 (UTC)

Overall, this has been a rather confusing discussion. I have read this section three times and I am still not sure what the issues are aside from the "dual-ISBN" books published by Tor and Baen in the 1980s. We all know that the same ISBN can be used by multiple ISFDB publications/printing, so that's not an issue. We all know that many books print the ISBNs of their "sibling", editions (slipcases, limited edition, library binding, paperback, etc) on the copyright page, so that's not an issue either. What are the other outstanding issues then? Ahasuerus 05:53, 8 February 2009 (UTC)

As I see it - people are suggesting that we don't create missing editions based on data in their hand. BLongley 20:54, 8 February 2009 (UTC)

I don't think I have noticed any suggestions along these lines in this discussion, but then it's gotten so long that I may have missed it. If so, please speak up! :) Ahasuerus 02:17, 9 February 2009 (UTC)

I create records for missing editions but make every effort to ensure that the book exists and that the set of metadata is all for that one publication. BLongley, you may be shocked to learn I look at www.fantasticfiction.co.uk, www.amazon.co.uk, www.amazon.ca, and sometimes even the DE, FR, plus JP versions of Amazon. I've run across books that mentioned a Canadian ISBN on the copyright page and can't recall a single instance where I was able to establish that the ISBN was used other than on the copyright page. They never show up in AbeBooks, www.bookfinder.com, on any of the Amazon sites, Google, etc. Thus I just document them in the notes. --Marc Kupper|talk 05:55, 9 February 2009 (UTC)

I'm not shocked. I check enough obviously relevant versions of Amazon until I can be fairly sure (although Japan is usually beyond me). And author websites (you may have noticed I've added dozens of such recently, even when they're just for Vampire-Shaggers.) And fan-sites. And whatever back-doors we have into OCLC/Worldcat data. And Abe and Alibris. But what seems to be being suggested is that "these publishers have known practices for Dual ISBNs, let's make a rule for all Dual-ISBN'd pubs". Narrow the Vote to just the two or three publishers that it seems everyone but me "knows" did this and I'm fine. Put it in Help, on the publisher Wiki pages, anywhere that will help prevent the practice. But restrict it to the ones you already know about. What happens when we get to US/UK ISBNs in the same book? I'd start another discussion. I do not want to go back to this and have to say "well, we only actually agreed on USA/Canada differences, for two or three publishers, forget the fact that it's been made a rule for all publications". I'm sticking my oar in NOW because I don't want this to turn into "First they oppressed the Canadian ISBNs, and I didn't speak up because I wasn't Canadian"... BLongley 22:12, 9 February 2009 (UTC)

Or that Editors and Mods will instinctively know when a second ISBN reference is actually just another reference to the same publication, rather than another edition. I can live with exceptions for certain publishers that are known to have put two ISBNs on the same publication without the other existing separately - but those should be the exceptions, not the rule. BLongley 20:54, 8 February 2009 (UTC)

I can think of only one reason not to create Publication records based on the additional (non-Tor-Baen-Canadian, that is) ISBNs printed on copyright pages and that is that the data can be suspect as we saw with the Merril collection above. However, I don't think that it's any more suspect than what can be found in most secondary sources, so I would encourage editors to use it to fill in any gaps and document their source in Notes. In some cases, especially when it comes to limited editions, it may be our only source of information. Ahasuerus 02:17, 9 February 2009 (UTC)

(unindented)Tor and Baen both experimented with a single publication with two different ISBN#'s. To be clear there is only one book not two out there. From my experience the only ISBN# anyone uses with any consistency is the US one. The Canadian ISBN# is mostly ignored by everyone including Canadians(which includes me). A separate listing would only confuse anyone looking at the entry and any searches for the book would come up empty. The Canadian ISBN#'s as far as I'm concerned should be relegated to notes where it belongs.Kraang 01:01, 9 February 2009 (UTC)

I think this. [2] Mar 1989, and that. [3] Aug 1988, show the Tor dual ISBN use and how I have started dealing with it. Also note that back of front cover ISBN formula is for returning covers for credit and the implication that the book is not really Canadian (IMO). I have not been mirroring the Can. ISBNs. It would be easy to mirror, but as a 'user' I would look for these to exist as Canadian not dual products. Sorry, if this does not meet your needs. Suffice to say, I am concerned as I use ISBNs first when searching for Amazon/Abe information and frankly the dealers have corrupted them over the years. My thought is if you can not find a Canadian ISBN in use, do not add it. This [4] was created from this SBN 52348500X on the bottom of the paperback [5] . I changed it to the ISBN used on the copyright page to 'divorce' it from the Amazon hardcover edition that is cited. As of yet, no one seems to have the hardcover or data to support it's printing. This is stated to show how a manufactured ISBN creates uncertainty. So, I lean to the double ISBN books, not having the Canadian ISBN used here until someone can show they were used. By the way, my two first examples, have the last digits printed on cover with Canadian and American pricing. Basically, I am only concerned with adding 'dated' ISBNs to the db, not new ones. New information from the publishing world will always need sifting, but as you can see the 'commercial' world can create publications that do not exist. Apologies for this confusion, Thanks, Harry. --Dragoondelight 16:29, 9 February 2009 (UTC)

(unindent) OK, it looks like we have reached consensus re: the 1980s Tor/Baen books which had two ISBNs printed in the same physical book. I think we are ready to change Help now. I will also start another section about other possible permutations that Bill mentioned above. Ahasuerus 22:09, 10 February 2009 (UTC)

Yes, I'm OK with Tor & Baen exceptions. I'm afraid I still can't quite understand Harry's examples, but I doubt I'll have problems with Tor and Baen, we seem to have pretty good coverage of those already (ebooks aside) and I don't think I'll add anything there. I'll pipe down now until I see people trying to use it for other publishers. BLongley 22:33, 10 February 2009 (UTC)

Science Fiction Series by Gernsback

This. [6]. I have No. 12, not my example, but before I enter it I would like to know if the example entered as an 'anthology' is correct by the present standards. I was leaning toward 'chapbook'. Just trying to get the ducks in line. Thanks, Harry. --Dragoondelight 00:09, 4 February 2009 (UTC)

See ISFDB:Help_desk#ChapBook_Anthology It's an Anthology Pamphlet. I had the same question on the same record 6 months ago. Kevin 00:59, 4 February 2009 (UTC)

Thanks, that is reassuring. I will enter it soon, but it's actual formatting is unique to it's age. The current entries must be from other sources. Thanks, Harry. --Dragoondelight 13:06, 4 February 2009 (UTC)

Anthology Builder - in or out?

An author just e-mailed me and asked if publications from http://www.anthologybuilder.com are ok. She's already put together a collection of her stories, all of which have appeared in dead-tree magazines. If this was Lulu.com I'd say "no problem" but the Anthology Builder it much like plugging a list of songs into an MP3 unit in that it allows anyone to take a pre-built anthology or collection and to customize it before printing a copy. You can change the cover art, layout of the title, type up your own introduction, add/remove stories, etc. The books apparently don't have an ISBN. I'm checking to see if they contain a code that would allow someone to order an exact duplicate of a book without needing to build it by hand.

Any thoughts on in or out? --Marc Kupper|talk 07:22, 6 February 2009 (UTC)

Considering that the data is probably of value to only one person I would think Out.--swfritter 22:56, 6 February 2009 (UTC)

It appears that crafted anthologies 'can' be referenced if 'published' to a library on the site (example Planet Interzone), but I agree.... the lack any hurdles beyond $14.95 invites anyone to craft an anthology, order one copy and 'earn' placement as an editor in the ISFDB. I'm all for being inclusive.... but that bar is a little low with no real world commercial culling or involvement.Kevin 00:18, 7 February 2009 (UTC)

On the topic of individual stories however.... I wouldn't be adverse to links in the story title records to show that they are available at anthology builder ... (example Looking Through Lace) but even then I would prefer to wait until the website is a year or two out of Beta (it is currently still a beta site) before we start linking to it. Links are fine and all... but they should be static links to singular or stable resources ... not a beta test or a start up operation. Also individual works here don't qualify as individually published chapbooks etc because they are only offered for publication, not actually published here.Kevin 00:18, 7 February 2009 (UTC)

If someone were to craft a particular interesting anthology...which sold in high enough number (Subterranean Press has print runs of 1000) (through whatever means) and/or which was later re-released with an ISBN, I would however whole heartedly support documenting the original anthology builder edition, (with a link to the anthology order page on that site) as the first edition of a work with its own separate pub record.Kevin 00:18, 7 February 2009 (UTC)

Out. MHHutchins 00:29, 7 February 2009 (UTC)

I stumbled across my first author website that referenced one yesterday, Jacey Bedford, and looking at her example there doesn't seem to be any sign of number of copies sold, if any. It's just a template: possibly an "8888-00-00" title. I'd say out. BLongley 12:06, 7 February 2009 (UTC)

We really haven't fully explored the use of tags as a database tool. Perhaps this would be a good use for tags assuming that there is someone who wants to maintain them on an ongoing basis as I have for Project Gutenberg. They can serve also as intermediate documentation until database entries are made. Has there ever been a discussion about using tags for personal collections? It would be kind of nice if we had tags that were invisible to others in order to tag items in our personal libraries? - although I would prefer pub level tags for that.--swfritter 21:04, 7 February 2009 (UTC)

I so VERY much look forward to the day when I can mark publications as 'I Own This', either publicly or privately, and then display / review / search within those items.Kevin 21:47, 7 February 2009 (UTC)

I would love to mark such publications too, preferably programatically - I'm sure I've already bought a book again because someone else verified it first. :-( My fault for not keeping my own database as well I suppose - but then again, why should I have to? In a true Web 2.0 world I should be able to match all the books Amazon know I have, with the ones I've verified here (Primary or Transient), plus anything I've done on Librarything (which is nothing actually), or have listed on ReadItSwapIt, and get reminded of books that I should have (and actually do) with a quick way to update them all. Ah well, wait till Web 3.0 I guess. BLongley 22:01, 7 February 2009 (UTC)

Out, IMHO. But it doesn't seem unreasonable to me to count an individual story's appearance on AnthologyBuilder as a sort of publication. --WimLewis 03:31, 8 February 2009 (UTC)

I agree with the "out-ers", but it makes you wonder what else may be around the corner courtesy of this wonderful digital age of ours. At one point in the late 1970s I came across a book which claimed that major libraries may be able to exchange data electronically in real time by year 2000. I remember thinking "That would be great, but doesn't seem very likely" :) Ahasuerus 04:33, 8 February 2009 (UTC)

Thank you for the feedback. I also heard back from both the publisher and an author on this. At present the publications don't have a code that can be used to either reference a book on the AB site. The author had printed one copy of her collection a while back and ordered another recently. As I asked her about the page numbers she was startled to learn that they are different between the two editions of the same book. Apparently the new edition is slightly smaller but has more pages. The AB people are still very much in beta but have added printing a publication code in the feature list pile. Thus for now they really are "out." --Marc Kupper|talk 05:03, 9 February 2009 (UTC)

Where did that pesky webzine go?

Helix, that is. I am assuming the author mentioned in the story is Janis Ian. Also mentioned is the fact that the archives would have eventually disappeared anyway.--swfritter 22:14, 7 February 2009 (UTC)--swfritter 22:14, 7 February 2009 (UTC)

Most deplorable. I was going to snapshot the site, but apparently I wasn't fast enough :( Oh well, there is still a megabyte of useful data left -- including some good biblios -- now safely on a CD. Ahasuerus 04:36, 8 February 2009 (UTC)

Janis is an idiot. This is the internet. Electrons are really hard to delete. Look here for a goodly portion of the website Internet Archive of Helix SF and look here for the story in question Mahmouds Wives. It wasn't really worth the effort on her part to try and remove it. Kevin 05:24, 8 February 2009 (UTC)

Looks like somebody deleted the Helix pubs and left orphan titles sitting out there. As long as we had enough documentation on the magazine page it might have been well to keep them - at least they served as justification for our decision not to document webzines.--swfritter 02:03, 11 February 2009 (UTC)

Is that the (an) acceptable solution for item published in 'out' publications by known authors (Where a work is likely/possible in a decade or two to show up in an 'in' publication as a reprint of some sort)? Document the publication and title, then delete the publication to leave the Title on the authors bibliography. Kevin 03:54, 11 February 2009 (UTC)

If that method is used it is probably a good idea to make a notation in the notes for the individual titles - which, of course, would be easier to do before removing titles from the pub.

This is not the appropriate place to make rude remarks about people who are much smarter and braver than any of us. Remember, what goes on the internet never goes away.--swfritter 00:45, 12 February 2009 (UTC)

This is my point exactly. ...what goes on the internet never goes away. Janis's professional representation would do well to learn this particular axiom. Kevin 03:22, 12 February 2009 (UTC)

It's probably not a good idea to comment on living speculative fiction authors, editors and publishers here. You never know when you may need to e-mail them to clarify a particularly thorny bibliographical issue :) Ahasuerus 02:13, 12 February 2009 (UTC)

I wrote a really long (and perhaps moving) response to this thread... but I have condensed it (in deference to the several adult beverages I have enjoyed this evening). I amend my statement. Janis's lawyers (or her agent, or her agents lawyers) are incompetent or un-savvy in the workings of the Internet. They should refund any and all fee's they charged for services in relation to the afore mentioned legal action as the results were substantially ineffective, and deleterious to her professional reputation IMHO but IANAL.Kevin 03:20, 12 February 2009 (UTC)

The Starry Rift

I just cloned my version off of a record that has so many different things that I did not feel right to blank it out. There are two main problems with the collection as is. One is it is set with a fictional short story, that others have termed an essay. The Title tell you this. It also is a split story in that the end is in the back, but I noted that. Do I leave it an essay or change it to a fictional short story? Second problem is that by showing the book as a collection. The Titles start with first tale, second tale and third tale. Now as a novel with three stories these would be termed chapters. A collection should have a ToC, and this does not. I am putting it to you, do you want three new variants created or do we just note and ignore the variants? Up to you as I am non-pulsed by this problem at this time. This [7]. Thanks, Harry. --Dragoondelight 00:08, 8 February 2009 (UTC)

I note that none of the 'essay's appear in verified pubs, so no-one else has verified it as an essay. If it is fictional, then change it as appropriate on your clone. This will update the other instances. The split story (in your edition)... do you believe it appears split in all other editions? If not, then on your clone add two new titles, 'Split Story (Part 1 of 2)' and 'Split Story (Part 2 of 2) and mark it as a serial, then change page number for the original version to 'remove' with plans to remove it from the pub later. If it appears split (and you have evidence) in all editions (check for reprints of the story first) then change the title of the original story to (Part 1 of 2), add Split Story (Part 2 of 2). (As a courtesy.. .you should go back and add Part 2 of 2 to the other editions, but you should only go down this route if you have evidence that all other editions split the story). On the 'First Tale: A Story' issue... it's a toss up, I would look for evidence of them being published as such elsewhere... and if so (or something else makes you think that your edition was re-edited) Variant the titles... otherwise leave well enough alone and just make a note about the story precursors being present in your edition. Perhaps in a few years someone can confirm it one way or the other. That's just my 2 cents and how I would imagine handling your issue.... I now sit back and wonder what our more experienced peers will have to say.Kevin 00:35, 8 February 2009 (UTC)

According to the Locus Index, this is a "collection of 3 loosely connected sf stories set in the Rift [universe] - not a novel, though it may look like one", so I have put the stories (and the collection Title) in a Rift series.

As far as the framing device goes, "in-universe" essays are typically entered as short fiction, so I have adjusted the Title Type. I also entered "At the Library" as a short story and set up variant titles. My SFBC edition has the same layout, so it's quite likely that the rest of the editions will need to be adjusted similarly. Ahasuerus 05:10, 8 February 2009 (UTC)

Thanks for the rescue. I love collections, but hate 'cloaked' collections. Without input I was stymied. Thanks, Harry. --Dragoondelight 13:40, 8 February 2009 (UTC)

US/UK ISBN issues

As per Bill's request, I have run some checks against Fixer's "library catalogs" database to determine whether there are any US/UK books that may be affected by the "dual ISBN" problem. The main cataloging format lets librarian put anything they want after the main ISBN, not just the data that publishers print on the copyright page, so Ficer's ISBN data is rather messy. Still, here are the results of a cursory investigation of some 200,000 ISBNs.

In a few cases, we have what look like regular "single ISBN" books which are marked "US" (and perhaps "UK", although I didn't look very closely once I saw what was happening) by cataloging librarians, presumably for clarification purposes. For example, Robert Giddings' J.R.R. Tolkien: This Far Land was first published by Vision (London, UK) and its ISBN was 0854783350. In 1984 it was reprinted by Barnes & Noble (Totowa, NJ, US) and the ISBN was 0389203742, while the copyright page apparently said "London: Vision ; Totowa, NJ: Barnes & Noble". Some librarians added "(US)" when cataloging the US reprint to distinguish between the Vision edition and the Barnes and Noble edition, but I don't think this designation appeared on the copyright page. In any case, this is a simple case of two different editions using two different ISBNs and there should be no confusion.

Other books, however, seem to use two ISBNs, one for the US market and one for the UK market, for the same physical book. For example, Wayne G. Hammond's J.R.R. Tolkien: A Descriptive Bibliography has two ISBNs associated with it, 1873040113 (UK) and 0938768425 (USA), which apparently match two publishers: St. Paul's Bibliographies (Winchester, UK) and Oak Knoll Books (New Castle, DE, USA). Similarly, The Dedalus Book of Portuguese Fantasy published by Dedalus (Sawtry, Cambs, UK) and Hippocrene (New York, USA), has two ISBNs, 1873982666 (UK) and 0781803861 (US). I couldn't find separate UK or US versions of these books, so I assume that they were jointly published by two transatlantic publishers, each one providing a separate ISBNs. Now that I am thinking about it, what's to prevent other co-conspirators from doing the same thing even if they are based in the same country?..

Now, if it turns out that there are no tangible differences between the versions distributed in the US and the versions distributed in the UK, then we would have the same "dual ISBN" problem on our hands. However, before we start panicking or at least thinking about different ways of handling this issue, we may want to make sure that we are indeed dealing with the same physical book. For example, the Library of Congress record for the Hammond book linked above indicates that the US price was $94.00 while it gives no price for the UK ISBN. If there is a recorded price difference between the US and the UK versions of the book (similar to the Tor US/Tor UK doppelgangers that Bill mentioned earlier), then the problem is solved and we can create separate records for the two ISBN using price as the disambiguating field. Some other identified "dual ISBN" books, e.g. the 2004 edition of Weis/Hickman's War of the Twins (ISBNs 0786932171 (US) and 078693218X (UK)), do have separate prices listed in the library record ($24.95 and £16.99 respectively), so there is still hope!

I should also point out that these dual ISBN books tend to be either academic or fairly obscure, which is probably why we haven't run into them so far. For example, one of the offenders is an Ancient Greek (sic) translation of Harry Potter and the Philosopher's Stone... Ahasuerus 00:15, 11 February 2009 (UTC)

Thanks for the summary. The Oak Knoll / St. Paul's connection is fairly easily explained. The Dedalus / Hippocrene one may be similar, and there are titles present here and here, for instance, where there may be some confusion still to come. Hopefully this is a small number of publishers that can be handled on an exception basis. BLongley 19:17, 12 February 2009 (UTC)

Multiple Sub Series Omnibus

Ahasuerus and I had a bit of discussion on this issue User_talk:Kpulliam#Miles.2C_Mutants_and_Microbes and he suggested I put my perhaps logical but unusual solution to the community as a whole. Discussion Duplicated below from my talk page. My standing argument is that the author herself states that this omnibus ...collects the quaddies in all their charm. Kevin 17:45, 14 February 2009 (UTC)

---

I see that you would like to move this omnibus from "Barrayar" to "Quaddies". As far as I can tell, only one of the three novels in the book is in the "Quaddies" series while the other two are Miles novels. Given this ratio, would you say that we want the Omnibus Title to remain under the Barrayar super-series? Ahasuerus 07:19, 8 February 2009 (UTC)

This omnibus collects all three stories (2 novels, 1 short fiction) with Quaddies as primary characters. The last sentence of the authors introduction states "But most of all, this omnibus, haunted as it is by the ghosts of books unwritten, never to be written, collects the quaddies in all their charm". So the intent of this omnibus was to collect (and identify) the Quaddie related material (even that material which may normally be classed in the other Barrayar sub-series). Kevin 07:42, 8 February 2009 (UTC)

Hm, interesting! I suppose the most logical way to handle this situation would be to put the two affected Miles novels in both sub-series, but, unfortunately, our software doesn't support that. If we moved the Omnibus to the Quaddies sub-series, then the fact that 2/3 of its contents is in another sub-series would be confusing. Perhaps leave it in Barrayar and add a Note to the Omnibus Title? Ahasuerus 03:04, 9 February 2009 (UTC)

I still disagree. The title (with "Miles") clues anyone looking for Miles Vorkosigan material that the volume has content related to him, but only the subtle hint of 'mutants' in the title clues folks looking for stories related to the Quaddies. By putting it in the Quaddie sub series, we flag it from the outset as quaddie related material. Folks looking for material concerning Miles literally cannot help but to trip over it throughout the Barrayar universe, but material related to the Quaddies is not so labelled elsewhere. Another way to look at is is that 2/3 titles contain fiction related to Miles, while 3/3 or 100% of the fiction is related to the Quaddies sub plot / story arc... in fact this collection tells that whole story arc as complete as it's going to get, while anyone looking for Miles material 'might' at most not stumble upon one novel and one short fiction work. Kevin 05:47, 9 February 2009 (UTC)

Well, the approach seems logical, but it's rather unusual, so perhaps we should bring it up on the Standards board to see what other folks think? Ahasuerus 06:06, 9 February 2009 (UTC)

Kevin's solution sounds as good as any other alternative. All of the individual contents are in the correct series so they will be easy to find.--swfritter 22:11, 15 February 2009 (UTC)

Sounds good, approved; notes added. Ahasuerus 02:49, 18 February 2009 (UTC)

Locus "interviews"

Well, I've reached that point in entering issues of Locus (#277, March 1984) that I dreaded. That's when Charlie Brown starting doing author interviews on a monthly basis. Here's the problem. They're labeled as interviews, but there is no credit given for the interviewer. I'm 99% sure it's Brown, but who can say that every interview in the past 20 years was conducted by him. I see that persons before me are giving him credit (about 150 records!), but if we follow the ISFDB standards it should be "uncredited". Here's another issue: they're not really interviews! There is no Q&A, but it's obvious that someone is prompting the responses. In fact, taken at face value, they are self-interviews, with an extensive (uncredited) introduction. In that case, shouldn't the interviewee be considered the author of the piece, and that it be typed as an ESSAY by the author? I just want it settled before I start entering them into the database. Should we make an exception to the rule for this unusual circumstance, or should we abide by the standards as they have been set? There is an "out": in the early issues of Locus, there was a note in the masthead that "all uncredited pieces are the work of the editor". Might we not invoke that statement to apply for later issues and giving Brown credit for the interviews? I don't look forward to changing those 150 pieces already in the db! Thanks. MHHutchins 19:44, 14 February 2009 (UTC)

The first question that comes to mind is "Why don't we ask Charlie Brown or, if he is not available to answer questions, Bill Contento or Mark Kelly, who edits Locus Online?" Bill and Mark are certainly available via e-mail and I have exchanged e-mail messages with both, so I don't think we have anything to lose. Ahasuerus 20:56, 14 February 2009 (UTC)

As an incredibly shy person, asking a "Speaker to Locus" to continue doing so is preferable to trying to construct a sensible message that will get through spam-filters, not sound too over-awed, and not too critical. Having said that, I've had no problems with Bill Contento - but that may be down to being a fellow member of the Secret Society of Bills. Mike may be less shy than me, and I don't think he'd ruin our reputation with a "Oi! Charlie! WTF did you think you were doing back then, you're making it a pain for us now!" sort of message - but it may be preferable to restrict the number of ISFDB editors bothering such people, if we're going to ask a lot of questions. BLongley 23:26, 16 February 2009 (UTC)

The staff of Locus has grown over the years and I think it is entirely possible and perhaps likely that others have conducted the interviews. Whether the questions are printed verbatim or not they probably should still be considered as interviews. I wouldn't be in a rush to change the existing ones because who knows when we will change our minds again. After the task of changing 25 years worth of John W. Campbell, Jr. editorial credits to John. W. Campbell, Jr. (as The Editor) and taking many of the existing editorials out of a series and putting the new titles in I can guarantee you it is not much fun.--swfritter 21:45, 14 February 2009 (UTC)

I think we've got a little off the point of the original question. :) Regardless of who actually did the interviews (and God bless 'em), ISFDB standards have us declare the pieces as uncredited if no one is credited in the publication itself. Right? Once someone (not me) has determined who did the interviews, they can create a variant. Until then I'll follow the established standard and enter them as "uncredited". I'm not going to change the existing ones immediately, but by the time I reach those mid-90s issues that have partial contents in their records I'll have formed a better idea about how the situation might be handled. MHHutchins 23:43, 16 February 2009 (UTC)

One of my (current) set of first principles on ISFDB editing is not to do anything that may well need to be undone. If it looks arguable, don't start doing it. So if you think that "all uncredited pieces are the work of the editor" might be the consensus, argue for that rather than do it. I'd just leave them out for now. There's supposedly no pressure on any editor to do more than they feel they need to do. I think the peer pressure on magazine entry is a bit (well, a lot really) higher than on books - I still don't feel much pressure to add "Maps" or other "Interior art" to books, but I can see how all the Magazine Series entries will need re-doing with canonical authors to get them to work "properly". And Artists. I know we have standards, but as many were set when there were only about half a dozen(?) editors and we have three times as many Mods now and I don't know how many more active editors, I think any standard should be questioned a bit - and a LOT if it means hours or days of rework. We'll eventually get to the stage where it's just too much WORK to change things however sensible the change is. (We may already have in some areas - see below.) BLongley 00:18, 17 February 2009 (UTC)

"uncredited" for now is the right approach - pseudonyms and series can wait for later. I am more than willing to let this be your legacy. And hopefully not your Waterloo.--swfritter 00:28, 17 February 2009 (UTC)

Concur; if we have an agreed upon standard -- in this case "uncredited" -- then we should follow it until and unless we decide to change it.

As far as Bill's point that "There's supposedly no pressure on any editor to do more than they feel they need to do" goes, it's true that partial publication data is better than no publication data and that every editor decides how much detail to include in Notes, although the "pressure to comply" also depends on what's left out, e.g. compare and contrast missing maps vs. missing ISBNs. In the case of Locus, a magazine devoted to the SF industry, interviews, letters and reviews, interviews are quite important and, besides, Michael is willing to enter them, he just wants to make sure that the how is correct. Ahasuerus 03:11, 18 February 2009 (UTC)

Fixing Pseudonyms

I see some nice activity here, but there's a few things I'd like to clarify:

1) Interviews. Greg Keyes for example. "Q&A: Greg Keyes" is still on the pseudonym's page and not available on J. Gregory Keyes's page. Unless someone knows how to create a variant for an interviewee, should there be some help changes to say that Canonical author should be entered as the reviewee rather than exactly as stated in the publication? Or can we live with the lost links? BLongley 22:58, 16 February 2009 (UTC)

If there is no other way, I would be inclined to enter the canonical name for the interviewee and add a comment to Notes. Ahasuerus 03:24, 18 February 2009 (UTC)

2) Tags. Greg Keyes for example, again. "Author Tags: fantasy (3)" (Only (1) on the canonical.). It might be worth leaving those on the pseudonymous titles if the pseudonym is only used for the keywords you're tagging. E.g. All Iain Banks (oh, another interview example!) could be non-genre and all Iain M. Banks could be genre. Or should we replicate all tags on the canonical titles so we can see that the canonical author wrote both? BLongley 22:58, 16 February 2009 (UTC)

I think it's important to keep tags under the canonical author's name, although it's easier said than done since you can't delete other editors' tags. An inactive editor's tags will remain there forever unless/until the software is changed to allow moderators to change/delete wrong/misleading/etc tags. Ahasuerus 03:24, 18 February 2009 (UTC)

3) Series. If the pseudonym had a series set up and the canonical author didn't, you have to put the canonical titles into the series or it won't show up on the canonical author's page. You don't have to remove the pseudonymous titles from the series to get both author displays to work. But the series display will double up titles if you don't, and I for one don't like that. Does anyone? I know removing series data from one and adding it to the other is more work though, and it's tempting to only do half the work. BLongley 22:58, 16 February 2009 (UTC)

I always remove Series information from Variant Titles, although it is admittedly tedious. Proper display of variant titles on Series pages has been an outstanding feature request for a couple of years now. Ahasuerus 03:24, 18 February 2009 (UTC)

Oh, and to go off-topic immediately - I'd encourage people to go off-topic and fix other things they find. For instance, I've separated the three "Dragonfire" series that we had intermingled. (And surprisingly, no McCaffreys were involved.) And if an author looks suspicious, post suspicions - if I see Carl Dreadstone "fixed" by making them all "Ramsey Campbell" titles I'll be a bit annoyed for instance. It's not a totally mindless exercise - but it may look like one at times, so I'd recommend wandering off and enjoying some proper research occasionally. This is why my contributions are spread over several ranges, not just on the unfinished A-C for instance. I have the attention span of a butterfly at times and need a change of routine, so I - did someone mention a butterfly? OOH! Must go find the butterfly! BLongley 23:13, 16 February 2009 (UTC)

I found out sooner than later that what looks simpler becomes ever more complicated the more you really look at. Along with the series problem (I've been removing them from the pseudonymous author's page, otherwise you'll see the series name but no titles), the problem of unconfirmed pseudonyms and house names has reared its ugly head again. I've stopped on that project for now, and if I go back to it, it will be those simple fixes. The most obvious being a slight change in the name, or a well-documented pseudonym. I've ran across very common names where the dates just don't seem right, or someone who's a writer and an artist (rare, but it's not unheard of.) MHHutchins 00:44, 17 February 2009 (UTC)

And not just very common names either -- who would have guessed that Mel Odom the writer (b. 1957) was not the same person as Mel Odom (artist) (b. 1950)?.. Ahasuerus 03:24, 18 February 2009 (UTC)

OK, we seem to have 3 votes for "remove Series information from Variant Titles". I think 2 (me and Ahasuerus) for "use canonical name for interviewees". 1 for "keep tags under the canonical author's name". (I'm a bit ambivalent about tags, they don't feel particularly right in a structured database but at least they're in the database rather than the wiki.) BLongley 19:48, 18 February 2009 (UTC)

Dzurlord---Not a novel

This. [8]. I have a copy and could easily verify it, but it is a RPG adventure using dice, etc. I believe that this DB does not keep these on file. I tried to delete, but of course no luck. Up to someone else. Thanks, Harry. --Dragoondelight 21:43, 17 February 2009 (UTC)

Gone. But what was the problem with you deleting it? Did you try to delete the title before you deleted the publication? BLongley 23:02, 17 February 2009 (UTC)

yes, I did not delete the publication first. It was a first attempt to delete an RPG product, so my thinking was bent. Thanks, Harry. --Dragoondelight 13:36, 18 February 2009 (UTC)

It can be a pain, agreed. You need all the pubs to go before you can delete the titles and just submitting the pub-deletes isn't enough, you need them approved too. And then if somebody's put the titles in an RPG series the series still won't go when all the titles are gone. I'd love some Super-Ultra-Killer-Obliterator tools at times but I guess that we'd need a few more checks, e.g. that a "Delete series/author/title and everything under it" submission needs to be approved by a different moderator at least. BLongley 19:38, 18 February 2009 (UTC)

One DJ, two ISBNs, two pub records?

I think the last discussion ended in a one-all draw, if you count me as one and "Rest of the Universe" as one. ;-) Or more seriously, that we can all live with a "one pub, relegate other ISBN to notes" for 1980s Tor and Baen books. I'm afraid the next publisher (or at least one publication) to consider has come up when I rechecked one of my older publications. See here:
It's one continuous scan of the DJ, I haven't merged images. Inside back flap ISBN (0-00-224637-6) doesn't match back cover (0-583-32905-5), or more importantly (IMO) what is on the book itself (which is 0-583-32905-5). I've left notes of course, but as Google is more likely to show the DJ ISBN rather than the publication ISBN, again I'm tempted to create a second ISBNed entry, cross-referenced to the first, to overcome the Amazon bias. Or we let Google show 122 results for one part of the DJ to 17 for the actual book, which seems wrong to me. BLongley 21:59, 18 February 2009 (UTC)

Anyway, just another example. I still generally support cross-referenced pub entries and exceptions on a case by case basis, but I'll listen to further opinions on how to deal with these oddities. BLongley 21:59, 18 February 2009 (UTC)

To elminate "bias" you could link 0-00-224637-6 to Amazon in the notes using <a href="http://www.amazon.com/exec/obidos/ASIN/0002246376">0002246376</a>. --Marc Kupper|talk 10:45, 19 February 2009 (UTC)

That seems to increase bias to me. The prefixes suggest that Amazon is more correct in what the ISBN should have been (00022 is right for Collins and its successors, 05833 should be Dragon Books, a children's imprint of Granada). I'd prefer to increase the number of Google results that point to what is actually on the book. If that's only on our site, so be it: I've never seen Amazon change an ISBN due to user feedback. Ah well, I guess I can add a publication Wiki entry as well in the meantime. And this conversation might appear on Google soon too. BLongley 20:06, 19 February 2009 (UTC)

(Unindent)I have supported not listing the Tor years ago Can. Ed. ISBN's because the books were not sold under that ISBN as far as we know. I just did some older Ace editions, pre-ISBN, and noticed that the sequencing 1,2, ISBN, etch showed the printing sequence in this db record, or the number was missing and I filled one in. Point is the publisher used a number system that I was using to help track publication. In checking, the Tor editions, I believe you will find that the same thing is occurring, the Can. ed. ISBN is one number greater as the five number publication sequence. Therefore, the publisher expended a number in his publishing sequence. Theoretically, by not recording the Can. ISBN as a mirror, etc, we could leave future users with publication gaps for no reason. The CAN ISBN is pretty useless as a product identifier, but does give a publishing sequence for that publisher. Therefore it has more value than just an ISBN, and this may be the 'good" reason to create the mirror ISBN's in the db. Sincere apologies Bill. Thanks, Harry. --Dragoondelight 20:45, 19 February 2009 (UTC)

I recuse myself from handling Tor and Baen editions with this problem, I'll let the experts have their way. And just add more notes. But I would like to know how anyone else that is not an expert can contribute, and how the experts will find and merge the problem titles. I once stated that anything with an SF-related ISBN should be recorded here - I retract that as it seems that Amazon have "ISBN"s for multi-packs and floor-displays and boxed sets. I'll delete the first two, I'm getting uncomfortable about the last. I don't think I want to record every combination of ISBNs of books within a boxed set within an ISBN though. BLongley 22:48, 20 February 2009 (UTC)

MartyD brought RCGNSTTMGR1985 to my attention for an unrelated issue but I see in the notes that the Canadian ISBN is in *three* places and the sequence thing Harry wrote about is also evident. I believe what Harry is thinking about is if we add pub records for the Canadian ISBNs then later someone could do a query for Tor and see if ISFDB has any gaps in the ISBN numbering that may point to publications we missed. I did this long ago for DAW and still do so to figure out their upcoming publications as they post these to Amazon six or seven months out meaning I've added everything up to August-2009 to ISFDB. --Marc Kupper|talk 02:00, 21 February 2009 (UTC)

You got my point, publisher numbers were expended, and I always wonder if I missed a good book, which was not adequately distributed or published. I also noticed that though Tor calls it the Can. ed. I can find no change that points out anything Canadian except there little blurb. It seems to be, pardon my ignorance, the same American numbering. Thanks, Harry. --Dragoondelight 13:00, 21 February 2009 (UTC)

For DAW I did the ISBN sequencing in a spreadsheet. If I started a Tor project I'd do the same. The ISBN linker can do the sequencing. Note that TOR's 812 prefix used for pb is actually 0-8125-####-# but they print it as 0-812-5####-# either in error or to make it appear they are one of the big boys that get a three digit prefix. I'd do search passes using Google meaning as long as the Canadian ISBNs are documented in notes we should be ok. They use several prefixes including a number of 3 digit prefixes meaning a comprehensive search would need to be automated. --Marc Kupper|talk 08:17, 22 February 2009 (UTC)

Bound Magazine Volumes

Should bound volumes of magazines be included? I'm asking because Arkham House collected the 10 issues of The Arkham Collector and issued them in hardcover without jacket. In this instance the book is hardbound and includes only the magazines (i.e. no additional copyright or title data). The title, The Arkham Collector, Vol. 1 and the publisher are printed on the spine. Several secondary sources give the date of issue as 1971 an edition of 676 copies and a cover price as $10. I've included this info on the magazine's wiki page, but am wondering if it should be entered in the database as well.--Rtrace 03:44, 19 February 2009 (UTC)

Technically it's an omnibus but ISFDB may be challenged when it comes to making an omnibus of magazines. In any case, yes, a publication record should at least be added and if you can't add magazines to the contents then add a note. --Marc Kupper|talk 10:40, 19 February 2009 (UTC)

According to ISFDB's definition this is not an omnibus (it doesn't include a novel). It's a new animal entirely as far as the ISFDB is concerned, even though magazine publishers have been doing it since at least the 1800s. (There's even fanzines that have been bound in hardcover volumes, perhaps the most famous is typed as NONFICTION.) Is there a need to create a new publication type? I don't think so, but even if others do, it would be rather low in priority when it comes to software changes. I guess we could change our definition of omnibus, but all past discussions ended in stalemates. It appears that much of the contents of this magazine is nonfiction, so why not go with that? It seems the easiest solution to me, and it could always be changed if a new publication type is created some time in the future. MHHutchins 22:29, 19 February 2009 (UTC)

The definition of omnibus in the glossary is: "An omnibus is a publication which contains within it at least two works which could legitimately be regarded as independent works." No reference to novels.--swfritter 22:35, 19 February 2009 (UTC)

Great definition. I argued for that moons ago, but the "novel" people won. I would also argue that by most standards the independent works would have to have been previously published. You can find another ISFDB definition on this help page. It states "OMNIBUS. A publication may be classified is an omnibus if it contains multiple works that have previously been published independently, and at least one of them is a novel.". MHHutchins 23:21, 19 February 2009 (UTC)

I like that glossary definition too. I dislike "previous publication" requirements though - sometimes the first appearance of a significant work is in such a title, and although "First edition" status is rarely given by booksellers on the basis of being in an Ace Double or suchlike, we're trying to record all publications in any form aren't we? Let the booksellers try and sell their books based on "First edition" or "First hardcover edition" or "First US edition". Let's just record what is, and what research says it should have been too, and any and all other discrepancies. And maybe we can sort out Help to be a bit more consistent, but that gets a bit more complicated when there's arguments over "Novel" anyway. :-/ BLongley 22:32, 20 February 2009 (UTC)

ARC (Advanced Reader Copies) Again

In going through several books today to clear up a coverart problem, I noticed that a number (I found four in a few minutes) of my Baen ARCs have their own ISBN. These ARCS are pre-release TP bound copies of the same printed pages that go into the hardcover editions (only one page was different in the one I just checked - that being the very first page instead of the standard 'title page' without the Authors name, it had additional info about the 'to be released book'.) The TP Cover is a full glossy reproduction of the HC cover on the front. The inside of the back cover has information about the upcoming book, ISBN, price, number per case, etc, while the back of the cover is blank except for a statement 'UNCORRECTED PAGE PROOFS', 'Please compare any quotes used for review against the bound book', and 'ISBN (ARC): 1-4165-1399-X' which is different from the HC ISBN of 1-4165-0923-3. What was the reasoning that ARCs should not be included again (There is discussion at the top of this page... and some people claim it's against ISFDB policy, but I couldn't find that policy in the 'Rules of Acquisition')? Personally I think any edition of a book that a reader may have in their collection or could find in a second hand store should be 'in' if other publications of the same work are also in. Thoughts in general? Thoughts about ARCs with ISBNs? Kevin 23:14, 1 March 2009 (UTC)

Allowing ARCs into the database would make the words "First Edition" meaningless. I'm not sure why it's not spelled out in the Rules of Acquisition as I came along far after those rules were established. Perhaps it never entered the mind(s) of the person(s) who wrote those rules that an ARC would be considered a publication of the title. I have many working drafts and finished, but not yet published, stories and novels by Michael Bishop in manuscript form. No one would ever consider that they be entered into the database. It is my personal belief, but probably shared by those who wrote the rules without making an explicit statement concerning them, that ARCs are not published books. You say "any edition of a book...should be 'in'". An ARC is not an edition of a book. It's a test run at best. One that's readable, even enjoyable, but it's not a publication. MHHutchins 05:14, 2 March 2009 (UTC)

The words "First Edition" are pretty much meaningless now. I'd need a country at least. And format probably, as first edition collectors seem to prefer first hardcover editions. As to ARCs being included - well, I'd exclude Review copies sent out for free or on loan, but if they're "Reader" copies for sale (as I believe some of the ones discussed last time were) and have used up an ISBN I'd call that a publication. Not necessarily a first, but then I don't care about which was first much. But to allow ARCs in without sorting out what the first title date should be is a bit premature - should pre-official-release publications be allowed to be dated before the title? Or do we back-date to the pre-release's date and confuse the first edition collectors? BLongley 20:10, 2 March 2009 (UTC)

I dare say that any print run of 100 or more copies of say 'Dickens' or 'Melville', or even Jules Verne would, be considered an edition (First or otherwise) of some sort, to some group of collectors and any serious bibliography of collectible works that intentionally left it out would be considered incomplete. Do we actually want to exclude valid, correct information because it might upset our sensibilities? What if someone wanted to exclude all works where the Author did not get paid (as per the contract)? Or all works which paid less than SFWA/Union wages/rates? The validity of a publication should not hinge on the appropriateness of that publication with respect to the authors renumeration. If a publishing house creates and distributes a work, and that publishing house is recognized, and the work is recognized....why not recognize the publication? As to dating, that is rather simple. Date the pub as any pub. Some ARCS have explicit printing dates listed... I have one from Ace with "Special Advanced Reading Edition / April 1989" on the copyright page. The subsequent mass market pb that was offered for sale changes that one line to read "Ace edition / November 1989" so in this case I have an exact date. For another example, the HC was dated Feb. 2006, Copyright 2005. The ARC says the same thing. Unless I have evidence that or I can verify that it existed in 2005 (for instance I might have a signed & dated inscription)... I would have to enter 0000-00-00 (and it may never be known). For a last example (hypothetical) HC is January 2007, copyright 2006; ARC says the same thing. It would be reasonable (barring information to the contrary) to enter the ARC as simply 2006-00-00 (with a note as to the route used to obtain the year) since we know the ARC preceeded the HC by some amount of time which puts it into 2006, but it is unreasonable to believe the ARC preceeded it by more than 12 months. In other words... we can date them with the same common sense we use to date other ill-defined publication dates. As far as marking them VERY clearly as ARCs in the database... we could simply require a note to that effect for each pub and prepend ARC before the publication type, perhaps with or without a dash. 'ARC-pb' 'ARC-tp' and 'ARC-hc'. That way even from the Title page it's obviously an exception and can be excluded based upon the personal desires of the person using the database (Just as some users ignore pb publications, or hc, or UK or US publications). I know this is long so I'll break now and see what other opinions crop up or if I've persuaded anyone. I just wanted to (attempt to) address the concerns to date. Thanks Kevin 01:48, 3 March 2009 (UTC)

To answer Michael's question why "it's not spelled out in the Rules of Acquisition" that ARCs are excluded, the reason is that it simply never occurred to me or to Al that anyone would want to include them. To the extent that we were concerned with them at all, we wanted to make sure that editors would not mistake them for regular editions and accidentally enter incorrect page counts, cover art, etc.

Having said that, the publishing business is constantly changing and we have been trying to accommodate these changes as they occur. For example, we used to exclude vanity presses, e-books and print on demand books, but now they are included under most circumstances. Similarly, if it turns out that the nature of ARCs has changed and they are used more like regular books -- e.g. by Baen, which seems to be selling pre-publication copies online these days -- then I am sure we can revisit the question.

Kevin's argument appears to boil down to "If a publishing house creates and distributes a work, and that publishing house is recognized, and the work is recognized....why not recognize the publication?" The tricky question in this case is how we define "distributes", which may cover a multitude of sins, e.g.:

1. in-house distribution
2. distribution to would-be reviewers
3. on-line distribution a la Baen
4. other forms of distribution (?)

Historically, #1 and #2 were not considered "distribution" proper, so ARCs were not counted as "real" editions/printings. #3 and #4 (whatever it may be) generally didn't exist, so the question never came up until quite recently. I'd like to see/examine more examples of #3 and #4 before changing our eligibility criteria. Ahasuerus 02:21, 3 March 2009 (UTC)

Regarding Bill's response to my comment: When I speak of "First Edition" I'm not referring to what a publisher prints on its copyright page. I (and any knowledgeable bibliographer/bibliophile) use the words to refer to the first publication of a title. It doesn't matter the country or the publisher. There is always a first. An ARC is a printing, not a publication. These advance copies are intended for reviewers, and most are not for sale (unless it's published by Baen, as I've come to learn.) And the number of copies doesn't even come into the equation. There's records here in the db for printings of 26 copies. Would any cinephile actually consider "for your consideration" copies sent to Academy members a release of a film? Like Ahasuerus, I'd need a lot more evidence to ever consider ARCs into the db. I may be in the Dark Ages, but I'm open-minded. Bring on the enlightenment. MHHutchins 04:37, 3 March 2009 (UTC)

Re: "number of copies" - I agree that is irrelevant. For a definition of "publication" I think I want it to be open to the public to buy. Private printings sent as presents, or for review purposes only, OUT. But I did add "Beedle the Bard" here when there was only one book available to the public. The fact that Amazon bought the one public copy for almost 2 million and then persuaded the author to do a more general release is irrelevant - that one copy was for sale to anyone, and was, IMO, a publication. Resale of review copies doesn't count, IMO: "publication" is when the public can buy it directly. Which makes some (but by no means all) ARCs in. BLongley 23:16, 3 March 2009 (UTC)

Taking the Devil's Advocates position... if we are to limit ourselves to items that are available for purchase directly from the publisher, then are all 'out of print' items now out? If we are not to take that position, then isn't anything for sale in second hand bookstore available to the public? Kevin 23:33, 3 March 2009 (UTC)

I assume that when Bill wrote "that one copy was for sale to anyone", he implied that any publication which was offered for sale when it was new should be included. However, we also list a number of publications which were never "sold" per se. For example, fanzines are often made available in exchange for "the usual", yet we index them, so that would be a counterexample unless we change the term "for sale" to "for sale or exchange".

In addition, a number of early "private editions" -- which were not sold to the general public but were sometimes done by subscription or otherwise effectively sold -- ended up printing dozens, if not hundreds of copies and are often included in standard bibliographies. I have a beautifully printed early 20th century "private edition" of Rabelais's Gargantua and Pantagruel and even though there were under 100 copies printed (75?), I'd like to include it when I get to that part of my collection. The "available to the public" standard becomes even shakier when we consider the fact that we already include books published by book clubs, which are sold to members only, and "library bindings", which are only offered for sale to libraries. Also, what about books which were distributed to the general public for free, e.g. as promotional material, and there were so many copies printed that they are indexed by Locus and other bibliographies?

Given these uncertainties, I think that we would need to qualify "available to the public" if we decided to use it as a criterion. Ahasuerus 00:44, 4 March 2009 (UTC)

I know you're not talking about "what a publisher prints on its copyright page", neither am I. I DO think the phrase has become almost meaningless now. Not just because of ebay sellers that state "HARRY POTTER AND THE PHILOSOPHER'S STONE: FIRST EDITION HARDBACK: NUMBER LINE 50 49 48 47 46"! In these days of worldwide simultaneous publication dates, it's obvious that a UK first edition beats a US first edition, but that doesn't seem to matter. If it did, I'd set up a publishing company in the Christmas Islands or suchlike - UTC/GMT + 14 hours. I've got "first editions" of most Terry Pratchett books, and most J. K. Rowling ones. With first print runs in the million+ range, nobody seems to care about true "first editions". When it gets down to the ones people care about, ah, then yes they do. A true first edition of "The Carpet People" or "Harry Potter and the Philosopher's Stone" commands huge prices. An Abebooks seller describes "A beautifully hand signed copy of the DELUXE LIMITED EDITION United States true 1st Edition 1st Printing" for "Harry Potter & The Sorcerer's Stone". First edition of a variant title in a special edition? 90% of all "first edition" claims are bogus or useless, and I'd say valueless. For current publishing practices - make that 99%. ISFDB isn't going to help much for First Edition collectors. Nor should it, IMO. If we can define what a true "publication" is we can put a date on it though. BLongley 23:05, 3 March 2009 (UTC)

I realized that you all may be imagining the old style ARCs with construction paper covers and large rough staples (In other words, things that barely resemble a modern book). Just to be sure we were debating the same thing, I took some photos tonight. Please see User:Kpulliam/ARC-Examples for some examples of the type of ARCs I'm talking about. I included Spine, Cover, and Copyight page pictures for examples from Baen, Ace, and Tor. Also Examples of #4 could include Baen's practice of selling an Arc or two on ebay to gage the market in order to determine the print run to order, or just to cackle in glee at the number of bidders. #4 could also include the practice of ARCs being given away to fans who perform services for the publisher (Throwing a Con Party) and also copies auctioned off at Con Charity auctions donated by publishers to promote the work / author. Thanks Kevin 05:40, 3 March 2009 (UTC)

Thanks, most informative! I have seen similar ARCs in used bookstores and at a "community library", one of whose contributors apparently had access to ARCs. However, I wasn't familiar with the varieties of differences between ARCs and traditional first editions.

Of all the uses of ARCs mentioned and alluded to in this discussion, two stand out in my mind. The first one is the use of the designation "first printing" for ARCs and "second printing" for the print runs sent to the stores, which was mentioned a few months ago. The second one is the use of separate ISBNs for ARCs, which I was unaware of. All other uses mentioned so far -- E-bay, giveaways at conventions, charity auctions -- do not seem to be quantitatively different from the traditional use of ARCs during the pre-publication process, but the first two seem quite different. Ahasuerus 17:45, 3 March 2009 (UTC)

Formatting ISSN numbers

Perhaps we should specifically mention in Help how ISSN numbers should be listed. Right now we have examples like ISSN: 1234-5678, #ISSN: 1234-5678, ISSN: #1234-5678, #1234-5678, and 1234-5678. The last is obviously wrong and the penultimate example gives no indication that the number is an ISSN which are significantly different from catalog numbers. I prefer the first. The "#" seems extraneous; it's primary purpose is to indicate that the number is not an ISBN and putting the ISSN: before the number accomplishes that purpose.--swfritter 23:04, 2 March 2009 (UTC)

I'd always thought ISSNs would trigger a bad Checksum warning, but I see from examples like this that they don't always. Any standard that doesn't trigger an unnecessary warning is fine by me. (I still lean toward full separation of ISBNs and Catalog Numbers so we can validate some automatically, and ISSN could be a third, but that seems as likely as separating Authors and Artists. Not going to happen anytime soon.) Adjust help to what actually works for now, I'd say. BLongley 23:53, 2 March 2009 (UTC)

And there is another variant: ISSN 1234-5678. I don't remember any checksum warnings when entering ISSN numbers.--swfritter 21:06, 3 March 2009 (UTC)

ISSN NNNN-NNNN is fine with me. As David mentions elsewhere it is the most likely format that a user will use for a search.--swfritter 17:24, 7 June 2009 (UTC)

Somebody, please, convince me why the ISSN should be entered at all. I can't think of one purpose for a general user of the database. It looks to me like another purpose for a field that's already being using for two other purposes: ISBN and catalog number. An ISSN is neither. MHHutchins 18:48, 7 June 2009 (UTC)

The ISSN is as I understand it, they key number for periodicals just as the ISBN is for books. So IMO it should be reciorded somewhere, whether in the notes or the field. A magazine wil have neither a cat # nor an ISBN anyway. I would favor separate fields in the long run, but that is not a top priority. -DES ^Talk 19:23, 7 June 2009 (UTC)

ISSN's do not identify a specific issue. One of the reasons I started putting them in downloadable ezines in particular was that some people, at least at one time, voiced concerns over their stability and the ISSN was an indicator, although not a guarantee, of a certain level of professionalism. Also, Bill gives an example of a user doing a search by ISSN.--swfritter 19:48, 7 June 2009 (UTC)

[Response to DES after edit conflict]I know what an ISSN is. The question is why would the average database user need to know it? An ISBN identifies a unique publication. An ISSN is assigned to a periodical, not a publication. Every issue would have the same number, so, again, what is the purpose of recording it in any record? I can think of information that would be far more useful in a periodical record than the ISSN. MHHutchins 19:51, 7 June 2009 (UTC)

[To Swfritter] I think you meant ISSN in your first sentence. MHHutchins 19:51, 7 June 2009 (UTC)

Yes I did. Other than what I stated above it is certainly no big deal to me. Notes are fine or even wiki level although that data is not as portable. It is definitely not a high priority for me to place data in fields just because they are there.--swfritter 20:09, 7 June 2009 (UTC)

There are 450 pubs with "ISSN" in the Catalog ID field.--swfritter 20:16, 7 June 2009 (UTC)

One benefit of the leading "#" is that various places in the code use this to bypass trying to use the value for ISBN purposes. Similar logic does not exist for any other form of non-ISBN. It probably should, but it doesn't. --MartyD 18:59, 7 June 2009 (UTC)

But I gather not at the time it decides to generate a warning? Is there any problem other than extra processing time?--swfritter 19:48, 7 June 2009 (UTC)

The # bypasses the code that would generate the warning, but that specific spot has additional checks (removes hyphens and spaces, then checks the result for 9 or 12 digits followed by one digit or X) and only warns if it could be an ISBN but the final digit-or-X does not match the check digit computation. Just (minor) extra processing cost there. But pubs having numbers/IDs without the leading '#' get Other Sites navigation links, for example. Compare #ISSN 0024-984X with ISSN # 0047-4959. I think there may be some more places that behave similarly. The leading '#' exclusion seems to be implemented universally, while testing/checking for other non-ISBN cases is not. --MartyD 20:36, 7 June 2009 (UTC)

Is there a site where lookup by ISSN is useful? So far, I've encountered ONE example of somebody seeking to search us by ISSN. And we'd have to fix the web API for his purposes. (Actually the web API looks in need of more attention than some of the rest of our software.) But I can see that it might be useful to download an entire magazine by ISSN. Not necessarily good for us though. BLongley 21:33, 7 June 2009 (UTC)

Keep in mind that ISSNs are used by some library catalogs and services, e.g. by OCLC's xISSN Web service, to group magazines and link them to their predecessors and successors. I am not sure how we could leverage this capability, but it's something to keep in the back of our hive mind. Ahasuerus 22:17, 7 June 2009 (UTC)

I think they should be recorded soemwhere in a consistant format. Since they do not identify partiocualr issues, perhaps listing on a wikli page for the magazien would be enough. But they should be recorded soemwhere, and it surely does no harm to put them in the ISBN field as things stand. -DES ^Talk 22:26, 7 June 2009 (UTC)

ISSNs have caused problems, see Development#Test_Results. It would be good to simplify use of the field to "ISBN, or #Catalog number" for developers. But don't make a judgement based on making it easy on developers. If it is useful to have ISSNs on individual publications, let someone speak up for them. I've no use for them but agree they should be recorded - magazines more than any other area depend on data in the Wiki rather than the database itself, but as ISSNs should rarely change (a change of title, publication format, or complete translation into another language seem to be the triggers) then recording at the magazine level seems more appropriate than publication level. Just make a decision on the desirability, we'll sort out the coding later. BLongley 22:47, 7 June 2009 (UTC)

Wiki pages do not have the same persistence as the database. I would put them in both the wiki and pub notes.--swfritter 00:41, 8 June 2009 (UTC)

Because there are technical issues involved with my preferred methodology of entering them in the Cat ID field, I no longer have a preference and do not care whether they are recorded there or not. No matter what decisions are made by those who have stronger opinions there are a significant number of pubs that will have to be updated.--swfritter 16:23, 10 June 2009 (UTC)

Titles for Maps as Content

This discussion originally started on the help desk as a how-should-I-do-this, and an initial "there is no standard" response evoked a discussion of what the standard should be. I have copied here that entry, except for the suggestion to move it to this page.... --MartyD 10:46, 6 March 2009 (UTC)

Start comments copied from ISFDB:Help_desk#Recording_Maps_as_Content

I have a copy of The Illearth War that contains a 2-page map that is separately credited and copyrighted. In addition, the Table of Contents has an entry for it:

What Has Gone Before.... xi

Map..................... xiv

I entered it here with a title "Map (The Illearth War)" based on that entry, the copyright page's "Map Copyright (c) 1977 by Lynn K. Plagge", and also the map's own "Map by Lynn K. Plagge" signature -- these latter two with the capital "M". In looking at the page for the artist, however, I find several other map credits using "title (map)", including "The Illearth War (map)" used here for a different printing of the same book.

Is there a preferred way to record maps, or is the case where the map is semi-formally labeled "Map" different from an unlabeled map? If the latter, should I make the new "Map (The Illearth War)" title a variant of the existing "The Illearth War (map)" title? Thanks. --MartyD 11:39, 4 March 2009 (UTC)

I don't recall any discussions of map standards beyond "Sure, let's include them as Interior Art", so, unless I missed something (and there have been a lot of discussions over the last 3 years), I don't think we have a standard at this time. Ahasuerus 03:33, 5 March 2009 (UTC)

I have put the "Make Variant" submission on hold since I suspect that we simply want to merge all maps that have been identified as identical. Any other ideas/suggestions? Ahasuerus 01:52, 6 March 2009 (UTC)

I kind of thought we would want to standardize as Map (TITLE) or Maps (TITLE) (depending on whether there is one or many) since that matches the form of Introduction (TITLE), Afterword (TITLE) etc....but thats just me wanting to make the listings look symmetrical. Kevin 02:24, 6 March 2009 (UTC)

I agree with Kevin here, and it follows along with the general rule (stated or unstated) about adding information that is not present in the title itself in parentheses. (Such as the name of the book or magazine to disambiguate generic titles from others.) But what if the map has a title? I came upon a few awhile back and I placed "map" in parentheses, here and here. Regularization (or regularisation, as our British pals, might spell it) should be simple. MHHutchins 03:47, 6 March 2009 (UTC)

End comments copied from ISFDB:Help_desk#Recording_Maps_as_Content

I think it's a mistake to try to include sub-type information parenthetically in the title unless it is needed to distinguish a subset of content within the same work. Unless "length" were replaced with a type-specific sub-type list, I think extra descriptive information should go in the notes. The fact that a map is INTERIORART distinguishes whatever title it is given from written works. So if we had My Novel with two different forms of untitled INTERIORART (say, some maps and some drawings) that were credited to two different artists, something like "My Novel (maps)" and "My Novel (drawings)", both of type INTERIORART.

I originally wanted to go with "Map" because it appears as such in the TOC. In this particular case, the map does have a title ("The Land") seen only on the map. If I used "The Land" (or "The Land (The Illearth War)"), the content list wouldn't match the TOC. Do we care? That said, the other two books in the trilogy use the same map, and now I'm thinking it would be better to name the map something all three books could share. A further complication in this example is that the second trilogy also presents a map the same way, except the map is somewhat different, and the credit switches from a separate copyright of Map and "Map by Lynn K. Plagge" on the map to no copyright and "From a map by Lynn K. Plagge" on the map. The TOC entry is still "Map", and the title on the map is still "The Land". --MartyD 11:31, 6 March 2009 (UTC)

And one little addendum: I just discovered that the British edition of The One Tree has "Maps" in the TOC and three separately titled maps (none of which is "The Land", btw). It seems a bit much to want to track a separate title for each map.... --MartyD 11:48, 6 March 2009 (UTC)

(unindent)Why not do it as is done for magazines? This. [9] . This result on title search. [10] . I thought this was the standard. I verified a novel recently with "maps by Shelly Shapiro" and there were four maps, not individually shown on any ToC in the body of the novel. There was also one map that I am fairly sure was used previously in preceding novels that was unsigned. In all fairness, without a signature I would not credit the first map or credit maps unless I stated a start page. Instead I noted that there were signed maps in the novel text and and unsigned one at the front. Thus if someone wanted to parse it, then I assumed they would Title (map 1,2,3,4.5). All in all I think it is too much to record. In the battle tech novels I would have created as many as 14 illustrations (1-14) and then 12 mechanichal drawings (1-12). To add to the problem the drawings are all "named", some signed, some not. Do we want that? Do we want to give credit for 'maps' to 4 signed, but one not? Obviously, no one wants Illustrations (1)(Belgariad 3). Add this in some series, especially, the same map is used in various novels, and is named. My solution credit all as one or not at all (as book title(map). A search for 'map' results in this [11] . Which is not useable past the first 100 or you have do specialized searches, which may not turn it either as you have no way of knowing if it was inputed into the system so that you find them all. I propose that we limit the use of 'titled maps/illustrations' to one novel title entry per type. Here is an example of an anthology in a series with both individual story entry with the story title used to title the illustration. [12]. This clutter may have achieved the 'magazine' level of display, but imagine the havoc of 'named chapters tied to illustrations' for a novel. I did this one this way, because each illustration was attributed, much as I did a Burroughs with two separate artists, with distinctive signatures. This can is full of worms. I believe a 'lesser' is better rule, if you have to have a rule, but in all cases book or story title then (1 etc) entries. Thanks, Harry. --Dragoondelight 13:44, 6 March 2009 (UTC)

I didn't mean this to cover ALL maps (or any other kind of art) -- I agree with your general point. Here, however, I have a map that is listed in the TOC, is credited to someone else, and is in the group of pages numbered separately from the main story. To me, it's a little bit like an "About the Author" -- I wouldn't bother listing it under normal circumstances (or perhaps address it in the notes), except that the pub explicitly presents it as distinct content. So if you're willing to concede the judgment of whether or not the map should have its own content entry, the question really is "When a map is being included as content, how should it be titled?". And as for searches, the "this produces too many results to be useful" argument, while certainly valid, equally applies to "Introduction" and to other often-used nondescript titles. I think the title is more important not for being used in a title search, but for viewing in both an author (artist) search and a content list within a pub record. --MartyD 15:01, 6 March 2009 (UTC)

It looks like no consensus was reached. Marty and I briefly discussed the issue earlier today and he agreed that the titles could be merged. Disposition: submission rejected, merge submitted and approved, can of worms lovingly put back in the closet until the next time... Ahasuerus 23:42, 31 May 2009 (UTC)

I routinely enter all maps as interior art, adn whatever title I give them I make sure the word "map" is included soemwhere. If there isa titel like "The Land" I generally enter them as "Map of The Land" or "The Land (map)" dependign on the exaxt format and how i think that day. If there is a copyright notice or sig, i credit the map to a person. -DES ^Talk 04:09, 1 June 2009 (UTC)

Webzine's Out! Individual stories from webzines in?

See this discussion for Idle Roomer. The information for the first publication appearance could appear in the notes. We would not have to worry about verifying the contents of a pub collectively although the title record would be an orphan of sorts and would require a couple of steps to create.--swfritter 23:22, 8 March 2009 (UTC)

I'm not saying it's the best solution, but the easiest might be to allow webzine stubs (as in non-genre magazines) for works that go on to be published elsewhere (But the entry as such should wait until it's actually published elsewhere). Kevin 23:35, 8 March 2009 (UTC)

When is an Audio Book not an Audio Book?

I've entered a few missing "Big Finish" publications today (well, quite a lot really) but some of them look a bit borderline. They say they're [Audiobook] on Amazon, and some of the authors do indeed write dead-tree publications too, and there are versions of dead-tree books in the series, e.g. Just War (although that was so heavily edited for the audio release that the Doctor doesn't actually appear in it any more). Quality-wise, they seem to have a bigger cast than your average audio-book (from the few I've heard), but not as many as your usual BBC radio play, and the sound effects aren't very good. Does anyone have any views on where we draw the line? Is the intent enough, does an ISBN make them valid, does later or earlier broadcasting on radio count, do we demand a dead-tree version first, should we exclude audio books "performed" rather than "read"? Or are any of the above fine so long as YOU don't have to handle them? ;-) BLongley 22:11, 15 March 2009 (UTC)

I happen to own the 'Audio Soundtrack' to both Star Wars: A New Hope on LP (The one with RD-D2 and 3PO printed on one side..all the way across, not just a label), and Monty Python's Holy Grail on tape (The Executive version if you must ask... otherwise known as 'The album of the soundtrack of the trailer of the film...'). Should they be in? If My LP with light saber Swooshes, and my cassette tape with real coconuts clapping together for horses are in, then everything is in. Monty Python is more like high art... not speculative fiction. Kevin 04:52, 16 March 2009 (UTC)

If it claims an association with a printed work... it should be in. If it claims an association with a 'universe' that includes printed works, that, alone, is insufficient. If we include 'radio-plays' and the like, we will spend years indexing the 30's through the 40's (and early fifties) equivalent of TV. Kevin 04:52, 16 March 2009 (UTC)

Well, in the Doctor Who Universe we have one "Radio Special" but that's backed up by a dead-tree version later. We have dead-tree novelisations of "A New Hope", and some unexplained ones. (I presume the illustrated, script, and illustrated script are dead-tree - not sure of some of the others.) I don't think I'd object to your LP of that, maybe to the Monty Python one (I own the same on vinyl and have never felt the need to add it here). Give it an ISBN though and I might reconsider... BLongley 21:49, 23 March 2009 (UTC)

Should a Hugo nominated semiprozine with an ISSN be in the database?

This year's nominations include Clarkesworld Magazine in that category. Only problem: it's a webzine. One of the arguments against webzines is that they don't have an ISSN number. Clarkesworld Magazine has one. It's time we moved into the 21st century. I intend to start entering issues of Clarkesworld Magazine in the next couple of days. Perhaps the ISFDB can have a small role in promoting contemporary s-f short fiction.--swfritter 19:18, 22 March 2009 (UTC)

I feel comfortable that a Hugo nomination is 'across the line' and 'In'. The ISSN also gives this publication an extra leg up and over that line. I say, go for it. Now that the barrier has been breached.... are there other webzines of suitable quality / documentation that we should consider them, in order to consider developing some guidelines beyond Hugo Nomination? - Thanks for keeping us honest. (Any rule that excludes a shortlist Hugo nomination for printed/read works is IMHO a rule that needs changing). Kevin 19:52, 22 March 2009 (UTC)

If you have an afternoon or two you will find some very long, convoluted, and contentious discussions packed away in the wiki. I consider Clarkesworld Magazine a perfect test case for setting standards for the inclusion of other webzines. For right now we could set a high standard, ISSN and Hugo nomination, and perhaps loosen up the standards even more later on.--swfritter 20:27, 22 March 2009 (UTC)

First entry. Interesting display issue with the audio version of one of the stories.--swfritter 20:42, 23 March 2009 (UTC)

You'll get the same issue with any publication containing multiple versions of the same story. Which I guess is a reason to not make "original", "expanded", "revised", etc versions variants of each other, just in case some severe "literary criticism" occurs and both are published in the same work for comparison. Is there a real need to make the audio version a variant of the text version? We don't usually do that for full audio-books, even when there's unabridged and abridged versions. BLongley 21:02, 23 March 2009 (UTC)

When novels are merged the publishing format is designated within the merged pubs linked to the title. In this case there is only title data. There is no way to designate the publishing format in title data and titles within a pub cannot be merged. I could just make a notation in the title data that there is an audio version but perhaps the author biblio display is more important. The variant title link makes it clear that there is an audio version. Like pseudonyms, we have had to make the variant title method a catchall for relationships for which the nomenclature is not totally accurate.--swfritter 22:19, 23 March 2009 (UTC)

How about indexing the audio versions as Chapbooks, separately published with a title "STORYNAME (Clarksworld March 2009)"? The audio versions are available through iTunes without visiting the primary publication source... so they are kind of separately published... Thoughts? Kevin 00:43, 24 March 2009 (UTC)

I wouldn't recommend Chapbooks as a solution to anything at present, they're severely flawed. However, if there is a separate publication in audio format only that might add support to making them a variant, or choosing a workaround Container-Type for the iTunes version that would allow the audio "binding" to show up, which would also make it clear that an audio version exists. There's several possible workarounds, but I'd never considered iTunes as a publisher before. BLongley 20:49, 24 March 2009 (UTC)

It's a No Go on audio versions until we have shortfiction format support. There are certainly some bizarre things that happen when two titles in the same pub are merged. Even though both show up on the edit screen, only one shows up on the display screen. And when one title is removed both are removed.--swfritter 00:05, 25 March 2009 (UTC)

By the way, I'm not against inclusion of Clarkesworld Magazine if the issues are stable. I was a bit more worried about some titles where contents were added between electronic and print versions. BLongley 21:10, 23 March 2009 (UTC)

Webzine pubs, both upon entry and modification, need to be thoroughly vetted by a senior moderator. If we had such a thing.--swfritter 22:19, 23 March 2009 (UTC)

I think you just got nominated as 'Senior' Kevin 00:45, 24 March 2009 (UTC)

I would like to avoid any kind of official hierarchy but our decision making process is somewhat flawed. There are a number of moderators who have been around for two years and I think we should trust their decisions as long as they make their cases and allow for input from others. If such decisions are badly flawed people will ignore them anyway. One thing we may need is a little more semi-official accountability. I guess I am the de facto ezine/webzine advocate and there are others who are knowledgeable in specific areas. Perhaps we should have a mechanism for identifying those editors and their areas of expertise.--swfritter 19:34, 24 March 2009 (UTC)

There's no possible acceptable overall hierarchy, I think. The "Bureaucrats" are, IMO, treated with a little more respect in some areas, mostly for their knowledge of why some practices and conventions started, but we've changed a lot of practices in two years. Likewise, a good editor often knows more about a certain area of expertise than any current Moderator. I think we're moving towards a flatter structure of "people that need guidance still" and "people that can guide (in some areas)". I once suggested that ISFDB:Moderator_noticeboard#Moderator_availability have an extra column for areas of expertise, as I for one have no intention of trying to claim to know everything. I would like to think that my fellow moderators know I can answer questions on British pricing and older British paperbacks better than most active editors, but there will be younger British editors that know more about recent British paperbacks than me for instance, and most American editors will know more about US prices and maybe magazines and I'll leave those alone. BLongley 20:49, 24 March 2009 (UTC)

How we claim or be acclaimed for specialities I don't know, I just suspect it would be good. I know it would be nice to know where to turn when somebody starts adding possibly valuable information that I cannot moderate, e.g. when someone comes here with knowledge of German editions of Perry Rhodan, or Polish Stanislaw Lem editions, or French editions of Van Vogt - it's all a bit vague about who knows what. And if it turns out no Mod is capable, then we probably need some new Mods - I see our current list is getting a bit stale. BLongley 20:49, 24 March 2009 (UTC)

(unindent) By the way, Swfritter, what happened to Rick Boatright? I'd hoped Baen would be mostly automated by now. :-/ BLongley 20:49, 24 March 2009 (UTC)

It takes me about 20 minutes to copy and paste the Baen's Universe pub data on a bi-monthly basis.--swfritter 22:37, 24 March 2009 (UTC)

Final call for Clarkesworld

Is there anybody having a problem with entering Clarkesworld in to the database? I will pull the data in a minute if there is. Otherwise, I will update Help to indicate that a webzine must be nominated for a Hugo award and an editor must commit to keeping the contents current in order to be considered for inclusion. I would like to get permission to link to their cover art and don't want to do so until I am sure the data is going to stay.--swfritter 21:21, 27 March 2009 (UTC)

I recommend that the Rules of Acquisition be updated as well. Please also but the information somewhere that this is experimental, and other 'noteworthy' webzines could be considered if they pass other undecided criteria in the future. It's like porn... I know it when I see it, but I'm not comfortable defining it. Kevin 01:11, 28 March 2009 (UTC)

I have added this sentence to the Rules of Acquisition concerning webzines: "Note: The Clarkesworld webzine has been added on a provisional basis primarily because it has received a Hugo nomination. Other webzines may be added based upon criteria yet to be determined, a primary one being the willingness of editors to update data in a timely manner." I might note that I am somewhat pessimistic about future additions. We are doing a spectacularly bad job of documenting current pubs. Even the Hugo nominated semiprozines are not well represented. Interzone is being done in a somewhat sporadic manner although this may be partially a result of the lag time between publication and actual receipt of the issues through snail mail. There is good work being done on Locus but we are missing most of the last six years. The New York Review of Science Fiction: we are missing most of the last four years. Weird Tales: totally dismal. As far as potential webzines: we already chased off at least one potential editor who wanted to add issues of Orson Scott Card's Intergalactic Medicine Show. Perhaps we should become the ILSFDB - The Internet Legacy Science Fiction Data Base.--swfritter 20:51, 30 March 2009 (UTC)

We're never going to be 100% up-to-date. Yes, current editors seem to be more interested in preserving the past than keeping up with the latest publications - but why worry when the current publications are advertising themselves? I know "Selected Upcoming Books" on our homepage is an apparent attraction. I don't particularly like it as it's all conjecture, subject to change. Even when its really been published, it's rarely verified soon. I can cope with us being a bit behind - if it took us 60/70 years to enter "Air Wonder Stories" why worry about 4 or 5 years delay for other titles? Being "current" is just another nice-to-have feature. BLongley 00:01, 31 March 2009 (UTC)

Foreign Language Editions of Translated Works

The Rules of Acquisition seem to suggest that foreign language editions of works of speculative fiction originally published in foreign languages that have been translated into English would be out (with the possible exception of the original publication?). However it isn't explicitly stated. I have a few French language editions of Verne (including one published in the US, with a Preface and notes in English), that I wanted to know whether I should enter them. Thanks. --Rtrace 22:50, 27 March 2009 (UTC)

You're reading Rule 4 "In - English language translations of works of speculative fiction originally published in foreign languages. In these cases, we will also provide information about the original foreign language work" as only including the original foreign work? I guess it depends on what "work" means in this context, e.g. just the first publication, or the first publication and all its printings, or the original title and all its publications and printings. I'm not too bothered if they're added so long as they go under the original foreign title where I'll never look at them. ;-) BLongley 12:40, 28 March 2009 (UTC)

SF Site - another stealth mode webzine

I just noted that there are stealth mode entries for SF Site which do not have a magazine wiki. Any more surprises that are going to embarrass us?--swfritter 22:53, 30 March 2009 (UTC)

Why be embarrassed? We don't support Manga or Comics or RPG Sourcebooks, but they're here. (I do try and Zap those occasionally, but if in doubt I leave them alone, and when I'm tired the "delete pub, approve, delete title" stages are sometimes too much work.) There's several Fanzines or Magazines that don't have Wiki pages yet, I'm sure. I've found obscure zine titles here and created Wiki pages for them before, sometimes I've just fixed one pub to make it look like a zine and someone else can take it further or delete it. There is a lot of crap in the database when you go off the normal routes: it's generally getting better now that Dissembler has restricted his activities, but I don't feel embarrassed by the presence of pubs that aren't officially welcome here. If somebody tries to claim them as a precedent, just reject the new submission and zap the precedent. Or use it - that title looks well-linked, but if it showed up reviews of titles we don't have I'd find those first. And probably not delete the webzine either - No, we won't support it, or organise it, but its presence harms nobody. BLongley 23:44, 30 March 2009 (UTC)

Foreign language translations of English works

I'm holding a submission adding a collection of Lord Dunsany stories translated into French. Looking at the Rules of Acquisition, I was shocked to see that such a collection would be allowed. If the submission were accepted, variants would have to be created for each story, which would be displayed on the author's summary page. Otherwise they would be listed as original pieces along with other shortfiction. And the collection would be listed separately under that category. Yes, I've seen some foreign language translations of novels in the database (which I don't like, but am willing to accept in some circumstances). These were entered under the English language title records and variants weren't created. Was it the intention of the person who wrote the Rules that all foreign language translations be allowed? Or just novels? I don't think that person (there's no way of knowing who) was aware of the results of a summary page displaying hundreds of variant titles for writers whose works have been translated into many different languages. And I'm not talking about works that originally appeared in another language. If the software were able to display search results based on language, then I wouldn't be so concerned. But we all know the software's limitations. I'd really like to hear some counter arguments. Maybe I'm just not seeing the situation clearly. Thanks. MHHutchins

I plead guilty to writing that section of the Rules of Acquisition. The intent was to include foreign language translations of English language books, i.e. novels, collections, anthologies and such, as Publications under the main Title record. It wasn't my intent to include short fiction, which, as you point out, would require lots of variant titles and would be unworkable. We have repeatedly rejected attempts to set up vts for short fiction translations, so I believe that we just need to clarify the Rules of Acquisition. Ahasuerus 05:26, 1 April 2009 (UTC)

No objections so far, so I plan to change the Rules of Acquisition based on the discussion above. Ahasuerus 22:54, 3 April 2009 (UTC)

Done. Ahasuerus

Add or not?

Came across four titles by Stanley Bennett Hough, written under the pseudonym of "Bennett Stanley" (who is not in the DB, yet). They are in Currey, but are likely non-genre, all from the 50s. Hough wrote mostly thrillers under his own name, sci-fi under "Rex Gordon". How many steps away from Spec-fict do we go with pseudonymous works? There are another seven titles by Hough that are thrillers. Add them? ~Bill, --Bluesman 22:48, 3 April 2009 (UTC)

I wouldn't. I'm not even sure we need "Stanley Bennett Hough" except as the legal name for Rex Gordon. BLongley 13:15, 4 April 2009 (UTC)

I concur, mostly. If he had written everything under "Gordon" they would be included and simply be in a non-genre section. It's hard to draw the line sometimes. McAuley writes 'scientific' thrillers that we include. Even Tom Clancy is in here. The Hough titles seem to fall in that area, though written in the 50s, so the proponents were always the US & 'villain-of-the-month' instead of Earth & 'alien-of-the-month'. I would prefer not to add them without more reason than 'we do it for X'. Sidebar: since Hough is the canonical and "Gordon" the pseudonym, why didn't the two Hough thrillers I did add (before the "WHOA" thought and this posting) show up on that page? Was all prepared to open a non-genre section... ~Bill, --Bluesman 16:22, 4 April 2009 (UTC)

No, in our terms Rex Gordon is the canonical author and he "Used These Alternate Names: Stanley Bennett Hough". BLongley 19:03, 4 April 2009 (UTC)

Many Abebooks entries have plot descriptions. It looks like many of the "Stanley Bennett Hough" are borderline as they concern themselves with third world world war scenarios while some under that name appear to be straight mysteries.--swfritter 16:17, 4 April 2009 (UTC)

If the author had never done any s-f at all I would tend toward exclude but would tend toward include if the titles have any kind of futuristic bent. Is the author significant enough as an s-f author that his non-genre works should be included? That is a pure judgment call and I am willing to leave that judgment to someone who is more familiar with his work.--swfritter 18:50, 4 April 2009 (UTC)

There's really only four or five definite SF works: the Historical Dictionary of Science Fiction Literature also mentions "Extinction Bomber" and "Beyond the Eleventh Hour" as "alarmist thrillers" and "The Yellow Fraction" as a "political fantasy". "Utopia 239" is borderline - nuclear war again, but a bit more imaginative. BLongley 19:12, 4 April 2009 (UTC)

Quotes

I have two submissions, which aim to change double quotes to single quotes in the title of an Aldiss story, on hold. Quotes are a tricky area in our world since double quotes wreak havoc with Author names, but they work OK in Titles and Pubs. I think our unstated policy has been to enter the exact kind of quote character(s) which appears in the printed title, which in this case means that we would want to create a variant title. Before I do that, though, I would like to double check that we are all on the same page re: creating variant titles where the only difference is the type of the quote character used. Somehow, it seems a bit excessive, but we have been known to do stranger things :) Ahasuerus 02:56, 7 April 2009 (UTC)

I don't believe that variants should be created based on a title having single or double quotes. This choice is not authorial, or even editorial. Most of the time it's decided by the book's designer or the magazine's art director. I also believe we've taken the original purpose of variant creation way beyond the point where it was necessary. Something like this just pushes it over the edge. MHHutchins 04:18, 7 April 2009 (UTC)

I've never been a fan of variants for variations in punctuation alone. That's probably why I left this title ("O Moon of My Delight!") alone - it's actually single-quoted in my two verified pubs of the The Airs of Earth but as Mike Christie and Scott Latham aren't answering questions I've never established whether the no-quote and double-quote versions actually exist. The no-quote doesn't technically exist in my verified Starswarm, there it's Sector Azure on title page - the sub-title of o moon of my delight (all lower-case, no quotes, no exclamation mark) only appears in the contents list and is single-quoted and InitCapped with no exclamation mark on copyright page. A Horrible title to work with indeed. But I think we're not all on the same page. In my opinion the missing O and My is worth a variant, a missing exclamation mark would be if such exists, presence or absence of quotes around the title would be, variations in the type of quote seems too much. If people want to take it even further and try and distinguish angled quotes as well I'd probably give up on editing and moderating such. I could just about live with straight single quotes/straight double quotes as a need for a variant but would definitely make such differences a very low priority. There's a bit of fixing needed here, but maybe not as much as All You Zombies where em-dash versus en-dash versus ellipsis, and single/double quotes differences and even spacing around punctuation seem to have been created and lost several times since I started here. BLongley 18:39, 7 April 2009 (UTC)

Mhhutchins - ditto his remarks. Also, agree with Bill that the exclamation point, if editor or author intended, is as far as we go. I take into account the TOC, page footer/header data, other title instances, and internal evidence in a story - and usually try to document my logic for using a particular title in the pub notes.--swfritter 19:49, 7 April 2009 (UTC)

It looks like the "least bad solution" is to decide that the addition/deletion of punctuation merits a separate variant title, but differences in the type of quotes used do not. If there are no other concerns, then we can change Help in a couple of days. (For what it's worth, I have just finished verifying the SFBC [UK] version of The Airs of Earth and they used single quotes as well, but it won't matter if we decide to ignore this distinction.) Ahasuerus 00:46, 20 April 2009 (UTC)

Template:TitleFields:Title has been updated and the two Held submissions have been rejected. I am not sure if my wording is clear to new users who are not familiar with this issue; improvements more than welcome! Ahasuerus 01:38, 26 April 2009 (UTC)

Initials without a period?

Continuation of discussion begun on Moderator Noticeboard. There are at least two authors that have a preference for not having a period appended to their middle initial. There are various problems involved:

• If the author has been credited both ways it will require defining one of the forms as a canonical name and the other as a variant name using the pseudonym process.
• Some magazines use the no period for all authors with middle initials which makes it difficult to determine whether this is what the author prefers.
• Much of the data for such authors has already been entered with a period whether or not that it was the way the titles in question were credited in a pub. This seems to be an implicit policy similar to the regularization rules we have for instances of "Ph.D.", "Jr.", etc.
• Other important bibliographic sources like Contento follow practices similar to ours.--swfritter 18:38, 12 April 2009 (UTC)

Please stop referring to my middle name as an initial. It is incorrect and just demonstrates to me that you are less interested in accuracy than you are in your own interpretation of reality.Shsilver 19:39, 12 April 2009 (UTC)

Lighten up. Is there anyone alive who doesn't interpret reality? :) MHHutchins 21:33, 12 April 2009 (UTC)

Welcome to the Wiki! First of all, a little background information. As Help:Screen:AuthorData explains, the ISFDB distinguishes between 3 different types of "names": canonical names, legal names, and names as printed on the title page. For example, our "canonical name" for Heinlein is "Robert A. Heinlein", his legal name was "Robert Anson Heinlein" and he has been variously credited as "Robert Heinlein", "Anson MacDonald", "R. A. Heinlein", "Lyle Monroe", etc. (When a person changes his or her legal name, we use the last one known.)

Canonical names are primarily determined based on the author's primary name used for genre fiction and are not always the same as legal names. For example, Murray Leinster's canonical name is "Murray Leinster" even though he published stories as "Will Jenkins" and "William Fitzgerald", which are much closer to his legal name, "William Fitzgerald Jenkins".

When entering a new record for a book, story, etc., we enter the name as it appears on the title page. If the name that appears on the cover, in the table of contents or elsewhere in the book is different from what's on the title page, we make a note of it in the publication record's Notes field. We then check whether the entered name is the same as the canonical name of the author. If it is not, we create a Variant Title for the book/story and, if this is the first occurrence of this form of the person's name in the database, we also create a pseudonym relationship between the newly entered name and the canonical name on file. There are a few documented exceptions to the "record exactly as printed on the title page" rule, e.g. we have a pseudonym record for "Robert Heinlein" and for "R. A. Heinlein", but not for "R.A. Heinlein" since we don't consider "R.A." significantly different from "R. A." and we also add a space between initials.

Given all of the above, there are at least three different questions here. The first one has to do with your "legal name" and it should be easy to answer. If your current legal name is "Steven H Silver", then we will record it as "Silver, Steven H" in the database. The second question is whether the name primarily used within the genre -- e.g. the way you are credited by DAW -- is "Steven H Silver" or "Steven H. Silver". The third (and most complicated) question is whether we want to change the past policy of always adding periods after period-less middle initials, which, as it turns out, may also be full middle names. If we do, then we will need to address the issues listed by Swfritter above.

Finally, there is no need to take any of this personally. We have over 50,000 Author records (where "Author" may mean editor, artist, etc) on file, so we have to deal with whole classes of issues rather than with individual records. Ahasuerus 20:54, 12 April 2009 (UTC)

Citing precedent. Even Harry S Truman had a hard time making his case.--swfritter 20:57, 12 April 2009 (UTC)

Now that the backup situation has been sorted out and I have some free time, let me see if I can address Swfritter's points above.

First, reviewing the evidence, it looks like the DAW anthologies credited "Steven H. Silver" while Steven's own fanzines and his SF Site reviews credited "Steven H Silver". Given this divergence, I don't think there is any doubt that even if we decide to record the period-less attributions as they appeared, we still need to capture the period-enabled version. If we don't, then our users -- most of whom likely know nothing about Steven H Silver except that his name has appeared on some DAW anthologies -- will find 0 records when they search for "Steven H. Silver". This means, as Swfritter suggested, that we will need to create variant titles and pseudonyms.

Second, I suspect that the thorniest issue here will be that "[s]ome magazines use the no period for all authors with middle initials which makes it difficult to determine whether this is what the author prefers". I guess a reasonable compromise would be to say something like "If there is no period after the middle initial(s), add it unless it's known that the middle initial is actually a full middle name". Given the limited scope of the problem so far, perhaps we could simply list the affected authors in Help and add more as needed? Ahasuerus 22:28, 19 April 2009 (UTC)

How about initials without periods or space? MHHutchins 01:32, 20 April 2009 (UTC)

As I have begun filling in the backlog of semiprozines it appears that is becoming a standard convention to not use a period for what is truly a middle initial and the combination of first and middle name initials. See [13] this for the second case. Do we use variants in such cases? I would sooner think it makes sense to normalize if we know that the author has a middle name longer than one character.--swfritter 18:17, 21 April 2009 (UTC)

[unindent]That's strange that two authors in one Australian semi-prozine would have initials with no periods and no spaces. Sounds like a case of editorial or printer's preference and not the author's. Especially in the case of K. V. Johansen, who has a page full of credits with initials. I think the names should be normalized until one of them contacts us complaining. :-) (You see, Mr Silver, we have a sense of humor around here. That's what the emoticons are for, in case we're too subtle for you.) MHHutchins 18:39, 21 April 2009 (UTC)

I'd regularise. At least James Pinckney Miller isn't going to get in touch anytime soon. BLongley 19:14, 21 April 2009 (UTC)

Dealing with Bad ISBNs

Don Erikson has found a very interesting example that made me question what I usually do, but also what other people have been doing. Specifically: Web of Light has a bad checksum and gets the big red warning. We never really agreed on what to do in these cases: putting a "#" in front in the ISBN field and adding Notes is one solution people use (it gets rid of the big red warning): and at other times when the "correct" ISBN could be found, some people would use that in the ISBN field and put the "as stated" ISBN in Notes (which gets rid of the warning and makes more links work). I think we're all agreed that we must record what is stated on the publication, and the only practical disagreement was whether something stated as an ISBN must be put in the ISBN field, however bad it was, when it could just as well be put in Notes. BLongley 19:36, 17 April 2009 (UTC)

In this case though, the Worldcat link works despite 0898650821 being invalid. If we "#" the number we'll break that link. We could move that to Notes and put in the "correct" (according to most of the other sites) ISBN 089865162X and get more links working - but break one. Oh, and we have the other ISBN as a separate pub too. BLongley 19:36, 17 April 2009 (UTC)

Suggestions, please? BLongley 19:36, 17 April 2009 (UTC)

You probably know already what my feelings are about this. I think we should record the ISBN printed in the book preceded by "#" in the catalog/ISBN field (to avoid the flag) and make a note that the ISBN comes up as invalid. If we know what the ISBN should have been (and that's going out on a limb I wouldn't want to be on), that number should be recorded in the notes as well. I'm less concerned that we link up to other online databases than recording what's in the book. MHHutchins 20:46, 17 April 2009 (UTC)

You're right, I did suspect what your feelings would be. But again, I'm confused about your intentions. I am still positively stating that we must state what is recorded in the book. Breaking the links just seems wrong though. Both on a relational database principle, and the hypertext principle. The links are there to help, and shouldn't be discarded lightly. Yes, they could be cut down a bit - I only use the Amazon one to see what Dissembler did wrong, the Amazon UK one occasionally to see where it could have been better, and the Worldcat one takes me into new areas. I check Abebooks and Alibris occasionally, but those seem to be polluted by Amazon anyway. We could cut 10 of the links immediately and I wouldn't notice. But when some of them DO work, why cut them off? BLongley 23:09, 18 April 2009 (UTC)

Oh, and what's your suggestion for the OTHER pub? BLongley 23:09, 18 April 2009 (UTC)

About the same way I felt about creating separate records for Canadian ISBNs. I don't like it. That doesn't mean I wouldn't go along with any consensus the group might come to. But it seems I'm the only one so far who has responded to your comment. Most of the moderators seem to have fallen off the face of the earth, and the non-mod editors sometimes stand back and let the few of us duke it out, even though they're perfectly welcome to join in. There's more than 20 mods, and only 5 or so ever participate in discussions. About the same number as active non-mod editors: Bluesman, Kpulliam, Dragoondelight, Willem H., Rtrace and MartyD. Should we start talking about drafting recruiting some new mods? Or maybe I ought to take Shatner's advice and "get a life." MHHutchins 02:33, 19 April 2009 (UTC)

I would agree with Mike on this(Bad ISBN policy) and I think some new recruits would be good, about half of the above mentioned are up to speed and a couple others are close behind.Kraang 02:57, 19 April 2009 (UTC)

If one of you could actually explain your personal Bad ISBN policies I'd be happier - I still have no idea what either of you would do about the other Pub that I found by going researching the problem rather than '#'ing the problem under the carpet. :-/ BLongley 22:02, 19 April 2009 (UTC)

MY BAD ISBN POLICY: If a pub I've entered comes up with an invalid ISBN (bad checksum) I will add a # in front of it and keep it in the catalog/ISBN field. I'll then record in the notes that the ISBN which is printed in the book is an invalid number. Sometimes (not always) I will make an effort to find secondary sources to see if they've come up with another working number. If so I will record it in the notes only. If I'm moderating a submission in which the checksum is bad, I'll ask the originator of the submission to verify that the number given is the number in the book, and precede the same as above. I do not create another record with a working ISBN. I do not correct the ISBN for the record I'm working on. As I've stated before, I'm less concerned about linking up to other databases than recording the book in hand. What's to prevent someone coming along with the identical book from creating another record with the bad ISBN only to find that it's already in the database with a "corrected" ISBN?

The other pub? I'd nuke it. MHHutchins 22:26, 19 April 2009 (UTC)

My Take

If there is a BAD ISBN printed on the book, which WORKS with Amazon and online booksellers, isn't it appropriate to record it in the ISBN Field 'without' the #. This triggers the checksum warning which can be explained in the notes and triggers the Amazon (Etc) links which don't need an explanation. Kevin 23:51, 19 April 2009 (UTC)

If there is a BAD ISBN printed on the book, which DOESNT WORK with Amazon etc, and there is a GOOD ISBN which can be derived/identified which WORKS, then it falls into the class of books with both a printed catalog number and an SBN which also works as an ISBN. In that case, the WORKING ISBN should go into the ISBN field so as to trigger the appropriate links, with a note explaining the substitution / correction in the notes along with the printed ISBN. The answer of "What's to prevent ..." is "Moderator's to catch it after its submitted and and Editor's to catch it after it is accepted", the same as to all the other screwy ways that editors can enter information that is not in accordance with our rules and standards. If you want an 'automated' solution, then we will need to wait for a code upgrade to handle multiple identifiers for a publication (Catalog number, printed ISBN, Printed SBN, Linkable ISBN, and whatever else we think of). Kevin 23:51, 19 April 2009 (UTC)

If there is a BAD ISBN, which DOESNT WORK with Amazon etc, and there is NO GOOD ISBN which can be identified or derived, then we should enter the BAD ISBN with a # and a note that the ISBN is wrong and why it is wrong. Kevin 23:51, 19 April 2009 (UTC)

The choices outlines above provide the most consistent USER Experience (Working links when at all possible (2/3 cases)), provides the printed ISBN in the ISBN field 2/3 cases, Documents (with notes) the inconsistency/problem with the printed publication is all cases. In all cases there should only be a single publication record in the ISFDB. When, as in one example, one ISBN works with some catalogs, and the other ISBN works with other catalogs, my preference is to go with the Amazon linked ISBN as the most 'useful'. Kevin 23:51, 19 April 2009 (UTC)

Kevin, I like most of the above until the "there should only be a single publication record in the ISFDB". The problem with zapping some duff pubs with "authoritative" sources (I use the term loosely as it appears to be Amazon) is that they tend to get put back by well-intentioned editors (or maybe bots). I'm more tempted to go stamp all over them with lots of warnings - e.g. "This book doesn't exist! It was actually published with a duff ISBN which only Worldcat and we seem to know about. Don't trust Amazon any further than you can throw them" etc. (Of course, in this case we wouldn't NEED the other pub as we'd have all the information on the record we all seem to want to keep - but I don't think it's desirable as an absolute rule.) BLongley 18:38, 20 April 2009 (UTC)

As to which link to keep, I'm not especially in favour of Amazon - yes, that's where Dissembler got a lot of the information here from. But Fixer should be doing more from Amazon UK soon. And Worldcat might be the more trustworthy one to "serious researchers". I suppose we could construct hyper-links within the notes for any useful Links that Al isn't giving us, but I'm always reluctant to suggest HTML in notes as it can go horribly wrong. And I wouldn't personally bother with fixing all of them - as I say, I only use three of them. BLongley 18:38, 20 April 2009 (UTC)

Keep in mind that the OCLC data is only as good as the data that they receive from individual libraries and library staff tend to be almost as interested in getting advanced publication information as Amazon.com/UK. It enables them to inform their readers about upcoming books, set up waiting lists for popular titles like Harry Potter, create budgets and do all kinds of other useful things. And so some of them enter lists of ordered books into their main databases, which eventually make it to OCLC and -- presto! -- you have the same problem with announced-but-never-published books as Amazon does.

There are two ways that libraries have of addressing this problems. The first one is to mark the ISBN as "canceled" (there is a special field for canceled ISBNs in the MARC standard) and the second one is to delete the record outright. Unfortunately, the "canceled ISBN" approach is not handled consistently even by powerhouses like the Library of Congress and deleting a local record isn't always propagated to the OCLC database(s). Besides, libraries have to decide when to stop listening to the publisher's assurances that the book will published "any week now, um, as soon as the writer delivers the manuscript, that is" and delete the record.

Having said that, there a couple of ways of telling when an OCLC record is bogus. First, unlike Amazon, libraries rarely enter the page count or the book size until they receive at least one copy, so an OCLC record with an empty page count and no size designation is highly likely to be vaporware. Second, if OCLC reports that a book supposedly published by a major publisher is available from only 1-10 libraries, there is something wrong with the picture. Finally, if you check OCLC's list of libraries that supposedly owned a copy at some point and all (or almost all) of them are not hyperlinked (i.e. do not report a current copy on file), that's a big red flag as well. Ahasuerus 21:07, 22 April 2009 (UTC)

(unindent) Do we have a consensus on the following?

1. When a working ISBN -- that is, one that has a valid checksum AND correctly links to various online sources such as amazon, other sellers and/or OCLC -- is known or has been derived, that number should be placed in the ISBN field and the invalid ISBN printed in the publication should be staed in the notes
2. When the only ISBN known is invalid -- has a checksum failure -- it should be recorded in the DB field preceeded by a # to avoid the warning msgs.

I would like to close this discussion with some agreement, and i think that all but one person was in agreement with the above. -DES ^Talk 16:26, 8 June 2009 (UTC)

If the consensus was that a working ISBN should replace the bad ISBN printed in the book, then it goes against one of the basic tenants of both ISFDB policy and bibliographic integrity: a record should reflect the book itself. When we start making concessions so that our records link with other databases, it will establish a precedent for any future compromises. I will go along with the consensus, but once it becomes policy, be prepared to confront those who will use it as a defense for other creative re-interpretations of data manipulation. MHHutchins 17:05, 8 June 2009 (UTC)

To be clear, I am not saying that the working ISBN should replace the bad ISBN, rather that both should be recorded, but the working ISBN should be in the field that enables the links, and the printed but invalid ISBN should be recorded in the notes (until/unless we have a new and separate field for it).

We should record what the book says. But what parts of that we put into DB fields and what part into notes is not so cut-and-dried, IMO. -DES ^Talk 17:15, 8 June 2009 (UTC)

Well, this is a database after all, and by definition, all the data fields should be searchable. The note field is not (currently) a searchable field. Is there any reason other than linking to other databases for placing a "working" ISBN in the ISBN/Catalog # field? I'd like to hear a stronger argument than that. MHHutchins 17:29, 8 June 2009 (UTC)

Actually the notes field is, in effect, searchable, although not via our built-in search forms. this Google search returns all pub records with "LCCN" in the notes (or anywhere else in the record, but notes is the likely place). There are 183 results. Similarly a Google search for a specific, "bad" ISBN will return any pub records with it in the notes. Perhaps a help page on searching the ISFDB via google would be a good idea. -DES ^Talk 17:38, 8 June 2009 (UTC)

Unfortunately you've just proved Google isn't a complete answer. :-( It should have returned 567 or more matches, as that's the number of "LCCN" matches in pub notes in the last backup I loaded, and I doubt somebody's deleted 384 since then. (Unless the people that prefer LCCCNs have been very active.) Still, I find Google search far more useful than Wiki search so a help page would be good. BLongley 18:05, 8 June 2009 (UTC)

To my mind making those (to me) very useful links work is a very strong argument indeed. Listing an ISBN that is valid -- i.e. one whose checksum is correct -- is IMO also a good reason. -DES ^Talk 17:38, 8 June 2009 (UTC)

DES, are you suggesting that the average ISFDB user would/should use Google to find a pub in our database? Perhaps when the advanced search on the ISFDB can be used to search notes, then my opinion on the matter might be tempered. MHHutchins 18:21, 8 June 2009 (UTC)

I don't think the average user will be looking for ISBNs at all. I think the average user will be either a) looking to see what books author John Doaks has written, or be looking for information on a given title. Such a user may then drill down to one or more publication records, and if s/he does so, having the links to OCLC, Amazon, and various other online resources will quite possibly be helpful. The more advanced user who is looking for a particular ISBN could quite possibly use google, especially if a help page is provided and linked to. Possibly when code changes are fixing/improving the advanced search function, a way to search the notes field will be made available -- obviously SQL can handle such a query. -DES ^Talk 20:26, 8 June 2009 (UTC)

I still support switching the duff and working ISBNs around where the net result is more useful links working. I don't think this breaks the "a record should reflect the book itself" principle so long as both are still recorded clearly. (Ok, technically the note is a record held in a separate table but you can establish which are used as pub notes rather than any other kind: and it's arguable whether they should ever have been "normalised" like that anyway.) It also doesn't break "by definition, all the data fields should be searchable" either - you can search them. Just not via the web interface, although if someone adds a feature request for that I'll look into it. All my other Advanced Search fixes are on hold at the moment though, as people realise the can of worms they might open. :-( BLongley 18:46, 8 June 2009 (UTC)

And we discussed this before: changing the last digit of an invalid ISBN doesn't necessarily create a "working" ISBN. It only creates a valid ISBN which is an entirely different kettle of fish. MHHutchins 18:23, 8 June 2009 (UTC)

Hopefully DES doesn't mean that's the only thing to check before deciding it's "working". One thing I still have a little doubt about is as stated when I started this iteration of the discussion: what to do when the bad ISBN creates useful links. Or when there is more than one alternative form that usefully leads elsewhere. But the alternative to a simple swaparound for the one good/one bad situation, the most common, is likely to lead to people trying to create the links (that the software would naturally generate if we let it) in notes, and the more hand-crafted HTML in notes I see the more nervous I get. So I still support the idea that DES is asking about, although we might have to clarify the help for the even-odder situations, and definitely give some guidelines on how to properly find the intended/working ISBN. And of course, if it's not findable and checkable we shouldn't use it. BLongley 18:52, 8 June 2009 (UTC)

Of course not. An ISBN not printed in the pub is not "working" or "correct" unless it can be used to access records that pertain, or appear to pertain, to the correct publication at multiple online resources. No one, so far as I recall, has ever proposed just "fixing" the ISBN check digit and assuming that this gives a "correct" ISBN. In some cases it does, in some it does not. That is why I said, above, "...a working ISBN -- that is, one that has a valid checksum AND correctly links to various online sources such as amazon, other sellers and/or OCLC...". Only when an editor has been able, via whatever research, been able to determine what the "working" ISBN is, should there be any thought of putting that "working" ISBN into the record in any way, IMO. -DES ^Talk 20:26, 8 June 2009 (UTC)

Look at this discussion and see how I dealt with a recent example of an invalid ISBN. (It prompted DES to re-open this discussion.) How would putting the "working" ISBN in the ISBN field make it a better record? MHHutchins 18:31, 8 June 2009 (UTC)

It would have a lot less dangerous HTML in the notes? Some people might still construct the OCLC link as that doesn't seem to demand a valid ISBN, but that's one link rather than three, so more likely to be right first time. Some of the other possible 12 links might work automatically too rather than have to be constructed one by one manually - I presume you didn't attempt the non-English ones for good reason but I suspect you also didn't check all the remaining English ones? (My apologies if you did, but didn't note that they didn't add any value.) I would support a feature request to reduce the number of links if we're not using them all, but I can't tell which people are using. BLongley 19:15, 8 June 2009 (UTC)

I don't feel as strongly as Bill does about the html in the notes, but three working sidebar links beats one IMO, and this is an unusual case where the "bad" ISBN actually works for one link, usually it doesn't work for any, and if you prefix it by # I don't think it works in the sidebar link even for OCLC. -DES ^Talk 20:26, 8 June 2009 (UTC)

Every now and then you come across a bad ISBN. I had one a few weeks ago, which resulted in this discussion. I think it's a good idea to use the #prefix and add the title to this list. Better than creating a working ISBN that (sometimes) has been assigned to a totally different book. Willem H. 19:07, 8 June 2009 (UTC)

That list was created as a temporary set of examples so we could look at it while deciding how to finally eradicate the problem via rational discussion like this. It was never intended to be a permanent list of exceptions. :-( BLongley 19:26, 8 June 2009 (UTC)

We (well, I know I am anyway) are NOT advocating a working ISBN that doesn't actually work. We want to encourage research into why it's a bad ISBN, and record the findings in the most useful and least dangerous manner. BLongley 19:26, 8 June 2009 (UTC)

After deriving a possibly "good" ISBN an editor should be expected to check that it actually links to data for the pub at hand. Unless this check is made and passed, a "corrected" ISBN should not be used. The specter of a corrected ISBN leading to a totally different book is a strawman -- no one is suggesting doing that, or allowing it. -DES ^Talk 20:26, 8 June 2009 (UTC)

(Unindent) Some things I think people should consider before demanding rigid rules for "ISBN / Catalog #" versus notes. We've already got several abuses or misuses or exceptions for this field anyway: books usually don't have a "#" in front of a claimed ISBN, but we sometimes add them: we don't show it as printed, we let the software hyphenate it: we allow people to turn an SBN into an ISBN: we allow a catalog number OR an ISBN in it: we even allow an ISSN in (although there's a separate argument going on over that - I'd be happy if we lost that though, some people might disagree but it's not being consistently used). And now I've found what a "prebound" book can mean, there's the issue of books that genuinely have a reason to have the ISBN of the contents and the ISBN of the rebinding. Swapping this field to notes and a useful note to this field seems trivial in comparison with the other problems we already have. BLongley 19:47, 8 June 2009 (UTC)

Still, if we implemented a feature request for "let each pub have an ISBN field, and a Catalog # field", would we all be happy to record working ISBNs in the ISBN field, and duff ones in the Catalog # number field? BLongley 19:47, 8 June 2009 (UTC)

I would prefer multiple ID fields with multiple Identifiers for those fields... Enter an ID, then pick which type of ID it is... if you have more than say 3 ID's, have an add ID button just like authors artists, and contents items. Let the first default to ISBN, and the second default to catalog, but there are lots of ID field items I would like to see. (And yes it would end this disagreement permanently). It would also require a re-write of the 'other sites' linking code as well. Each site would look for it's preferred ID. We can add other sites that don't use ISBN. This would also GREATLY improve our support for ebooks etc. Kevin 23:52, 8 June 2009 (UTC)

That's a great idea. A month ago I would have sloughed it off as pie in the sky. With the recent movement toward updating software, I can see it happening. MHHutchins 00:12, 9 June 2009 (UTC)

He beat me to it, but I was going to make a point similar to Kevin's. The ISSN discussion had me thinking there should be a set of ID types with standard meanings (e.g., ISBN-10 or LCCN) and a type of "Other ID". Values can be filled out as known, and sources of the values documented in the notes. Software can then make use of these values as appropriate. An errant value printed in the book could then be recorded as an Other ID. The list of IDs could be a dynamic list, as Kevin suggests, or a fixed list along the lines of the types of verification. --MartyD 00:18, 9 June 2009 (UTC)

The notion that we really have two separate data elements here, i.e. "ISBN as printed" and "ISBN as implied/used by others", was mentioned a few months ago, but we couldn't do much about it at the time. If we implement multiple book identifiers, we could certainly make "implied ISBNs" a selectable choice and resolve this issue once and for all. It would also help with recording SBNs vs. ISBNs, recording books with 2 different ISBNs (see discussion above), recording catalog IDs in addition to ISBNs, etc. It will likely require database changes, so it may not be as easy to implement as some of the other fixes that are currently in the pipeline, but at the rate we are going I would guess that it should be doable within a couple of months. Ahasuerus 00:57, 9 June 2009 (UTC)

More moderators

I agree that we could do with more Mods. But one of the major requirements is being able to communicate well: and on that basis we should sack half the current Mods. :-( Still, I think you've identified some good people: Bluesman communicates well but doesn't really understand variants yet. Dragoondelight adds a lot of detail. Sometimes too much for my taste, but that might help someone else merge things with more confidence. Willem H. is certainly getting there, Rtrace and MartyD and Kpulliam too. If they can step up to the plate and come argue some things a bit more then I'm happy to pass on the blackjack that Ahasuerus lent me. BLongley 22:02, 19 April 2009 (UTC)

The latest cohort of active editors is certainly both active and well-meaning, but a cursory review of their Talk pages suggests that they may need a bit more exposure to the underlying complexities of our data structures and policies (e.g. house pseudonyms) before we give them the magic wand. Some are getting quite close, though, so perhaps a concerted effort to address their "blind spots" may get them to self-sufficiency quickly. Ahasuerus 01:37, 20 April 2009 (UTC)

I'd like to aim a bit higher than "self-sufficiency". We're not overly-burdened with unmanageably-long submission queues, I think where we are lacking is high-level communication where we can get help fixed, policies clarified, group projects not just started but finished, etc. I frankly don't know the current capabilities of some of our mods, we seem to lose communication with many after "self-approval" status. BLongley 18:51, 20 April 2009 (UTC)

That's a reasonable point, but keep in mind that the "self-approval" status makes the editor's life much easier. With the ability to experiment, easily perform multi-step operations and revert your own mistakes, editors can better appreciate some of the more complex problems that we face -- not to mention that they will have more time for finishing projects :) Ahasuerus 03:42, 21 April 2009 (UTC)

I fully admit that if/when I attain moderator status, I will be cruising the wiki at least little bit less than I do today. I won't be 'killing time' on a complex submission awaiting a mod to accept stage 1 of the submission. Also, last weekend I had a 24 hour or so wait for many edits to get approved. Kevin 04:59, 21 April 2009 (UTC)

Wow, Kevin, I'm sorry to hear that. The only explanation I can think of is that one or more may have been placed on hold. (But if that were true, the moderator would have left a message on your wiki page.) This past weekend (18-19th) there were almost 1900 submissions made and accepted. The weekend before (the one you're speaking of, I think) (11-12th) there were almost 1400. I was away on the 11th, which may account for the drop, but there were submissions being accepted every hour of the weekend by other moderators. You had submissions accepted by ChrisJ., Ahaseurus, BLongley, Rkihara, and Kraang throughout Saturday into Sunday morning, and the last submission you made was accepted by me at 12:46 PM on Sunday. The next one of your submissions accepted was on Monday night. It's hard to believe any submission stayed in the queue more than an hour. I've been asking for more pay, but I've not threatened a strike or even a slow-down. We gotta get a union 'round here! MHHutchins 06:12, 21 April 2009 (UTC)

I'm happy to sign the petition for that pay raise. Neither to complain nor to advocate for personal self-approval, I've experienced the "killing time" syndrome, too. It's worst when I have but an hour or two and COULD complete a complex thing in that time, but the activity then has to stretch over multiple days of 1-2 hour availability. Sometimes I have found myself unable to remember all of the things I had wanted to do. I don't see a solution, though. It is the nature of moderated submissions. --MartyD 10:25, 21 April 2009 (UTC)

Well, I've never experienced "killing time", even when I was not a moderator. I always had several projects in the air that didn't require immediate attention by moderators. Even now I get bored easily and switch between projects at whim (just look at the list on my page at the projects that I'm in the middle of.) Sometimes to this day, other editors' submissions remind me of pubs, authors and publishers that need clean-up. One thing leads to another and I have to look back at my browser history just to figure out what led me to where I'm at!

While waiting for submissions to be accepted, you can add covers, either links or upload images. Back in the non-mod days, I'd keep an open notepad (the windows program) for reminders while waiting for acceptance of submissions. I still do for copying and pasting notes to pubs when I have phrases that are repeated often. Poor Ahasuerus finally threw his hands up and made me a moderator so that he wouldn't have to deal with all of my submissions filling up the queue. I was fortunate in that I'd experienced so many different areas of the database that it came natural. That's why you'll see comments from me about what might or might not be in the help pages, because I seldom go into them, except to point out certain areas to new editors, or to back up a statement that I'm making to prove a point.

I apologize for any delays in the acceptance of submissions. We're doing the best we can and trying to live a normal life at the same time. MHHutchins 18:09, 21 April 2009 (UTC)

It was the first or second weekend in April that it was worst, but it's a general pattern. If I start entering edits in the morning, by the time someone starts approving them, the backlog can be 6+ hours long. While I'm sure that many edits get approved very quickly on a last in first out basis, I often don't see my first edits approved until late Saturday night or sometimes until the next day. I wasn't complaining per se, as in 'This is unacceptable' I was only commenting that there are sometimes lengthy delays in the approval process. As Marty mentioned above, that's just the facts of any moderated approval system. As to the subject of 'killing time', often (especially when its a funky edit) I want to see how the results appear, and determine if A 'I am doing it right' and B 'Did it result as I intended'. Also, for those of us still working on entering our collection (as opposed to working on the database in general) there is the fact that I have pulled a shelf of books and won't return them until I'm satisfied with the results. So the stack sits there until I finish that author. Heck I even understand it when I see an active Mod self approving and not digging through the backlog, we all want time to concentrate on our own problems and areas of interest and not deal with everyone elses. With that said... to recap, I wasn't trying to complain, or be deragatory in any way, just pointing out that there are delays in the system (I think someone posted elsewhere today that there isn't even a function to 'see' when an edit is submitted, so you folks may not have even been aware of the issue). Thanks Kevin 22:25, 21 April 2009 (UTC)

The time of submission does appear in the queue (it's exactly one-hour behind my local time), but once the submission is accepted it's removed from the queue and placed onto the "recent integrations" list which only shows the time of acceptance. So the only list that shows accepted submissions doesn't show how long that submission sat in the queue. I do find that there are times that moderators will pretty much deal with their own submissions which leaves all the non-mod edits to the more, shall we say, conscientious mods to pick up the slack. I personally can't stand to see a long queue, so despite having plenty of projects to work on, I make a conscience effort to clear out submissions between my own. Other mods can see that and lay back as we don't want to step over each other while working with the edits. Some editors are set in single-step one-record-at-a-time mode, so it's easy to work with those submissions. We mods, or at least I, forget how cumbersome multi-step edits actually are when you're used to approving your own edits without any delay between the steps. You say the first weekend in April was slow, and now I remember. I was in Charleston SC, away from my computer and the database all weekend. Now I understand how something like a 24-hour wait could have happened. Heads up, I'll be on vacation the second week in June, another one in August, and the first two weeks in November. MHHutchins 04:40, 22 April 2009 (UTC)

I have penciled in those dates on my calendar to drop great piles of edits in the queue during your absence, so that you will feel even more appreciated and adored upon your return. Cheers! Kevin 12:48, 22 April 2009 (UTC)

Yup, Michael has been incredibly productive over the last couple of years with 68,638 submissions and 90,311 approvals to his credit. That's 217 submissions/approvals per day and I am surprised that a human can sustain this level of activity for that long. Fixer is telling me that he would like to be like Michael when he grows up... Ahasuerus 14:09, 22 April 2009 (UTC)

One of my goals was to put us humans ahead of the robots in adding new publications. I recently succeeded. UP WITH PEOPLE! DOWN WITH SKYNET! MHHutchins 17:06, 22 April 2009 (UTC)

Hmmm, when I was counting I was just trying to overtake the founders. Now that's been done though, maybe I should stop supporting Fixer's entries? ;-) BLongley 20:14, 22 April 2009 (UTC)

Shh, don't give Fixer ideas! He has half a million records in his two Amazon-derived databases alone and if he sees it as a challenge, he can very easily beat the rest of us! As a matter of fact, the only thing stopping him from creating submissions this afternoon is that he is in the middle of a database reorganization with 352,000 records out of 500,000 processed so far. But once he is done with this task and learns how to create combined submissions based on records from both databases, watch out! Ahasuerus

Well, at about 20 a day I'm happy to cope with his submissions - I just have to do 21 myself to keep ahead. I guess I should provide an update on how he's doing though. BLongley 21:06, 22 April 2009 (UTC)

You have nothing to apologize for. Nor does any other moderator. But, really, what's this "normal life" stuff? I suppose the next thing we'll be hearing is that you sleep sometimes, too.... --MartyD 03:49, 22 April 2009 (UTC)

Yeah, I catch a few winks before the warden notices and sends the electric charge to my chair. MHHutchins 04:40, 22 April 2009 (UTC)

Final Decison on Helix

See this discussion. There are two issues:

• Resurrect the issues of the webzine Helix that were removed. These were based upon the primary source which is no longer available.

I would say yes since they were in the system for a fair amount of time and there are other similar cases which I don't think should be removed.--swfritter 21:26, 18 April 2009 (UTC)

Let them return, but only with notes about their unavailability. Hopefully their inclusion won't be a precedent for adding other defunct webzines or even current ones that don't match the criteria (see my answer below.) MHHutchins 21:36, 18 April 2009 (UTC)

• Use secondary sources to create what data we can for the three issues that were not entered before the webzine was pulled.

In (with notes about the unavailability). Kevin 06:26, 19 April 2009 (UTC)

An ambivalent yes with many disclaimers on the pub page and the pub notes. The secondary sources are fiction reviews from sources which I consider to be reliable.--swfritter 21:26, 18 April 2009 (UTC)

Only because we're allowing the other issues. In the future there should be a greater threshold for inclusion of webzines. First they must be a paying market. And must have published stories by recognized authors in the field. That's not as subjective as it might first sound. It should be rather obvious to most of us. I don't think the 'zine's being Hugo-nominated is a high enough criterion. All it takes to get on the Hugo final list is a few dozen nominations from friends and allies. MHHutchins 21:36, 18 April 2009 (UTC)

But Hugo Nomination will drive users to look for information about the publication. Are we recording history or not? If users will come here looking for it, don't we have some desire as a community to feed that desire for information? If a work is Hugo Nominated people WILL come looking for information. Kevin 06:26, 19 April 2009 (UTC)

I don't think Hugo nominations have that much drawing power. Off the top of your head, what were the fanzines and semiprozines nominated last year? MHHutchins 01:49, 20 April 2009 (UTC)

I don't know and don't care (or remember) who was nominated (or won) last year (When I care.. I look it up). The question to answer is "When a user is looking at the list of Hugo nominees, for this or any other year, and they want to know more about that work/author/publication/zine/series, should they consider the ISFDB a likely source of more bibliographical information?" The answer to that question is the answer to whether it should be a tripwire for inclusion. Kevin 02:48, 20 April 2009 (UTC)

I have no problem with Helix returning if a Mod is sponsoring such. Nor any other SF webzine. I don't personally have any problem with stub entries for first publications of a work, and a lot of webzines did provide such. I might create a stub entry for an issue of "Playboy" that had some SF content later reprinted in something more normally included, and I wouldn't object to somebody doing the same for a webzine. If they want to go further and include non-SF I'd have to say that's out (just like I'd expect a complaint about entering INTERIORART entries for Playboy, or Coverart Images :-) ).

OK, purely-SF webzines might get entered entirely on that basis, but so long as we don't encourage that I can cope with them. By leaving it to someone else. We don't do webzines. If someone will deal with them, let them. Deletion is a major pain here, and should be reserved for RPGs, Comics, etc. Deleting true written SF is not something I have an interest in. BLongley 22:42, 18 April 2009 (UTC)

In. (Hugo Nomination (work or publication) is a reasonable and understandable line in the sand) Kevin 06:26, 19 April 2009 (UTC)

I'm fine with the re-inclusion of Helix for now, but I am still thinking about where we want to draw this particular line in the sand. There are all kinds of SF webzines out there and they present all sorts of problems due to data mutability (as we have seen with Helix), so we may be opening a Pandora's box. Ahasuerus 00:50, 20 April 2009 (UTC)

Along with setting a high standard for inclusion, the moderator who accepts that webzine's inclusion should also become guardian over the records in the database. I believe this concept arose from either Swfritter or BLongley, but I would go along with it. In other words, each webzine's inclusion should be debated among the editors and allowed on a case-by-case basis. That should keep Pandora's box from bursting open completely. MHHutchins 01:49, 20 April 2009 (UTC)

I will start on the Helix resurrection tomorrow. If the semiprozine category is discontinued this year, as is highly possible, I think such publications either become eligible in the fanzine category or their editors will have to compete in the Best Editor (Short Form) category. Hopefully more of these publications will have the good sense to provide downloadable versions. More and more of the deadtree semiprozines are doing so.--swfritter 17:57, 21 April 2009 (UTC)

Section Titles in Story Titles special case?

This discussion originally started on my talk page but seems worth a broader discussion, even though it's likely a rather isolated case. --MartyD 10:51, 21 April 2009 (UTC)

The original question pertains to the titles I used on some of the content in Lord Halifax's Ghost Book and Further Stories from Lord Halifax's Ghost Book. For example, Haunted Rooms: Head of a Child or Apparitions: The Ghostly Passenger.

Start comments excerpted from User_talk:MartyD#Further_Stories_from_Lord_Halifax.27s_Ghost_Book

[... A]re the section headers actually in the titles of each piece, i.e. does "Apparitions: The Ship in Distress" have that complete title on the page on which the story appears? We ordinarily don't prefix story titles by section titles. Thanks. MHHutchins 03:17, 11 April 2009 (UTC)

[...] The stories in them don't have the same title page as do the other stories or sections. The sections are sort of mini-collections in their own right, but I couldn't see entering them that way. One, The Renishaw Coffin, is done differently from all of the others. It has an inner title page listing the credits for the three contained works and has a single story-style entry in the main TOC; its contents do not appear in that TOC and are given simple italicized headers and no further attribution information. For all of the other sections, each has an inner title page like so:

The Section Title

First Story Title

Second Story Title

Third Story Title

The first story then appears on a page with the section title right-justified at the top (as is done for the normal individual stories), and the story's title is below it, centered, with small-font attribution information below that (for a normal story, this paragraph appears below the title on the title page). Subsequent stories in the section are separated by a centered star, under which appears centered just the story's title and under that the small-font attribution information. The section title appears in the TOC with a trailing hyphen and no page number, and the contained stories are listed indented under that and with their own page numbers.

So I interpreted the inner title page for a section's being "The Section Title: First Story, Second Story, Third Story" and then just broke up the instances. Maybe I should go back and redo them as collections, but they're not really works. If I remove the section portion of the titles, there's no record of the section despite its title page. Is there something else I could have done that I missed? [...] --MartyD 11:22, 11 April 2009 (UTC)

End comments excerpted from User_talk:MartyD#Further_Stories_from_Lord_Halifax.27s_Ghost_Book

Opinions on what should be done with them? FWIW, here is Locus1's handling of similar content (same stories in a different omnibus). --MartyD 10:51, 21 April 2009 (UTC)

One possibility to explore is the use of series to connect related stories. I don't really know if would work in this situation or not.--swfritter 17:24, 21 April 2009 (UTC)

Unfortunately, series is about the only way to deal with it considering the limitations of the current software. Locus has many more types than we have here: forewords, biographical material, vignettes, indexes, miscellaneous material, facetious articles (my personal fav), etc. Another situation arises when a story or two from the section is either a reprint or will be reprinted, and having to deal with how it is subsequently titled. I don't like the creation of variants in situations like this, so I would opt for not including the section name in the title. (I hated what happened to all of Bradbury's Martian stories when variants were created adding the years in the titles.) So the bottom line is, if the story and section are both printed on the first page of the story then use them both. If they're only in the TOC, don't. That's the rule as it currently applies, but there's always wiggle room... Or being persuasive enough to get the rules changed. MHHutchins 18:20, 21 April 2009 (UTC)

The Martian Chronicles was a special case, and if the Years hadn't been on a title by title basis (e.g. if they'd been SECTION titles instead) or didn't actually add some value (nobody had written synopses for those titles that included the year it was set in, and I suspect nobody ever will), then I'd have probably have left them - the work involved in creating and maintaining the variants is too severe. It should be the last option really - in this case I don't know how many variants would be needed, it's a viable option for things that are almost unique though. I have seen section titles created within a publication as markers for what the ToC says - but those are not really essays or shortfiction (unless referring to an introduction, maybe). Series looks the best bet to me. And we have got a lot of defunct series that need reusing as we can't delete them. BLongley 19:50, 21 April 2009 (UTC)

Yeah, I still run across those "essays" that were created as section titles (remember the nightmares for the Arthur C. Clarke "series"). You bring up something that I'd forgotten about. Where do we go to look for bad series so that we can reuse them. It was relatively recently that I started marking my bad series with my name, but I've used them up. Have you run a script or is there a simple search? Thanks. MHHutchins 20:25, 21 April 2009 (UTC)

I haven't run my "find empty series" script in a while and the few dozen empty series record with "delete" in their name were reused a couple of months ago, although I created "delete or reuse (was R. L. Stine's Ghosts of Fear Street-2)" yesterday night after cleaning up some of Stine's ghost series (no pun intended!). The remaining ones tend to be nested series, so unless you need to create a nested series whose structure happens to match the structure of a currently existing defunct nested series, they can't be reused. Parenthetically, this is something that we may want to add to the "self-sufficiency" list. Ahasuerus 21:09, 21 April 2009 (UTC)

I've just posted my script to find useful non-entangled series, and renamed a few to "to be reused" for other editors to pick up via normal searches. It seems some editors thought it would be a good idea to try blanking the series name out - sorry, that doesn't work. And some people chose stupidly short series names - e.g. "1". Unsurprisingly, searching for that series doesn't turn up a unique name. I'll fix a few more of the silliest. BLongley 17:53, 22 April 2009 (UTC)

OK, 74 blank or typoed or truncated series are now available for reuse. There's about 250 more, then we'll have to move onto series that are in a hierarchy. But that should be enough for now? BLongley 20:02, 22 April 2009 (UTC)

There are another 52 with DELETE that are also available for reuse, but only 16 have no parent series. Kevin 20:11, 22 April 2009 (UTC)

We could probably rename some of those to make it clearer which have too many children or siblings. But it's not a major issue - we've got about 7000 series, so there's not too much pollution of search results yet. I just don't want to increase the problem. We'll get that naturally - it's when the series name is too short that you can't find it, or so long that it gets truncated, or just plain typoed, that we get a minor problem. More careful entry would be good, reuse of duff series is better than creating new ones, etc. I'm glad that I can switch this from "only people with database skills can help" to "here's a bunch of things anyone can help with" though. BLongley 21:00, 22 April 2009 (UTC)

One other semi-gotcha I should have mentioned: Each of the two pubs has an Apparitions section. The others are named distinctly even when similar (e.g., Dreams and Portents vs. Prophetic and Other Dreams. Can one have two series with the same name, or would that lead to something like "Apparitions (Lord Halifax's Ghost Book)" and "Apparitions (Further Stories from Lord Halifax's Ghost Book)"? --MartyD 10:21, 22 April 2009 (UTC)

The system won't prevent you renaming series to have the same name, but they're unworkable afterwards. E.g. I've just created several called "To be reused" - but if you tried to add a title to that name, which one would it go in? To reuse a series, rename it to something unique and use the unique name. I wouldn't recommend duplicate series names for anything you want to keep, only in cases like this where you want to find a series to reuse. BLongley 18:10, 22 April 2009 (UTC)

"Censored" Poe titles -- dashes or underscores?

Sorry if this has been discussed before, but I couldn't find it. In working with some Edgar Allan Poe collections and titles, I've come to the cases where the titles omit names, using one or more ruled lines in their place. Someone had previously entered titles using underscores, a practice I had copied without thinking. But in one of the books I have, and also in this seemingly authoritative Poe Society site, it looks like (em)dashes are used, not underscores. Is there a standard for which form of punctuation -- dashes vs. underscores -- to use? Also, if using dashes, is there a standard practice for using two hyphens vs. one emdash? --MartyD 12:30, 25 April 2009 (UTC)

Concerning the situation where a title or name or date has been omitted, I don't think we have a standard here, but whatever we choose it must be obvious that it's an omission and not the break that an emdash or ellipsis implies. If I had to choose, I'd go with the underscore. (Anyone got an MLA Style manual?) Also the two poems titled "To___" you might want to disambiguate by adding the first line in quotes and within parentheses. MHHutchins 15:08, 25 April 2009 (UTC)

Using the first line in quotes to disambiguate identically titled poems (and adding Notes at the Title level if there is additional information available) sounds like a good idea. If we don't add this level of disambiguation, the titles will more than likely get merged at some point. This is not uncommon in the world of poetry and we may want to add it to Help. Ahasuerus 18:06, 25 April 2009 (UTC)

About the difference between hyphens and emdashes, many editors haven't learned how to create emdashes so they use two hyphens. But they both have the same implication and shouldn't be used to create variants. If you see the use of two hyphens, feel free to change them to an emdash. MHHutchins 15:08, 25 April 2009 (UTC)

A good reason to use dashes. I am going to disambiguate all of the ambiguous "To ___" Poe titles with parenthetical first-line portions as suggested. Seems straightforward and should help with merging. I have a source that provides that info. --MartyD 22:50, 26 April 2009 (UTC)

Title date clarification

If a variant title has a clear first-publication-under-that-title date, should the variant title really be given the first publication date under ANY title as stated in the EditTitle help:

The original publication date of this title. [...] For works that have had variant titles, the date to enter is the first under any title and any pseudonym; variant titles do not have their own dates.

These two sentences seem to contradict each other. If the original publication date of any title is used for all titles, how can one tell the date this title first appeared? --MartyD 15:40, 25 April 2009 (UTC)

Personally, I agree with you, Variant 'Titles' (New Titles) should have thier own dates... and Variant Titles (New Author Names e.g. unmasking a pseudonym) should have their own dates, while Variant Titles (A slight change in name preference/title punctuation (Periods, initials, em-dashes, Ph.D, honorifics etc) should conform to the originally published date. That said (My personal opinion above). The correct way to read the help is that when it says 'Title' above it is referring to the 'Title Record' and the 'Root' or 'Parent' Title record is controlling. "The original publication date of this title" indicates the first date and publication appears under the parent title record. While the second sentence you wrote uses the word works when it could be re-written to say "For Title Records that have had Variant Titles, the date to enter is ...." - Again, I agree with you, but unless there is a site wide agreement to change this practice, them's the standards. Kevin 15:56, 25 April 2009 (UTC)

I agree with you both that the first date of a title's appearance should be the date of that variant record but only when the variant is a change in the title of a book. I say that because both records are displayed at the same time. But there's one big gotcha when it comes to content records. When that title record appears in a content listing it will only show the date of that record and not the original date. Here's an example where a title record and its variants have different dates: the Brian W. Aldiss story A Romance of the Equator which was first published in 1980. In the record for its publication in Asimov's January 1983 you'd think it was an original publication because it's got the same date as the issue. The "Brian Aldiss" variant appears in this 1986 collection and the story is dated 1986 which make it appear to be a new story. These are all the same story, the only difference is a change in title or author credit. Shouldn't they all have the same date? MHHutchins 16:31, 25 April 2009 (UTC)

I should add that the ambiguity in the help page should be corrected. Instead of "original publication date of this title" it should read "original publication date of this work". MHHutchins 16:39, 25 April 2009 (UTC)

First of all, thanks, I will stick with accepted convention of using the earliest date of publication under any title (and will go fix a few things I just did this morning before noticing what the help says). I can see from your example there's a subtler issue of what-makes-a-title-a-title. It's almost as if author names also need dates. But anyway, I'd argue the pseudonym relationship takes care of variations in author name, and I wouldn't use the variation in author name to treat the title as "different" (I'd consider the need for a variant title an artifact of the database implementation). To me, the title's text's changing is a different case. Consider A Paean vs. Lenore, two titles for the same poem. "A Paean" was published in 1831, "Lenore" in 1843. No one will ever find a "Lenore" published/printed in 1831 because that title did not exist until 12 years later. Yes, if I look at a collection containing "Lenore" (1843), I would not know the poem was first published in 1831, but if I went to the title record OR to the author record, I would see it is a variant of "A Paean" (1831). In a way, it's sort of just a display issue (if the content record showed the parent title AND date parenthetically....). Easy for me to say. :-) --MartyD 17:03, 25 April 2009 (UTC)

(Slightly edited after Marty's reply preceeded mine) Michael has given a perfect example of two things. First, that minor changes in author spelling, and in titles do not warrant a new date. Second that for items that 'deserve' a new date, we are letting a 'software display issue' cause us to enter incorrect data. Heinleins The Rolling Stones is a good example of a work where a variant deserves a new date. Looking at the display summary at the top, seems to indicate that Space Family Stone Variant appeared simultaneously with the original publication, but in reality it appeared many many years later. Anyone visiting the Space Family Stone title record will think that there is a missing publication from 1952. Is it appropriate to cause allow a display issue to cause us to overwrite the first appearance of a true variant title? Do we want to take a stab at defining 'Minor' vs. 'Major' Variants and changing our rules? Kevin 17:11, 25 April 2009 (UTC)

This goes back to part of my original response. I personally feel that the title record for Space Family Stone should be date 1969, the date the title was first used. It would appear on Heinlein's summary page as a variant and the date would show when it was first used, AND all of the pubs under it would appear chronologically under the parent title record. I try to get by doing this as much as I can (until someone comes along and changes it back!) MHHutchins 18:40, 25 April 2009 (UTC)

Agreed - So are enough of us, of this opinion, that the rules should be changed to reflect this reality? Kevin 18:56, 25 April 2009 (UTC)

Not until we have agreed on a methodology for documentation. Notes + a tag? In addition the determination of whether a title is significantly different needs to be based upon verifiable documentation - not just the opinion of the editor and it must be documented at the title level. Most of the time this will be available within a pub under the wildly variant title; otherwise a source should be cited. These instances need to stand out; otherwise the methodology will be emulated when it is inappropriate.--swfritter 20:01, 25 April 2009 (UTC)

Hence my comment about distinguishing between a major and a minor variant. If we can define a 'Major' variant which is worthy of its own date, then we can put that definition in the help so we can stop emulation where it isn't appropriate. One possible identifier of a major change in my mind is 'New or Dropped Title Case Words' which will ignore a, an, the, etc type title changes. Kevin 20:08, 25 April 2009 (UTC)

Are we talking differences in title or difference in story content? The current standard is totally appropriate if the stories are essentially the same in content. That is the very definition of variant title. The significant issue is how to treat a variant title where the text of the story is significantly different.--swfritter 21:17, 25 April 2009 (UTC)

My original question had nothing to do with content changes, just completely different titles. I am going back and making sure the dates all match the first publication of any title and am adding a note in the title record documenting when and where the current variant was first used in publication. Much as I might like the standard to be different, I wanted to conform, and I can't see simply changing the rule without figuring out how to deal with all of the existing data entered under different rules. --MartyD 21:36, 25 April 2009 (UTC)

Thanks for the clarification. The discussion got a little bit off track. It would take an incredible amount of work to change the existing data in order to implement a new standard. What you are doing is fine.--swfritter 22:05, 25 April 2009 (UTC)

There seem to be at least three different issues here. We'll need to split this into separate discussions, I think. 1) A variant author name alone is not always enough to change a date. For instance, all Iain M. Banks titles should match their Iain Banks equivalent, as only one version actually exists. But it might be appropriate to separate title dates when, for instance, Kurt Vonnegut dropped the "Jr.". Or Dean R. Koontz dropped his "R.". Or Stephen R. Donaldson dropped his "R." Some are date differentiated, some are just US/UK differences. (Other countries can be considered.) 2) A completely different name for the same work is probably worth recording a different date for so people don't go looking for a prior edition that doesn't exist. A minorly different name (e.g. when the Brits don't use an exclamation mark and the Americans do) is probably not worth it. 3) Significant textual differences should be distinguished at title level, and a date change might be appropriate - even if one country continues reprinting one version and another reprints the revised version. The date the versions diverged is significant. BLongley 22:34, 25 April 2009 (UTC)

1) Yes - the date changes only when the name change has significance, insignifcant name changes do not warrant a new title date. 2) Yes - as I described above. 3) Yes. Kevin 15:03, 26 April 2009 (UTC)

If a title is not changed but there is a significant change in the text a non-variant title is sometimes created with a title something like "Title X (expanded)" and the later date used. But if the title is changed and the text is substantially altered it is more likely to becomes a variant title with the date of the parent title. A 'minor' versus 'major' variant relationship should actually be a database designation but that is not likely to happen soon. The only way we have of doing it now is through text in the title enclosed in parenthesis. "Space Family Stone" would have to be modified to something like "Space Family Stone (The Rolling Stones major revision)" in order to justify giving it a different date. That is a particularly cumbersome and ugly method. Another method would be documentation in the title notes to justify a later date.--swfritter 18:06, 25 April 2009 (UTC)

Yes, document it in the title notes. I have no idea if my "Space Family Stone" is significantly different from "The Rolling Stones" but I know what my books are titled and putting the latter name into Amazon UK shows up this R&B band that apparently has some notoriety but no SF relevance? ;-) BLongley 22:34, 25 April 2009 (UTC)

On the Stone variant, there is no revision involved, only a title change. Kevin 15:09, 26 April 2009 (UTC)

It's almost as if there is a missing "Work" object. Author::Work, Work::Title, Title::Publication. A substantial change in the text could then be a derivative or revised Work with its own set of titles. The Work could have the canonical author, the canonical title, and an original date (could even be before the first publication date, if known). A Title would then represent the label under which a Work was published, and Publications would be instances of that Title. Thank goodness no database changes at all required for that.... --MartyD 18:21, 25 April 2009 (UTC)

For what it's worth, back when the project started in 1995, we debated how deep the hierarchy would need to be. One contender was a 3 level "Title - Edition - Publication" scheme, but we decided that it would be too cumbersome and, besides, it's not always easy to tell whether a book is a new printing or a new edition. Eventually, we decided that Titles and Publications should be sufficient for most practical purposes.

Much later, ca. 2006, we realized that we had no easy way of capturing relationships between revisions, abridgments, fix-ups and similar beasts (aside from Variant Titles) and asked Al to create a new Relationship field. He still has it on his list of things to do, but it kept getting pushed back as other projects were deemed higher priority. Hopefully, it is something that he will be able to address when he becomes available again, but even then it will likely take some time. Ahasuerus 18:33, 25 April 2009 (UTC)

[unindent]Assigning the same initial date of publication for variant titles and their parent title is a core concept. Pseudonyms and title differences are not relevant except as a basis for creating variant title records. The answer to Martyd's initial question is that the date(s) of appearance of the variant titles can be determined by looking at the list of pubs in which they appeared.--swfritter 22:54, 25 April 2009 (UTC)

Yes, but if you look at, say, the parent title record, you cannot tell. The list of variant titles provides each variant with a parenthetical date that is the same as the date for the parent record you are looking at, and the list of publications does not tell you which title is used in which publication. At least for works appearing in collections. See, for example, Siope — a Fable. What I would expect to see the first-use dates associated with each variant title listed. --MartyD 11:14, 26 April 2009 (UTC)

This is a display issue that needs to be resolved through software, something that is not likely to happen soon. Data integrity should take precedence over display issues. It is a design flaw that variant titles even have dates and story lengths associated with them. The only place they should be available to edit is from the parent title record. The method we have now totally breaks the rules of normalization - data that needs to be synchronized should not be stored in multiple places.--swfritter 13:35, 26 April 2009 (UTC)

Unfortunately, our software has a number of gaps where it allows similar discrepancies. In many cases the only thing we can do is write a script to find potential issues and them review/fix them manually -- obviously a suboptimal solution, but the only one we have until Al returns. Once I finally retire (which I have been threatening to do for years, so don't hold your breath), I may try to learn enough Python to fix these issues. Until then (or until someone else steps up to the plate) Al and Roglo are the only two people who have actually touched the code and neither one is available at the moment... Ahasuerus 21:43, 26 April 2009 (UTC)

And it seems we disagree on what data "needs to be synchronized" anyway. As long as we have gaps in our publications to go with titles, then we need to have a way to record when the title varied without the pubs being present. We either need a lot more effort on "stub" publications (so we can resurrect the variant dates by reference to publications we have), or we keep the current situation for now. I've already seen 2 or 3 edits today where people are following the current standard (mostly - why keep the Year and Month but not the Day when you're recording a weekly publication?) but the only way we'll recover the variant dates we've lost so far is by reading notes. And I for one haven't written a bot THAT intelligent. BLongley 22:13, 26 April 2009 (UTC)

Every time this issue has come up in the past the issue that could not be resolved is what to do in the case where the canonical title has been published at a later date then the story was published under it's original title. The only way to resolve that issue is by making the earliest occurrence of a story the canonical title. It is obvious from the Help excerpt stating that "variant titles do not have their own dates" that the original intent was that there be only one date associated with a title. It's unfortunate that we do not have the software resources. It has resulted in many conflicts and discouraged many valuable editors. It would be nice if there were someone to blame but everybody is doing the best that they can.--swfritter 16:41, 27 April 2009 (UTC)

[After edit conflict] See my note below which I hope will bring an end to this discussion. Same text, but different author or title, should have one date (the first), which is the current standard. Changes in text can not be resolved until a major software change. MHHutchins 16:54, 27 April 2009 (UTC)

[unindent] Here's an extreme example from Poe that's at least interesting to think about. It seems most of his poems were revised multiple times. "Imitation" is thought to be the original work for "Dream Within a Dream", and there's an intermediate "To — —". If you look at them in chronological order: Imitation (1827) then To — — (1829) then Dream Within a Dream (1849) you can see it. (If you want a more detailed chronology, see the notes here). I have currently set these up as variants, and in reviewing that, I noticed my very own pub in hand Great Tales and Poems of Edgar Allan Poe has two of them -- "Imitation" and "Dream Within a Dream". And the detailed notes linked above also mention: “Imitation” — 1969 — The Collected Works of Edgar Allan Poe, vol. 1: Poems, ed. T. O. Mabbott, Cambridge: Belknap Press of Harvard University Press (1:75-76) (Mabbott prints the three main versions of this poem as separate items, although he notes in the descriptive text that they are related to one another.). Variant titles just isn't quite enough here. In fact, I'm thinking it will be better to go with three separate, unrelated titles in this specific case, just because there are instances where multiple "variants" are published in the same collection, and notes will be the only place to record the linkage. --MartyD 12:08, 27 April 2009 (UTC)

There's two issues here that are being conflated. Dating for changes in author credit and title, and dating for revised or variant text. The first is more easily resolved than the second. As Ahasuerus and others have pointed out, the second issue will only be resolved when changes in software allow "connections" between records based on text not titles. And that's not going to happen soon. Marty, there are plenty of records here of new works based on older works where variant creation has not been done. If both title and text changes I don't see any problem of each being a parent record, with notes to explain the connection. Leave them separate until the software allows you to connect them. MHHutchins 16:49, 27 April 2009 (UTC)

I hear you and am not arguing. I only meant it as an interesting example, not as a campaign. I realize in this case the variant approach was -- and still is -- wrong (I wasn't aware at the time I linked them up they are so different). I will be separating them (and giving them their own dates) and will check the other ones, too. I think this radical a degree of revision is more the exception than the rule. --MartyD 18:22, 27 April 2009 (UTC)

When I first began editing ISFDB data I had a full head of hair. No longer the case.--swfritter 16:58, 27 April 2009 (UTC)

Ah, I hadn't considered previously existing baldness to be an advantage... So glad I'm starting with a leg up! :-) --MartyD 18:22, 27 April 2009 (UTC)

Proposed update to Help concerning works signed by initial

I think this is standard practice for most of us but Pub Help concerning Authors should be updated to so indicate; Proposed text: "If a work is signed with initials and there is sufficient evidence as to the author's identity the author's canonical name should be entered. The substitution should be noted in the publication notes."--swfritter 22:09, 27 April 2009 (UTC)

I'm agree with the proposed change. MHHutchins 22:24, 27 April 2009 (UTC)

I take it we are talking about 'Artwork' (Cover or Interior) signed by initials? Kevin 00:03, 28 April 2009 (UTC)

I understand it refers to both art and text, but I'll let Swfritter clarify his proposal. I think it should be whole hog. MHHutchins 00:50, 28 April 2009 (UTC)

I'd also understand it to refer to text as well. The kind of thing that comes to mind is editorials, prefaces, & other, well, editorial matter. (I'm in strong agreement with this--both sentences, BTW.) -- Dave (davecat) 01:39, 28 April 2009 (UTC)

I advise a bit of caution here. One notable example is Kevin J. Anderson and his pseudonym K. J. Anderson which was intentional. If we go down this path... we should be very clear that this is for 'editorial', 'publisher' imposed, ease of writing, and abbreviated signatures, etc type initials. Otherwise we may obscure an intentional pseudonym by ignoring the printed page. Kevin

I don't think Swfritter means every use of initials. He's referring to the use of initials solely, as he says "signed with initials." This has nothing to do with title page credits. And the change in the help pages will make that clear. MHHutchins 03:27, 28 April 2009 (UTC)

A light Dawns.... Thanks! Kevin 03:51, 28 April 2009 (UTC)

In the case where the initials are for someone explicitly credited on the title page (e.g., the author or editor), should the name from the title page be used instead of the canonical name? --MartyD 09:58, 28 April 2009 (UTC)

Canonical saves a variant. But you could argue that "B. W. A." initials on the intro to a "Brian Aldiss" book would be more correctly expanded to "Brian W. Aldiss". I wouldn't especially object to the title page version or even copyright page version or common sense version if those make more sense than mere initials, so long as they don't lead to creating yet another pseudonym. E.g. I think I've seen "D. N. A." for "Douglas Adams", but as we don't have "Douglas N. Adams" I wouldn't create it. BLongley 17:57, 28 April 2009 (UTC)

There'll be plenty of wiggle room for some time, though - many artists don't even have a canonical name yet. BLongley 17:57, 28 April 2009 (UTC)

Typed as Bill was saving - Good points. This would currently apply to art and text but I think we need a separate section for art - someone else can come up with proposals for that. Initials could also credit a pseudonym. How about: "If a work is signed with initials and there is sufficient evidence as to the author's identity the name most logically associated with those initials should be entered even if it is a pseudonym or variant spelling. Internal evidence within a publication should take precedence. The substitution should be noted in the publication notes." For now I think that will also work for art since it allows external sources to be used if there is no internal evidence.--swfritter 18:20, 28 April 2009 (UTC)

I agree with the proposed scope, i.e. introductions, editorials, covers, etc, but I suspect that we need to spell it out more clearly or else some well-meaning editor will try to change the authorship of all books by, say, A.E., e.g. The Avatars: A Futurist Fantasy, to "George William Russell". Ahasuerus 01:48, 29 April 2009 (UTC)

Restrict it to "short essays (introductions, editorials, etc.) and artwork"? Perhaps with the added provision: "In rare cases, example, authors are more commonly identified by their initials and their credits should be entered as such." Would the addition of the second part make it unnecessary to restrict the title types to essays/artwork?--swfritter 18:15, 29 April 2009 (UTC)

Maybe specific exclusions would be easier to describe (and clearer) than a restrictive set of inclusions. I.e., "Not to be used on the credit for the main work/collection or on the credits for primary works in collections. Judgement should be exercised in cases where secondary works appear in a non-collection, with a bias toward using the same treatment that would have been applied had the publication been a collection." Is there a term for "collection" that encompasses Anthology/Collection/Omnibus without being one of those three terms, so strict constructionists can't pick nits? --MartyD 14:42, 30 April 2009 (UTC)

I think it was Bill who introduced the term "container publication types" to describe Anthologies, Collections and Omnibuses. Their distinguishing characteristic is that they don't display their corresponding Anthology/Collection/Omnibus Titles and, unlike Novels, won't let you edit them in the Contents section. The original idea was to avoid confusing our users, but we have found that it tends to mislead new editors into thinking that "container" pubs do not have associated "container" Titles. It also makes it hard to find problems with missing (or extraneous) "container" Titles in "container" Publications.

By the way, we also have to be careful with the term "sufficient evidence". For example, copyright page data may well be "sufficient evidence" if you know the author's legal name and/or who owned the rights to his work at the time of the publication, but for 90%+ of our users this evidence wouldn't be sufficient. Also, consider John Sladek's The Sublimation World, originally published as by "J. G. B******". I am sure we can devise a paragraph that will cover these permutations, but it may not be easy to make it both comprehensive and clear. Ahasuerus 15:11, 30 April 2009 (UTC)

"Doc Smith" variants SQUARED + Serial reprint in Novel format

Apologies, this will have "clarity" issues. I am in pursuit of serial vs novel publications of authors like the "Doc". The problem is that there are huge variations in text content between some serials and the novel, not to mention possible revisions afterward.

I just purchased Skylark Three, 1-4346-7008-2 [14] from Amazon and first problem is I got a different cover than the "three" covers shown at Amazon. It gets worse.

Title page: Author is Edward Elmer Smith. {OUCH!}. Skylark Three (over) Sequel to the Skylark of Space (over) The Tale of the Galactic Cruise Which Ushered in Universal Civilization. "For two years readers of Amazing Stories have literally clamored for a sequel to the famous story, "The Skylark of Space" which appeared exactly two years ago. Except that "Skylark Three" is more thrilling, more exciting, and even more chockful of science than the other.(not comma) Dr. Smith tells about the story in his author's note far better than we can do." Illustrated by Wesso. BiblioBazaar.

Copyright page: Copyright data is they copyright it to themselves, BiblioBazaar, for 2007.

Actual novel text is exactly the same as the Pyramid edition May 1968.

Author's Note ends with Edward Elmer Smith, PhD. No date but seems current to 1930.

Epilogue, is not in Pyramid, and ends with this "I present to you my successor as Chief of the Galactic Council, Richard Ballinger Seaton, the fourteen hundred sixty-ninth of Earth" (over) The End. To me, the unhappy person, different ending, different story.

Some Remarks on the "Skylark Three" and About Errors. A Compliment to Dr. Smith's Stories. (over) Editor, Amazing Stories, with letter following. (Signed) John W. Campbell, Jr (over) Cambridge, Mass. Followed by an editor comment in parenthesis. Last line is "We hope to hear from Dr. Smith,-(long dash) Editor."

My summary is the novel printing is a reprint from the out of copyright Amazing Stories first publication. Including commentary, headers, author note and letter.

BIG Problem is this is all unstated, but quickly recognizable. This precedes the novel printing and though it is not significantly different in the novel text, it does predate the novel printing dates (though it is all unstated).

Do I enter under 1930 or 1948 version in the DB. How do you show the correct author name. Etc. I actually have a headache from this. Edward Elmer Smith obvious first author name used, then Edward Elmer Smith, Phd. from author's note, then Edward E. Smith (1948), then E.E. to "Doc" variations back/forth. None are true pseudonyms, they are only variations of one name with a title stuck into the name sequence or after it.

Here is the VBP as I see it. The out of date copyright reprints are free falling through the reader market today, they are high profit and will deluge the 'authorized' versions in a short time. To me a serial reprint novel is different.

I am also having connection problems to the DB. Thanks, Harry. --Dragoondelight 14:16, 28 April 2009 (UTC)

Just to make your headache worse, consider the Project Gutenberg pub (view the actual text from PG ebook # 21051). (This was a re-release in 2009, presumably fixing some error(s); the original release was in 2007. I wonder whether these clowns had the gall to take the PG version, print it, & then claim copyright for themselves!).
We really could use some kind of software change to allow distinguishing real variant works from mere slight variations in titles, & real pseudonyms from mere variations in form of authors' names. (This has recently (& confusingly) been discussed above, under Title date clarification.) I'd hate to have to design an adequate system myself.)
Faced with your pub, my own approach would be to enter the information as it appears, with copious notes documenting most of what you recount above. I'd then probably make it a variant of something for the sake of the form of the author's name.
Can't help with your connection issues, I'm afraid. (Not that I've helped much with anything else, either.) -- Dave (davecat) 15:35, 28 April 2009 (UTC)

It's probably best to make it a variant of the 1948 printing with good pub notes. Any other solution I can think of would also give me a headache. It's a common practice for companies to sell PG books - they show up often on Fictionwise in ebook format. Wildside Press also does print versions. About the only thing they can truly copyright is any additional content. It also works in the other direction. This title is actually scanned from a 2006 print edition with the extra material removed.--swfritter 17:43, 28 April 2009 (UTC)

Entered and awaiting my second pass. Thanks for all the input. Thanks, Harry. --Dragoondelight 23:01, 28 April 2009 (UTC)

Malzberg's Pseudonyms

Good old Barry N. dabbled in many genres, from straight sleaze (as Mel or M.L. Johnson) to a 14 book series Lone Wolf (as by Mike Barry) with at least four others that I can find of indeterminate content (Gerrold Watkins; Francine di Natale; Claudine Dumas; Lee W. Mason) plus the recognized (by this site) K.M. O'Donnell. All of these are listed in Currey, and though the Johnson books would introduce some very interesting covers, the sleaze can go elsewhere. I'm really baffled what and what not to add to the DB. The Barry-type series can be found for various authors and we include them. The rest probably don't belong. Any opinions?? ~Bill, --Bluesman 00:18, 4 May 2009 (UTC)

We're not particularly prudish here, or Andrew J. Offutt and others would have the "Spaceways" series deleted. But I don't see any particular reason to add non-SF sleaze even if someone is above "a certain threshold", as we put it. Author notes for the rest maybe? BLongley 00:42, 4 May 2009 (UTC)

Author notes where? --Bluesman 02:14, 4 May 2009 (UTC)

Bibliographic Comments: Author:Barry N. Malzberg . BLongley 17:48, 6 May 2009 (UTC)

And our biggest bibliographic question is still what does the "N" really stand for? BLongley 00:42, 4 May 2009 (UTC)

Nathaniel (Contento); Nathaniel (Wikipedia).--Bluesman 02:14, 4 May 2009 (UTC)

Norman, according to an author-completed questionnaire published in R. Reginald's Contemporary Science Fiction Authors II (1979). MHHutchins 03:21, 4 May 2009 (UTC)

Now that I am thinking about it, this may be a good argument in support of moving Non-genre series to their own area... Ahasuerus 01:18, 4 May 2009 (UTC)

And shift A. N. Roquelaure as well? I think those books may be here as "Fantasy" has a double-meaning, not just because Anne Rice is above the "certain threshold". BLongley 02:09, 4 May 2009 (UTC)

I believe all of Malzberg's non-genre work should be included because he's beyond the "certain threshold". If we include anything by Nicholas Sparks, surely everything by Malzberg should be included. MHHutchins 03:21, 4 May 2009 (UTC)

Have you read "I, Lesbian" ?? Me either, but there has to be a 'sub-threshold'....... Methinks we include far too much dreck already, but then that is taste as opposed to "bibliographic nicety". --Bluesman 04:46, 5 May 2009 (UTC)

Or "bibliographic completeness" perhaps. MHHutchins 18:58, 6 May 2009 (UTC)

I just read all the produce descriptions for his books and there are no ghosts, blood suckers, time travelling or anything else that would lead me to beleive any of his books belong here. Have I missed something and if not would you like me to delete the whole thing? I'm getting very good at the deletion thing, it gives me my fix for the day! :-)Kraang 03:45, 4 May 2009 (UTC)

Try "Tactics of Conquest" --Bluesman 04:46, 5 May 2009 (UTC)

I have no objection as I've found no spec-fic in any of his work. Anybody have an idea how his work got here in the first place? MHHutchins 16:24, 4 May 2009 (UTC)

The covers!!! Great Covers!!! Doesn't that trump content?? ;-) --Bluesman 04:46, 5 May 2009 (UTC)

Or perhaps to attract female editors???? MHHutchins 18:58, 6 May 2009 (UTC)

Just to clarify: I am all for including Malzberg's (and Silverberg's etc) non-genre fiction, but I am a little worried about displaying it at the top of the page. It seems to be a good argument in favor of creating a separate section for Non-Genre Series right before the regular Non-Genre Titles. Ahasuerus 04:09, 4 May 2009 (UTC)

Top or bottom, if the 'Section" is non-genre, does it matter? I do like that Non-Genre should be somewhere in the nether regions. --Bluesman 04:46, 5 May 2009 (UTC)

In the experimental changes I'm making, Nongenre Series will appear before Nongenre if a series with only non-genre and later types of titles exists. Series at the top will be only series having at least one Novel, Collection, or Omnibus (which is as it is today). BTW, I notice in testing that Malzberg has two series of Short Fiction that currently do not appear in his bibliographic display (Writer's Heaven and one entry in Riverworld). --MartyD 10:05, 4 May 2009 (UTC)

"Later types"??? Does this mean that writers devolve????? ;-) --Bluesman 04:46, 5 May 2009 (UTC)

I meant "types appearing farther down in the display". But perhaps it was a Freudian slip.... --MartyD 10:18, 5 May 2009 (UTC)

The change mentioned above has been deployed, and Malzberg's bibliography reflects the new Nongenre (and also Short) Series sections. --MartyD 10:12, 8 June 2009 (UTC)

I believe the A. N. Roquelaure works certainly postulate a society that never existed in our history. Kevin 01:02, 5 May 2009 (UTC)

So does the Constitution.....--Bluesman 04:46, 5 May 2009 (UTC)

Kevin, if it's mentally conceivable and it's physically possible, it's been done. MHHutchins 19:02, 6 May 2009 (UTC)

I wasn't referring to the acts of the novels, rather the socio-political framework that serves as the background for the story. Kevin

I guess it depends how big a "society" you are talking about. There are certainly Goreans, and enough of them that they can have splinter groups. BLongley 21:54, 6 May 2009 (UTC)

Essay Series - might need "Serial" instead?

How protective are people about SERIAL being for Shortfiction only? BLongley 22:42, 6 May 2009 (UTC)

Only there's some multi-part Essays that I can't put into a series as the series name is too short. For instance, "Neither the Beginning Nor the End of Structuralism, Post-Structuralism, Semiotics, or Deconstruction for Science Fiction Readers: An Introduction" has part 1 of 3, 2 of 3 and 3 of 3. If I try and create a series of that title, it gets cut off at "Neither the Beginning Nor the End of Structuralism, Post-Structu". It's a bit late to go beat Delany around the head with a wet haddock to get him to use less pretentious titles before they're published. I suppose we could abbreviate the series to "NtBNtoEoS, P-S, S,or DfSFR: An Intro" but I for one wouldn't really recall that easily. Suggestions please? (I've already got the one about taking a large aspirin and going to bed, which I shall do now.) BLongley 22:42, 6 May 2009 (UTC)

It certainly would not show up in the right section. How about "Neither the Beginning Nor the End . . ."? or "Neither the Beginning Nor the End (NYROSF 1989)". The titles for the three articles will still be visible beneath the series header.--swfritter 23:08, 6 May 2009 (UTC)

And what am I supposed to do about the four-part graphic story I will soon be entering? Serial will work I guess because there is an author and and artist.--swfritter 00:03, 8 May 2009 (UTC)

Webzine Proposal - Subterranean Online - And a possible Inclusion benchmark

I would like to propose Subterranean Online as the next webzine for inclusion in the database.

• Pro: Stable Site, based on a major independent SpecFic Publisher.
• Pro: Pays above market rates ($0.05+/word) suitable for qualification for membership in the SFWA per thier requirements. (Rates ($0.07+/word) determined from Issue 4 of the print version available as a download here, see the editorial on page 3). ++Possible Benchmark
• Pro: Evolved from a print operation to online, and has a physical history. See Magazine:Subterranean
• Pro: I will make PDF or JPG documentation of the 'Table of Contents' of the webzine and upload them to the ISFDB as 'permanent' documentation of the state of the webzien at the time of entry, and link them from the notes of the publication. ++Possible Standard to Document Webzines?

On an aside, I think that any webzine, paying market rates should be considered for inclusion. If a new author submitting an application to the SFWA, cannot point at an ISFDB bibliography as documentation of their published bibliography, are we 'serving' our user base properly? - Thoughts? Kevin 05:09, 15 May 2009 (UTC)

OK by me. Great idea to download 'Table of Contents'. I should do that for Clarkesworld. I'm all for the inclusion of Subterranean if you agree to keep an eye on Clarkesworld in case I cannot continuing to enter that title.--swfritter 15:55, 15 May 2009 (UTC)

This benchmark (paying market) was one of the criteria I mentioned in a previous discussion about webzine inclusion. Subterranean is quite professional and appears to be stable as well. I say it should be included. MHHutchins 17:09, 15 May 2009 (UTC)

I used the latest issue of Clarksworld as a Guinea Pig. I think this may be sufficient documentation for almost any webzine (non-downloadable). I used a Firefox addon, which lets you screengrab the entire page (even if it doesn't fit on your screen). Surprisingly enough, it's called 'Screengrab!' Thoughts? (Would you like me to do the rest of the Clarksworlds?)Kevin 23:52, 15 May 2009 (UTC)

It's basically free advertising so it's unlikely that a webzine owner would complain but it's possible there might be copyright problems; primarily if artwork is captured. There are other tools like snipping tool (free with vista and Windows 7 but no xp version) and snagit (not free) which allow you to capture part of a screen. All we need is the table of contents and perhaps something to indicate the issue #. Although, if we keep the data in a place where the average user cannot access it there probably won't be a problem no matter how much of the screen we snag. Oh, and by the way it is Clarkesworld - from the publisher's name. Even Bill Contento made the same mistake in his latest sf magazine cd. Since it's my responsibility I can do the screen grabs for Clarkesworld as soon as we figure out how much to capture and how to keep the captures organized.--swfritter 00:25, 16 May 2009 (UTC)

Ahhh - With an E. As to a lesser screenshot, I tried printing to PDF and got the text only version. Quite frankly it's ugly and doesn't do the webzine justice. Just taking a screengrab of the TOC frame is doable with the extension I've got, but again, it just feels like a job half done. With an entire page grab it provides all the same copyrightable stuff on the cover of most magazines in print (artwork, masthead, fonts, etc). On that basis, I feel pretty comfortable on fair use of the coverpage/frontpage as a whole. It provides context, attribution, etc and it's an undeniable record of the issue 'as it appeared' to any user on the archive date (just as our other magazine covers present the item for accurate identification). Kevin 00:55, 16 May 2009 (UTC)

I agree but I know there are others who worry more about that sort of thing than I do. We could also get permission from the webzine publishers not only to grab the front page but also link to artwork. Clarkesworld has a nice cover gallery.--swfritter 01:08, 16 May 2009 (UTC)

I might also note that I avoided contacting Clarkesworld about linking to artwork because the Clarkesworld project is still a test case. Kevin's idea of screen grabbing the table of contents seems to me to make the inclusion of webzines not only acceptable but desirable.--swfritter 14:46, 16 May 2009 (UTC)

I do feel that by properly 'documenting' the website, any webzine that defines it's releases as 'issues' (some sort of grouping of fiction/essay postings by date range?) is now elevated to the historical value level of a mimeographed fanzine, and should be considered for inclusion. Today's 'kids'/Younger fans don't photocopy newsletters... they set up a google groups email list, and put up a webpage. The same, love, excitement, and dedication to Fandom that used to smell of mimeograph fluid or toner, today smells like electrons. Kevin 15:03, 16 May 2009 (UTC)

It looks like three positive responses and no negatives. So Kevin may want to work up a proposed change to the Policy page. I have already saved off the contents pages for Clarkesworld. As soon as we have a methodology for storing them I will do so.--swfritter 17:23, 21 May 2009 (UTC)

I'll write something up tomorrow night. Kevin 04:10, 22 May 2009 (UTC)

I am still dubious about webzines -- I think the reasons we had for excluding them before still hold water. I think the disappearence of Helix merely shows that we should not have included that webzine. An ISSN does help. Some sort of commitment thjat the archivces would be maintained indefinately (or at leasst for the foreseeable future) would help more, IMO. -DES ^Talk 04:33, 22 May 2009 (UTC)

You probably missed the recent nightmarish Helix debacle and a lucky man you are. I certainly have reason to be even more dubious than you. The nature of webzines requires extra care in their entry. If we had fully documented Helix in the first place there wouldn't be a problem. Helix is in many ways an anomaly. There are a number of webzines which have been stable for many years. Our coverage of contemporary short fiction, both print and electronic, is woefully inadequate and I am hopeful we can attract enthusiastic editors like Kevin to help fill in the gaps. In the end it is going to be editor/moderator judgment and commitment which is going to determine whether a given website should be included.--swfritter 13:42, 22 May 2009 (UTC)

DES, compare and contrast Webzine and Fanzine. Please highlight the reasons why Webzines should be excluded and Fanzines should be included. Bonus question, If every last copy of an issue of a fanzine is destroyed or lost, does that mean it should not be included or even that it should be removed since it has now been 'un-published'? Kevin 03:15, 23 May 2009 (UTC)

It is a given that every Webzine will disappear forever at some time. There is always the possibility for anything physical to survive.--swfritter 13:36, 23 May 2009 (UTC)

I don't accept that premise. I propose that all pulp era magazines, will finish self destructing within another 100 years due to their physical construction (Barring extra-ordinary preservation attempts). Do we really want to limit our definition of 'in' based on the expected lifetime of the publication Medium? Should we perhaps limit ourselves to publications printed on archival quality paper only? Are all audio books to be excluded, because vinyl, cassette tape, and CD don't have the same expected lifetime as paper? Are all magazines to be excluded due to their flimsier construction? I really don't think the Medium matters anymore as an in or out criteria. Kevin 14:38, 23 May 2009 (UTC)

If my judgement were followed, fanzines would be excluded, and so would self-published works by authors with no professional credits. And it is possible for pulp-style magaziens to be preserved pretty much indefinately, and i kn ow at least one library whose special collections department contains lots of SF, both magazien and early paperback, and does take significant preservation steps. It is not so much the medium as the single-source nature of webzines that I object to. A non-dowenloadable webzine efectivly exists in a single copy, and when it's gone it is gone. Even if we fully docum,ent the bibliographic metadata, no one can ever comnpare the actual text with the text of any later published versions of its contents. Anyway those are my views. But having said that, i will yield to consensus, just as I did with fanzines, and self-published works. -DES ^Talk 14:57, 23 May 2009 (UTC)

I would definitely draw the line at twitterzines.--swfritter 15:07, 23 May 2009 (UTC)

I was about to agree (It's absurd right?) then I recalled there is a fad in Japan? recently for reading novels via text message or something similar... each message is a chapter. See Cell Phone Novel at wikipedia. It could happen. Kevin 03:05, 24 May 2009 (UTC)

UNINDENT I present for your critiquing judgment the first 4 issues of Subterranean Online - Click the ISFDB links from 2007. The first doesn't have coverart, and the second doesn't have coverart credited. I uploaded the three cover's to the Wiki, and I have also documented the Table of Contents (on the bibliographical notes wiki page) with screen shots of the TOC, which in this case also shows the coverart. No story text is shown or archived. A link to each issue is provided in the publication record, and from the WIki Bibliographical notes page. Thoughts, comments, and suggestions please. I would also like feedback on handling the audio books they began including. At the moment I added (Audio Download) to the title, and varianted it from the original story publication. We will probably run into this problem even more going forward... But coding for audio contents in a non-audio format publication is not going to be an easy fix. Kevin 03:05, 24 May 2009 (UTC)

I've received some very helpful comments and reminders of items, thank you all. I also created a Template:Table of Contents entered as {{Table of Contents|PUBTAG}} and an alias {{TOC|PUBTAG}} for use in documenting the front page or TOC of a webzine. As the abbreviated Cover template, I (we, you) can add it as a callout in the file comments when uploading, just replacing PUBTAG with the ISFDB Publication Tag. Comments, etc are also welcome on those talk pages, here, or elsewhere. Kevin 03:13, 26 May 2009 (UTC)

After reviewing the latest crop of Webzines, I am afraid I am forced to agree with the inclusionist camp. When I last checked, most Webzines were 1/4-professional at best and for the most part outside of the genre's mainstream. Now, however, just one randomly selected Webzine issue may include stories and articles by Charlie Stross, Elizabeth Bear, Mike Resnick, Lewis Shiner, Michael Bishop, Thomas M. Disch, Joe R. Lansdale and Neal Barret, Jr. If we are to exclude this issue, we will then have to add "First published online in ..." notes to these very Titles when they appear in other venues or, what is arguably worse, miss them completely if they are not reprinted.

Something else has changed in the Webzine world over the last decade too. I distinctly remember some of the early (1990s) Webzine issues as unstable, with articles added after the fact, which made it hard to catalog. Now, however, the leading ones seem to be as stable as any prozine -- at least as long as the Web site exists.

As far as inclusion criteria go, "stability" and "a paying market" seem reasonable. Generally, I am not a big fan of using "major awards" as a criterion, but I suppose we could use it as a stopgap measure until we come up with better ideas. Capturing tables of context is certainly a very good thing to do, now that we have a place to put them. Ahasuerus 00:14, 7 June 2009 (UTC)

See the discussion the discussion Kevin and I have been having in attempts to bring something to the table that we will think is acceptable. I think my wording is getting a little too doctrinaire. Kevin and I are pretty much in agreement on five other webzines other than the two we are already doing but I think a couple of the ones he would like to see included are just not up to standards. I hate the idea of committees for deciding which webzines are included but it might come down to that.--swfritter 00:39, 7 June 2009 (UTC)

Here is a spreadsheet with links to about 30 current webzines of widely varying quality. This does not include webzines that also have an ebook version. There are also 30 or more print or ezines that would also pretty much automatically qualify.--swfritter 01:05, 7 June 2009 (UTC)

As for short fiction in general. According to Mark Lewis of Locus the number of published stories increased from 2,109 in 2007 to 3,037 in 2008. As to where you are going to be reading them in future. Perhaps on one of these.--swfritter 13:56, 9 June 2009 (UTC)

(Complete Novel) in Webzines with Rule Hidden in 'Date'

The rule to enter whole novels (defined at items separately published, and/or the word count definition) in magazines as a Serial with the appended (Complete Novel) I think was created to document that 'Novels' in magazines often were re-edited for magazine publication, and for that same reason get their own date. (Inexplicably, this rule of entry for a type is documented only in the 'Date' area of the help, and not under novels, complete novels, or serials.) Three (longish) questions. Kevin 16:26, 25 May 2009 (UTC)

Should this rule apply to Novels (Separately published, and/or defined by word count) that appear in webzines? Webzines (due to the low low 'printing' costs) typically do not re-edit for length. If for some reason they do shorten the work, they typically seem to label it as an excerpt, which we already know how to handle. I propose the answer is no, this rule should not apply to electronic magazines. Thoughts? Kevin 16:26, 25 May 2009 (UTC)

Am I correct in understanding this rules origin? What purpose does it serve to add 'Complete Novel' beyond to note that this 'might' be a magazine edited variant? (Follow on question - Will this go away once we have multiple variant types in the future?) Kevin 16:26, 25 May 2009 (UTC)

Should we move (or duplicate) the instruction for '(Complete Novel)' to the title area, and the serial area of the help new pub instructions? Can we clean up the language some as well, it's rather ambiguous: "Serial installments of a work are always given the date of the magazine in which they appear even if the work has been published previously in book or serial form. Novel length works (40,000+ words) printed as a single installment in a magazine are treated as serials and given the date of the issue in which they appear" As written it equates to 'Serials get the new date', 'Even if in book form', and 'novels (by wordcount) get the new date too. Perhaps "Novels previously serialized, published as separate books, or that meet the word count definition all get the new date. If the novel in this publication is unserialized (i.e. appears in a single issue), then it still gets the 'SERIAL' type applied, and it gets the (Complete Novel) appended to the title" or something similar Kevin 16:26, 25 May 2009 (UTC)

Rules are made to be broken as long as the reason for doing so is fully documented and justified. I would have no problem with entering such a title as a NOVEL and merging it as long I was 99.347% sure it was the same animal. The entry of new (Complete Novel) serials is actually rather rare. Most of these appeared in the old super-size pulps which are mostly in the system already.--swfritter 14:49, 3 June 2009 (UTC)

Sorry I missed the question when it was first posted last week :-( The original reason for the "(Complete Novel)" rule was twofold. The first part was, as Kevin wrote, the fact that there are often significant textual differences between the magazine and the book versions of the same novel. The second part has to do with the way bibliographers and collectors traditionally entered book publications vs. magazine publications. If you open one of Clute's Encyclopedias, (i.e. of Science Fiction or of Fantasy), you'll see that he gives two dates for each serialized novel, one for the original serialization and one for the first book appearance. As far as I can tell, this tradition first arose in the book collecting world, where magazine publications were not considered "true first editions" due to the ephemeral nature of many early magazines and cheap pamphlets (think Eugene Sue's original serialization of The Wandering Jew in 70+ pamphlets). I am not sure how we want to apply this reasoning to the world of Web magazines, but it's something to consider. Personally, I am a magazine collector as much (if not more than) a book collector, so I have always been vaguely uneasy about this approach, but it seems to be well established in the field. Our current display logic, which shows both the first book appearance as well as all serializations of each novel, does a pretty good job of explaining what's going on. (Except for the lexical match issues, which Al was working on when he became unavailable, of course...) And yes, improving Help and making it easier to use is always a goof idea. Ahasuerus 21:28, 3 June 2009 (UTC)

Was that a Freudian slip? Searching for help in our Help pages can make you a little goofy! MHHutchins 23:37, 3 June 2009 (UTC)

If the publication of a novel-length work in any other publication (magazine, collection, anthology) is subsequent to the book publication and it's the same animal with few editorial changes I think a merge makes sense. The other way around wouldn't. I recently changed a merged novella to a Complete Novel serial and unmerged it giving the second printing the date of the issue in which it appeared. It was one of those Amazing Stories reprints and was very likely 100% a duplicated story and it probably would have made more sense to have left it merged with the other Complete Novel serial. A recent Zelazny collected works book had reprints of serial parts and they were merged with the appropriate magazine appearances since there were likely few textual changes. The standard in place probably makes sense more than 99% of the time. The biggest problems with treating exceptions in a different manner is that editors may not realize that they are exceptions and they may use them as examples when it is not appropriate.--swfritter 00:49, 4 June 2009 (UTC)

Unfortunately, determining whether "it's the same animal with few editorial changes" is not always a simple proposition. For example, Magazine:Famous Fantastic Mysteries was notorious for its severely abridged reprints, but some reprints may have appeared with very few cuts. I suspect that it's safer and easier to use the "(Complete Novel)" format and a separate publication date for all magazine reprints. Of course, the line between "magazines" and "anthologies" is not always as bright as we might want it to be -- see, e.g., Destinies -- but that's a different can of worms. Ahasuerus 01:19, 4 June 2009 (UTC)

Date standards

I had thought that a title date indicated the date of first publication, allowing for entry errors and imprecise secondary data. But in this comment, this followup, and this later followup from the "Magazine editing improvements" thread on User talk:Swfritter, Bill Longley says that a publication date, even one known to be long after the initial appearance, is preferable to 0000-00-00 as a title date, and that title dates should be interpreted as "no later than" the given date. I strongly disagree.

I was under the impression that 0000-00-00 means "unknown" and that if we truly do not know when a title was published, it is better to leave the title date saying so, rather than incorrectly stating a date on which the title is known to have been reprinted.

I have entered a number of reprint anthologies, some from library copies, some from OCLC data. In such cases I know the pub date of the anthology, and that the contents were initially published many years earlier, but sometimes have at hand no information on the original publications. When i don't, I always make a point of entering 0000-00-00 for the contents dates rather than letting them default to the pub date. Often a later merge or further research allows accurate dates to be specified. In the mean time neither users nor other editors are fooled into thinking that I had evidence of a date of initial publication. (Of course if the book indicates the date a story was first published, i enter that)

I recall suggesting that the ISFDB support date ranges, and in particular a "not later than" and "not earlier than" date setting, to deal with this, and with the case of a book that has a copyright date but no publication date. But until/unless such features are implemented, I urge that we adopt the standard that a title date means the date of first publication as accurately as it is know, and when it is not reliably known to the nearest year, 0000-00-00 should be used, rather than the date of a reprint publication.

Note that Help:Screen:EditTitle#Date says "The original publication date of this title" and "If you know that a book's first edition is prior to the date in the title, the best thing to do is find a copy of the first edition, and enter that publication. Then the title record can be updated to match the earlier publication. If you're unable to locate a copy of the earlier book, it is generally not a good idea to use a copyright date without further checking." (text which is also included in Help:Screen:EditPub#General contents.) This seems contrary to Bill's ideas quoted above.-DES ^Talk 20:21, 3 June 2009 (UTC)

Of course, I strongly disagree with DES. It seems clear to me that the fact that the software assigns a default date to contents based on the publication date means that that it is intentional that a known date is preferable to an unknown. If DES is fighting what the software does, it's clear that he disagrees with that, and it's right that he should bring it up here (as I asked). BLongley 21:35, 3 June 2009 (UTC)

0000-00-00 does indeed mean "unknown", but contents for a known-dated pub are either known to have been published on that date at least, or the editor may have found evidence of earlier publications and use those dates instead (preferably not just a copyright date though). Either way, we have primary or secondary evidence of a date. BLongley 21:35, 3 June 2009 (UTC)

If we go for 0000-00-00 as the date for contents of unknown providence then we'll have to adjust many titles where we have not yet firmly established a first printing date: e.g. one user is finally dating Edgar Allan Poe titles to exact dates in non-genre magazines, I've found some for H. G. Wells in a similar manner. I prefer we keep what we have as "best known date" in the meantime. BLongley 21:35, 3 June 2009 (UTC)

I also like the suggestion that ISFDB support date ranges like "not later than" and "not earlier than" date settings, but that's a big change that I'm not sure current developers and testers could do - yet! Still, we have time. BLongley 21:45, 3 June 2009 (UTC)

In response to Bill's comment about the software behavior, It seems clear to me thqt the edit feature of defaulting contents dates to the pub date was intended for original collections/anthologies, where all dates should be the pub date, and for introductions and other added matter that should defualt to the pub date. If this is not correct, then the help needs to be changed, but i think that the software exists for us, not us for the software. I would want an explciit statement by Al or Ahasuerus, or someone else involved in the design that this was a matter of design intention, not just something that fell out of the implementation, befoire accepting it as evidence of ISFDB policy. -DES ^Talk 22:01, 3 June 2009 (UTC)

I have no idea how you find it "clear" that the software behavior was intended for original collections/anthologies only. I find it clear that we're providing a definite date of a publication of a title. That will usually not be the first publication, but it helps when we're merging. Should all versions of a title be unknown till then? I see no benefit in that. Of course, we can ask Al or Ahasuerus if your perception matches theirs, but frankly, we're mostly on our own now and even if the intent was different, there's now several years of "I can live with that" from everyone else to overcome if we change something as big as this. BLongley 22:38, 3 June 2009 (UTC)

I don't see this as a change. i don't think anyone but you has been intentionally setting title dates to pub dates when the publication is known to be a reprint, or when there is no evidence that the publication is the first publication of the title. I see a fair number of titles with 0000 dates that have known pub dates, which I take as evidence that at least some editors aren't doing what you suggest. I see the help saying that title dates should be dates of first publication, and no hint that they can also be the date of any later publication in dafault of better data. I see no indication to the user that title dates have this alternate meaning that you suggest. I think that leaving the title date as "unknown" would be an indication that further research is needed. Note that the title record in any case displays the list of pubs, and each pub will (if possible) have a date of its own, so the first pub on the list gives the first known publication of the title, without falsely implying that we know the actual first publication date. From my point of view it is you who are proposing a major change -- I am simply proposing that we continue to have the title date field mean what our help and every discussion of the matter that I can recall always said that it meant.

As to why it seemed obvious to me what the software purpose was: such a purpose would mean theat the software was in line with the standards described in the help, rather than at odds with them, and the case of an original anthology or collection is a common one, as is the case of adding editorial matter to a collection when the fiction contents are reprinted, and will often have dates set by merges in any case.

To other active editors -- what has been your practice when entering a reprint of a title known to be a reprint when the original pub data is not at hand? What have you assumed when seeing a title date? That it was publihed on that date, or that it was published on or sometime before that date, posisbly decades before? -DES ^Talk 22:50, 3 June 2009 (UTC)

My practice for entering:

1. Single-author collections where I don't have dates for the individual stories: let the system default to the date of the pub. I've done this for hundreds of pubs for lesser known authors where stories may have appeared in obscure publications, or not, as I'd have no way of knowing where, if at all. This particularly applies to entries for collections published in the late 19th and early 20th century that I've entered using Tuck and Bleiler as a secondary source. After entering the new pub, I will merge any titles with titles already in the database.
2. Multi-author anthologies is the same, let the system default to the pub date. THEN I go back and merge stories with those records already in the database, choosing the earlier dates. If there is no current record to merge it with, I do my damnedest to find the original publication date, using various secondary sources, both in-print and on-line. If I can not locate the original date, I leave it as the pub date. In 74000+ edits, I've entered 0000-00-00 as a content date maybe fifty times. MHHutchins 23:29, 3 June 2009 (UTC)

Merging has probably fixed (or broken, to DES) a lot more dates than plain pub-entry. The help suggests "Generally select the more specific one or the earlier one, but checking may be needed." I don't recall ever keeping a 0000-00-00 date on a merge, though I've occasionally changed the merged title date further back afterwards according to research. BLongley 17:53, 4 June 2009 (UTC)

I'm with Michael and Bill. If I know an earlier date, I will enter it (And often enter the original pub if I am so inclined/informed). If I know of no earlier date but I suspect or know an earlier date exists but I do not know what that date is, I will allow the publication date of the collection/anthology/*zine 'stand' as better data than no data. The concept of 'Least Surprise' requires that we include some date (if any date is known) because the work should have been published in order to be indexed. A date of Unknown is Surprising at the title level 99.9% of the time. Kevin 00:36, 4 June 2009 (UTC)

We have known dates for pubs, but that does not always mean known dates for titles. I have seen many pubs with pub dates specified bu unknown dates for titles. An unknown date for a title is a big flag saying that research is needed on the title date, IMO. I am not suggesting going back and trying to change all dubious title dates to unknown, but I am suggesting NOT changing unknown dates to pub dates, particularly not automatically without human oversight, and i am suggesting that the standards for future entry be made clear, that a pub date should NOT be entered as a title date unless there is reason to believe that the pub is in fact the first publication of the title, and that 0000-00-00 should be used in place of the default (blank) contents item date when it is clear that the publication is a reprint. -DES ^Talk 22:01, 3 June 2009 (UTC)

As to why there are title dates of 0000-00-00 with known Pub dates under them, I think a lot of ANTHOLOGY and COLLECTION titles are that way because fixing the pub-date won't automatically fix the Container Title date. Forgetfulness, ignorance or laziness rather an intent to keep an Unknown over a best-known. BLongley 17:59, 4 June 2009 (UTC)

I fully understand that date range support is not coming right away, but IMO it should be on the list of long range desired features. -DES ^Talk 22:02, 3 June 2009 (UTC)

I think i have made my views pretty evident above, and i will try to refrain from further comment awaiting responses from editors other than Bill and myself. -DES ^Talk 22:55, 3 June 2009 (UTC)

I'm also in full support of a 'Not Later Than' and 'Not Earlier Than'. We could add halfway support for this (With preference for 'not later than') by simply adding a couple of columns to the title table, Boolean, 'Not Later' and 'Not Earlier', and have it as two input check boxes next to the date field in data entry. This only allows entry of 'one' part of the date range, but I believe that is sufficient for most of our unknowns with the least programming. Kevin 00:41, 4 June 2009 (UTC)

Originally, the software didn't default unspecified Title dates to the Publication date, which caused editor grumbling since editors would occasionally (well, OK, make it "often") forget to enter Title dates in original Anthologies and Magazines and then have to go back and correct the resulting 0000-00-00s. Al changed the default behavior to make our editors' life easier and not because he was making a policy statement.

Having said that, I think that there are good arguments on both sides here. I agree that the ultimate solution would be to change software behavior to allow (at least) "no later than", but that would require fairly significant changes to the display, storage and sorting logic.

My own practice has been mixed. If I find, e.g., an otherwise unknown-to-ISFDB story by a long deceased author which appeared in a 2007 anthology, I will make it 0000-00-00 since making it 2007-00-00 would suggest a recently surfaced "trunk" story and that's just misleading. On the other hand, one time I was researching an early 19th century German author and I couldn't find any editions of a particular novella prior to 1842 even though there was circumstantial evidence suggesting that it first appeared some time between 1838 and 1841. I think I ended up setting the date to 1842-00-00 and adding a Title level Note explaining my reasoning since I figured that 1842-00-00 was more useful than 0000-00-00 in this particular case. Ahasuerus 01:37, 4 June 2009 (UTC)

I agree with Ahasuerus. One thing I noticed during the course of the Poe work is that incorrect, non-0000-00-00 dates are harder to spot than "unknown". Dates 100+ years after the author's death are rather obvious candidates, but dates within his lifetime or just after his death are not. That said, having the title's date reflect the earliest known-to-us publication seems a reasonable compromise between nothing and full research. Not to open a rat hole, but just out of curiosity, what was the thinking behind not capturing copyright dates (presumably as a date separate from the first publication date)? --MartyD 10:23, 4 June 2009 (UTC)

The average database user can be easily confused when it comes to discerning the difference. Copyright dates are not publication dates, and especially so for publications at the end or beginning of the year. I've seen books published in December copyrighted with the following year, and books published in January with the previous year's copyright. Many paperback publishers (e.g. Popular Library, Ace in the 60s and 70s, etc.) will not have publication dates at all, only the year of copyright, even if it's a first edition or printing. Hell, even some hardcover publishers don't give publication dates (Putnam). As for having a separate field for the copyright date, it would only be useful as a fallback date on pubs without publication dates. As such, it would cause more confusion than it's worth, especially when it comes to displaying pubs chronologically. Titles can be copyrighted several times, sometimes for legitimate reasons (textual changes), but sometimes for no reason whatsoever. Baen and early Tor books were notorious for printing new copyright dates, whether the works had actually been newly copyrighted or not. I think the database designer (Al von Ruff?) was correct not to include a field for copyright data. MHHutchins 15:18, 4 June 2009 (UTC)

I am afraid you are vastly underestimating just how clueless we were when we began designing this beast in 1995, Michael. Suffice it to say that it took us some time to decide that Titles and Publications should be different entities and we didn't give dates much thought at first. The unspoken assumption was that the first publication date was always the same as the copyright date (what else could it possibly be??), which is why some fields in the database still say "copyright_date". But we got better over time :) Ahasuerus 15:44, 4 June 2009 (UTC)

It's the obligation of each generation of isfdb editors to repair the data of the previous generation. In return they are expected to leave subsequent generations with even more problems. Now that I am working on pre-20th century works I agree with Ahasuerus on dates. In many cases the appearance of a story in a durable book is of greater lasting significance than its initial appearance in a generally unavailable publication.--swfritter 18:16, 4 June 2009 (UTC)

I thought I would point out this discussion about Magazine Serial Dating which occurred a couple of years earlier. At that time it was decided that serials would be given date of magazine publication. I would suggest a modification of the date range idea, listing all known publication dates serially or maybe graphically as a distribution. This has the advantage of showing how writers go in and out of favor with time.--Rkihara 15:03, 4 June 2009 (UTC)

The worst example I have seen lately of a serial being different from the novel is Catherine Asaro's "The Quantum Rose". I am reading the Skolian empire novels and started reading the serial version in my collection thinking I could save some bucks. Turns out is essentially a second draft of the first half of the novel version. Luckily I was able to paperbackswap the book version.--swfritter 19:32, 4 June 2009 (UTC)

Server outages

Split from above thread to avoid instant topic drift. -DES ^Talk 21:17, 3 June 2009 (UTC)

(By the way, I this is my fourth attempt at making this entry. The first three were lost due to browser problems, mis-clicks, and ISFDB timeouts. The server has apparently been intermittently slow the past few days, in my experience. -DES ^Talk 20:21, 3 June 2009 (UTC))

The server has been in a strange mood lately, refusing connections whenever it feels like it. The problem usually persists for anywhere between a few seconds and a few minutes and then goes away. Firefox (and perhaps later versions of Internet Explorer) will remember your posts and allow you to resubmit them later if the connection is temporarily unavailable, but I always copy my posts to the clipboards before I hit "Submit" just in case. Ahasuerus 21:11, 3 June 2009 (UTC)

I have started doing that again. Worse is when the outage is on a db submit. Then, at least with IE, all is usually lost. -DES ^Talk 21:17, 3 June 2009 (UTC)

I seem to be having more problems when I first get into the system, especially if my first destination is a wiki page.--swfritter 21:51, 3 June 2009 (UTC)

Firefox is your friend. When the connection times out, you can reload and re-enter all submitted data. Kevin 00:42, 4 June 2009 (UTC)

Alan Moore and Other Graphical Monsters

Now that Fixer has learned how to create submissions by author, I asked him to do a couple of "easy" authors with exactly 10 "unknown-to-ISFDB" Amazon books to their credit. (The caveat is that these books have to be either in Amazon.com or Amazon UK, but not in both in order to avoid problems with intra-Amazon differences, which I am still working on.) It so happened that the first author that Fixer found was Alan Moore, a prominent comics guy with a large following, which probably explains why we have many of his comics (and a few legitimate covers, essays, etc) in the database.

My first reaction was to delete all of the comics, but then I remembered that we had had discussions about Gaiman and Moore (e.g. http://www.isfdb.org/wiki/index.php/User_talk:Rkihara#Common_Ectoids_of_Arizona) and at least some editors wanted them to stay because they "transcend the genre". However, our current coverage is very selective, so I guess we need to decide whether we let the rest of their work in (and define the rules for other comics folks) or delete their non-SF. Personally, I would very much prefer to delete them since I don't see a way to let some comics in without opening the floodgates. Ahasuerus 02:28, 13 May 2009 (UTC)

His name was on my list of authors(along with Elf Quest) to be cleaned up or deleted. At the moment he is linked to Gaiman and several others of the comic and graphic novel genre who have some real spec-fic work. If an author has a large body of spec-fic work we should include the non-genre, but in most of these cases the non-genre far outweighs the genre work. As far as I'm concerned he and the rest that don't meet a certain threshold should be deleted. Keep the spec-fic get rid of the rest.Kraang 03:31, 13 May 2009 (UTC)

I think it's dependent on the individual work. (Note, Moore did do at least one novel Voice of the Fire, which was shortlisted for an award). One example of a comic/graphic novel that is 'in' in my book is Kevin J. Anderson's Veiled Alliances - which was released in conjunction with Anderson's 7 volume series Saga of the Seven Suns. Once I accept that some are in, then any thing published in book form becomes at least worthy of consideration. I've been working on Anderson's bibliography for a while now, and there are 6-8 other titles I would like to add, but they are excluded as being graphic novels. (I put Veiled Alliances in because it was directly tied to a major work) Kevin 03:43, 13 May 2009 (UTC)

Moore's writing has filled comics, graphic novels, movies based on the comics/novels, and then novelizations of the movies. Should we exclude the short story, because it was published with lots of pictures? Most graphic novels at least qualify as a piece of short fiction. One of Moore's works that you propose to delete is V for Vendetta. I think it meets all the relevant criteria of SF. Kevin 03:43, 13 May 2009 (UTC)

The only reason I'd keep V for Vendetta[15] is because of the review in a spec-fic magazine, if he produced lets say a half dozen real novels it's no reason to keep the hundred other things. Remove the rest. I agree with you on Anderson only because of his large body of spec-fic work. The vast majority of authors don't qualify for this honor.Kraang 11:20, 13 May 2009 (UTC)

I think we're doing quite well at keeping Moore's comics out. We seem to be only allowing graphic novels. So we're not flooded with 10 comics for "V for Vendetta" (or more if you include the original British comic strips) and 12 for "Watchmen" and scores of "Saga of the Swamp Thing"s. We don't even have the collected Swamp Things that look like Graphic Novels but aren't complete in themselves, just being several consecutive comics in an ongoing series collected together. I can live with the Moore's we have already - they're spec-fic, and if you're going to include items because they're reviewed then we should probably include the things that won awards we cover, which brings a few more in. (I would fix the artist as co-author problem though. What next, Inkers and Letterers?) I think Alex Raymond is probably more out (as an author) than Moore and Gaiman. BLongley 17:51, 13 May 2009 (UTC)

We can't even know what things have been reviewed. I know I saw a post somewhere that reviews of graphic novels aren't being entered in all the magazines. So then I don't think we can use 'it was reviewed' as the only inclusion trigger. Kevin 22:25, 13 May 2009 (UTC)

That's right. We had this discussion a while back and decided that only books and magazines -- as opposed to games, movies, comics, etc -- get a separate REVIEW record. Other types of reviewed material are entered as ESSAYS (see the Reviews section of Help:Screen:NewPub). Ahasuerus 03:01, 14 May 2009 (UTC)

Actually only fanzines are listed as essay entries in the New Pub help. See Here for discussion of SF related music reviews being omitted, and here for discussion of graphic novel reviews being omitted. Those were just the two easy ones to find. I'm sure there are others because I said in the one that i had found the old discussions somewhere. Kevin 03:27, 14 May 2009 (UTC)

Sorry, I was a little unclear/confused. What I meant to say was that articles that review games, comics, movies etc are entered as Essays at the article level, but the games, comics, movies, etc that are reviewed do not get individual Review records. Ahasuerus 03:45, 14 May 2009 (UTC)

I think I understood you... My point was that we have absolutely no text record (in the index or in notes, or in the wiki) of many items reviewed in verified magazines, if the item reviewed was considered out at the time of verification based on the form of the work (Graphic Novel, Filk Songbook, Filk CD/Cassette, Convention Ephemera, etc). (shrug). From one point of view, it's a travesty (If the in/out rules change, then all those verifications (and which ones?) need to be revisited at some point). From another point of view it's noise in the database when its present. I just want to be absolutely clear that we understand what isn't in the database (and it is correctly excluded per the Help:NewPub as written today). Since those reviews have been excluded (even in essay notation) there is no way to document, even with an SQL search of the all fields in the DB if a review ever existed. I'm not saying it's great, I'm not saying it's horrible, I'm just saying that's the way it is. My point in all of this was simply that Reviews are going to remain incomplete, so they should not be the only trigger for inclusion. (and that ladies and gentlemen is me getting stuck on a tangent - Try the veal, I'm here all week.) Cheers - Kevin 04:27, 14 May 2009 (UTC)

(Unindent) Just to clarify - of the submissions currently on the queue I'd reject the "From Hell" issues, I'd only want the complete work. I'd reject the "Swamp Thing"s as part-works of an open-ended series. I don't know the "Miracleman" titles but suspect they're out, it looks like an ongoing series. The only one I'd keep is the "V for Vendetta (New Edition)" as it's a complete spec-fic work and it did do well in the 1991 Prometheus Award as a Novel (link needs doing). So I'd still mostly reject. BLongley 20:16, 13 May 2009 (UTC)

oO far I have seen 'Reviewed', 'Linked to another work', 'Written by a SpecFic author (above a certain threshold)', and 'Nominated for a SpecFic award'. Are there any other possible inclusion triggers? Do any of these triggers strike anyone as insanely promiscuous (opening the floodgates)? Kevin 22:25, 13 May 2009 (UTC)

If we lived in a world of only two choices selections would be much easier, but we don't. For the last two months I've been working in this grey area and have no problem with leaving somethings in if they fall within a certain range for inclusion(triggers). Just because Alan Moore is popular or the flavour of the day doesn't mean his other graphic novels should be included. If we include his why not everyone else's who has one short story and fifty comics, graphic novels and rpg works? I don't see the floodgates opening if somethings are included because of "triggers".Kraang 22:58, 13 May 2009 (UTC)

How do people feel about Buck Rogers in the 25th Century: The Complete Newspaper Dailies? (And I don't mean the questionable author.) As it's pretty prominently listed on our home-page as coming soon, and I think it's out. BLongley 18:57, 15 May 2009 (UTC)

I think it's sitting somewhere near the line, but on the inside of the line. (Pro - Historical, authoritative, plot/story arc based; Con - Newspaper daily (a step below even comic books)) But then again, The Collected Works of Buck Rogers in the 25th Century (collected comic strip) appears to have sat nicely in the database for 2 years or so not bothering anyone. Kevin 21:59, 15 May 2009 (UTC)

More grist for the mill... I would like to see a consensus, instead of a 'wandering away'. Perhaps we should revisit this on Rules and Standards. I found volume two of this set of Harlan Ellison's Graphic Novel adaptations reviewed today. [Harlan Ellison's Corridor Vol 1] and [Harlan Ellison's Corridor Vol 2]. What else have we ignored over the years because the colorful pictures scared us away? Kevin 06:07, 17 May 2009 (UTC)

So far I see Bill and Kevin arguing in favor of conditional inclusion of graphic novels while Kraang and I are against it. Anyone else want to chime in? Ahasuerus 20:52, 4 June 2009 (UTC)

Marginally in for historical reasons for Buck Rogers as per original discussion. Isn't s-f also known as "that Buck Rogers stuff"? Buck Rogers is the evolutionary link from genre s-f to popular s-f. Without Buck Rogers there would be no Star Wars. As a derivative creation Flash Gordon would be out.--swfritter 13:43, 5 June 2009 (UTC)

(unindent) I'm not absolutely sure what is meant by "graphic novel" or more exactly where the line is being drawn between GNs and comics, on the one hand, and novels with a fair number of illustrations, on the other. i recall that the original publication of Niven's The Patchwork Girl was called a "graphic novel" That had one full-page illo to every 4-6 pages of text, and has been republished without the art and stands up just fine, thank you. In fact i would use that as a test -- if a work is a viable work of fiction with all the art deleted, it should be IN, and if not, if the art carries the essential story, it should be OUT. But I don't feel strongly about the matter. Of course once we allow such things in we have even more problems with merges "is the art really the same?" questions, such as have lead to the IMO unwise reluctance to merge coverart records. -DES ^Talk 22:07, 4 June 2009 (UTC)

Perhaps a threshold based upon authorship by an s-f writer who has a significant number of entries? General inclusion would require a vary elaborate set of criteria.--swfritter 13:43, 5 June 2009 (UTC)

I prefer general exclusion of graphic novels, with exceptions for a) those we have recorded awards for (desirable, IMO, to explain the award a bit better), and b) a graphic novel version of a title we already have (livable with rather than desirable, IMO). I'd also like to continue total exclusion of comics (pictorial story magazines rather than books), and preferably of collections of comics unless the collection is a graphic novel in itself, e.g. V for Vendetta mentioned above would be IN as a complete work that won an award, Collected Volumes of issues of Saga of the Swamp Thing would stay OUT as they're just four comics bound together. I see there is some support for collected Comic Strips so there may be a c) "above a certain threshold of importance" although I can't think of one I'd propose, but I can live with others supporting a few like that on a case-by-case basis. BLongley 18:09, 5 June 2009 (UTC)

Note that I consider passing the "the work is a viable work of fiction with all the art deleted" test as a sign of being an includable work anyway. We have such titles as Eric where the original edition might look more like a Josh Kirby artbook than a Pratchett novel(la?), but that survived into a less illustrated version, and I'm pretty sure the audiobook didn't have much art. ;-) I've had no qualms merging those despite different artwork. (Actually, I've also had no problems with Unabridged and Abridged versions of audiobooks being merged, but that's a separate discussion and might need Fixer to learn something new.) BLongley 18:09, 5 June 2009 (UTC)

The "viable work of fiction with all the art deleted" standard seems to be reasonable. By the way, there may be some confusion re: what the term usually describes. The way it's defined by Wikipedia, "A graphic novel is a type of comic book, usually with a lengthy and complex storyline similar to those of novels. The term also encompasses comic short story anthologies, and in some cases bound collections of previously published comic book series (more commonly referred to as trade paperbacks)." As Kurt Busiek recently put it, "It's tough, it is, walking that line between being a debased comic book writer, unloved and sneered at, and being a graphic novelist, the new hip thing and wellspring of the future. If I could only figure out the difference..."

This is not the same as the "lavishly illustrated novel" phenomenon of the 1980s (as practiced mostly by Tor), which remained a work of prose fiction, which could stand on its own even without all the extra illos. Granted, some authors (e.g. Gordon Dickson) spent a lot of time working with the artist and making sure that the art was as faithful to the book's spirit as possible, but it's still not the same as a comic book, er, I mean a "graphic novel". Ahasuerus 00:01, 6 June 2009 (UTC)

The only difference I can see between a comic and graphic novel is the way it's printed. The comic is in a magazine format and the graphic novel in a paperback or hardcover format other than that there the same beast. Would it be wrong to assume the general consensus is to delete the offending publications with the exception of reviews(with note about inclusion), citations(with note about inclusion) and if the author has a much larger body of spec-fic work(with note about inclusion)? As for illustrated novels they should stay. The illustrations are not the main part of the novel just an added piece to make the it more interesting.Kraang 00:54, 6 June 2009 (UTC)

I don't agree with all this 'Note about inclusion' stuff as a requirement, perhaps a good idea, but not a requirement. Do we require a 'note about inclusion' for non-genre work of a major author? Do we require a 'note about inclusion' for Downloadable electronic publications? If we can agree on standards, then we put the new standards on the ROA. We shouldn't have to make a note about every item that someone thinks is questionable. We should be able to define standards that are objective against which most questionable pubs can be compared. A note should only be required when an item is literally straddling the in/out line. Kevin 01:20, 6 June 2009 (UTC)

We already have a non-genre category, no note required. I agree with you about standards but not all people read the rules of inclusion, so a short note for some authors works does not strike me as unreasonable. I've been adding these as I work my way through the non-genre SF 2004 and it's no extra work.Kraang 01:51, 6 June 2009 (UTC)

(unindent)Let's see if we can agree on something here. I suspect that there may be two categories that we have consensus on. First, if we define "illustrated novels" as novels which have lots of illustrations in them, but which remain effectively unchanged if all illustrations are removed, I think we will all agree that they should be "in".

Second, there is a new kid in town by the name of "light novels" -- see this Wikipedia article. Some of them have an unusual smallish format: 10.5 cm by 15 cm whereas mass market paperbacks are 11.5 cm by 18 cm. They don't rely on illustrations to tell the story -- although many are related to manga in some fashion -- but they tend to have short sentences and short paragraphs, making them somewhat similar to manga/comics. I think we will all agree that they should be "in" if their subject matter is SF, but we may want to make it explicit in the Rules of Acquisition. Ahasuerus 04:02, 8 June 2009 (UTC)

I agree with Ahasuerus that both of the categories he describes should be IN. (Note that the term "graphic novel" has sometimes been used for what he calls "illustrated novels", mostly before the current meaning of graphic novel" had solidified. This could cause confusion.) -DES ^Talk 15:09, 8 June 2009 (UTC)

That's exactly why I am trying to make this distinction explicit :) Ahasuerus 20:18, 9 June 2009 (UTC)

I don't see the term "light novel" being used much in English, and the description sounds rather as if the Japanese are to some extent reinventing the classic children's novel. Since we already include SF aimed at very young readers indeed, I don't see the need for a clarification to the RoA for "light novels" but I see no harm in one either. -DES ^Talk 15:09, 8 June 2009 (UTC)

(unindent)Here is another possible compromise solution based on the discussion above. My two main objections to including comics (either au naturel or in their "graphic novel" disguise) have always been that:

• we would need to capture a lot more data elements -- like "penciller", "inker", "colorist", "letterer", etc -- to do this medium justice
• calling comic books "novels" would mislead our users into thinking that they are regular prose fiction. Few things are as irritating as jumping through hoops to buy an obscure Ron Goulart title listed as "fiction" in a catalog only to find out that it's really a graphic novel.

I am still a little apprehensive about the first concern, but I think the second one could be addressed by putting comic books under Non-fiction, which we already do for books that collect illustrations, covers and other forms of art. Granted, it's not a perfect solution since comic books are fictional, but it seems to be a lesser evil than listing them as Novels. We could further restrict eligibility by agreeing to include, e.g., only "major", "fiction award-winning", etc works, -- a set of criteria which I find inherently arbitrary, but can live with as long as we call them Non-fiction. What do you think? Ahasuerus 20:18, 9 June 2009 (UTC)

Comic book enthusiasts do have their own datebase. I do not know their criteria for inclusion but we are justified in excluding most of what they include.--swfritter 23:15, 9 June 2009 (UTC)

Wow, they have improved quite a bit since I last checked them out! Lots of additional features, e.g. here. It's good to see that they are doing such an impressive job, which, to me, suggests that there is even less need to overlap with their field. Hm, we already have a link to Authors' IMDB pages, why not add a link to their Comicbookdb page?.. Ahasuerus 23:28, 9 June 2009 (UTC)

You guys keep thinking inside the box. (Though I like the Author ComicDB link). The development gloves are off. There is nothing but our own hesitation preventing us from adding 'Graphic Novel' as a new type of Fiction (Alongside Novel, Novella, Novelette, SS, etc). There is also nothing stopping us from adding new classes of Authors if we decide to go that route except time and effort. Do I want to work on 'The Incredible Hulk Saves Cleveland' - Not really. Then again I also don't participate much in entering Fantasy works either. All we have to do is decide to do it, and make it happen. All we have to do to let GN's in and keep 100,000 comics out is define Graphic Novel (Could it be as simple as Page Count?). You say there's a ComicDB - Looks great (Hey They even have 'Variant Cover' support - Woot!). They are set up for the complexities of multiple jobs for making a GN. Then perhaps we could limit ourselves to the 'Writers' and perhaps the cover artists once we define a Graphic Novel. Maybe we could even link to their record for people who want more information. Just because we can Say "It's already out there", doesn't mean we can't catalog the part we care about. Kevin 23:37, 9 June 2009 (UTC)

We have enough on our plate to keep us busy for the foreseeable future. Ahasuerus is right there are other sites that handle these genres/formats. The one thing I've noticed about comics and graphic novels is authors listed on many sites(Abebooks, Amazon, etc.) can't be trusted, may pubs are identified to Moore for example and when you dig into it you find it's based on something by Moore. We should stay with what we know and leave the comics and GN's for others.Kraang 01:30, 10 June 2009 (UTC)

It's not a Genre. It's a format. Kevin 03:05, 10 June 2009 (UTC)

Well, it's certainly true that no one is proposing that we start cataloging non-SF comics, so the disagreement is over the medium/format rather than over the genre. A brief overview of how we handle other media shows that we have decided to include some newer media like audio recordings and e-books (which seemed close enough to traditional books) while excluding other media like movies and TV shows (which add a lot of other data elements.) SF art is generally "in" either as Cover Art, Interior Art or Non-fiction books although original (i.e. original) paintings are out. Comics/manga are a thorny borderline area since they can come close to written fiction in some cases, yet generally belong to a different world with a different set of data elements. Ahasuerus 03:39, 10 June 2009 (UTC)

LibriVox in?

Based upon the same criteria as Project Gutenberg with similar entry standards? Most of their stuff ends up at Project Gutenberg but in a somewhat untimely manner and not in zip files. There are some great items that are not at Project Gutenberg.--swfritter 14:14, 5 June 2009 (UTC)

Looks to me as if they should be IN on the same terms as any audiobook publication would be. -DES ^Talk 15:26, 5 June 2009 (UTC)

I take it that LibriVox versions are Internet-only? I suppose that from the policy standpoint it's a cross between audiobooks (which have a physical existence) and PG. Nothing wrong with that and they are archived by www.archive.org, so I guess they should be "in". By the way, at least some of their TOCs are available in XML format, e.g. here, so we may be able to automate the submission process using our Web API. Ahasuerus 16:11, 5 June 2009 (UTC)

Dowloadable mp3 and ogg vorbis formats plus playable from the internet. It doesn't look there are genre designations in the XML and matching with ISFDB entries might be a problem since, like PG, they often use the author's legal name. It seemed to me that they were pretty much automatically in but wanted to be sure. Without the foundations laid by all those who worked on the PG and audiobook standards that might not be the case.--swfritter 16:26, 5 June 2009 (UTC)

I might note that the Project Gutenberg data is also available in XML. See the bottom of this page. An ongoing problem though is that many of the titles are short stories which would fit most appropriately in a CHAPTERBOOK-like container.--swfritter 21:46, 5 June 2009 (UTC)

Hmmm - Internet Archive has a nifty 'embed' player in html. On a Lark I added the embedded player (without explanation of what it is) to Plague Ship. I admit no matter what it needs some explanation if we were to leave it there, and I'm not sure how often we would want to do this, but take a look at the title page. I'll remove the embedded player later tonight... this was just a test for comment. Kevin 22:17, 5 June 2009 (UTC)

Shockwave blocked here, by my choice. I can't say I'm keen on the example, we already allow far too much unchecked HTML, IMO. But it's a coding nightmare to ban it all everywhere we already allow it. BLongley 22:33, 5 June 2009 (UTC)

"Noscript" blocked it for me as well. Ahasuerus 23:06, 5 June 2009 (UTC)

I just thought it was a fantastic example of what we could do on this type of Internet Resource. Kevin 22:58, 5 June 2009 (UTC)

Additional functionality is always a good thing, but embedded Shockwave links to another site which we have no control over strike me as a potential security issue. (It's amazing how many people are running on unpatched Windows/IE out there and not just the clueless newbies either.) If we decide to implement them, we may want to give our users more control over whether they want to follow the links. Ahasuerus 23:06, 5 June 2009 (UTC)

A clickable link is one thing, embedded html that serves content unwarned is another. i would strongly oppose putting such "links" in our db records. Not that it appears to dispaly anything for me, and I have sound turned off normally. -DES ^Talk 23:39, 5 June 2009 (UTC)

Embedded link player removed. As I said before, it was simply a test for comment, and to see if it would work. Someday in the future when we have gobs and gobs of space, I would love to start actually hosting some things here at the ISFDB, but that's more on the 5+ years horizon. Kevin 01:12, 6 June 2009 (UTC)

ISFDB 4 maybe? What are we calling the new version? ISFDB 2.5? Troglodyte that I am I not even fond of embedded HTML beyond "<br>" in the notes although I would like to see a few more fields specifically designed for web links.--swfritter 14:15, 6 June 2009 (UTC)

We are still running version 2. The big change between 1 and 2 was rewriting the application in Python and MySQL and that hasn't changed, although we may need to do something about CGI to improve performance. As far as HTML goes, there are various open source rich text editors out there, but they are typically heavy on Javascript, which opens a different can of worms. Ahasuerus 14:47, 6 June 2009 (UTC)

I don't object to embedded html for bulleted lists and links, those are pretty straightforward and every browser supprots them. Note that to display HTML on a wiki page you msut enxloe it in a nowiki on/off tag pair. Use of a code tag is not a bad idea if the html is more than a single tag. -DES ^Talk 21:45, 6 June 2009 (UTC)

I don't see very many types of HTML in notes. Over 95% of people seem content with bold, italics, ordered and ordered lists, and an occasional anchor reference. The only one I recall causing severe problems to our software is the broken end comment tag on some submissions where Magazine navbars weren't copied correctly. The most likely danger is that somebody will find it amusing to submit links to offensive images, and they can do that in the coverart anyway. So it's the mods that would be affected first - have any of us been TubGirled or goatsed yet? I haven't here, but have elsewhere and it's not pleasant. BLongley 23:43, 6 June 2009 (UTC)

Ebooks have ISBN's for each format

Morning! I just contacted Martin Folkes at Mushroom Books, through a Yahoo group. I wish to enter the ebooks that meet the DB standards. Here is the discussion; RE: [Kregen] Mushroom ebook catalog number or ISBN

Hi Harry,

All the Mushroom eBooks editions have a seperate and unique ISBN for each individually available format, as stipulated by the ISBN Agency. Each paperback and hardback edition of the Cycles also has an individual and unique ISBN.

There is a complete list of Mushroom Publishing ISBNs at:

[16](sorted by title)

[17](sorted by ISBN)

Alternatively, each individual book page on mushroom-ebooks.com has ISBNs listed for each ebook edition.

Thanks for doing this.

Let me know if you need any further information.

Martyn

Martyn Folkes Mushroom eBooks www.mushroom-ebooks.com

The question is how to precede. I am pushed in the ebook format because the Dray Pescot series is only in ebook format after #37. Thanks, Harry. --Dragoondelight 15:28, 7 June 2009 (UTC)

Many publishers use the same ISBN for all formats of a given ebook -- I think fictionwise does so, for example. Baen generally uses the same ISBN for a paper publication and an ebook (all formats), PG of course, does not assign ISBNs at all. But if Mushroom assigns separate ISBNs for separate formats, then i would favor entering them as separate publications -- at least if you are willing to do the extra work. Alternatively, we could simply note all the ISBNs but one in the notes, but IMO that is less than optimal. -DES ^Talk 15:42, 7 June 2009 (UTC)

• First, The stipulation that each electronic format have a separate ISBN is not really a stipulation of an agency, it's part of an ISO Standard If I Remember Correctly; and the proper interpretation of that standard says that (paraphrased) 'Each publication available separaetly should have an ISBN.' In other words, if you sell a 'multi-format' ebook, and one purchase gets you access to all 14 formats, then you only need a single ISBN for the electronic editions as a whole. Publishers that sell encrypted ebooks have sometimes been putting different ISBNs on each encrypted format (since you only buy access toa single format). Kevin 15:55, 7 June 2009 (UTC)
• Second, There is no 'enforcement' agency for ISBN's; the requirement for an individual SKU/ISBN/EAC Barcode is a requirement of the retail industry to list your book in the various supply systems... (In other words, no ISBN, = No sales at Wal-mart). Kevin 15:55, 7 June 2009 (UTC)
• One key thing to remember is that in the US, ISBNs cost money, and in blocks of less than 1000 it can cost more than $30.00 each IIRC. ISBN's in Canada are Free, so Canadian ebook publishers might be a little more 'free' with the ISBN assignment. Kevin 15:55, 7 June 2009 (UTC)
• Now that I've set aside a few mis-perceptions... Have at it. If you want to input separate publications with separate ISBN's I believe we have discussed this before. Excepting the dual US/Canada ISBN publications, ebooks with different ISBNs assigned to different formats can be entered as separate publications. In all honesty though, Except for making the store links work, and creating extra work, there isn't much point at this time. The Store links only work for ebooks at Powells and maybe one or two other links. With all the development work going on, we may have the ability to assign multiple ISBNs to a single Pub record in a few months or later this year. That might be a better solution than separate pub records for unencrypted ebook formats. Just a thought. Kevin 15:55, 7 June 2009 (UTC)

As I was saying while two people were saving and some of this might be redundant. From my experience it would seem that a lot of publishers are ignoring this standard or it may be a developing standard that has recently undergone some changes. Some publishers use the ISBN from the physical version and I have downloaded different versions of multi-format ebooks that have the same ISBN. About all you can do is enter the ISBN of the pub you purchase and state the format. The question is whether to enter ebook (PDF), assuming it is the PDF version, as the binding or you could enter ebook as the binding and state that it is the ISBN for the PDF (or other type) in the notes. My own opinion would be ebook (PDF). If purchasing the book allows you to download all version types you could really get ambitious and enter them all. I don't expect you to do that but you might want to mention in the notes for each pub that it is available in multiple formats each with a separate ISBN. Oh well, I guess this is no worse than zillion printings we are obligated to enter for a given pub. Thanks for the info.--swfritter 15:58, 7 June 2009 (UTC)

I quite agree that the "standard" is not generally adhered to. Inded some publishers retain the same ISBN for revides, repritned, and price-changed printed books, others do not.

As to the bindign field, until now we have been entereing all ebooks as simply "ebook" with the format or formats in the notes, and that is what the help currently calls for. I would continue such practice. -DES ^Talk 18:38, 7 June 2009 (UTC)

I submitted an add pub for 'Transit to Scorpio' with the general information available from Amazon, Fictionwise and Mushroom eBooks. Please check. I plan to enter the others up to the ones I actually have copies of. Hopefully this will keep the information at book level in case there are changes to how the DB wishes to display ebook or their variants. Thanks greatly, Harry. --Dragoondelight 20:02, 7 June 2009 (UTC)

This is now 291203 for anyone to look at. -DES ^Talk 22:21, 7 June 2009 (UTC)

Reading Time

291203 list an estimated "Reading time" as "Reading 186-260 min." in the pub notes. i think this is unwise. People read at widely different speeds, so this can not be an objective measurement. List the fiel size, as this pub does. If a word count is available, listing that in notes is fine IMO. But not a "reading time" unless it is supplied by the publisher, and noted as such. Do others agree on this point? -DES ^Talk 22:22, 7 June 2009 (UTC)

See the listing on fictionwise (near the ratings bars). I think the site being used is a partner site of fictionwise.--swfritter 22:30, 7 June 2009 (UTC)

I think it is supplied by the publisher, or retailer, but is meaningless and I wouldn't support it even in notes. In an audiobook where you have to listen to someone for that amount of time (or fast-forward) it's probably worth recording. Not really my area though, as I still prefer dead-tree stuff. BLongley 23:00, 7 June 2009 (UTC)

I see no objection to including it in the pub notes. It's obviously an estimate, just as the number of words in a story or novel is an estimate. Are we here to protect the self-esteem of those who would feel inadequate if they don't read that fast, or to booster that of those who would laugh at such slow readers? It's part of the info provided by the publisher. Maybe someone has determined the average reading speed and that forms the basis of their estimate. It's not like someone has placed a note that "it took me three weeks to read this book". If we're going to start policing the notes, we have to set up some rules. I've seen some that would make your head spin. And after all, it's only an Alan Burt Akers book. Ha! Ha! (He continues to laugh as he stumbles off-stage.) MHHutchins 22:10, 10 June 2009 (UTC)

Since it can not be an objective measure, i would at least want it clearly stated that this is "part of the info provided by the publisher" or whatever the source is. -DES ^Talk 22:18, 10 June 2009 (UTC)

Sounds like a reasonable request. MHHutchins 23:31, 10 June 2009 (UTC)

As long as it is supplied by the publisher and noted as such, it seems like a harmless addition. Ahasuerus 01:20, 11 June 2009 (UTC)

In a sense it is no different than giving a page count, which is often unavailable for ebooks. They all came from the Fictionwise as the kindle information and kb size is from Amazon. Obviously kindle advertises Amazon. Do you really want a little blurb to advertise them in preference to half a dozen other vendors. Reading time, page count and size are all caveats to reader concerns. As a note, I totally do not understand the concern, while I attempted to both serve without promoting vendor sources. If readily available reading time and size may just get one person to read something and therefore is worth it, but it is not worth trampelling the earth to find. I want people to pick up or tune in and 'read'. For some reason, the reading time appeared on the search title result page, but does not on some others. I imagine that most people will not be offended and possibly some will be challenged. The 'words' entry serves much the same purpose, but may have some utility as a history entry. Let it go DES. Thanks, Harry. --Dragoondelight 10:43, 11 June 2009 (UTC)

As for the source, some is publisher, some the primary source recommended by the publisher, Fictionwise, and some from Amazon, a deeply imbeded force in the field. I frankly felt more chagrin at touting a Kindle ASIN than anything else. Thanks, Harry. --Dragoondelight 10:43, 11 June 2009 (UTC)

As for attributing, what is an industry available entry, just envision me following advice and putting Mushroom ebooks, Fictionwise, or Amazon or a combination behind each entry. Nyet! Thanks, Harry. --Dragoondelight 10:43, 11 June 2009 (UTC)

Frankly, on the personal level, the publisher has offered even more cooperation, but I will not take him up on the offer, even though it would give me a free example to use. Too much possibility of compromise, neh? I am entering the sequence for history/continuity sake and hope that some other advocate will review/verify them. I am walking a fine line and do not want to promote vendors, beyond the obvious. Thanks, Harry. --Dragoondelight 10:43, 11 June 2009 (UTC)

Thinking about it, I would love a deluge of new people stating the offense they find in reading time being included! Might be a new editor or contributor willing to man the barricades. LOL Thanks Harry. --Dragoondelight 10:43, 11 June 2009 (UTC)

If the note read something like Stated: "Reading time 150-210 minutes" or like "Reading time 150-210 minutes" (from publisher) or some other format to make it clear that we are merely repeating what is stated by someone else. As for giving blurbs, I can't count the number of records with "data from Amazon" or "pub month from amazon" or "price confirmed by several listings at ABE" or the like. When our data source is a bookseller, we should IMO credit that source and not worry about "promoting vendors". -DES ^Talk 12:22, 11 June 2009 (UTC)

The reason i object to this is that it is not, even in principle, objectively verifiable from the publication. Pretty much all of the other things we list are, or ought to be, verifiable to anyone with access to a copy of the publication. But how can a statement of "reading time" made as if it were factual, be verified. If we are reporting that the publisher or someone else identifiable states the Reading time to be such and such, that can be verified. Just as when i list a manufacturing statement like "Printed in USA" I always put it in quotes, because what i am reporting is that this is stated on the copyright page or DJ or somewhere on the book. IMO "Reading time", if reported, needs to be in effect quoted and attributed. -DES ^Talk 12:22, 11 June 2009 (UTC)

When a matter is really not under discussion, then what do you do? I am still totally unconvinced, but in the interest of "if your concern" is really that terribly important to you. You have my permission to eradicate each and every reading time. Please start with the SRA series of the 1960's after you finish this. Thanks, Harry. --Dragoondelight 14:10, 11 June 2009 (UTC)

An even better suggestion: stay away from ebook records. That's what I do. :-) When I think of the thousands of records that don't belong in the database, added by non-sentient entities, I get a headache. So I try not to think about them. You've got editors in a tizzy if a genuine work of speculative fiction has, my god, pictures! You've got military fiction set in the future that reeks to high heaven of nationalistic jingoism. There's romance novels with a ghost or vampire thrown in for a little kinky sex. And these actually do belong in the ISFDB! When these aspects of the db start to trouble me, I steer away from them and cultivate my own little patch of the db (apologies to Voltaire). I'd suggest the same to any editor who would be concerned over something as insignificant as a publisher-stated reading time. And the same goes for those with passive-aggressive reactions to the others' concerns. Peace out. MHHutchins 14:48, 11 June 2009 (UTC)

The bald truth is it is silly and the sad truth, no one will clear the back log, so I can finish the process. It amounts to blackmail over an opinion. So as when I reminded him that editing can lead to a change in the facts, this has lead to keeping the facts from appearing. DES and everyone had a chance on the first submission, and he made it apparent that it was there. I submitted 10 more and would not argue the use of both author and pen-name. I had the publisher declare that it was only the pen-name and then changed the ten or so bad entries to the pen-name. He then waited a day or so and popped me with Reading time. He also wanted a public debate, which I still feel no one else really felt strong about either way. Effectively he has dragged the topic from my personal wiki to the public wiki for his purpose. He then stung me to go from my personal wiki and answer in the public wiki. My argument qualitatively was as good or as bad as his. He then proposed a compromise, much like he did when I was forced from this to [This Publication Record]. A compromise has the inherent aspect that something is wrong with 'Reading Time'. This coming from a moderator who insisted at some length that I not remove Assumed and Stated commentary, even when the original book did not have the necessary information to quantify in such a manner. I draw notice, that he thought nothing wrong with those indefinites being used, though most I feel would say they beg a source or signature. I now find my self blackmailed into an argument and do not accept that a compromise by myself is necessary. Instaed, I suggest he master edit "the newly offensive descriptive terminology out". I will refrain from using it, as any beaten dog would, but I will not compromise and semi-validate this obtuse interpretation. Why is the kb, okay? Should not the word count, be tagged for authority. All in all, this has been a baited argument that has jammed my ability to complete the series additions I started on. I see no great mistake, but I can see the avoidance to pass my work out of a false respect. It never works to placate someone, when they blackmail you into it. I will not validate a wrongness, as I have told DES before. No matter what he may or may not desire he has taken an active role in what the content shall be, again without having that data in his hands. Take the challenge, get the material, and I will gladly give place to a moderator to do the needful job of entering data that was published for the German market, written by a prolific English author. That it is in ebook format is of no consequence, but it is important that data be made available. This whole encounter shows the tendency for argument without positive result. It has been spiritedly pursued, and no retreat other than submission to power seems offered. Compromise is a doggerel term for acceptance of power. My suggestion, is that it has all been unneedful and that other moderators do the checking of my submissions. If it was desirous to show that I am the untamed/untameable Barbarian so be it. For my part, I say better to stand proud and unbowed than with the tatters of a compromise that is hollow. Thanks, Harry. --Dragoondelight 20:24, 11 June 2009 (UTC)

Be aware, I no longer will take part in this kind of argument. I will not be roundly summoned before an indifferent court and then prosecuted by anyone who can not himself recant his position. A discussion has value, only if there is agreement, and any one who must suffer consequences till he agrees is the victim of force. Thanks, Harry. --Dragoondelight 20:24, 11 June 2009 (UTC)

I raised the issue the first time i saw a submission. When I saw several more such submissions, I held exactly one such submisison, and simply refrained from moderating any other submissions in the series, and I think other mods have approved all such sub,missions. I attempted to make the case for my views calmly, logically, and without personal insult. I strongly resent this being called "blackmail". I have never tried to enforce my views over a consensus here, although i have attempted, sometimes strongly, argument for my views to try to form or change a consensus. So far as i can see, the majority of those who have commented have disliked "Reading time" to one level or another, but I don't think a clear consensus has yet formed on the issue. You say "and no retreat other than submission to power seems offered." As you yourself point out, i suggested a compromise. This was not an absolute demand -- none of this was an absolute demand, we don't make such demands here I hope -- but a suggestion. If you thought the suggestion unwise or unacceptable, you were free to say why. I don't recall you indicating why you felt that 'Publisher states "Reading time XXX"' was unacceptable compared to 'Reading time XXX'.

You mention the co-authorship of pesudonym and and author which also occurred in several of these submissions. On htat I merely noted that this was unusual and asked you what the publication said. I at no time so much as attempted to dictate how the records should be entered, i merely delayed until I had time to hear back from you one way or another. You, after double checking, decided to change the data for these records to match the actual publications, as i understand the matter. Querying an entry that seems odd, possibly mistaken, is what a mod is supposed to do, i thought.

You say that I "dragged the topic from my personal wiki to the public wiki" and that I then "stung me to go from my personal wiki and answer in the public wiki." I raised the issue on a general forum because a question of standards, of what data is entered in what format, is normally settled here by public discussion, intended to lead to consensus. Two people cannot decide for the rest without giving others a chance to voice their views. It is precisely because no one person, even a mod, should be able to force his views on others that we have public discussions of such points. I notified you of the discussion because i thought you might not have seen it, and i wanted your views to be taken into account, not to "hale you before a court" or to "prosecute" you. The question was how to handle a data element not seen before on the ISFDB (to my knowledge), not to praise or condemn any person here.

The "[This]" Discussion is long over. I took the position then that your format was likely to be confusing to users. I don't think I eve held a submisison over the issue, nor changes your notes format to one i preferred. And had the general weight of opinion agreed with you, i would have dropped the matter.

No one has said anything so much as implying that you are some sort of "Barbarian", to my knowledge. I certainly have not. I don;t think I have said anything in this debate, or indeed ever on the ISFDB wiki, about your personal virtues or flaws, or those of anyone else, except for spammers and vandals, which you are surely not. I have dis agreed with your views at times. i have disagreed with Bill Longly, Mike Huthchins, and even with Al von Roff, sometimes strongly, but always over matters of substance, not personality, IIRC. I do resent what I perceive as an unprovoked personal attack against me, in the comments above.

All that said, the issue of a "Reading time" note is hardly a vital one for the ISFDB. I have made my views on why such notes should be omitted or modified clear above. I will not restate them. I will approve the one submisison of yours I have on hold. I will not hold future submissions over this issue, nor edit merely to change such notes. (By the way, there is no "master edit" feature available that I know of. If I were to set out to remove those notes i would have to do it one pub at a time, just as anyone else would.) If a clear consensus develops on such notes one way or the other i will abide by it. In the mean time, i will neither add nor remove any such notes, nor object further over anyone else who does so. Does that satisfy you? -DES ^Talk 21:06, 11 June 2009 (UTC)

As I wrote on Harry's Talk page a few minutes ago, terms like "blackmail", "popped me", "baited argument", "jammed me", "stung me", etc. ... raise the temperature of the debate and turn it personal, which never helps. We have plenty of things to discuss and resolve here without personal attacks. Ahasuerus 23:47, 11 June 2009 (UTC)

I've approved, without comment, many editor submissions where I don't agree with the way that editor is entering things but I also see that the disagreement is between my way of thinking about something vs. the editor's. Don't sweat the subjective stuff. If a submission appears to be factually wrong, factually ambiguous, or violates one of the umpteen thousand ISFDB rules a moderator can hold and query, approve and send a note to the editor, or approve and silently correctly the issue. On inspection, the record under discussion seems to be ok other than it seems to be violating one sentence that's over eight thousand words deep into Help:Screen:EditPub. Discussion of that sentence would be topic drift from the exciting Reading Timethread and so I'll start a new R&S section about it. --Marc Kupper|talk 02:27, 12 June 2009 (UTC)

This never should have gone beyond the "Are your sure you want do this? (Y/N)" stage. A similar discussion concerning a different submission by Harry that I questioned resulted in a humorous exchange over an issue just as trivial. DES has nothing but the integrity of the data in mind. Harry has always been a pleasure to work with. As I would have said if the discussion hadn't gone in the direction it did: Supplying reading times is a common practice with many ebook distributors. Most ebook readers, the ones most likely to view the record, would be familiar with the convention.--swfritter 15:10, 13 June 2009 (UTC)

Title Prefixes and pointless sub-titles

I can recall some discussion that title prefixes that merely indicate a series should be omitted from the title and/or pub records. For example, should Operator 5: #2 The Invisble Empire be listed as simply The Invisble Empire? Or consider On Basilisk Station. On Both cover and title page this was identified as Honor Harrington: On Basilisk Station, but we don't record it that way, and booksellers and libraries generally file it under O, not H. But there is no mention of omitting such prefixes inn the help, nor of when to do so.

Similarly, such sub-titles as "a Novel" or "a Romance" are often omitted. But the help again says nothing of this practice, and indeed uses Lilith: A Romance as an example of a sub-title.

When should such prefixes and sub-titles be omitted? Should the help be updated to confirm to and standardize our practice in this regard? -DES ^Talk 22:14, 7 June 2009 (UTC)

I think proper titles can usually be established fairly easily from primary verifications. In the meantime, I tend to live with them as I for one would never have come across "Operator 5" if you hadn't mentioned it. Now I see it, this needs some de-shouting attention. But I always prefer to remove Series details and useless suffixes like ": a novel" from the title, while letting people record whatever nonsense they like at publication/review level, e.g. Good Omens: The Nice and Accurate Prophecies of Agnes Nutter, Witch. Much better to have squabbles over one book rather than many. BLongley 23:19, 7 June 2009 (UTC)

I look at this and despair. What's the point of creating a variant for a subtitle if people aren't going to use it? BLongley 23:29, 7 June 2009 (UTC)

Since getting to VTs from the main Title page is not easy at the moment (something that is likely to change soon once your change goes live), it may have been just a technical issue. Ahasuerus 23:41, 7 June 2009 (UTC)

As far as Help goes, we have a template (TitleFieldHelp) which states "If the title has a subtitle, you don't need to enter it; this is sometimes a judgment call as to whether a change of font or a colon indicates a separate subtitle or just some creative license on the part of the typesetter." We really need to reconcile our diverging Help pages and templates :(

Substantively, I don't think that having separate VTs for every reprints that adds or subtracts ": A Novel" is useful, but the "Dr. Bloodmoney" example above would, IMO, merit a VT. Ahasuerus 23:44, 7 June 2009 (UTC)

{{TitleFieldHelp}} appears not to be used anywhere, while {{PublicationFields:Title}} and {{TitleFields:Title}} are used in Help:Screen:EditPub, Help:Screen:NewPub, and Help:Screen:NewNovel. Both say "If the title has a subtitle, enter it, with a colon and a space used to separate the title from the subtitle." one goes on to say "For example, the 1986 edition of George MacDonald's "Lilith" has "Lilith" on the title page, and below that, in a smaller font, "A Romance". This should be entered as "Lilith: A Romance"."

I would like to add wording such as "Generic subtitles such as "A Novel", "A Romance", "A Mystery" or the like should not be entered. Neither should subtitles that merely indicate the series, such as "Book 5 of the Scrolls of the Lost" or "A Jack Jones Adventure". Likewise, title prefixes which consist of the series name should not be entered. In all such cases, make sure that the series is properly recorded on the title record, and optionally add a note about what the title page or cover actually said." -DES ^Talk 00:06, 8 June 2009 (UTC)

My take: Title subtitles that indicate series should not be entered. Title subtitles should otherwise always be entered on new (previously unknown) titles. Title subtitles on new publications of existing titles should be optional, but when distinctive should/could result in a title variant. If the original publication is later verified with a subtitle, then the parent/variant relationship should be reversed (non subtitles publication belong to the variant now). Also don't forget, many of the pre 20th century titles seem to come standard with a subtitle (or two), and Lilith: A Romance is just a well known example. I personally think it's best not to get too specific here, either in the help or the templates, etc. Whatever rule we come up with, there are 20 examples of titles where that rule should be ignored. I prefer however not to institutionalize rules which cause us to stray very far from the printed page. Subtitles which indicate series, already have a home at the isfdb, 'series'. Subtitles should have a place in the 'Title' field. Kevin 00:29, 8 June 2009 (UTC)

Generic sub titles, such as "a Novel", can lead to multiple search returns for the same work; pointless variants; failure to find duplicates on the "Dup Candidates" screen, and are not, i think, in the vast majority of cases, helpful. Distinctive subtitles are a very different matter, IMO. And I think that making our help as specific as possible, while indicating where there is room for judgment and variation, is usually a good idea. (Note that the templates are being used to build the help pages -- in effect they are the help text. They are used because the same text appears in several help pages.) -DES ^Talk 00:55, 8 June 2009 (UTC)

Leaving them out will lead to zero returns for a specific and accurate search "Lilith: A Romance", which should have succeeded. Kevin 01:15, 8 June 2009 (UTC)

Just a note that libraries have been struggling with this issue for a very long time, so the current cataloging standard has a special field for "subtitles". It wouldn't be easy to implement in our software, but it's something to consider. Ahasuerus 01:27, 8 June 2009 (UTC)

ISFDB publication records and citing sources

It looks like there's a weakness in the publication related help in that it does not emphasize that the source of information for a publication record is the physical publication itself. If some other source is used for some or all of the information then it must be cited.

With that in mind the forth sub-bullet of Template:PublicationFields:PubNote should be revised in that it currently says "You do not need to record the source of data if it is taken from one of the standard bibliographic sources used in verification..."

As there is no physical publication for eBook media this means 100% of the information in an eBook publication record would need to cite its sources. Ideally they would be as links to web pages along with the information the editor chose to extract from that page.

Using Transit to Scorpio as an example we'd have

• Mushroom eBooks the publisher's web page and is the source for the title, author, Year, and Publisher fields. The page also includes some miscellaneous information
  • LANGUAGE: English
  • SUBJECT: Fiction, Science Fiction
  • WORDS: 65000 words (approx.)
  • AGE RANGE: T-General).
  • Finally, the page lists eight media formats, each with its own ISBN and price. (I don't know if that means eight publication records, or maybe one, depending on other R&S threads.)
  • The page links to pages for two distributors Fictionwise, and Sony Connect.
• The Fictionwise page provides additional information which from ISFDB view is optional. These are:
  • File size: eReader (PDB) [209 KB], ePub (EPUB) [204 KB], Rocket/REB1100 (RB) [185 KB], Adobe Acrobat (PDF) [971 KB], Palm Doc (PDB) [211 KB], Microsoft Reader (LIT) [177 KB], Franklin eBookMan (FUB) [241 KB], hiebook (KML) [481 KB], Sony Reader (LRF) [238 KB], iSilo (PDB) [173 KB], Mobipocket (PRC) [215 KB], Kindle Compatible (MOBI) [251 KB], OEBFF Format (IMP) [271 KB]
  • Words: 65248
  • Reading time: 186-260 min.
• The Sony Connect page is source for the publishing date, February 28, 2007, of the Sony Reader eBook format. It's also the source for
  • Filesize: 428.70 KB
  • As a matter of curiosity, the "Add'l Info" tab lists "Mushroom Publishing" as an imprint but does not state who the publisher is.

I've posted this as an R&S discussion item to see if there is consensus with regard to publication records and citing sources. --Marc Kupper|talk 02:54, 12 June 2009 (UTC)

Somehow this got edited out when I was writing the initial comment. The main publication help pages include Help:Screen:NewNovel, Help:Screen:NewPub, Help:Screen:EditPub. I have not done an all out search to see if there's more. --Marc Kupper|talk 02:58, 12 June 2009 (UTC)

I disagree with the core premise. The ebook itself is a 'fine' source of information and is assumed and presumed to be the source of unattributed information. Only information that comes from somewhere other than the ebook file itself requires any attribution. If the ebook is published/available from a single source (Gutenberg, Baen, etc) then information not in the ebook, but directly from the single source download location, is also in a gray area that may not need attribution. Also attributions can be implied. See NTRSTLLRNV2005 as an alternate (shorter, simpler, less wordy) attribution of ebook information. Kevin 03:09, 12 June 2009 (UTC)

I agree with Kevin that the ebook itself can be a source, just a printed pub can be. When I have a downloaded copy of an ebook and i report that it shows the author on the title page as "J. R. Author" instead of the canonical "J. Random Author" that is primary sourced data. (Inded i did this for soem of the PG editions of works by Le Fanu, who is commonly known by at least three versions of his name.) Ditto for an artist credit, if the ebook included embedded art and a credit for same. In at least some cases, the ebook itself include publication dates (the Project Gutenberg ebooks pretty much all do). Word counts are also primary data, if you check them on the actual text. So are page numbers for those ebooks that indicate page numbers as some of the PG ebooks do). -DES ^Talk 04:49, 12 June 2009 (UTC)

However I agree with Marc that we are not emphasizing the need to record sources enough, and that the statement in the help that he cites -- "You do not need to record the source of data if it is taken from one of the standard bibliographic sources used in verification..." should perhaps be changed. I often see "month of publication from locus1" or "Price from Curry" even though those are "standard bibliographic sources". IMO this should be encouraged if not considered "the one true way" I don't see a need for a web link for things like this, or "ISBN confirmed from several listings at ABE" or "Pub month from amazon.com" or "interior art credit from OCLC" however. -DES ^Talk 04:49, 12 June 2009 (UTC)

Maybe it'd be better if I write about something I know about. :-) I've never seen an eBook and so for some reason I'd made the assumption that a web site like Mushroom was the cover plus metadata and when you buy a book you got just the text. When I saw Kevin's remark I realized most (all?) eBooks must have title pages, copyright pages, etc. I struck out my comment above. Do eBooks state the price on the front flap or back cover? If not, then the source for the price should be noted.

I'd started this thread as it appeared to me that the Transit to Scorpio record was constructed from publisher and distributor listings but was not citing them as sources. I then saw that the help pages were largely silent on this. --Marc Kupper|talk 07:52, 12 June 2009 (UTC)

Well the citation issue applies just as much to paper books.

If we adopt this as policy, should mods start holding submissions that seeem to use secondary sources but fail to cite them. For example there is this submission that I jsut approved. It changes the year of pub of 262158 from 1956 to 1955, and the page count from 208 to 207. it appears that this data was from OCLC, or perhaps from Curry. Would such an edit be held under the proposed modified practice? -DES ^Talk 18:44, 12 June 2009 (UTC)

Nope. (and my gut says we shouldn't adopt this principle). What ever happened to the principle of least effort welcome? Come enter a pub or enter 10. Document to the level you feel comfortable with, and if you leave something out or incorrect, someone else will come along and make it better eventually, and after that someone else will primary verify it. (This excludes Records with a Primary Verification.) Yes this reduces the value of Secondary verifications without notes to document what you got out of the secondary source. Yet this is still the way we should operate. Kevin 01:05, 13 June 2009 (UTC)

But the added effort to an editor, who knows what source she has used, to type "(pub month from Locus1)" or "(price from Curry)" is so minor, and the added effort potentially imposed on Simone else to recheck possible secondary sources until support is found is so much larger, that I think this is no worse that saying "please don't enter '24, enter '$24.00' in the price field". if editors don't enter currency symbols we don't apply your "principle of least effort welcome" nor do we if people enter "hardcover" instead of "hc" nor if the enter "SFBC-Doubleday" instead of "Doubleday / SFBC". In all such cases we gently correct new editors, and expect them to get it right after a few corrections. I don't see why listing sources need be any different. We recently refused to accept a bunch of submissions that added cover art credits until the editor listed his sources. -DES ^Talk 01:32, 13 June 2009 (UTC)

It is 100 times greater effort to require free form text explanations. Price is a data field where conformance to an exact standard is required to make searches and analysis work. Similarly with name formats, titles, publishers etc. Precision is key because we rely on software to identify matching/differing items. Sourcing however is an art, takes added effort, time, desire, and ability which not all editors can, will, or are required to have. You may have held a new users submissions for cover art credits (I don't actually recall if it was you or not), but I would have accepted into the database every single one of them on any publication that did not have a primary verifier. (IIRC, most of his updates did already have a primary verifier - which was what caused the problem.) Kevin 03:37, 13 June 2009 (UTC)

Until a pub record has a primary verifier, I have, and will continue to accept any reasonable submission of information. Once it has a primary verifier, I will rely upon that person to 'check' or confirm any additional data added to that record. If that verifier is unwilling or unable, I may or may not attempt to check or confirm the additional data, depending on the history of the submitting editor. Kevin 03:37, 13 June 2009 (UTC)

As a moderator, our job is to protect the database from malicious and obviously incorrect information. Our job is to safeguard records that people have put the extra time and effort into, to verify from a primary source. Our job is to provide an environment that leverages all the effort that is donated to 'fill up' the database with the best information we can get donated. Our job is not to squeeze every last bit of detail, time and effort from a volunteer who has already donated their valuable time and knowledge. Our job is not to verify every submission, check, cross-check, and double check the facts submitted (unless we have a reason to believe them to be in error). In short, all donated data is correct until proved otherwise. All editors are innocent until proved guilty. I will not (and we cannot) require each volunteer to 'prove' their sources, unless I have sufficient reason to believe it to be incorrect. Kevin 03:37, 13 June 2009 (UTC)

The whole concept of crowd-sourcing is the same as open source software. Many eyes, make light work. In contrast, heavy requirements make it heavy work again. Kevin 03:37, 13 June 2009 (UTC)

I think you are way way overestimating the effort and skill that we would be asking of editors. Typing a few words is not an onerous burden "100 times" greater than getting a code correct or adding a currency symbol, or so it seems to me. It does not IMO take much "art" to simply record the source that the editor has already found and used. I am not suggesting, and I don't think anyone is suggesting, requiring or even strongly urging editors to go out and do cross-checking or use additional sources. I am just suggesting asking (probably not requiring) editors to briefly and quickly note, in any form they like, where they got their info. This does not seem to me like a burden likely to drive of most editors, it is not nearly the level of source that Wikipedia is now demanding for many entries, for example. -DES ^Talk 14:14, 13 June 2009 (UTC)

(unindent)It would appear that there may be a few separate issues here.

The first one is just how much data is required before a submission is considered valid and can be incorporated into the database. Historically, we have accepted bare bones submissions based on the notion that different contributors are able/willing to contribute at different levels and, besides, it's better to have some information on file than no information at all. As Help:Screen:Moderator currently says, "Incomplete data, however, is not an issue. If an editor fails to enter everything that we would like to see captured, this is not a mistake on their part. Of course we would like to encourage editors to enter everything, but as they are volunteer labour, it is much preferable only to intervene for mistakes of commission, not of omission." I don't think anyone is questioning this approach although sometimes we have to ask follow-up questions, e.g. if the resulting publication record looks like it may be a duplicate of a pre-existing record.

I quite agree. -DES ^Talk 14:14, 13 June 2009 (UTC)

The second issue has to do with verifying the data submitted by our contributors. Help:Screen:Moderator provides guidance in this area and suggests some bibliographies to check as well as mentions the circumstances when the approving moderator may want to check them. Judging by our Talk pages, many moderators cross-check all or almost all new submissions against OCLC, Locus, etc, especially when the circumstances appear to warrant it (and we all develop a "sixth sense" for suspicious data over time.) I don't think we have had any problems with deliberately false information submitted so far (knock on wood), but some areas are trickier than others (how many randomly selected people would know how the number line mechanism works?) and even experienced contributors have been known to run into problems with some of the more, ahem, challenging aspects of our software's behavior.

The third issue is whether, as Marc put it, "the source of information for a publication record is [assumed to be] the physical publication itself. If some other source is used for some or all of the information then it must be cited". I believe that this statement generally reflects our intent, e.g. we have modified Dissembler and Fixer to state which secondary source the data came from and as of what date. Similarly, we have asked our contributors with specialized knowledge in the cover art area (mostly Don Erikson) to leave a note whenever they identifies the cover artist based on his or her style.

As a general observation, secondary sources can be very useful both as the sole source of Publication information as well as additional sources of certain data elements like "month of publication", cover art attribution, etc. However, when we use data from secondary sources, we need to document where it came from. If we don't, subsequent verifiers will have a hard time matching their copies against our records. For example, the other day I was verifying a paperback and noticed that we already had what seemed like the same pub on file, except that there was a cover artist credited while my copy had no signature (much less an explicit attribution) on the cover. Did it mean that I had a different printing or that, more likely, the cover artist had been identified based on some secondary source? I went ahead, verified the record and added a note to the effect that the cover art wasn't credited and that the source of its current attribution was not known. We have quite a few records with similar Notes and it would save all of us a fair amount of work if we had asked the original editors to indicate where the information had come from. Similarly, if we knew that certain old records had originally come from Amazon.com, we would be in a much better position to judge how reliable certain data elements are.

Re: the comment that "[s]ourcing however is an art, takes added effort, time, desire, and ability which not all editors can, will, or are required to have", I don't think that clarifying that the data came from, e.g., Amazon.com, OCLC or Locus-1 is particularly difficult. BTW, Wikipedia also started out with a rather relaxed approach to sourcing but, after experiencing various kinds of data quality issues, adopted a much more stringent standard.

Overall, I believe that this is an important area and that, although all moderators have somewhat different styles, we need to make sure that conceptually we are all on the same page re: secondary sources and their attribution. If different moderators pursue different approaches, it will cause no end of confusion and heartburn. Ahasuerus 05:27, 13 June 2009 (UTC)

I'll offer up my own real life example. I was thrilled after many years of use, to realize that the ISFDB was 'open' for submissions. (I had emailed Al a couple of grievous/obvious errors in the 1999-2004 time frame IIRC). When last year I realized he had finally progressed the system to the point I could contribute I was excited. I acquired a new secondary source in order to contribute to the database (Bleiler's 1978 checklist). My original intent was to enter the bare facts presented in that volume and leave the skeleton for others (or myself, later) to flesh out at a later date. So I started happily entering data and truly contributing new titles, new authors, etc. Then I was requested to enter more/different data, and pointed at Worldcat as an additional secondary source, worthy of documentation. My Type in project (that I had committed time and $$ to) had become a Documentation project where I was expected to perform extra labour (Yes, Bill, if you read this I spelled it that way just for you). Some people might have been deterred. I would have been deterred at a different point in time, life etc. I also know that I have not done Bleiler entries at times because I wasn't in the mood to do the extra research, but I was in the mood to contribute and just enter bare bones data. But I didn't want to seem 'not fully participating' or however you want to term it.... so I did nothing / something else (not ISFDB related at times - Gasp - Shock and horror). This is my own experience where a request to enter extra data caused less data to be entered. Kevin 06:51, 13 June 2009 (UTC)

To my knowledge - There are no regular (weekly?) contributers under the age of 30 (but I could be wrong). (I'm 35 for a couple more weeks, and I feel like I'm the youngest person regularly 'here' oft-times). I am fairly certain, that we have no regular contributers under the age of 20. Yet, in the aprocrophial? quote attributed to John W. Campbell, the "Golden Age of Science Fiction is '12'". Kevin 06:51, 13 June 2009 (UTC)

The idea was first popularized by Randall Garrett in 1956 (see this exchange) and condensed to "The Golden Age of science fiction is twelve" by Peter Graham ca. 1957. Ahasuerus 16:53, 13 June 2009 (UTC)

The Internet is a Young Medium, no matter how long we may have been using it. Kevin 06:51, 13 June 2009 (UTC)

Well, that was true when the internet was first opened to the public in the early 1990s, but the age of the average internet user is fast approaching the average age of the population at large, at least in the developed world: see this 2006 US graph and this 2008 UK page. Of course, kids have more time to spend on the internet, computer game, text messaging and other "cool stuff", but I doubt many of them would be interested in something as dry and "uncool" as bibliography. At a guess, probably the same percentage that has always been interested in it :) Ahasuerus 16:53, 13 June 2009 (UTC)

I want to harness that young energy. I want drive-by data dumps this summer from kids out of school and surfing the net. I want young fans in dark air-conditioned rooms this summer lovingly cataloging their books because they are important to them (I imagine the pure joy and feeling of purpose I could have experienced had such a website existed when I was 14). It's a Joy not a Job to contribute here, but only if we invite those who have the energy to to contribute what they feel capable of. The ISFDB will get a significant upgrade in perspective when we get our first MOD under the age of 20, but that won't happen if we require professional level editors/documentation for all contributions. Editors at all levels of capability, as long as they submit reasonable edits must be welcome. Kevin 06:51, 13 June 2009 (UTC)

I don't suggest insisting that editors be asked, much less required, to do additional research or consult addition sources, merely to note, simply, what sources they have in fact used, other than the actual pub. And even that I would put in the form of a query to the editor, rather than a submission hold, i think. You refer to your own experience and seem to imply that you were requested, even demanded, to seek out and record additional info and enter it. But the msg you like to (from me) said "When you have checked a publication against [various online catalogs] it is helpful if you include the appropriate reference number or catalog number." This was not intended to mean "you ought to check such a catalog" but merely "if you have already checked a catalog, please note the fact and record the linking info)" and I sent it because some of your entries made it clear that some info was coming from such catalogs. i may have mis-estimated how often you were already doing such checks.

You say above "All editors are innocent until proved guilty." Of course, but then i hope we don't ever judge editors (except for the rare spammer) "guilty" of anything. You also say "[A]ll donated data is correct until proved otherwise." Here I disagree. i often approve submissions without double checking if they are from editors whose quality work i recognize. But if an edit is from a relative newcomer, or if some element in it seems suspicious (e.g. a price that is out of line with the pub date, an initial publication long after the author's death, an image from a site not on the permissions list) i will do at least some research, and if I can't confirm the data, i may hold the submission and query the editor. I would expect that most moderators would do likewise. Help:Screen:Moderator discusses in some detail which kinds of submission are more likely to "need research" than others.

In short, i think you are seeing more of a change here, and more of a demand on editors, than anyone intends to make, and are overestimating the effects of such a policy change. -DES ^Talk 14:14, 13 June 2009 (UTC)

Now that I have read Kevin's example, I see what he meant by "[s]ourcing however is an art, takes added effort, time, desire, and ability which not all editors can, will, or are required to have". Certainly learning to use online catalogs and various (and sometimes obscure) bibliographic conventions used by librarians can be a challenge and a hurdle for new editors. (Although, in fairness, Web interfaces tend to be much easier to learn than the old roll-and-scroll stuff that the Library of Congress used just 20 years ago. Why, I used to walk fifteen miles, uphill, in the snow, both ways, just to check a single record there!) I wasn't suggesting that we require our editors to do this kind of research when they submit data unless they are so inclined. I was merely pointing out that adding "Data from Bleiler-78" to the Notes field would make our data much more valuable to the general public and easier to work with for future editors while requiring very little additional work from the submitting editor.

BTW, this was the main reason why we added a list of Verification Sources, but not every source can be added to the list. Besides, different sources provide different levels of detail, e.g. Tuck lists prices while Reginald doesn't, so sometimes we need to clarify where different data elements came from.

Having said that, the approving moderator may choose to use other sources to cross-check the submission, but that's a whole different story.

As far as age goes, speculative fiction bibliography seems to attract all ages, e.g. Bill Contento, R. Reginald, Jerry Boyajian and Phil Stephensen-Payne all started when they were fairly young. We did have a few younger contributors at one point, e.g. Grendelkhan, and others are in their 30s, but most active contributors seem to be in their 40s-60s. Of course, being retired helps :) Ahasuerus 16:22, 13 June 2009 (UTC)

Ahasuerus put it better than i did. I fully agree. -DES ^Talk 16:47, 13 June 2009 (UTC)

(For pubs without primary verifications) Recommending (In the help) and nudging (on a talk page - "Hey, can your drop a quick note into your edits about your sources "Price from Locus", "Contents from Worldcat", etc? - It helps the next guy out." are all great things to do. Holding submissions, seeking clarification of every unsourced edit submission, and generally requiring it is a whole different animal. We have two databases in one. Primary verified records, and all other records. We need to take care that our requirements for verified records do not spill over into unverified records. If we add secondary verifications to the Moderator flag on the approval screen, I would even support moving secondary verified records into the more stringent category. I would whole heartedly support updating the help to clarify more stringent entry and sourcing requirements for verified publication records. Kevin 16:55, 13 June 2009 (UTC)

(unindent) Since everyone seems to agree that requesting primary verifiers to record sources in future is acceptable, i made this change to Help:How to verify data. Does anyone object, or want to reword this? -DES ^Talk 19:13, 17 June 2009 (UTC)

Proposal: I would change "You do not need to record the source of data if it is taken from one of the standard bibliographic sources used in verification." and the following sentence in Help:Screen:EditPub#PubNote (and the other help pages that get text from the same template) to say:

"If some of the data elements are not stated in the publication, but are derived from secondary sources, this should be mentioned. If possible briefly indicate which element is derived from which source you consulted. However, no editor is required to do additional research to find data not in the source(s) at hand, nor to search for the sources of data already in the record, although you may do such research if you choose to. Entering what you have, and briefly indicating where it came from, is sufficent. The abbreviations used in the verification source list may be helpful for this purpose. Thus a note might say "Month of publication from Lotus1" or "Cover price from Curry" or "Page count from OCLC" if the editor had consulted those sources, or "Book only states year of publication, source of month unknown" if no source at hand gives the month.

.

Implicitly (or we can make it explicit if you like) mods would be expect to use "Recommending and nudging" to encourage -- not require -- editors to supply such data on non-verified records. Any objections? -DES ^Talk 19:13, 17 June 2009 (UTC)

I would go along with these clarifications of the standards, but feel they should be in regard to the verification of pubs rather than directed at newbie editors. Which brings up another question. Have the standards for verification ever been codified? MHHutchins 20:27, 17 June 2009 (UTC)

No. I know I've asked at least three times in different ways. And there seems to be no possible future-proof solution either: so we have Verified Collections and Anthologies with no contents. (OK, we at least know it exists.) Or illustrated books with no interiorart credits. ("Hey, the fiction is recorded, who said we were the Internet Speculative Fiction and Artist DataBase?") I know people want to Verify so they can say "I've got this!" and there are proposals already submitted to allow multiple verifiers, but I think we need the "I've got this!" statements to be demoted to such (in unlimited numbers) and Primary Verifiers should be the ones willing to help in the ever-more-anal improvements of pubs. I know I've seen people remove Primary Verifications rather than answer questions: I can't blame them really. I've had to do that when I can't find the book again, even if I'm sure I haven't actually swapped it yet or lent it to somebody. I think we need to make "Primary Verification" a bit less possessive, but we'll have to support that by letting people retire gently from discussions over such. BLongley 21:37, 17 June 2009 (UTC)

Actually i think the standards for verification have been codified, but many older verifications don't match them, and no one currently enforces them on new verifications either. Currently Help:How to verify data (which in think the the codification asked for) says "Primary verification should mean that all fields are complete" and then lists five exceptions, all for things that some pubs do not list. It also says "If the work is an anthology or collection, you should enter all contents before marking the pub as verified. This includes entering page numbers, except when no numbers are available..." I think that is fairly explicit. It now also says "If some information about a verified publication does not come from the book itself, this should be indicated in the notes, and some indication of where the information did come from is desirable." I don't know that we could make that much stronger, although we could make an explicit list of the things that should be entered and checked against the book-in-hand.

I don't think it is unreasonable to request, gently (not demand) minimal sourcing info from even relatively novice editors. Just an answer to "Where did you find that info? Not sure? Well please try to note it another time, it helps the next guy."

I agree that making primary verification less possessive would be good. I think that multiple verification would help. Do you recall seeing pubs that were verified within the last 1 1/2 to 2 years with missing contents? i think that recent verifications have at least mostly come up to the standards of Help:How to verify data -- do you disagree? -DES ^Talk 22:15, 17 June 2009 (UTC)

Nongenre shortfiction

I recently encountered a nongenre novelette by Fredric Brown in a non-genre anthology (of "pulp" mystery fiction). Now Brown is pretty clearly over the "certain threshold" so that his non-SF work should be included. But we currently have no good provision for recording non-genre short works included in non-genre pubs. Should i just let this go? Or should it be recorded somehow? I know there is a proposal for a non-genre shortfiction type. When/if that is implemented, i could record this, but omit the other contents, just as is done for a "general interest magazine", I suppose.

Any comments on this situation and how it should be handled currently? -DES ^Talk 18:53, 17 June 2009 (UTC)

I personally do not like shortfiction being separated based on IN or OUT. There are many works in the suspense/thriller category which have no fantastic element, but would be considered IN if included in a spec-fic anthology, and OUT if reprinted in, for example, a mystery anthology (those "edited" by Hitchcock come to mind.) It's all a matter of context. I'm not in a position to make the subjective judgment, and wouldn't feel comfortable for others to be making that assessment for me. The author's oeuvre should be used to determine IN and OUT, not individual pieces. Unfortunately, Fredric Brown isn't a good example, because his body of work includes as much (or even more) mystery than spec-fic. A better example would be Avram Davidson. I'd rather have all of his short fiction regardless of the genre in one listing rather than for it to be subjectively segregated. As for omitting contents from general interest magazines and non-genre anthologies that are obviously not spec-fic, that's fine with me, if you make a note that only spec-fic has been included. (Isn't that already being done with mags like Playboy, Argosy, etc?) MHHutchins 20:17, 17 June 2009 (UTC)

So, to get back to your question, I don't think the Brown piece should be included, if obviously not spec-fic. BUT, if the same piece appeared in a spec-fic anthology it would be included. Again, it goes back to context. Sorry if this sounds wishy-washy. There are not absolutes. MHHutchins 20:21, 17 June 2009 (UTC)

The piece in question is in no way spec-fic, it is a classic hard-boiled pulp-style detection novelette /novella (50 pages of a large-format TP), with a reporter playing detective, included in a large anthology of other such works, mostly by authors who, as far as i know, wrote no spec-fic. (i do not that we already have some 25 non-genre novels by Brown on file, so some editors don't seem to take the same approach you do. But suppose for argument the author had been Avram Davidson. Just how would you suggest this be handled currently? How would the story be entered that indicated that it was both non-gene and not a novel?

By "general interest magazine" I meant exactly mags like Playboy, Argosy, etc (see General Interest Magazines

Currently if I enter an anthology that has both spec-fic and thriller but clearly non-spec-fic (not by authors "above the threshold") I omit the non-genre short-fiction, with a note that the contens are intentionally incomplete, only SF having been entered. (I would tend to include non-genre short fiction by "above the threshold" authors, but agian there is the problem of how). I would do this no matter how SF the anthology as a whole was. I feel that IN or OUT should be judged for shortfiction on more or less the same basis as for novels. I don't know that we've ever clarify that standard explicitly. Others might disagree with this practice. -DES ^Talk 21:59, 17 June 2009 (UTC)

If the story had been by an author that is known primarily for his spec-fic, and this is a rare mainstream or genre other than spec-fic story, there is no easy way to "slip" it into the database, unless one were a mod. Michael Bishop has published a fair number of non-spec-fic stories in the past decade, and, as a mod, it would have been easy for me to create pub stubs just to include just his pieces in the db. But I've not abused my position to do so. (Another reason, even a selfish one: I wouldn't want the ISFDB record be a substitute for the work I do for his online bibliography.) MHHutchins 22:50, 17 June 2009 (UTC)

I don't quite see how that would be abuse -- do you think another mod would not approve such, if you were not a mod? Would you reject such if you were reviewing an entry by a non mod. We have a documented and approved method of creating stubs for non-genre mags, and I don't see why a similar method can't be used for non-genre anthologies with some SF content. What I don't see is a way to enter the non-genre short fiction that I think should be IN (by whatever standard). I hate to just not enter such, but I can't see a good way to enter the short fiction works of even "overwhelmingly" SF authors. -DES ^Talk 23:19, 17 June 2009 (UTC)

Yes, we have the practice of creating pub stubs for non-genre sources that contain one or more spec-fic pieces, but that's not we were talking about. We shouldn't be creating pub stubs for non-genre sources just to add a single non-genre piece. That's another matter, and yes, I would reject any such submission as outside the ROA. What moderator would accept such a submission? MHHutchins 02:59, 18 June 2009 (UTC)

As for your practice of not entering contents that you consider non-spec-fic from an anthology that is primarily a genre publication, I can't disagree on grounds of ROA. You've got me there. I would argue that there are book-length works included in the database that the ROA would not include, but are here only because they were reviewed in a genre publication. By that (perhaps circular) logic, I would include all contents in that publication based on the context criteria I wrote about above. If my stand on books vs. shortfiction is contradictory, so be it. Like Walt Whitman, I contain multitudes. MHHutchins 22:50, 17 June 2009 (UTC)

It hasn't come up often for me. And it is in a grey area. -DES ^Talk 23:19, 17 June 2009 (UTC)

My take - Above threshold author, Any work, any where, just try not to drag too much non-spec-fic in with it (I support non-spec fic anthology entered per the rules used on non-genre magazines). Please make it a Non-Genre Short Fiction work (Once that gets added). For non-genre work in a Genre collection - In. The editor sold it to a Spec-Fic Audience, and it has merit, but feel free to mark it as a Non-Genre Short Fiction work (Once that gets added). Kevin 00:20, 18 June 2009 (UTC)

I can mostly agree with that. what if anything would you do until Non-Genre Short Fiction is implemented? -DES ^Talk 00:27, 18 June 2009 (UTC)

Thinking that non-genre short fiction is 120 days out or less, I would just enter the anthologies, and put a note "non-genre" in the title of the title in question. Come back to it when you can. Kevin 01:36, 18 June 2009 (UTC)

I must be in the minority in believing there's no need to create a separate listing for non-genre shortfiction. As I stated above, if it's included in a spec-fic anthology or magazine, it should be listed with the other shortfiction. If it's in a non-spec-fic anthology or magazine, it should not even be in the database. Why would anyone be entering anything in a non-spec-fic anthology or magazine that's not spec-fic? Allowing them into the db would open the flood gate for all the fiction published in Argosy, All-Story, Blue Book, and the like. (The Fictionmags group would love that.) MHHutchins 02:59, 18 June 2009 (UTC)

In the past, we were pressured by SF authors to include their non-SF -- regardless of whether it was book length or shorter -- since they argued that it was important to present a unified view of their bibliographies. (I am sure that self-promotion was never part of the equation! ;-) I don't think we have felt this kind of pressure in the last few years, so we can step back and look at it from our users' POV. How much value does it add to list all of Brown's mystery stories? Simak's westerns? Kuttner's military stories? Leinster's hundreds of non-genre stories? Is it something that we really want to do or do we want to limit non-genre fiction to book length Titles even for major authors?

I don't know the answer to this question, but we probably want to answer it before we implement support for non-genre short fiction. Ahasuerus 03:13, 18 June 2009 (UTC)

(Unindent) I was taking the RoA at face value. They currently say (#9):

In - Works (both fiction and non-fiction) which are not related to speculative fiction, but were produced by authors who have otherwise published works either of or about speculative fiction over a certain threshold (see below). This includes any non-genre works published as standalone books as well as non-genre short fiction...

and they have said this for over a year, i think since before I became an editor, so I thought that decision had been made long since. If we really want to exclude non-genre short fiction by "above threshold" SF authors, then we ought to revise the RoA -- once we decide what we want. I don't feel very strongly about the issue, but there is something to be said for the "complete bibliography for all significant SF authors" idea. Mind you: "IN" does not mean "Must dig up all examples next week", of course. Anyway I started this by seeing an example story that is IMO clearly IN by the above rule, but which there seems no good way to enter, pending a non-genre short fiction type being implemented. I will delay any such entries until/unless such a type is implemented. -DES ^Talk 11:53, 18 June 2009 (UTC)

Entering Reviews of items that are not 'Books' or 'Magazines'

This issue came up before I became a mod. I wasn't thrilled with the interpretation of the rules, but I went along with them. I've since made some noise, and it turns out that there is a variety of "What we do" when it comes to entering a review in the database, when the item reviewed is not a Book, not a magazine, and not a fanzine, or is otherwise 'out' by the Rules of Acquisition. With our recent development work, the discussion has also come up, because we are considering a new (or modified form) of the review record to better handle items like this. Below is the discussion from Michael's talk page. Kevin 23:27, 19 June 2009 (UTC)

That only points out the futility of trying to clear the list of "authors in the database due only to reviews" (or whatever it is called). All it takes is one title for the author not to show up on the list, even though the title in the database is not actually the title under review! My concern about changing this to an essay is that somewhere down the road graphic novels may be IN, and there would be no link to this title as being reviewed. Ah, well, just another reason to re-verify several thousand magazine issues. MHHutchins 03:27, 15 June 2009 (UTC)

(Shrug) - I'm satisfied that posterity can find the unlinked review/essay if we ever need it. The bigger concern is all those reviews totally undocumented in verified pubs. It take 5 seconds per issue to look for unlinked reviews. It will take complete re-verification everytime the RoA change to find undocumented items. Kevin 04:20, 15 June 2009 (UTC)

By "undocumented" do you mean "unlinked" pubs or just pubs that are not in the database? Sure, it takes 5 seconds to find the unlinked reviews, but sometimes several hours to create the pubs and link them to the reviews. Believe me, it happens more often than I care to admit when you're adding mags that have more than one hundred reviews. MHHutchins 04:48, 15 June 2009 (UTC)

By undocumented I mean that some verifiers are removing / deleting / not even entering reviews in any form (not even notes) for items that are out. Therefore, it's like the review never existed, and if the work is ever considered 'in' or the review gets a different treatment (non linking review) it doesn't exist for us to even find by reviewing the magazine publication record. Kevin 03:27, 17 June 2009 (UTC)

That's not good. Even when the review is of something that is obviously out (like a film or an audio recording), I still create a record for it. This comes down to a previous topic about how much detail should we expect of editors. The "standards" for magazine entry have changed so much over the past few years, and only recently have became somewhat stable. I'm not looking forward to going through hundreds of issues if the "standards" change again. MHHutchins 03:48, 17 June 2009 (UTC)

Verifiers deleting stuff certainly isn't good - are you sure that's happening? BLongley 18:05, 17 June 2009 (UTC)

That's not what I said, or, if misunderstood, it's not what I meant. I believe Kevin was talking about editors choosing not to record certain aspects of a magazine. Personally, I haven't deleted anything in my verified pubs. I've gone back and changed records from reviews to essays, and even added stuff that wasn't standard when I first entered the issue. That's what I meant by going back over issues that I've already verified. MHHutchins 18:23, 17 June 2009 (UTC)

No but it is what Kevin said, above. Whether anyone has actually been deleting reviews of things that would be OUT by the current RoA I don't know. My inclination would be to convert such to essays if I changed them at all, but I might well never enter such a review, only entering the overall review column as an essay, if I were entering contents for the first time. If this is a problem i don't see how to deal with it, unless/until we implement non-linking reviews -- and that won't prevent other future standards changes. -DES ^Talk 18:48, 17 June 2009 (UTC)

DES is right, the question is directed at Kevin. While we're stomping all over your talk page though, have any of you seen people removing data? I know I've seen Ron Kihara suggesting he would do so for a Filk review, and DES has just tried to dissuade such. BLongley 19:39, 17 June 2009 (UTC)

As a moderator I've not seen any submissions where an editor was deleting content records, whether the pub was verified or not, or even if it were their own verified pub. MHHutchins 20:02, 17 June 2009 (UTC)

As to standards for Magazine Entry, the level of detail seems to go by project. E.g. the recent addition of SFRA Review seems limited to Speculative fiction reviews only (but at least clearly states that). BLongley 18:05, 17 June 2009 (UTC)

I mean the systemic removal, deletion, and erasure of non-in item review records. Kevin 23:24, 17 June 2009 (UTC)

• See Ron's Talk User_talk:Rkihara#Stray_Robert_Coulson_Reviews where I queried about some stray reviews, (That I must assume had previously been in the magazine record - asked about them, and was told "Accordingly, I've left out all reviews of music in pubs I've verified". I don't know who removed the reviews from the magazines, but they used to be there, and were then removed (but not deleted until I found them floating in the ether). Kevin 23:24, 17 June 2009 (UTC)
• See Ron's Archive Removing Reviews where he previously removed some 'out' reviews from a magazine because the item reviewed was out, but neglected to delete them, instead putting a note in the title for the stray title. In that instance, the decision appears to have been to retain them for Magazines and Fanzines. Kevin 23:24, 17 June 2009 (UTC)
• See Help:Screen:NewPub#Reviews under reviews where it states "Note also that only books and magazines are entered; if the column reviews fanzines, you don't need to enter the review records for these, only the ESSAY record. Non-sf works should be entered but if an onerous number of non-sf-related works are reviewed in a column you are entering, discuss the situation on the Bibliographic Rules page to decide what can be eliminated. " (Emphasis Mine). This has been interpreted to mean.. "Only Enter Reviews for Books and Magazines" and for all other works, Only the Essay Record is required. So if there is a Review Column... it gets entered as an essay; then reviews of books and magazines in that essay get review titles, but the excluded items (not books, not magazines = Graphic Novels, Music CD's, Filk CD's, Song Books, anything in a form that is currently 'out' - Even if written / authored / sung by a Well known SpecFic Author) are not required to be mentioned in any way, shap,e or form. ( I swear I found a user talkpage discussion back in April that spelled it out, but I can't find it now Kevin 23:24, 17 June 2009 (UTC)
• See Fanzine Reviews discussion from 2006 that originated this text. Kevin 23:24, 17 June 2009 (UTC)

I see now that several people agree with me that this is a travesty of 'lost data', or at least believe it should be different. I tried bringing this up last month, but it got lost in the noise of Graphic Novels. Search Rules_and_standards_discussions for 'Travesty'. If I was king for a day, I would change the help and the rules so that if the item reviewed even smells of SpecFic, it deserves a separate essay entry that is findable Just in case we ever move the line of In/Out again, they can be found and properly linked/indexed/entered etc. Kevin 23:24, 17 June 2009 (UTC)

[unindent] You have no argument from me. And I wouldn't know what other mods are deleting, changing, removing, editing, etc. 'cause I got enough on my plate not to be looking over anybody's shoulder. Maybe this discussion should be taken to one of the community pages. MHHutchins 03:08, 18 June 2009 (UTC)

It's an RoA discussion perhaps. "What does this line mean", "What should it say to mean what we think it means"? Kevin 23:55, 18 June 2009 (UTC)

Huh? MHHutchins 02:17, 19 June 2009 (UTC)

Sorry - Sometimes my inner stream of consciousness skips the translator. I guess it's a discussion for the Rules and Standards Boards. Probably starting with a discussion of 'What does this line mean"? to you, moving to "How should it be re-written, in order to mean the same thing that we (as a group) think it should be saying". Thoughts? Kevin 03:41, 19 June 2009 (UTC)

I think there's always room for misinterpretation, but hopefully the standards are written in such a way to avoid ambiguities. Even then, you can't be 100% sure that no one reads it differently than you. Getting back to the issue at hand, would you like to move the discussion to the Rules and Standards page? I pretty much agree with everything you're saying, and maybe one or two moderators (from those discussions you pointed out) may want to air their opinions. MHHutchins 04:43, 19 June 2009 (UTC)

I think moving this discusssion to the Rules and Standards page would be a good idea -DES ^Talk 04:47, 19 June 2009 (UTC)

Conversation above this point moved from User_talk:Mhhutchins#Science_Fiction_Eye.2C_July_1989_-_Review_of:_Brought_to_Life:_A_Graphic_Docudrama Kevin 23:27, 19 June 2009 (UTC)

UNINDENT - I didn't realize it when I first tried to catalog various places in the wiki that referred to removing or ignoring some reviews, that most of them were from Ron Kihara. That's just what I happened to find. I know there were other conversations I previously read with similar interpretations, I just couldn't find them again when I went looking. I left a notice on Ron's talk page to come talk about this question, and in no way intended for this to become a discussion of any particular person's past actions. Only a discussion of, where we want to go from here. Kevin 23:27, 19 June 2009 (UTC)

I'm proposing that we change the language Help:Screen:NewPub#Reviews under reviews where it states "Note also that only books and magazines are entered; if the column reviews fanzines, you don't need to enter the review records for these, only the ESSAY record. Non-sf works should be entered but if an onerous number of non-sf-related works are reviewed in a column you are entering, discuss the situation on the Bibliographic Rules page to decide what can be eliminated." Kevin 23:27, 19 June 2009 (UTC)

I know what I want, but I'm unsure how the group as a whole feels about it, and unsure how it fits into the language above. I personally want every item reviewed, that even 'smells' of SpecFic to get an essay title mention, or one of the 'new' non-linking reviews once we create them (if we create them). I feel we should be documenting reviews of Music CD's, Songbooks, Plays, Musicals, Movies, Graphic Novels, etc. I feel that eventually we will likely revisit several of these 'types' of work, and later judge them to be 'in', and by not recording the fact that a review exists, we invalidate the supposed completeness of any record we are completing. Thanks for considering this, Kevin 23:27, 19 June 2009 (UTC)

I see your point. I have seen that some editors do not enter reviews of individual works of short-fiction, although the help would seem to suggest that such should be entered. I tend to agree with Kevin here, although I would support an option to enter a review column with a note that reviews of items not currently IV had been omitted, if the editor thinks that entering them all as essays would make the record cluttered. There have I think, been discussions about a "full" and a "concise" display, with things like reviews and interior art omitted from the concise version. Such a feature might help with the clutter issue.

I also recall some suggestion of a "subject" relationship or property. This could be used to link a biography of an SF author to the proper author record, and a critical study of a particular work to the work in question. (An obvious example would be The Road to Middle-Earth.) I'm not sure if allowing an essay to have a "subject" would be a way of dinessign a non-linking review -- probably it is better to just fully implement a non-link review, if we want one. -DES ^Talk 23:57, 19 June 2009 (UTC)

Note, I think songbooks would be IN under either RoA #8 "In - Works about speculative fiction..." or RoA #1 "In - Works of speculative fiction originally published in English". Most Filks and other SF-related songs are IME either "about" SF or are themselves works of SF at least as much as the poems we now record. And there is no RoA or other policy that explicitly excludes them, and i for one would at least as soon have them as fanzines. I have bought songbooks, and can't imagine buying a fanzine. Plays are already in, at least if printed. Music CDs are currently out by convention, but nothing in the RoA distinguishes them form audio books. Performed dramatizations, whether audio or video, are now OUT and i think should probably stay OUT forever, but scripts for such should IMO be IN, and reviews of such i think would be valuable to capture. I won't enter the Graphic Novel debate/minefield, but again reviews might be of value even if these are OUT. (Both movies and comics/GNs have their own bibliographic fora, and have specialist needs we would need a lot of change to handle well, IMO. But those fora are not likely to index reviews in SF mags or books, so someone wanting to find where, say Star Wars was reviewed when it first came out is out-of-luck unless we record such.) -DES ^Talk 23:57, 19 June 2009 (UTC)

There's a lot of stuff to read and digest here (no pun intended) and I'm trying to disengage from the ISFDB and get on with things I abandoned when I started entering stuff here, but I do feel a responsibility for the work I've done and if there's a unified consensus about how something should be done, I will go back and correct it.

The problem as I see it is that there is no clear consensus, and the rules tend to be conflicted. I generally do not delete items in verified pubs, and typically enter only additions or corrections, even if I feel something should not be there. I do not enter reviews of music, graphic novels, comic books, fanzines, movies, games, computer or otherwise, and puzzles or quizzes, and sometimes remove those if the pub is not verified. As mentioned elsewhere, listing reviewed items in those categories generates authors and titles tied to nothing but the reviews. I enter all columns as essays no matter what the subject matter, even those for games, fanzines, and so on, but I don't enter the reviews. In the case of fanzines, as many as two dozen may be reviewed in a single column, some of which may have survived only for only one issue. I don't enter reviews of graphic novels, since they are generally out and there are a couple of editors who have deleted hundreds if not thousands of them (I've noticed that Gaiman and Moore seem to get a pass). I don't enter letters, but don't delete them if they are entered. For a lot of stuff the rules are vague. I enter all original art, but I've noticed that some people don't enter unsigned or uncaptioned cartoons, or art that fills less than half a page/column. I enter all essays signed or not, but some editors omit essays of less than a half column.

I know that there has been a suggestion to document some reviews as essays, or to put them in the notes, but this is only manageable for a modern magazine. With everything I leave out, many pulps have near a hundred individual entries. Even recent magazines would become cluttered. IMHO, if you add 20 fanzine reviews, then 10 game reviews, 5 movies, and so on, and you'll have a mess.--Rkihara 06:07, 20 June 2009 (UTC)

We had this discussion a while back and decided that reviews of movies, games, etc (which are common in some magazines, e.g. Science Fiction Age) should be entered as Essays, one per article, rather than as Reviews, which resulted in Help changes. I am sure we can find the original discussions in the archive pages, but the gist of it was that we didn't want to create Review records that couldn't be linked to ISFDB Titles without creating movie/game/etc Titles and we definitely didn't want that.

I can see how some users who are looking for reviews of movies, games, etc may benefit from creating one Title record (either Essay or Review) per movie/game/etc reviewed, but I don't think the benefit outweighs the extra work and the extra complexity that would be added. We can certainly add this information to the optional Title level Notes field for now and see where it takes us. Ahasuerus 02:18, 21 June 2009 (UTC)

For Movies, Games and "things you can't own" (Performances (Plays, Musicals, etc), websites (but not webzines), Museums, Tours (Sci-Fi Cruise?)) which I think we can all agree will never be in, I have no problem with an Essay record, where the essay notes indicate the items reviewed, and the essay title is just 'Review Column MAGNAME ISSUE DATE'. What I would prefer we avoid in the future is 'Review Column Essay', no notes, and then the books reviewed... leading to absolutly no record that some other items were reviewed. If it ever comes up, we can revisit notes (perhaps we can suggest a key phrase "Other Items Reviewed:" or similar to be included in the notes), and essays for research or for expansion to non-linking reviews or essays, if someone cares to donate the labor. My concern was the loss of all record, and the invisibility of the loss. Kevin 03:34, 21 June 2009 (UTC)

For anything (except Movies, Games, etc) which has a SpecFic story embedded in it (Graphic Novel, Filk CD, Music CD, Dramatization, etc) which is currently out, and there is some possibility that it could be 'in' someday (or rather if there is some disagreement today about it being out), I would appreciate an individual essay or non-linking review. (But I can live with notes as described above... again so they can be revisited if the need arises, or the extra labor is volunteered). Kevin 03:34, 21 June 2009 (UTC)

Paperback sizes and codes

See the discussion at ISFDB:Moderator noticeboard#Pending edits, from which some of these comments are adapted. -DES ^Talk 14:11, 24 June 2009 (UTC)

The current help is pretty clear I think: it says "Trade paperback. Used for anything significantly larger than a paperback. 7.5" x 5" is a common size, but there exist many variant sizes larger than this..." (bolding added) and "Paperback. Typically 7" x 4.25"; low-height paperbacks such as Ace Books from the fifties are about half an inch shorter" I don't think that says that 7.5 in is the minimum height for a TP. When using OCLC or other library records that give measurements in cm, I have been assuming that 19cm is probably a tp, and 19.5 or higher surely is. (n.b. 7 in = 17.78 cm, which will be reported as 18. 7 3/4 in = 19.685 cm, probably reported as 20)

However, it used to be the case (at least in the US market) that a height significantly over 7 in was associated with different (and higher quality) binding techniques, significantly higher price points, and a different distributions system. But grocery stores in the US (part of the traditional "Mass Market" system) are now selling some books in a "reading format" basically a tall (1-2 in extra) MMPB, priced and marketed like a MMPB. So far I haven't seen any sciecne fiction in this format, but there probably has been some paranormal romance and light horror. Kraang notes that this book from Pyr is 7.75 in (19.685 cm) tall. He also notes this earlier edition which has a higher price, and is 23cm high accordign to OCLC, and generally seems to be a "traditional" trade-paperback. If we use "tp" for both then the difference may not be clear to users.

None of the traditional differences between MMPBs and TPs hold true for the "tall" pbs we are now seeing. They are designed, priced, and marketed like MMPBs.

We may want to create an "opb" (oversized paperback) for such "tall" MMPBs, or we may want to stick to "anything larger than pb is tp" which for height probably means anything over 7 1/4 or 18.5 cm would be tp.

I think we need to decide what the standards will be going foreward. We can then worry about specific past cases separately, IMO. -DES ^Talk 14:11, 24 June 2009 (UTC)

I'm all for standards. Any standards we care to define for size is fine with me. As size is not a subjective discussion (i.e. an 'In' or 'Out' discussion) it should be fairly simple to define, and fairly simple to apply. How about pb = 7x4.25 or less, Tall Paperback tpb = Mx4.25 where M>7, Deep Paperback dpb = 7xN where N>4.25, Trade Paperback tp = MxN; Additional Option: Large Trade Paperback ltp = Larger than normal hardcover dimensions (either height or depth) whatever those are, like the XXXX for Dummies books, or some Programming books etc. Size codes like this should satisfy all desires to separate based upon shelving requirements (Height, Depth, and Oversize) Kevin 23:37, 24 June 2009 (UTC)

I think that "tpb" would be a mistake, as it is too easily confused with "tp". I also think this is more different categories than there is need for. "tp" has always covered books with widely varying sizes, so I don't see the need for "ltp". Most vendors list actual sizes so anyone considering purchase will be able to see if the size matches, or we could record actual size in notes for unusually sized books. To the best of my knowledge overly deep books are not common enough to warrant creating "dpb". I don't see any need for more than one new class, to handle books somewhat larger than 7x4.25 in (18x11 cm), say up to 8x5.5 (or whatever) that are similar to traditional MMPBs in binding, distribution, and price point. If they are not created and marketed like MMPBs, just call them TPs and be done with it. That's my view. -DES ^Talk 23:52, 24 June 2009 (UTC)

How would we identify 'Created and marketed like MMPBs' from secondary sources, library records, and books in hand with objective reproducible results? Kevin 00:02, 25 June 2009 (UTC)

We can't always, but then even now we can't always tell whether an item is a pb, a tb, or an hc. Our best guide is price points, when these are available. When he have just a size and no price, we often can't even now tell the difference between a tp and hc, as tps are normally bound from the same pages as an hc, and so have the same size.

If you say that construction and marketing doesn't matter, then what is the point of a new code. Simply clarify that anything larger than 7x4.35 is tp, and call all the new "tall" format books tps. The only reason to distinguish them from tps is that they differ, not in size (they at least overlap the traditional tp size range), but that they are larger than standard MMPBs while sharing construction and distribution/marketing and pricing with them. it used to be that size correlated well with construction, distribution/marketing and pricing. It no longer does so as reliably. That is, IMO, the reason for considering a new code. We need to decide whether the code is to indicate size only, or something more. -DES ^Talk 00:10, 25 June 2009 (UTC)

Perhaps we need to start capturing and reporting physical dimensions? Those are at least objective, and are reported by most library catalogs and by Amazon and some other online vendors. -DES ^Talk 14:17, 24 June 2009 (UTC)

xpb(extra dimension paperback) Where the "x" refers to the extra dimension of the book be it height or depth(not both) but the book is priced and marketed for the mass market. I don't see it being used a lot but it would classify what's being published by Orbit, Gollancz and others. Users would then be able to distinguish the different publications listed here. Most of these books have already been entered by Dissembler as "pb" since that's how Amazon classifies them.Kraang 00:46, 25 June 2009 (UTC)

I have no objections to that. FYI fixer/dissembler bindings are quite unreliable -- I have found hardcovers imported as pb, albeit rarely, and I have found MMPBs listed as tp often. -DES ^Talk 01:34, 25 June 2009 (UTC)

I think I'd have highlghted a different offending word in current help. "Trade paperback. Used for anything significantly larger than a paperback." To me, a book that disrupts all my paperback organisation work is significant. That's not just shelf height, it's shelf depth too as I have to double-stack now. As it currently stands, the definition of "pb" works for me - at worst, some old Ace paperbacks will be overstated but they still fit. "tp" has always been the catch-all category for anything bigger. Might be hc size, might be tp width but pb height, might be A3 artwork albums, it's so far been used for anything bigger than normal - in any dimension except thickness (and I'd love to be warned about books like that anyway, but I can guess that from page-count). I'm happy to accept a sub-division of "tp" and will go find all the "tp" books I have and reclassify them - but if there's a redefinition of "pb" you will have ruined one of the major aspects of ISFDB for me. I do not buy anything larger than a "standard paperback" unless forced into it by it being only available that way, and being really, really, desirable. If you make "pb" vaguer I cannot trust ISFDB for book-buying guidance anymore. Keep "pb" as is, redefine "tp" at will. Or demand rigid rules for "tp" and invent something else that won't affect me much. BLongley 22:36, 25 June 2009 (UTC)

We seem to happily added new things to the database why is this so different? I'm not suggesting changing the "pb" defintion just adding a new classification "xpb" which would deal with the new size MMPB. One other thing to keep in mind is the ISFDB is used by hundreds if not thousands of people not just one person. Generally I'm not this agressive but on this issue I won't back down until it's resolved.Kraang 23:32, 25 June 2009 (UTC)

I have no problem if you're not changing "pb". But the original problem was that you defined one particular pub as "pb" whereas I think it's not - I'm happy if it becomes "tp" or "xpb" or any of the other options. I definitely agree that "tp" could be usefully broken down into other categories, but as all such are warning signs to me I'm probably not the best person to ask about how to do that - but go for it, we're not actually at odds with each other at all. Peace? BLongley 20:22, 26 June 2009 (UTC)

The advantage in including price/marketing as a factor for determining "tp" is that our definition would actually reflect the actual definition of the term in the real world. Any paperback edition which has a dimension exceeding a certain height or width would then be some sort of a large paperback, an "xbp". In any case, I find Bill's definition for "pb" based upon storage requirements also is valid for me. Anything more than two or three codes would probably create more confusion.--swfritter 23:40, 25 June 2009 (UTC)

Poll

(unindented)Are you in favour of the "xpb" classification or against it?

Up or down vote YES or NO

1. Yes-KraangKraang 00:13, 26 June 2009 (UTC)
2. No-until we know what specific changes will be made to Help and Standards and how this will affect the definition of "tp".--swfritter 14:57, 26 June 2009 (UTC)
3. Vaguely Yes - as in 'it seems to need sorting out from "tp" and "pb"'. And vaguely "No" as it shouldn't be a final solution, if people have some definite ideas about the various "bigger format book" options there are. "xpb" would be an interim solution that should make it easier to find examples of controversially-sized publications. BLongley 21:33, 26 June 2009 (UTC)

   But if people want to separate based on marketing, that's completely different from the size issue, and I doubt it can be resolved here. US "Mass Market" publications are more likely to be "bookshops only" here in the UK, and "TP" editions are the ones more generally available to the "mass market" here. BLongley 21:33, 26 June 2009 (UTC)

4. Tentative No. If we can't make a useful decision based on some combo of marketing/construction/price point; if the only factor to be considered is size (as Bill claims above, with some plausibility), than I see no value in a new code, and would prefer to call anything larger than a standard MMPB a tp, as the current help calls for. This would mean designating the "large-format MMPBs" as TPs. I think this is somewhat misleading, but a new code without a distinction based on more than size would IMO be worse. -DES ^Talk 21:52, 26 June 2009 (UTC)
5. Firm NO as explained below. -DES ^Talk 04:31, 27 June 2009 (UTC)

Post-poll discussion

• I would be in favor of a new xbp code (although i would prefer "opb") to be defined as a book distributed, priced, and constructed like a MMPB but with dimensions up to 8.5 x 5.5 in. TP would now signify any book with significantly higher price points or constructions standards, or which did not use the mass-market distribution system (in the US at least), or which was even larger than the limits for an xpb/opb. If people feel that would be too tricky, than I would favor retaining our current classifications, and simplay saying that anything notably over the standard MMPB size {7 in (18 cm) x 4.25 in (11 cm)} would be a tp -- say anything at least 7.25 in (19 cm) x 4.55 in (12 cm). I think it is too soon for a straight up-or-down vote on a single proposal. -DES ^Talk 15:42, 26 June 2009 (UTC)
  • To those saying "This will disrupt my shelving": do you really buy books based on ISFDB listings without looking as the vendor's detailed description? Most vendors, including Amazon and ABE, list size in cm on the ordering screen. The ISFDB is not, i thought, primarily intended as a giant book ordering system, although it can and does help people find books to buy. -DES ^Talk 15:42, 26 June 2009 (UTC)

I do actually find the ISFDB useful in its "pb" usage as I mostly buy obscure, older, cheap, paperback second-hand books that do not have a detailed Amazon listing that includes dimensions, and find that the Amazon seller, ABE or Alibris listings are mostly just copies of the Amazon data - they usually don't make the effort to expand the information for cheap books. (Half of the books I buy online are 1 penny plus postage - as postage is a standard £2.75, the profit is actually made from the fact that it will usually cost under £1 to post. And if I order multiple books at the same time from the same vendor, they're laughing.) Yes, of course I could ask each vendor to clarify exactly what they're selling before I place an order - but if I had to ask them about books clearly identified by ISBN each and every time, I think that would point out that what we record isn't useful. As it is, I buy several hundred books a year (averaging more than one per day, last time I checked) and if I had to check each and every one in advance I would probably drop that to mere dozens. Yes, I've accidentally bought a few TPs that I wouldn't have bought if I'd known, but being mostly right here saves me a lot of effort in checking, and keeps the dealers busy and (hopefully) profitable without having to go check each book and answer emails over something that will only make them a few pennies. BLongley 21:00, 26 June 2009 (UTC)

Really? Well if that's how you use the ISFDB (or one way in which you use it) then perhaps others do the same. I buy a lot of books, but not nearely that many from online sellers, and although i do separate MMPBs from others for shelving (TPs are shelved with HCs) I haven't found it to be a problem, and i will routinely buy used TPs if the price is right.

Perhaps we ought to consider recording actual physical dimensions, as OCLC and other library catalogs do. Could help avoiding suprises. In the meantime, i hope the xpb or opb idea above will handle the matter, provided we agree on how to distinguish an xpb/obb from a tp. -DES ^Talk 21:28, 26 June 2009 (UTC)

As with Chapterbooks/Chapbooks a term with a fairly specific meaning, in this case "tp", has far too general a definition in the ISFDB. The problem now is that there are a huge number of titles that would have to reclassified based upon any changes in standards. I would prefer to focus on stories and authors rather than get bogged down in detailed descriptions of the physical characteristics of the publications in which they appear.--swfritter 00:59, 27 June 2009 (UTC)

What do you see as the specific, real-world, meaning of "trade paperback"? I would say it primarily indicates a specific distribution channel, secondarily indicates quality of manufacture, and only in a tertiary sense indicates the size of the pub. I would also say that that specific meaning is largely US-centric, as the correspondence between publication sizes and distribution methods is diffrent in the UK, and perhaps elsewhere. -DES ^Talk 01:06, 27 June 2009 (UTC)

The help pages indicate the size of a "tp" to be 7.5 x 5 inches which I believe is incorrect. All of the trade paperbacks I own are at the extreme minimum 8 x 5. and are all priced as you would expect them to be $12(US) to $16(US) on average. These new MMPB are all between 7 and 7.75 inches high and 4.25 to 5.25 inches deep, these are priced the same as regular paperbacks in the US and the GB versions are also priced the same as their ordinary paperbacks(7 to 8 pounds which includes the vat tax). My experience tells me these books are in the extreme minority at the moment so there would be very little to change but looking forward these type of books could become more prevalent. Orbit and Gollancz are the only GB publishers that I know of but I'm seeing these tall size paperback more and more.Kraang 01:43, 27 June 2009 (UTC)

My personal definition of a trade paperback is any perfect bound softcover book larger than a traditional (As defined by the last 60 years publishing, 1949-2009) mmpb. I don't care if it could have been pulped, or stripped, or sold in store type a, or store type b, or sold at one price point or another. I think it describes a type of book I have in my hand. I further believe that any new definition we agree to should be identifiable from the published work alone (In hand, in a cave, with a candle, and maybe mmpb for comparison) by any individual who has reasonable level of experience holding pb and tp books in their hands. Don't forget, we have to have standards that we can explain to anyone who wants to contribute, and that can be determined solely from the book in hand, 30 years after publication / printing. Kevin 02:17, 27 June 2009 (UTC)

The powers that be in the publishing industry call these new books MMPB, the larger versions "Trade Editions" are you suggesting they're wrong. Your also correct that for the last 60 years the standard "pb" was 7 x 4 and it still is but like everything else new thing arrive and we can't live in the past. The book trade is changing like everything else and like my business if I only looked back I'd join many of my fellow dealers(not a book dealer) and be gone. It's time to add something new, don't be afraid of change embrace it.Kraang 02:43, 27 June 2009 (UTC)

I understand that things change, and I'm all for being inclusive in the database. The problem here is that (as you have described) the definition of the phrase mmpb has changed. Unfortunately what was previously a single definition had two different words to describe it. One could use either 'Paperback' or 'mmpb', and generally know that you were discussing the same animal. When you used the term 'tp' you also generally knew what kind of thing you were describing. Now... due to some marketing shifts and manufacturing price points.... the term mmpb is being applied to things that look like tp. The definition has changed, and our old comfortable world where we used to use pb or mmpb to describe the same thing is going away. Now we need to face the reality that what used to be one definition with two words, is splitting. Do we continue to use the old word (pb) to describe things that look different (A bad idea IMHO), or do we define new words (one possible solution), or do we ignore the change in one definition (mmpb) and revert to using only a sized based description system. (Which is what we have today in sheeps clothing, and which could simple be expanded and clarified, and could involve defining new words). My goal in posting was not to say that a new category was a good or a bad idea. It was simply to redirect and dissuade any discussion of basing that category on marketing or pricing concerns. Kevin 03:48, 27 June 2009 (UTC)

A size based description system can be applied retroactively, evenly, and objectively. Any definition which requires knowledge of how a book was marketed on a wholesale level is doomed to failure. Kevin 03:48, 27 June 2009 (UTC)

(Unindent) A purely size-based classification can indeed be applied objectively, although not uniformly because in some cases we lack data. But it is also pointless. The think that unites what we call TPs, and distinguishes them from what we might perhaps call xpbs is not size. If we aren't willing to makes some attempt, however flawed, to capture the marketing/distribution/construction/price factors that generally go together to make a tp distinct from a mmpb, and that leave the new "larger pbs" more closely aligned with the mmpb, then there is as far as i can see no point to having one category for heights from 7.25 to 8 in, and another over 8. In fact, you've convinced me, we can't objectively capture the true difference, only the size. So I now firmly oppose any new category or code, let us stick to the descriptions as written in the help now, anything significantly over 7 x 4.5, say anything at 7.25 or 4.75 or higher, is a tp for our purposes. I might add that i hold in my hand as I type this, a trade paperback (non-fiction, ISBN: 0-89558-804-1) clearly marketed and bound as such ($9.55 in 1991). It measures 7.75 x 4.75 in. It is not an xpb. But then nothing is, since there is no definition of such a thing. (I might add that with a book-in-hand quality of binding alone would make the distention I was suggesting above, as would price. The only problem in telling the difference is when working from secondary sources) -DES ^Talk 04:29, 27 June 2009 (UTC)

It might actually make sense to use a script to replace all instances of "tp" with a more generic code to avoid confusion and accurately reflect current practice.--swfritter 13:35, 27 June 2009 (UTC)

What "more generic code" would actually be of meaning to our users at large? The code "tp", while not absolutely matching the outside world's use of 'trade paperback", does match it in the overwhelming majority of cases. Anything that changed all those pubs to some other code would be, IMO a major mistake. Indeed I no longer see a reason for any change at all. -DES ^Talk 13:51, 27 June 2009 (UTC)

If you see no need for a change, could you (or anyone) suggest some different wording for 'Help' to make our current situation more well defined? Kevin 13:58, 27 June 2009 (UTC)

The help currently says:

Trade paperback. Used for anything significantly larger than a paperback. 7.5" x 5" is a common size, but there exist many variant sizes larger than this, all of these should be simply recorded as "tp". This includes large format artbooks in paper covers, for example.

I would clarify it to say:

Trade paperback. Used for anything significantly larger than a paperback. Thus any book that is at least 7.25 in (or 19 cm) tall, or at least 4.5 in (11.5 cm) deep, and is not a hardback, is a tp. 7.5 in x 5 in is a common size, but there exist many variant sizes larger than this, all of these should be simply recorded as "tp". This includes large format artbooks in paper covers, for example.

Unless we are going to include factors other than size, i think this is the best definition we can fomulate. -DES ^Talk 14:12, 27 June 2009 (UTC)

Do we need the qualifier "significantly"? Shouldn't the term be wide instead of deep?--swfritter 15:26, 27 June 2009 (UTC)

"Significantly" is meant to indicate tha differences produced by differeing trim don't matter, that a hieght of 7.05 in is not to be counted as different from 7 in. But it could be dropped. "Deep" fits if you think of the book sitting/standing spine-out on a shelf, perhaps "wide" fits better if you think of it laying cover-up on a table. Either term will do for me - discussion above had used depth. -DES ^Talk 15:37, 27 June 2009 (UTC)

I think there is reduced confusion and less redundancy by removing "significantly". I think "wide" made sense back in the days when paperbacks were actually displayed in stores so that you could see the cover. Unless anyone else has an opinion "deep" is OK with me.--swfritter 14:11, 28 June 2009 (UTC)

To me, "deep" in context of a book suggests "thick". -- Dave (davecat) 22:28, 29 June 2009 (UTC)

It's typically best when considering one definition, to also consider the wording in the sister or neighboring definition, so I've reproduced pb below. Kevin 14:51, 28 June 2009 (UTC)

Paperback. Typically 7" x 4.25"; low-height paperbacks such as Ace Books from the fifties are about half an inch shorter, but this does not need to be noted.

With this in front of us now, I'm going to recommend that pb be amended as follows Kevin 14:51, 28 June 2009 (UTC)

Paperback. Typically 7" x 4.25" or smaller, though trimming errors can cause them to sometimes be slightly (less than 1/4 extra inch) taller or deeper. For books taller than 7.25" or deeper than 4.5" see Trade Paperback (tp). Low-height paperbacks such as Ace Books from the fifties, and similar publications are about half an inch shorter. Variations in size below the typical 7" x 4.25" do not need to be noted unless significant.

Please suggest any improvements, I'm sure it could be better. Kevin 14:51, 28 June 2009 (UTC)

I'd be happy with this pair of improvements to help. If there's still some confusion over "wide" versus "deep" then just say we don't mean "thick", as that is mostly covered by page-count. This whole kerfuffle started over whether a "significant" change included a "significant" change to width/depth, or only to height. We seem to have got that sorted, so I'm fine. BLongley 21:43, 28 June 2009 (UTC)

Perhaps we could work in the term "shelf depth" for tp?--swfritter 15:19, 30 June 2009 (UTC)

(unindented)It would appear the majority wish to leave things as they are so I'll drop the matter, but please rewrite the definition so it's less confusing. Thanks for participating!Kraang 00:53, 28 June 2009 (UTC)

Hopefully the proposed changes to help will actually help? BLongley 21:43, 28 June 2009 (UTC)

Wouldn't it be a good idea to notify us that Help has been updated when an issue has apparently been resolved? Was the addition of A4 and A5 magazine sizes discussed somewhere other than this page?--swfritter 15:32, 4 July 2009 (UTC)

I apologize, I somehow forgot to post that notice. I think i recall discussion of A4 & A%, and I recently entered several magazine issues that, like others already entered for the same publication, used A4. If anyone thinks A4/A5 are a mistake, it is easy enough to take them back out. -DES ^Talk 15:51, 4 July 2009 (UTC)

It does seem a little inconsistent to limit ourselves to two different paperback sizes while expanding the magazine designations. But I think there are probably more consistent industry standards for magazines. That might be because most magazine publishers have to farm out the printing tasks and are limited to a finite number of sizes. Browsing through the internet I could not find a central listing of the various sizes but if there were one it might make sense to adopt that list. Our magazine binding standards are quite often ignored which means they need to either be more strongly policed (dropdown list?) or expanded.--swfritter 20:28, 4 July 2009 (UTC)

The help does seem to be a bit confusing now. Why is "Digest" listed twice for instance? And audio help mismatches what we actually have: "audio CD" and "audio cassette" far outweigh the suggested conventions. "mp3 audio" is a bit ahead of "audio MP3 CD" still though, but we have "audio (MP3)" catching up. Maybe somebody is following the standard documented in help (that I don't recall anyone agreeing on) and someone is following common practice instead? It might be nice to pin this down or we'll never make it a simple dropdown list. (Which I think will still have to include "Other, see notes".) BLongley 22:37, 4 July 2009 (UTC)

None of those were change by recent changes, they have been like that for months if not years. The two "digest" mentions are for its 9rare) use on books or book-like objects, and for its common, and rather different, use on magazines, as I understand the matter.

See #Audio Binding conventions below

Back on topic - A4 and A5 are actually paper sizes, and work very well for Fanzines, being the easily available options for self-publishing Europeans. If size is the definition, then many BSFA publications and some magazines like Interzone fit perfectly with paper-sizes. Even some odd TPs like the Josh Kirby Posterbook fit - that's A3. (Fold an A3 in half, you get A4: fold an A4 in half, you get A5. There are oddities like Foolscap though, but they're uncommon and getting rarer, unless that actually matches a US definition.) BLongley 22:37, 4 July 2009 (UTC)

At least some magazines are listed by Locus, and from locus here, as A4. See Magazine:Odyssey (1990s) for an example. -DES ^Talk 23:00, 4 July 2009 (UTC)

Anyway, if we're going to pin this down then a better definition of "Bedsheet" and "tabloid" and "octavo" and "quarto" is needed, at minimum. The first two make me think of newspaper sizes (although I know I'm really thinking of "broadsheet" for the first) and the latter makes me think of Shakespeare. (And isn't covered in help anyway.) We might need the newspaper sizes for Non-Genre publications though. And we could clarify pbs and tps better - for instance, many pbs could be called "A" format (wouldn't include the Ace smaller pbs though) and tps "C" format and we can split the tps into "official tp size" and "bigger than normal book". But we seem to have 197 formats currently, and even allowing for a possible intent to break-down tps a bit further, and a lack of direction on audio and ebooks, there seems to be no need for more than about 16. BLongley 22:37, 4 July 2009 (UTC)

I won't post all 197 options (I suspect most of us have got used to "hc" rather than "hb" or "Hardcover" and would just clean the others up, for instance) but I'd happily support (another) attempt to standardise them. We DO have the coders available to restrict things now, we just don't have a realistic help page to work from, or a nice big discussion of why "Limited" should or should not be part of the format. BLongley 22:37, 4 July 2009 (UTC)

Oh, and please remember that we haven't sorted out the " in publication format bug yet (as far as I know) so posting in "inches" abbreviations is likely to accidentally lose such at the first ". BLongley 22:37, 4 July 2009 (UTC)

I don't see any reason to post an exact size in inches in the format field anyway, adn I would change such a designation if I found it in a pub that wasn't primary verified.

I prefer to enter all ebooks as simply "ebook" with specific formats in the notes. So many ebooks are in multiple formats these days, anything else seems foolish to me. I don't know enough about the older magazine formats to have an opinion that is worth anything there. Some OCLC and LoC records still use "8vo" and "12vo", so it helps if we know what those mean, even if we aren't going to use them here. -DES ^Talk 23:00, 4 July 2009 (UTC)

Ebook Binding conventions

split off from #Post-poll discussion, above. -DES ^Talk 04:37, 5 July 2009 (UTC)

I prefer to enter all ebooks as simply "ebook" with specific formats in the notes. So many ebooks are in multiple formats these days, anything else seems foolish to me. -DES ^Talk 23:00, 4 July 2009 (UTC)

Now that we know different ebook formats are "supposed" to have different ISBN's there is more of an argument for including the format as part of the binding description. But considering the chaotic and often non-standard manner in which they are actually assigned and often not assigned at all it's probably better not to bother. "audio (MP3)" is creeping up because I am working on LibriVox pubs and actually took the time to consult Help. My own interpretation is that "audio (CD)" includes both regular CD's and mp3 CD's while "audio (MP3)" refers to internet files but other interpretations would also make sense.--swfritter 00:27, 5 July 2009 (UTC)

"Supposed" according to whom? None of the major ebook publishers do this. Baen doesn't; fiction wise dosen't; Project Gutenberg surely doesn't (of course they don't use ISBNs at all, but they don't assign separate etext numbers for different formats). Until there is some sign that practice is following any theoretical standard here, I don't think it need matter to us. Also, given how many evooks are sold in multiple-format editions, where if you buy one format you buy several (all of Baen, most of PG, most of Fictionwise) I don't see the point in recording what is essentially the same thing many times over. -DES ^Talk 00:41, 5 July 2009 (UTC)

Per the ISO Standard defining assignment of ISBN's, anything sold separately is supposed to have a separate ISBN but there is no enforcing body, there is no penalty (by law or economic means), in fact in the US there is an economic bonus (reduced cost) for ignoring this part of the standard. It is no where near remotely applied outside of the Major Major publishers as written. Kevin 04:04, 5 July 2009 (UTC)

And even if that standard were to be followed, lots and lots of ebooks are not sold separately on the basis of format, and so would not get separate ISBNs -DES ^Talk 04:40, 5 July 2009 (UTC)

When they're not sold separately I'm happy with "ebook", formats can stay in notes. When they are sold under different ISBNs, I'd prefer separate publication entries for each ISBN. Those could stay "ebook" (ISBN differences should be enough to stop accidentally "Dup Deletion") but I could be convinced that separate formats might help distinguish all the ebooks on the summary pages. BLongley 22:27, 5 July 2009 (UTC)

You have a point. W hen different ISBNs have in fact been assigned to different ebook formats by the publisher, then multiple pub records seem warranted, as we have no good support for multiple ISBNs in a single pub record. In such a case, whether to list the bindig as "ebook" or, say "ebook (PDF)" or some such might be debated. I do hope that if the specific format is put into the binding it will be in a form like "ebook format" or better "ebook (format)", so that a binding search on "ebook" will find all of them. -DES ^Talk 22:38, 5 July 2009 (UTC)

I could support ebook Multi-Format for the combo isbns (And Gutenberg, Baen, etc) and ebook Format1 and ebook Format2 etc for the publishers that use separate formats (usually encrypted). And for those that don't encrypt, but that do sell under multiple ISBNs I could even go for ebook Format1 unencrypted or some similar designator to let me know that it's an unencrypted format. Kevin 22:53, 5 July 2009 (UTC)

Audio Binding conventions

split off from #Post-poll discussion, above. -DES ^Talk 04:37, 5 July 2009 (UTC)

...audio help mismatches what we actually have: "audio CD" and "audio cassette" far outweigh the suggested conventions. "mp3 audio" is a bit ahead of "audio MP3 CD" still though, but we have "audio (MP3)" catching up. Maybe somebody is following the standard documented in help (that I don't recall anyone agreeing on) and someone is following common practice instead? It might be nice to pin this down or we'll never make it a simple dropdown list. (Which I think will still have to include "Other, see notes".) BLongley 22:37, 4 July 2009 (UTC)

If we standardize the audio conventions, one way or another, I would favor a scripted mass update to convert equivalent formats into a standard set. -DES ^Talk 23:00, 4 July 2009 (UTC)

I recommend caustion for any automatic merging of audio formats without due consideration of the formats involved. Audio does not require a CD, or any physical medium. MP3 could be on CD, on DVD, or just MP3 as a downloadable file. Kevin 23:57, 4 July 2009 (UTC)

Of course. I am recommending merging only forms identical in meaning, but different in form. For example if we have both "Audio CD" and "Audio (CD)" on file, and we standardize on one of these, the other can, i think, be safely merge with it. If we have "Audio (MP3)", "Audio MP3" and "MP3" on file, those all mean the same thing, and can be merged. I agree that "Audio MP3" can't be mass merged with "Audio MP3 CD"; but "Audio MP3 CD", "Audio (CD MP3)", and, "CD MP3", and "MP3 CD" should all be safely mergable. -DES ^Talk 04:37, 5 July 2009 (UTC)

Ahasuerus is being extremely cautious with mass updates (understandably) and it might be that this will be another case where we are expected to manually fix the aberrations, if we can agree what the aberrations are. In which case, adjusting Help to what we actually do is far easier than changing what we should do on all existing entries. (E.g. Although "hb" would be a far more obvious abbreviation than "hc" for hardback books (as we call them in the UK), I'll stick with current conventions. We seem to have an agreement on what those mean.) Separating the CD formats looks useful, as does clarifying any or all "audio" formats on physical media. Not my speciality at all though - I don't do audio-books - but if I did I'd want to know which would fit my cassette-player, which my CD player, which were available for my MP3 player, and which were DRMed to the hilt that could only be played on Microsoft-Reader or such. BLongley 22:49, 5 July 2009 (UTC)

"uncredited"

Stumbled across [this] page, and must admit to giddy shock/dismay/hysterical laughter..... not that there is a page that lists all the "uncredited" miscellanea, but that someone or some software oddity has entered that "uncredited" has been used as an alternate name by two artists. It gets better! "Uncredited" has used alternate names!!!!!! How did/does this kind of daft entry ever get posted? Can it be fixed? Does "uncredited" have a union? ~Bill, --Bluesman 03:00, 3 July 2009 (UTC)

Info in=info out. If we tell the database that the author of a work is "uncredited", it can't interpret that as an uncredited work, but as the work of an author whose name is "uncredited". So it's natural that it would combine all of this "author"'s credits on one page. The alternate names just means that someone set up a pseudonym for a work that is uncredited but he has strong evidence who the author/artist who created that work. I wouldn't do it that way. I'd simply set up a variant for the work and not create the pseudonym. Maybe the person who did this had another interpretation. MHHutchins 04:06, 3 July 2009 (UTC)

This will be one of the first pseudonyms up against the wall when the revolution comes, if I have my way - uncredited, anonymous, unknown shouldn't be allowable pseudonyms for or of anyone else, IMO. BLongley 23:27, 4 July 2009 (UTC)

Okay, some of that is mine. At least, on quick glance, all the Perry Rhodan in which uncredited is used. My conscience truly bugs me on the issue. Forrest J. Ackerman is definitely the uncredited, but he did not bother to state it anywhere in those PR pubs. Can I put the genie back and go change it to Forrest J. Ackerman and then just note that he did not know that the ISFDB format would take his crediting away? LOL Something like "Forrest J. Ackerman took over the Perry Rhodan franchise and this was his vehicle" even though he was remiss with most of the accepted publication rules. More pointedly, FJA was a notorious creator of pseudonyms and I believe also followed the practice of taking over readers/commentators names to make comments not directly attributed to himself. At least one contributor to Perry Rhodan is two to four individuals, depending upon what FJA did. Thanks, Harry. --Dragoondelight 00:36, 6 July 2009 (UTC)

When no one is credited explicitly, then our records ought to say so, IMO. If there is very solid evidence of who the actual author is, that I think the author can be listed as such, with a note indicating that the author was not credited in the normal way, and the source for awarding credit. If there is any plausible doubt, I would use uncredited as the author, and list any indications of authorship in the notes. -DES ^Talk 01:43, 6 July 2009 (UTC)

If I understand Bill correctly, what he is saying is that putting "uncredited" in the Author field is not the best design possible and ideally we would have a separate field for "uncredited". Is my understanding correct, Bill? Ahasuerus 02:07, 6 July 2009 (UTC)

I'm not sure what Bill meant -- I think he was perhaps just surprised that there were cases that a "real" name would ever be listed as a pesud of "uncredited" or vice versa. It does sound odd until you understand why things have been entered that way -- as a way to indicate that texts were in fact not credited, but we nonetheless know who wrote them with fair assurance. Whether this is the best way to handle that case, either with the current software or with enhancements, might be debated. We need to handle at least two fairly common cases: a) text (such as an into or appendix, occasionally a work of fiction) is not credited and there is no good evidence as to who in fact wrote it; b) a text is not credited as such, but there is fairly clear evidence attributing it to a know author (or authors). Handling both of these so as to make each clear, and each distinct from works credited in the normal fashion, is not trivial. an "uncredited flag" would handle case b) well, but leave case a) untouched, if I understand how such a thing might work correctly. -DES ^Talk 02:16, 6 July 2009 (UTC)

We, Bill Longley and I, were not talking about the use of "uncredited" in the author field. We, Bill and I, were referring to someone making "uncredited" a pseudonym of another author, instead of simply making the work a variant when the party is known but otherwise uncredited. Please read Bill's comment again: "...uncredited shouldn't be allowable pseudonyms for or of anyone else..." How could one read that as "uncredited should not be used in the author field"? Come on, people! A comment sitting here for two days and then it gets pounced on like a frog on a junebug. (Yes, I'm a Southerner.) MHHutchins 02:43, 6 July 2009 (UTC)

Now to expound on the issue as it was first presented by Bill the Bluesman, i.e. "uncredited" as a pseudonym. There must be hundreds of works in the db that are not credited, yet a variant of the work has been created, based on reliable sources, of who the author of the work. For instance, James R. Frenkel, was the implicit, but uncredited, editor of Dell's Binary Stars anthology series. Variants were created and now the works appear on Frenkel's summary page and not on the uncredited page. Yet, Frenkel was never made a pseudonym of "uncredited". I've personally assigned variants to dozens of works that are uncredited in publication, using this method, and not created a pseudonym. At the moment there are two authors, actually artists, deigned to have used "uncredited" as a pseudonym, and five authors, actually collectives, for which "uncredited" is considered the canonical author. So until we can break pseudonymous relations, let's consider not allowing "uncredited" to be used as a pseudonym. I think both Bills would agree with me. MHHutchins 02:56, 6 July 2009 (UTC)

That makes some sense, sorry for having contributed to any confusion. -DES ^Talk 03:05, 6 July 2009 (UTC)

Mike has expressed my views perfectly. I'd be quite happy to code exceptions for "unknown", "uncredited" and "anonymous" to prevent further Make Pseudonyms. (And "Anon" and "Anon." if those aren't going to be regularised.) I don't like "anonymous (Leonard C. Lewin)" either. BLongley 19:00, 6 July 2009 (UTC)

Hmm. If you don't like "anonymous (Leonard C. Lewin)", what do you do when a work is in fact credited to "Anonymous" on its title page and cover (as some are) but it becomes known, either at the time or later, exactly who the author is? Enter the work simply as by Anonymous and create a variant? -DES ^Talk 19:13, 6 July 2009 (UTC)

Yes. Seems fine with things like Exploring Cordwainer Smith. BLongley 19:28, 6 July 2009 (UTC)

Anon., Idem, Ibid., and Trad.

wrote much that is morally bad.

Some songs and some chanties

All poems on panties

and limericks too, i must add.

-Anon

(The above not strictly on-topic, but I just couldn't resist. -DES ^Talk 19:17, 6 July 2009 (UTC))

Pseudonym assignments

Moved from ISFDB:Help desk#Thomas Barber, Jr, Thomas Barber, Tom Barber versus artist Thomas Barber -DES ^Talk Situation: Cover artist signature "Barber" is for the DB purposes prominent and well noted, unfortunately I seem to find no reference to him outside the db. I notice there is no "Barber" separate entries, so the canon name is the issue, except the easiest choice has a current artist, [18] it may be him, but I am doubtful. There is also a current Tom Barber design site that may conflict with that name. Here is the wikipedia for Tom Barber (artist) [19], which reflects back on this DB. Basic purpose of this DB, is to record authors and there names. Other book process contributions have been allowed to share spacing. First Argument: The DB is recording names as pseudonyms, hence they are using up our choices for real writers. My argument is the DB is not designed to create the numerous alternates for additional naming sequences for artists and others. The copyright pages are loaded with wrong spellings and creative naming. It is probable that the artist's often have their names shortened or even dropped by other persons with no authority to do so in the business/commercial sense. In other words, we have no basis to show that the alternate names should be recorded as such, as there is no evidence that the artist wishes this in most cases. Yes I know, Boris (Boris Vallejo), Romas (Romas Kukalis) and Donato (Donato Giancola). Yes the unusual works, but as in my case above, it does not encompass the totality of the associated contributions created by non-writers. Second Argument: How to make a clear and easily found distinction. I know current argument is to include artist with published works such as Josh Kirby as authors. I have no argument with that due to the unusual name. This does not work in the Barber case. Solutions; 1)Instead of creating pseudonyms for this type of entry, I suggest choosing/creating a cannon name, individually changing the cover or interior art entry to that, but clearly noting it in the specific book entries. This would not lock the name entries into the pseudonym relationships. 2) Need is then for a note field below the summary bibliography field. Something that allows things like this. Thomas Barber, Thomas Barber, Jr., and Tom Barber are the same artist. There is an artist Thomas Barber, but his displayed works seem disimilar to the cover artist. Thanks, Harry. --Dragoondelight 19:44, 4 July 2009 (UTC) (which I put on the author's comment page, but which is therefore invisible to the user and therefore does not meet the DB needs). or Thomas Barber, Thomas Berber, Jr., and Tom Barber are ths same artist by the 'observable data in this db' for the years 1975-83. Please check previous entries for matching data, such as working for specific publishers in that era, etc. 3) Another solution, would be to create a designator such as (A) for author entries in the Artist field or those with 'interior art' in contents. Note this would be needed to be able to form the pseudonym relationship in cases like Josh Kirby.

I can easily do the author pseudonym process for this artist, but this and some other similar artists, inhibit me doing so. I am not vested in any argumentative position, just in stating the liabilities that seem to be growing with time. Obviously, you are all aware of this type of situation. Sorry I have no real answers, but I find I must point out that Wikipedia followed the ISFDB lead, but is it what you want. Thanks, Harry. --Dragoondelight 22:38, 4 July 2009 (UTC)

In my view, the DB was precisely designed to handle multiple pseudonyms for individual artists and authors. I also think that who chose to spell or shorten a given name in a given way, or whether they were legally authorized to do so, or whether the name is as the artist )or author) wishes, are all quite irrelevant for our purposes here. The questions are: how did the name actually appear in any given publication, and by what name will readers and fans best know a particular author or artist. Those don't require determining intentions and desires, often mamny years in the past. -DES ^Talk 23:17, 4 July 2009 (UTC)

Sufficient research should be done to confirm the identities involved when creating a pseudonym relationship. The results of that can be saved on a wiki page for those who are interested. But once the facts are confirmed, i think that a pseudonym is exactly the way to go. Some artists or authors who have similar or even identical working names may need disambiguation. We have handled that in several cases already, and can continue to do so, IMO. -DES ^Talk 23:17, 4 July 2009 (UTC)

There is an outstanding feature request to make it more obvious when the bibliographic comments wiki page for a given author is non-empty. By the way, when creating a new pag4e in the Author: namespace, i urge you to use {{tl}AuthorHeader}}.

There has been recent discussion about splitting authors and artists, and the conclusion was not to split them, although that can always be re-opened. See ISFDB:Community Portal#Splitting Authors and Artists. -DES ^Talk 23:17, 4 July 2009 (UTC)

I suggest that this discussion would be better moved to the Rules and standards discussions page. -DES ^Talk 23:17, 4 July 2009 (UTC)

Moving it is fine with me, since we have established two differing approaches or viewpoints. Like I said I can do it with pseudonyms, but I am curious or concerned that there is a need to establish limits or checks/balances, especially some system of telling a user the where/what/how that is being used at the db to provide this data. I would rather the ISFDB establish a clearer guideline, than Wikipedia and Google making their assumptions. I frankly do not expect a quick fix. Thanks, Harry. --Dragoondelight 11:22, 5 July 2009 (UTC)

Double think, I am unsure of the proper method of moving this. Thanks, Harry. --Dragoondelight 11:24, 5 July 2009 (UTC)

Enf of text moved from ISFDB:Help desk#Thomas Barber, Jr, Thomas Barber, Tom Barber versus artist Thomas Barber -DES ^Talk

I'm not quite sure what you mean by "some system of telling a user the where/what/how that is being used at the db to provide this data". One can search the author/artist table for names, on visiting a given name it will be clear whether it is pseudonym, and if so of what other name. An enhancement already being worked on will indicate in search results if a name is flagged as a pseudonym. I am also not clear what sort of "limits" you suggest -- pseudonym creation is already supposed to be done only with fairly clear evidence, since it is hard to undo (although it should gt easier soon), and moderators do check why a pseudonym is being established unless it is pretty obvious (establishing "J. R. Author" as a pseudonym for "J. Random Author", say). What do you think the ideal system would do, or prevent? -DES ^Talk 13:57, 5 July 2009 (UTC)

The direct problem with Barber triplets is there is one active current artist, who uses the Thomas Barber variant. With common names and with time, this will re-occur. What I am suggesting is that some kind of active period of years be shown to give a quick visual limit to the user. Then if someone adds one out of that spectrum, then we know it needs checking for accuracy (of course, it can be a reprint). All this would do is narrow the field of our observation. In fact the above example is my presumed reason that the Wikipedia lists Tom Barber and not Thomas Barber. My point is that they are shaping the data and ingeniously throwing it back on this DB. Since I can not find out, and have no idea what a comprehensive/complete search would be, what criteria determines this DB's selection of a core/canon name. I can read a headstone, but I can not read tea leaves. Thanks, Harry. --Dragoondelight 13:07, 6 July 2009 (UTC)

There is no very formal method at the moment. In most cases it is obvious: most authors publish under only a single form of their name, or with a single form overwhelmingly the most common. There it is easy. No one argues that Mark Twain should really be Samuel Clemens. When it is not quite so obvious, and a book or other work comes in under a different name or form of name than other works by the same person already in the db, the editor entering the new (to the db) work must make a choice as to whether to record a pseudonym, thereby fixing the canonical name, and if so which name to make canonical. If an editor is unsure, s/he may well post on the Community portal, or elsewhere, and invite discussion. If a mod sees a Make-pseudonym edit, s/he will usually check out the situation. If it seems at all unclear, and particularly if there seems not much work in the db yet by the author or artist in question, the mod may choose to hold the edit and start a discussion. In an number of cases the suggestion has been to create a variant to document the relationship, but to hold off on creating the pseudonym until the commonly used name is better established, particularly if it is still early in the author's career. This is largely because variants can easily be broken and switched, but a pseudonym relationship is currently hard to undo (one must delete or change the author of every work by the pseudonym, and then re-enter or change back). When and if we get an easy way to change or remove pseudonym relationships (currently on the feature request list) that advice may change. In any case, the decision is made by single editors or groups of editors in discussion here, with no hard and fast rule but the general one "use the name by which the author is most commonly known". For disambiguation, the only rule is "make the author name unique". Dates are commonly used for this, but aren't always know, or may not be different enough to help. An ad-hoc judgment is made in each case. So far this seems to have worked reasonably well, although particular cases may be a problem. -DES ^Talk 15:51, 6 July 2009 (UTC)

The one biggest problem with our Disambiguation process is that the original name doesn't work anymore. Consider the regular case of disambiguation... using years. Right now we have John_Grant_(1949-), John_Grant_(?-), and John Grant. If anyone visits the normal canonical name John Grant there is absolutely no indication that there are other primary names to look at. Consider also the case of "Role" disambiguation, which we use when there are no dates of birth, death, etc known. Since there were only a couple of works, I cheated and made two authors psuedonyms of the shared canonical name. See Donald Macpherson, which disambiguates between Donald_Macpherson_(artist) and Donald_Macpherson_(novelist). But other than that singular problem, I believe we have a system that (once you can establish whose work belongs to which canonical name) you can easily put a unique identifier in the 'Author' data field. Kevin 16:02, 5 July 2009 (UTC)

Kevin, your multiple example is partial this problem, but is the greater problem through time that I am worried about. More John Grant and Tom/Thomas Barbers will come forward. I keep thinking that a backstop based on years needs to be used to separate them. The "Lord Harlifax" example is great because it involves the change of legal name, at the user's option, but infers that the importance of the individual is such that one follows his family line. Specifically if the artist continues to be totally anonymous then a timeline stop seems the only viable backstop. Thanks, Harry. --Dragoondelight 13:07, 6 July 2009 (UTC)

Not Forgetting. There is a Primary Problem with our recording of non-author/editor crediting. As I stated above, the copyright page crediting is obviously not done with circumspection. The recording of the 'as is', if there is any, really is not the same as creating pen-name pseudonym by an author or editor for their work. I can NOT rationalize a misspelling as a pseudonym candidate, only if the artist/other contributing person did it themselves would it make small sense, even then it would be 'catering' to that whim, and would be a disservice to the recording of the data for users. In the above example, we have no evidence that the artist Barber wished to be known as the diminutive Tom, other than it is a good indicator of personal interaction. My contention here is we owe the 'artist and others' best and clearest 'true' name identification. That recording diminutives is a disservice as it dilutes the impact of their work. The recording of misspellings super compounds this. Therefore I am saying avoid the singular name recording of these people, use the full name as it is the common use methodology of recording names. Yes, there are exceptions, for instance, Josh Kirby, as I can not remember Joshua. Thanks, Harry. --Dragoondelight 13:07, 6 July 2009 (UTC)

What about an artist project work page with a listing of names/variations, which people/us can check for greater detail. I think as individuals we think an operate in differing modes and one person will find data others will miss. Once someone thinks we have a fair working model of that artist/other individual then bringing it to group attention and then creating the canon/specific identity. Thanks, Harry. --Dragoondelight 13:07, 6 July 2009 (UTC)

Last item/position. I still feel that there is no reason to create variant/pseudonyms for most artist/others. I know this is/will not be popular, but I think the physical changing to the canon/specific name at book entry level and the noting that it was done is a better alternative. Granted the greatest artist or multi-talented ones deserve the pseudonym crediting, but the common/unlauded artist needs clear consolidation to give their efforts the greatest user appeal. Thanks, Harry. --Dragoondelight 13:14, 6 July 2009 (UTC)

If I understand your (Dragoondelight/Harry's) concern it is that it will, over time, become more and more common for multiple authors or artists to have the same name, so that we need to disambiguate them, and the user looking for a work by such an author or artist will not know where to look, and the user finding the bibliography of one such person will not know there are other records to check; that finding John Grant there is no obvious clue to the existence of John_Grant_(1949-) and John_Grant_(?-). This is indeed a problem, at the very least we need some sort of "see also" indicator on such records. Kevin is correct above on that point, IMO.

I think you are also concerned that by recording and using pseudonyms, we use up the "namespace" and make such collisions more likely. Possibly so, but they obviously will occur in any case, and i don't think that our recording of pseudonyms makes the matter much if any worse.

I gather that you also think that it is unjust to the author to record a work under a name, or form of name, that the author (or artist) never approved or chose to use. In particular you don't think that recording works under names that resulted from editorial imposition, or mere clerical error by a publisher or editor is a good idea -- you would prefer to use the "correct" (legal? most commonly known?) form of the name on each publication, and record only in a note that the name had been changed. Have I understood your idea correctly?

Here I disagree. I think we must consider who the db is likely to be used. Consider a hypothetical case. The mid-list writer "J. Random Author" writes Spaceship to the Stars! which is published by Careless Books, inc (a division of MegaCorp). The book appears with "J. Rundom Author" on the cover, title, and copyright pages, due to a typo in the publisher's office. J. R. never intended to be known as "Rundom", and never authorized this spelling. However The book was in the stores before he found out, and he couldn't even gt Careless to recall and reprint unsold copies, or do much of anything. Some years later, an SF fan (Emma User) finds a copy of the book in a library book sale. Emma wants to know what else "Rundom" has written, and goes to the ISFDB. Working as we now do, a search on "J. Rundom Author" will return Spaceship to the Stars! with a note that it is a variant title of Starship! by "J. Random Author" (printed in a restored and corrected edition by Goodguy Small Press 12 years later), and a link to Random's bibliography, which will also list the work he did as "J. R. Author", and his early work as "Stochastic Creator". If we had, instead, "corrected" the record for Spaceship to the Stars! to show that it was written by "J. Random Author" (as in fact it was) Emma would not have found anything on a search for "J. Rundom Author", and would have just assumed that the ISFDB was incomplete. Perhaps worse, Emma might have created a new record from the book-in-hand. Not knowing of the typo, she would have entered it as by "J. Rundom Author", and any future fans would have found this record, and might well have assumed that "J. Rundom Author" was a "one-book wonder", and not found the other 30 books and 50 stories listed on the "J. Random Author" page. This is not a good result. You say above that "...common/unlauded artist needs clear consolidation to give their efforts the greatest user appeal" and i don't disagree, but as long as the actual books don't all use a common canonical name for each author and artist, I think we best provide that "clear consolidation" by recording variants and making sure that all variants properly appear on the page for the associated canonical name. Perhaps there is a better way, but i don't see it.

You suggest that we should indicate, for each author/artist, the range of years during which that author is active, so that if a book or other work comes along well outside the range, we can suspect that it is by another person of the same name. That would be a good idea, but I see some questions. One is: at what level will this be shown? On the title or publication-level listings, or only at the top of an author's bibliography page? Or in search results? The other, more significant problem is, IMO: where would the dates come from? Would they be automatically generated by the first and last work we have on file? taken from the birth & death dates in the author record? Entered and updated manually? If the dates are automatically generated, a single late reprint skews them and makes them of limited value. If generated from title rather than pub dates, a single posthumously published work does the same, as does a single work actually belonging to a different author of the same name incorrectly attributed. Indeed this so skews the reported dates that it appears to validate itself once it is in the db. Since most authors' period of active work is only a section of their lifetimes, birth and death dates don't help much (although a new work long after a death date merits additional checks). Manual entry and validation of 'active dates' would solve all these problems, but that makes one more thing to enter, check on, and keep in sync as new data is found and authors start and stop work. I'm not sure if the gain would be worth the effort, but it might be. If you see another and better method of determining an author's active period, please say so.

I am interested other people's reactions to your suggestions, so I won't writ more on this at the moment. -DES ^Talk 15:35, 6 July 2009 (UTC)

[after edit conflict, so this is a response to Harry, not DES] I disagree that pseudonyms for artists should be handled differently than those of authors. We should be just as diligent in recording printed credits. If someone came upon a work by Tom Canty, and wanted to search the database for other works, would we not be doing a disservice by not crediting it as such in the database, even though he's also been credited as Thomas Canty? As for misprints, how can the user know that it's a misprint? Janny Wurts is credited as "Janny Wurtz" on her very first cover. If a person only has that particular book and does a db search, shouldn't they be told that "Wurtz" has many more works published under her correctly spelled name?

One area where I feel that Harry would agree with me, is when a pseudonym is created based on the signature on a work of art, because there is no printed credit. We wouldn't, I hope, create a pseudonym when initials are used in introductions and afterwords by authors, so why do the same for an artist's signature? Yes, the signature can be used for identification, but the record should be credited to the canonical name of the artist when identification is made (and noted as such). We shouldn't create a variant based on a signature. Someone has actually created a pseudonym of "J. G." for Jack Gaughan. Would any user of the database actually think that Gaughan used this as a pseudonym? Yes, Emsh should be a pseudonym, because that was the way he was credited in print, not just as a signature. And Kelly Freas should be a pseudonym of Frank Kelly Freas, but why is Freas a pseudonym of Kelly Freas and not Frank Kelly Freas? I think most of us can agree that Frank Kelly Freas should be the canonical name, and variants should be created so that all of his work would appear on the same page (once we've broken that odd pseudonym of a pseudonym business.) Just my two cents. MHHutchins 15:56, 6 July 2009 (UTC)

I agree with the above, although I think we actually have created pseudonyms in one or two cases for sets of initials used to sign introductions and the like. This is probably a mistake, unless the initials were the only credit in a book, and a user might plausibly be expected to search for the writer under those initials, a rare case, it it has ever occurred. -DES ^Talk 16:07, 6 July 2009 (UTC)

[after another edit conflict with DES, damned we're talkative today...] Responding to the issue of running out of names: the odds are great that two (or more) authors/artists will have the same name, but the percentage will be quite low. So we deal with those situations when they arise. There are at least ten different people who have published books as "Michael Bishop" or a form of the name. Thankfully, with the exception of one idiot, all published outside of the spec-fic genre. But in my online bibliography of Bishop's work, I acknowledge that fact and present those titles so that the user knows that these works are specifically not the work of the author at hand. So authors will have the same name. So we deal with them. This should not be a factor in determining the creation of a pseudonym. MHHutchins 16:09, 6 July 2009 (UTC)

For the record, I have no problem with Author/Editor pseudonyms, so using them as examples to prove the need to do so is 'missing' my point. I think we have a slight difference in perception, mainly because I have done a lot of series crediting of artists and others. Sicne I do overdue it and try and get any person, reasonably (ha!), mentioned, my position on misspelling has many examples to formulate the basis of the artist/other problems. In most cases the hardcovers are best artist sources. The rest are back and forth and there is some mixing with cover design, and design credits. It is apparent that much of the 'copyright page' material is last minute or without due review. Thus I see that you have to weigh what it is said, they often credit the wrong artist, as a signature of another is apparent or in series with a certain type of 'design' look, the artist is being 'cut/trimmed' sometimes it is almost deliberate. Map crediting is also often wrong to so hard to read clearly that misspelling errors appear. So, I believe everyone has had to do some checking. I am therefore trying to say, a take it like it looks and created pseudonyms for that is not the best method, I think the rules should 'clearly advocate' by stating that checking the db for correlations in name and spot correcting is encouraged. If an 'editor' feels uneasy, check, wild misspelling errors to simple ones, should be 'Noted' and then go to the best fit. When questionable then it should be Noted, but not credited. Clear example Roc series "Battletech" one of the very first IA credits was this [20] which is a phonetic variation to [21] which is true name. I left it as is and made a personal choice that I would no longer credit solely by the copyright statements, as some of the art was not 'singed'. My rule of thumb became all or majority for crediting and no more generic crediting. My fairness gauge says that DL did a lot of initial work, but others must have did a great deal later as it became apparent signature "loose" was not on everything. Also, the art was repeating and therefore would become like quoting constantly the same canard. I do the same with chapter ornaments, if it is not visually and noticeably different, then only one credit. The one true guide is that style to my eye may be co-opted on IA, but signature patterns, mostly partials (often dated), are the key to individual effort. Basic IA attributions for the DL case is that I could have often found five named/titled art objects that repeated constantly through the series. I would then have created say 40 Intruder, or Overlord or Union starship entries. I feel quilty for not, but think that this type of IA needs rule guidance to curb that kind of entry. I true gamer, would/could have. The same is true of battle maps and spatial maps. Does anyone wish to have over 200+ DL art of basically less than two dozen line drawings. For me, I decided that creating more entries was DB abuse. Also general credit for everything was IA artist abuse.

Sorry went skew some there. I am still holding out for sensible research for the 'does not look/feel right' and then when turning what appears to be correct going with it, with notes about how/why. I think the editor needs some guidance to check for variance, especially with signature, but also with oddities. I find no value to creating 'Coukolis' pseudonyms for Kukalis. Anyone, checking for misspelled variants of that type will be checking that error anyway. The common user gets nothing out of most misspelled error pseudonym. I know even the dictionaries do/have creatively done this, but it does not really help to compound errors.

Point back is the Barber example needs consolidation and there is potential for conflict if some kind of limiter is not applied such as a bold statement that they are grouped for clear user insight and that a common grouping factor such as year from/to applied and the DB knows that there are tangible defects, and is awaiting further data gathering/scrubbing. 'Barber' may also be involved in romance, western, etc fields and grouping with/between other db's will give us a 'real' person, though we are protecting the present 'identity' from being occluded from the real present day Thomas Barber, whose work looks good enough for SPF to me. Besides some quick catch 'the DB is aware and working on it' we need a specific or possibly numerous specific spaces for recording scanned 'signatures'. Personal of the top of the head, Michael Whelan, has 5 variant signature/initiallings. A few artists used spelling/font variants like Ken W. Kelly as in Celly. Clyde Caldwell uses a runic symbol, that needs to be brought to the 'editor level' attention. Sorry for the hash, but the infusion, especially of series data which just does not conform easily to stated/wished DB formatting needs airing by me. I build these kind of conflicts, until I am released or I have duly aired it to others to bear the burden of my sins. LOL. Thanks Harry. --Dragoondelight 20:35, 6 July 2009 (UTC)

Perhaps my concern is more to create an IW in IN Work status for many artist groupings, instead of creating pseudonym relationships. Give some warnings, some special info blocks (modified image spots), and a IW listing that editors can go to and build the data flow till we actually have some 'real' individual to build further on. Comparing Ellissa, Ellisa, Elissa Michel who has real data to include, there is little problem, other than some wastage, in the pseudodnym system, but doing something like that for Tom/thomas/Jr Barber seems excessive, yet he deserves some delineation of shared data bits to bring him from the shadows. and I resent the wikipedia pointed out that it needs be done. Let us get ahead of their curve balls. Yes, I know this programing and I am prepared for very long wait time, but getting it to a this 'Process' needs work can only be achieved by this evolution. Thanks, Harry. --Dragoondelight 20:45, 6 July 2009 (UTC)

Graphic novel/comics exceptions

I know that Graphic Novels & Comics are generally OUT, but some limited exceptions have been made. Not havering worked much in this area, I'm not clear what those exceptions are or on what basis they are mode. While working on ISFDB:Chapterbook cleanup, I encountered Angloman. I started to delete it, but then put my own delete on hold, not sure if this should count as an exception or not. Advice, please? -DES ^Talk 16:51, 7 July 2009 (UTC)

It looks like someone went to a lot of effort (or maybe just a lot of copy'n'pasting). I'd tend to leave it as harmless (maybe add the cover from the Amazon Canada site, fix the price) as I've no idea if the authors are notable enough to warrant complete inclusion. (And Morrissette should probably be an artist rather than an author.) And there appear to be five shortfiction contents needed if we correct it, but I wouldn't expend that much effort. BLongley 18:50, 7 July 2009 (UTC)

From the outside sources, it appears that Morrissette did the illustrations and also co-wrote the text, probably as a secondary author, but thank God we don't try to record percentages of contribution by co-authors. -DES ^Talk 19:37, 7 July 2009 (UTC)

In which case, our "written by Shainblum and illustrated by Morrissette" title notes are inaccurate - we have no mention of co-writing. The 1996 sequel seems to be vapourware, or unrecorded here. Not worth a lot of attention really though. We've thousands of other records in need of more attention. BLongley 21:15, 7 July 2009 (UTC)

No review, no awards, deleted it!Kraang 00:21, 8 July 2009 (UTC)