ISFDB:Community Portal/Archive/Archive05

From ISFDB
Jump to navigation Jump to search

This is an archive page for the Community Portal. Please do not edit the contents. To start a new discussion, please click here.
This archive includes discussions from

Archive Quick Links
Archives of old discussions from the Community Portal.

1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 · 9 · 10 · 11 · 12 · 13 · 14 · 15 · 16 · 17 · 18 · 19 · 20 · 21 · 22 · 23 · 24 · 25 · 26 · 27 · 28 · 29 · 30 · 31 · 32 · 33 · 34 · 35 · 36 · 37 · 38 · 39 · 40 · 41 · 42 · 43 · 44 · 45 · 46 · 47 · 48 · 49 · 50 · 51 · 52 · 53 · 54 · 55


Duplicate Titles Created By Adding Contents to a Pub

After adding the contents to this book [1] I noticed that all the stories which were already in the isfdb have been duplicated. That is to say, the original instance of them with whatever pubs it already had, and a new one that only has the pub I just edited, and thus requiring an orgy of merging. Is there any way to add contents to a book without this happening? (I just looked at the instructions for adding a new pub to a title, and it looks like the same problem would occur, i.e., creating a whole new title, rather than adding the story to an existing one.)Jefe 15:41, 11 Jan 2007 (CST)

There is a way around this in some cases, but contents do still have to be manually merged in some situations.
First, if you're adding a new publication of an existing title, if you choose to clone one of the publications, instead of "Add Pub to Title", then you will get a screen that looks like the addpub screen but it will be already filled in with much of the data. Adding page numbers and submitting/approving the result will cause an automatic merge for each content item.
If you're adding a pub and there are some differences, it's often most efficient to clone, then use "Remove Title" where necessary, and finally "Edit Pub" and add contents as needed, manually merging these last items.
Cloning has limitations, though. You can't use it if you want any differences in the title of the book, or the author's name. If you have a "Brian Aldiss" pub in your hand, and the existing title in the ISFDB is by "Brian W. Aldiss", then you're going to have to enter it from scratch.
There's an outstanding feature request to allow selection of stories to automerge with on submission of new pubs, but I don't think it's easy and so it may not arrive in the near future.
-- Mike Christie (talk) 20:24, 11 Jan 2007 (CST)

Eric Ambler's "The Dark Frontier"

from Wikipedia: "Based on the development in weaponry of the year 1936, The Dark Frontier was one of the first novels to predict the invention of a nuclear bomb and its consequences. Ambler evidently had no knowledge of what producing an atomic bomb may involve (even professional physicists at the time had only a vague idea). The book makes no mention of uranium or any other radioactive material, and makes instead the assumption that setting off an atomic bomb would involve a considerable electric charge. Still, Ambler could be credited with having become aware, before many others, of this coming weapon which was to have such a profound effect on the entire world, and his depiction of scientists in a secret hideout buiding such a bomb could be considered a premonition of the Manhattan Project - and he correctly surmised that refugees from Nazi Germany might get involved in such a project."

What do you think? Hayford Peirce 16:37, 12 Jan 2007 (CST)

I believe the Nicholls encyclopaedia omits "future war" books on the grounds that the genre is too vast. However, specific works like this sound like they're too close to the border to be worth the effort to delete. I say leave it in if it's in; don't bother adding it if it's not unless you're working on Ambler's biblio (assuming he wrote anything else that would qualify). Mike Christie (talk) 17:34, 12 Jan 2007 (CST)
No, this is the only thing he ever wrote that could remotely be called SF. It's pulpish and melodramatic and it's practically unknown -- I don't think he liked being associated with it. It isn't listed at the moment. It isn't really "future war" -- a standard spy-thriller or innocent-in-danger thriller that Ambler later perfected, but with the atomic bomb McGuffin. I'll wait to see if anyone else offers an opinion. Hayford Peirce 18:41, 12 Jan 2007 (CST)
Reginald lists it, but no one else does (including Bleiler's "The Early Years", which is fairly inclusive). Reginald says the first edition is the 1936 Hodder & Stoughton. If it's good enough for Reginald, it's good enough for me. Alvonruff 19:32, 12 Jan 2007 (CST)

Redundant info and editing process

Some of what I'm going to say I've also written a little earlier under "Why can't we put Series info into the initial New Novel data input?" and I *do* understand that this is an evolving project in which newer (and maybe better) ideas have been overlaid over what has gone before, sort of like Windows 98 evolving out of MS-DOS1, groan. Anyway, here's my gripe: If you go to "Jonathan Gash - Summary Bibliography" you see a list of 20 books in the Lovejoy series. If you click on some of the earlier ones, such as The Vatican Rip, you'll see a Note that says "The Lovejoy series link etc." And then, a little lower, under Publications you'll see the name of the book. If you click on that, another page opens up and here, once again, is the same Note with the same link info. If, on the other hand, you go back to the Summary Biblio. and click on a later book such as The Possessions of a Lady, you'll see the same Note on that first page, but NOT under Publications if you click on the name of book there. This is because I never know where to put in this info -- and also because there are TWO separate info inputs for this, one under New Novel, and then, later, under Publication info or Title info, I forget which. So, since I have to wait for a Moderator to approve my initial input of the New Novel, by the time I come back to put in the series link info I can't remember if I've put it in Notes under the first edit or not, or whether it should *also* go in Notes under Publications. What I'm getting at is that it would sure be nice if we had a *single* info template into which *all* the info could be put at the same time. Title, author, name of series, number in the series, publisher, date of publication, hb or pb, Type, etc. etc., with synopsis, notes, links to Wiki, links to reviews, EVERYTHING, quoi, on that same page. So that there wouldn't be any more back-and-forthing (at least for me and other non-moderators) between New Novels, Authors, Titles, Publications, etc. etc for editing. Probably can't be done, sigh, but it would sure be nice! Hayford Peirce 18:04, 15 Jan 2007 (CST)

I think the core problem here is that you have to wait for approval. I do think that the notes don't all belong in the same place -- there are title-specific notes, and publication-specific notes, and so on. In this particular case I think I would place the note on the series by preference, which would require you to do it only once. However, there is no note on the series record. This would be a reasonable feature request. In the meantime, the title record is the only reasonable place to put the note.
I also think that if you get the distinction between publications and titles clear, it will be a lot less confusing. Take a look at a title with lots of pubs, such as Le Guin's The Farthest Shore. I think you'd agree that it would make no sense to put the note on each pub here. If you visualize a pile of different editions and different printings of "The Vatican Rip", then your note naturally attaches to the whole pile, not to each book in the pile. Does that clarify it? Mike Christie (talk) 19:07, 15 Jan 2007 (CST)
I have seen the two "Note" fields confuse other editors as well and I recall using the wrong field a couple of times myself when moderator editing was enabled in May 2006. Perhaps one way to address the confusion would be to disambiguate the two fields by giving them separate labels? Ahasuerus 14:36, 16 Jan 2007 (CST)
Would Series Notes, Author Notes, Title Notes, and Publication Notes work or do you think something else needs to be different? Marc Kupper 15:08, 16 Jan 2007 (CST)
I think that should work unless somebody can come up with catchier labels. Ahasuerus 16:52, 16 Jan 2007 (CST)
Those ought to do the trick. But I thought that Series Notes are *not* available -- but that we all wish they were. Or am I wrong about this? Hayford Peirce 16:59, 16 Jan 2007 (CST)

Some things I've noticed in recent submissions

This is really a note to the other moderators, though others may find it interesting too, to see what problems are coming in on the sub queue.

Here are a couple of things I've seen. Several of these are from one user, Dsorgen, but I've seen them from others too.

  1. People submitting new contents when it would be more efficient to clone. I'm hesitant to reject these, so I've been holding them for one recent case (User talk:Dsorgen#A_Hole_in_Space), with an explanation of what's going on.
  2. Deletion of "jvn" in the story length. I'll add an action item to clean up the def on the story length field.
  3. Editing titles in collections, not realizing that this will edit the title for all cases
  4. Entry of first printing data from a later printing. I have some of Dsorgen's on hold for this now, and will write him a note about it shortly. See this one for example.

Got to go; more later. Mike Christie (talk) 19:14, 15 Jan 2007 (CST)

  • One more would be editors trying to remove-title the parent title of a publication. --Marc Kupper 19:51, 15 Jan 2007 (CST)

shortfiction storylen

Sometimes I run across a storylen of “sf” or “shortfiction.” Where do these come from and what does it mean in terms of the story length? Marc Kupper 19:51, 15 Jan 2007 (CST)

It's the default when you don't select a story length for a SHORTFICTION content type. I generate a lot of them as I have been entering magazines, which are notoriously unreliable in their description of length, so I never use a length. I have sometimes done so when the story is just two or three pages long, since that's pretty clearly a short story, but other than that I've been leaving it blank. In fact I don't really like this field, though I see its utility -- it has an objective definition but it's not one we're very well equipped to apply. Still, there are external sources such as Hugo qualification lists that can be used to be objective. Mike Christie (talk) 22:06, 15 Jan 2007 (CST)
FYI, there are some minor and apparently harmless bugs in the way the software handles "shortfiction" titles, especially in the Title Merge area. One of these days I'll figure out exactly what's going on and submit a bug report. Ahasuerus 14:39, 16 Jan 2007 (CST)
I was on Locus recent and saw a storylen of vi. I wondered what that was and found on their their Abbreviations page the following storylen codes. Based on the following table I set up a spreadsheet where now I copy/paste the ISFDB Contents from a publication display and the spreadsheet extracts the page number and computes the storylen code. There is some pain because edit-pub does not sort the contents making for a manual hunt/edit. Marc Kupper 15:19, 16 Jan 2007 (CST)
Code Description
vi vignette (under 4 pages, under 1,000 words)
ss short story (4-20 pages, 1,000-7,499 words)
nv novelette (21-45 pages, 7,500-17,499 words)
na novella (46-100 pages, 17,500-39,999 words)
n. novel (over 100 pages, over 40,000 words)
Unfortunately, the number of words per page varies depending on the binding/font/publisher, etc. Also, in the pulps, a story could be easily split in two or even more sections, appearing, e.g., on pages 50-60 and then 80-84, so the algorithm is rather unreliable :( Ahasuerus 16:58, 16 Jan 2007 (CST)
BTW, what’s the source of the ISFDB storylen codes? They are different than Locusmag for example nt, nv, na. Marc Kupper (talk) 17:47, 20 Jan 2007 (CST)
Locus ISFDB Description
vi N/A vignette (under 4 pages, under 1,000 words)
ss ss short story (4-20 pages, 1,000-7,499 words)
nv nt novelette (21-45 pages, 7,500-17,499 words)
na nv novella (46-100 pages, 17,500-39,999 words)
n. N/A novel (over 100 pages, over 40,000 words)

Pratchett question

Over at BLongley's talk page he has raised some questions about some Pratchett books. I have to run for now; if anyone else has time to head over there and comment and clarify things, please do. I'll try to catch up later tonight. Mike Christie (talk) 18:32, 16 Jan 2007 (CST)

I've done some fixing over some period of time and edits... there's a lot of possibilities still. Like separating the plays into the series they are plays of - Discworld versus Johnny Maxwell, etc. The Maps should go under Discworld I think, there's no other maps. There's a new can of worms arriving possibly, with the Discworld Diaries - how can you categorise a book that that is SUPPOSED to be mostly blank pages? There's more to add...
Feel free to leave it with me to mess around with if you like. I promise not to add the jigsaw puzzles. ;-) BLongley 17:09, 22 Jan 2007 (CST)

Short fiction in series?

I am writing a "How to work with series" help page, and realized that I had never seen short fiction in a series. I looked at a couple of obvious candidates, such as Merovingen Nights, and found that they've been handled by putting the anthology in as the series member. I tried adding "The Day Before the Revolution" to Le Guin's Hainish series, and found that in worked, but is not identified as short fiction, though there is an "SF" tag on Le Guin's biblio page.

Do we have a preference for whether short fiction gets included or not? I would think it should be included, though if a collection of stories is written in internal chronological order and does not overlap other series elements I could see it being included en masse.

I am going to write the help to suggest that short fiction be included; let me know if there's a reason not to. Mike Christie (talk) 07:35, 17 Jan 2007 (CST)

Series support is still evolving. A few weeks ago Al added additional logic to display shortfiction (marked as "SF", a rather ambiguous abbreviation in our context) pieces on Summary Bibliography pages, but the same logic hasn't been added to the Series Bibilography pages yet. Also, Summary Bibliography pages only display short fiction pieces as part of a series if there is a book length work in the series as well. If the series consists exclusively of short fiction pieces, then they are not displayed as a series - cf. the Doctor Bird and Operative Carnes series listing and S._P._Meek's summary page. I would check with Al to see where he is taking this whole area before writing anything up. Ahasuerus 12:32, 17 Jan 2007 (CST)
I already created Help:How to work with series. I'll edit it to conform when I hear from Al. I'll also add a note now explaining the limitation you cite. Mike Christie (talk) 12:39, 17 Jan 2007 (CST)
The other problem with short fiction that appears in a series is that such stories are displayed out of chronological order in the short fiction list (this is true regardless of whether a series has books in it as well). See Michael Swanwick's summary biblio for an example, where the stories in his three (listed) short fiction series appear (in no apparent order) after the last non-series story (being "Lord Weary's Empire (2006)"). Jefe 16:36, 17 Jan 2007 (CST)
Yup, it's a side effect of the rewrite of the series logic that is still incomplete. The problem was reported a while back, but I don't see it among Open Display Bugs. I'll add it to the list, thanks! Ahasuerus 19:28, 17 Jan 2007 (CST)

Moderator qualifications

I have just created Moderator Qualifications as a place to discuss what it takes to be a moderator, and also to host any discussions of creating new moderators from existing editors. It's a draft; it seemed like a good idea to have a permanent location for this -- prior community portal discussion has been archived.

Looks good to me, Mike: particularly the bits on welcoming new editors, which you seem to be showing the perfect example of! BLongley 17:21, 17 Jan 2007 (CST)

I suspect the next moderator we create should come from the UK timezone; we have nobody on that side of the pond right now. Mike Christie (talk) 11:05, 17 Jan 2007 (CST)

While rapid approvals help some editors that know what they want to do and how, I think the "giving advice" is the more important bit. And I hope you're not looking towards me just because I'm in the UK - I do learn pretty quickly, but that's led to me being Moderator/Craftsman/Admin for 7 sites already (3 I'm still actively responsible for) and I don't multi-task perfectly. But give me 6 months and IF I'm still active here it's the sort of role I'd be interested in. (If the programmer jobs aren't available - I do that sort of thing for a living, just not the way it's done here.) BLongley 17:21, 17 Jan 2007 (CST)

PHP Error (AKA: What the flip does this mean?)

I'm trying to add the contents to "Witpunk" anthology edited by Claude Lalumiere and Marty Halpern. I dump in the first page (page #, title & author). I get to the end and submit with the result of a paga junk and the following message:

+++++++++++++++++

AttributeError: 'list' object has no attribute 'value' 

     args = ("'list' object has no attribute 'value'",)

+++++++++++++++++

What's up with this? I'm NOT a PHP guy. Oh, crap, I've forgotten how to comment HTML too. Jeez, I'm getting old...

Dave Sorgen 20:44, 17 Jan 2007

This may be a newly introduced bug as I ran into the same thing tonight. See Open_Editing_Bugs 10089 Python error adding essay with two authors. The "fix" for now is when adding contents to only include one author the first time and then do an edit-pub (after approval of the first edit) to add the second author. I did some more testing and the bug only seems to apply to the first title record in the contents and hit me because my anthology had two editors and the first record was the introduction. It seems I can add-title and add the introduction with both authors as a second or later title. Marc Kupper 03:30, 18 Jan 2007 (CST)
Hmmm... Seems to work as you suggested (make the two-author entry #2 or later). I've built part of the entries for the anthology. I'll inspect later to see what happens, then build the rest. Too bad PHP didn't localize the problem a bit better; I might have figured this one out. Sigh... The update to two editors is still pending for the item as a whole.

--Dsorgen 15:38, 18 Jan 2007 (CST)

Blue Mars

I have an edition of Kim Stanley Robinson's "Blue Mars" which I was trying to use to verify the ISFDB entry; most of the data is the same, but the page numbers are wildly off -- mine has 609, the isfdb has 576. The Locus index agrees with me, but I'm loath to go and change the page number in case there's some weird variant version of this edition running around (it seems unlikely but who knows?). Should I enter my copy as a new pub and leave the existing one, or edit the existing one?Jefe 13:27, 18 Jan 2007 (CST)

Edit it to agree with your copy. If there's a new one, it'll show up as a separate entry eventually. It's reasonable to do a little searching on used.addall.com or somewhere similar, when you suspect an extra edition exists, but in this case you're unlikely to see page numbering recorded so there's little you can do. Some things would indicate a different copy; a change in price, for example. But if you verify it, and add a note indicating the printing, someone else can add the other version and they'll have enough data to distinguish it from yours. Mike Christie (talk) 14:19, 18 Jan 2007 (CST)
This brings up something else I've been wondering about. Since we're trying to record every printing of every edition, would it make more sense for there to be a field labelling something like "Edition/printing" for this information, rather than just putting it in the notes? My concern is that many editors are not going to think to put in this data if there's no explicit location for it (I know I didn't for the first several books I entered). If this data is left out, it might actually lead to lost data, or at least lost work. For example, I entered data for Gene Wolfe's "Calde of the Long Sun" from the second printing of the first edition. If I leave that information out, somebody with another printing of that edition may either 1) not enter their copy because they think it's already been done, or 2) edit my entry to show their printing information.Jefe 15:41, 18 Jan 2007 (CST)
Take a look at Feature:90091 Edition label for publications where something along these lines is discussed. I started adding the "Stated third printing" to the notes because I was running into exactly the situation you describe, where you can't tell if you're dealing with the same printing. That led to this feature request, which would probably be better described as "Series functionality for publications". Mike Christie (talk) 15:51, 18 Jan 2007 (CST)
I’ll have to chime in “me too” on this one too. I have been adding publication-notes about the printing number but each time I do so I wonder if future editors will understand my notes. Recently someone tried to change one of my notes (good catch Mike) so I edited that particular publication’s note to (hopefully) make it clearer. I have the same problem on the DAW list page.
In my case what I’m trying to do is to say “This is a confirmed sighting of the nth printing. If you have the same printing then it’s quite likely we have identical publications. If you have another printing but otherwise your publication seems identical to mine (same cover, price, and all that) then great, just add your printing # to the list of confirmed sightings. Over time we will end up with a list of exactly which printings were used for this price/cover/etc.” I don’t want to say all of that in each publication’s notes and shorten it to “12th printing” (73 words compressed to two!) which, unsurprisingly, mystifies future editors.
On the Blue Mars page number thing. I personally would add a note saying that ISFDB had the YYY page count and that you corrected it to ZZZ based on your physical copy. That way if someone shows up with a copy that has YYY pages they don’t just overwrite your page count but instead can clone the pub, fix the page #, and update the note to explain their copy. Marc Kupper 23:36, 18 Jan 2007 (CST)

SF Works in Non-genre Publications

I saw on one of the introduction pages that sf works in non-genre (or mixed-genre) publications are fair game for the isfdb, but I don't see anything in the help pages about how to enter them. I.e., should I just enter the genre content and leave the rest of the contents out? This seems to make sense, but we will end up with anthologies that have only one story listed.Jefe 13:27, 18 Jan 2007 (CST)

I came at this from the other angle: there are stories listed such as Evelyn Waugh's "The Man Who Liked Dickens" that are clearly not genre and I bitched about it for a while. But it's a long-established custom that an anthology with even 1 or 2 genre stories get listed -- as well as *all* the stories in that book or publication, whether they are genre or not. Hayford Peirce 13:38, 18 Jan 2007 (CST)
Hayford's right that the custom was to enter all stories in anything that gets entered. However, we did have a discussion about this a couple of months ago, and the most recent consensus was that we should only enter the genre material in the contents. See Bibliographic_Rules#Stories_published_in_non-genre_magazines; unfortunately I can't find the original discussion -- perhaps Marc or Ahasuerus will remember where it was. The reasoning was, as I recall, that we would find ourselves with vast amounts of data entry to do from (e.g.) Argosy, which would clutter up the indices. I think that's still the right answer. If that's not reflected in the help, it should be (unless we want to reopen that discussion).
Of course, as Hayford mentions, occasionally a clearly non-genre piece will show up in a genre magazine; that's cause for indexing. Whether we then go back and index all other appearance of that work hasn't yet been discussed, but I don't think we would do that. Mike Christie (talk) 14:27, 18 Jan 2007 (CST)
The discussion linked above states that works in non-genre magazines can be found via the regular search apparatus. However, I can't seem to figure out how to search for magazines that don't have a wiki page. I'd like to enter a story published in "The Black Cat" but I want to make sure that the issue isn't already in the database (the story itself doesn't show the mag as a pub, but that's not conclusive). I tried various things in the advanced search, but couldn't even get mags that I know are in the db to show.Jefe 16:01, 18 Jan 2007 (CST)
Part of the problem is that titles for magazines are grouped by year. To see what I mean, look at this title for Astounding. This is done by merging all 12 titles created by each issue that year into one title, and editing it to the form you see there.
The reason is to avoid hundreds of issues showing up in the biblios of the editor of the magazine. See Campbell's biblio to see what the result looks like. The drawback is that there is no longer the expected title/pub relationship for magazines. Also, it's not done consistently, as it's done for the look of things, so some magazines don't have this treatment yet -- do a title search on "Infinity Science Fiction" and you'll see what I mean.
The other thing to be aware of is that a MAGAZINE publication has an EDITOR title type. So if you go to Advanced search and put in Astounding for the title, you can put in EDITOR for the Title Type and you'll see all the magazine records without the books that have "Astounding" in. As far as I can see, based on these searches, there are no editions of "The Black Cat" in there now, so you are OK to enter this one. Mike Christie (talk) 16:49, 18 Jan 2007 (CST)
On my home page I have these links
I use the Google links to search for content on the ISBDB sites and for "The Black Cat" it does get mentioned on the Wiki side at Magazine:1900-1919. There are 58 pages on the ISFDB side meaning you'll need to add more keywords, perhaps a story title from your issue, to better see if your magazine is listed. It's not perfect as Google does not index ISFDB in real time meaning it would see recent changes. Marc Kupper 23:46, 18 Jan 2007 (CST)

(unindent) You have to be careful with Google, folks. It's good at what it does, but perhaps a little too good in our case. It indexes out of date ISFDB Wiki pages that are currently orphaned or at least no longer linked from the main pages.

As far as finding Magazine publications goes, you typically have to search for them using the Publication Search form. It is at the bottom of the Advanced Search screen. Please note that you can use it to search for Publication titles, but many other, more complex searches (ISBN, publisher, etc) are currently broken.

As as a general observation, our magazine implementation has gotten rather involved with the addition of the EDITOR title type. Mike Christie put a summary of the current EDITOR functionality that I wrote a while back on his User page and it's not for the faint of heart. We may want to add it to the "Advanced topics" section of the Help pages for now, but in the long run I hope we can make it more transparent to the users. Ahasuerus 10:41, 19 Jan 2007 (CST)

Creating Magazines

I'm having two problems with creating new magazines. One is a documentation issue: there's no documentation for the "tag" field on the NewPub helpscreen. The other is that you don't seem to be able to enter a magazine without an editor. (Well, the error I'm getting is "*** ERROR: No authors were specified", so I assume that's the problem.) I was trying to add the first publication of RAH's "It's Great to Be Back" in "The Saturday Evening Post", and I can't imagine that we care who the editor was. (Plus, I don't have a complete copy of the magazine -- I just have a photocopy of the story I made for a research project).Jefe 16:12, 18 Jan 2007 (CST)

I'll add documentation for the tag field; I'd forgotten about that. Essentially it's an alternate unique id. It's actually possible to create a duplicate tag, but then you'll find that you can only retrieve one of the pubs that has that tag, so you don't want to do that. Usually on picks a string of consonants from the magazine title and appends the date, in some format. The tag for the November 1957 issue of "Infinity Science Fiction", for example, is "INFNTOCT57". The date at the end pretty much ensures uniqueness.
All pubs have tags; for everything but magazine the tag is generated automatically, and the uniqueness is ensured by the ISFDB scripts. I believe the reason that the magazines don't do this is to make it easier to build the magazine index pages. In the case of "The Saturday Evening Post", I'd pick a string like "STDYEVENPST27JUL1947".
For the editor, I'd probably enter "unknown". Using "uncredited" or "Anonymous" would imply that you had the publication in hand and the editor could not be determined, so those aren't appropriate. Mike Christie (talk) 21:29, 18 Jan 2007 (CST)
The general rule of thumb is to remove all the vowels to make the tag (THSTRDVNNGPSTJUL1947) though I still have no idea why it's done that way rather than a more human friendly TheSaturdayEveningPost-Jul1947 (which fits in the 32-character tag field). Yes, use "unknown" for the editor but also add a note to explain the story is getting entered from a photocopy and that the editor's name may well be available if someone has the original publication. Marc Kupper 00:02, 19 Jan 2007 (CST)
Ahasuerus has pointed out to me that apparently if you leave the field blank, it fills it in for you (using the algorithm Marc suggests). I didn't know it did that. Mike Christie (talk) 19:26, 19 Jan 2007 (CST)

Single-author omnibuses

For single-author omnibuses, is there any reason not to enter them as collections? I'm thinking of things like this, this, and this -- one collection, one omnibus, and one I entered as a collection. Does the OMNIBUS type tell us something more that we can't tell from a COLLECTION type? Mike Christie (talk) 19:29, 19 Jan 2007 (CST)

We typically use the term "omnibus" to describe a book which contains at least two or more previously published novels. This definition of the term (one of a few given by the OED) is favored by many genre bibliographers and booksellers, but is not always properly understood by the general public. It emphasizes the reprint nature of the edition and usually implies that it is inexpensive since omnibus editions have been typically downmarket reprints since the 1930s when they were quite popular. Ahasuerus 19:52, 19 Jan 2007 (CST)
OK, I'm convinced. I think one generally knows an omnibus when one sees one. Mike Christie (talk) 21:26, 19 Jan 2007 (CST)
It gets interesting when it's an omnibus that also includes a short story. For example, I decided to call The Winds of Darkover & The Planet Savers an omnibus though it also includes a shortstory. The publication [The Planet Savers/The Sword of Aldones http://www.isfdb.org/cgi-bin/pl.cgi?THPLLDNS9C1980] also got called an omnibus though it includes the same short story. But, someone entered The Planet Savers as a collection. The story The Planet Savers itself is exactly 100 pages putting it right on the border between novella and novel while The Waterfall which seems to get included with planet-savers is a 13 page shortstory. If I was entering that last pub I probably would have called it a novel that contains both the main novel and a short story. Marc Kupper (talk) 18:08, 20 Jan 2007 (CST)

Editors of collections

I started entering some Anderson and ran across a book Marc has updated with some DAW data: The Book of Poul Anderson. If you read the notes, you'll see that Marc has quite reasonably credited Elwood as author/editor since he is credited on the title page. However, I suspect that his role here is that of packager, or perhaps freelance editor/agent. He may have selected the stories and obtained copyright and so on, but this is no more than any other editor at a publisher does for any collection.

However, going by the rule that we enter what's on the pub, I think we have to record Elwood's participation in some way. Does he deserve an "Editor" level entry, though? I am in two minds about this and would like to hear other opinions. Marc, did you base your decision simply on the fact that he's on the title page? Mike Christie (talk) 21:26, 19 Jan 2007 (CST)

regarding the decision process for The Book of Poul Anderson. I first gave thought to removing Anderson's name as he's not really on the title page (his name is in medium type in the upper/right corner) but figured that would create a bit of a disconnect bibliographically plus his name is on the spine where DAW normally puts the author’s name. As for Ellwood – that he was on the title page carried a lot of weight but that alone was not enough to get him listed as a “co-author.” After spotting him on the title page I flipped through the book for more evidence and saw that he wrote the foreword plus got the copyright assignment. I then checked each of the stories in ISFDB to see if Anderson had written any new material for the collection. There was none and it’s possible Anderson had next to zero input on the book as I figured he could not resist sticking in a preface had he worked with Ellwood on the project. Thus I was comfortable with listing Ellwood as a contributor to the publication though also added a note explaining the roles. Marc Kupper 02:17, 21 Jan 2007 (CST)
There have been numerous requests to add support for editors of single author books, be they novels or collections, especially since Tor and some other publishers frequently make this information explicit on the copyright page these days. However, there is nothing in the software (that I know of) that supports this functionality at this time.
The current practice is not to list editors as "co-authors" and mention them in the Notes section instead, but there are quite a few cases (that mostly come from Amazon.com) which do. For example, see this Avram Davidson collection. One of the Publication records lists Davis and Silverberg as co-authors and the other one doesn't.
If we decide to create a feature request for this, we will need to figure out whether we want to add "single author editor" support at the Title level as well as at the Publication level. I am pretty sure that we generally don't want novel editors to be listed at the Title level since we would then have to split many Titles unnecessarily. After all, "essays" like introductions and afterwords do not a new Title make (in our little world).
However, I can see how editors could be claimed to be so closely associated with the selection of works for collections that they belong at the Title lefvel as well we the Pub level. For example, the two Leinster collections published in the US and the UK in 1976 were really two separate Titles with a different story selection, one by Brian Davis and one by John J. Pierce. As an aside,we currently list the two under the same Title, which needs to be unmerged.
Finally, copyright information is generally not a good source of editor attribution. Too many works are "works for hire" in this day and age, which can mean that the copyright belongs to a third party, not to the author or to the editor. Ahasuerus 01:37, 20 Jan 2007 (CST)
My first novel, Napoleon Disentimed, had "Ben Bova's Discoveries" in large letters above the title on the cover, and my name in small print on the bottom (my wife was upset about that, hehe). There was an intro by Ben on the very first page, then his BB Dis. on the title page also. My second and third books both said "Ben Bova Presents" on the cover. (I was at an enormous book-signing at the S.F. Hilton once and a guy came up with one of books and, as I signed, gushed about how my novel about the spaceman in orbit was his *favorite* book of all time. I puzzled about this for a while but thanked him heartily and he went away happy. It was only afterwards that I realized he thought I was Ben Bova and that he had just gotten a new Ben Bova book! Wonder what he thought when he examined it more carefully -- he told me he wuz a schoolteacher from Indiana....) So, anyhow, what do we do (if anything) above the Bova presence on/in these three books? Hayford Peirce 11:39, 20 Jan 2007 (CST)
"Presenter", e.g. "Isaac Asimov presents", is presumably a separate role. The Library of Congress uses various MARC21 codes for contributors, e.g. illustrator, editor, etc, but we don't have a similar structure, so for now it just goes under Notes. Ahasuerus 17:21, 20 Jan 2007 (CST)
There is a title data type called EDITOR that could be employed. This would allow the editor to be acknowledged for collections while not having his or her name appear in the main title record. In fact, perhaps the EDITOR type or something similar (CONTRIBUTOR?) could get expanded to handle other background people such as cover designers, translators, etc. where ISFDB may not have a specific pigeonhole available. We use [Add Title] and set the title itself to match the pub’s and in (parentheses) explain the role. In the bibliographies you add a "Contributed to:" section and you’d see a neat list of titles and the role that person had. Marc Kupper 01:43, 21 Jan 2007 (CST)
Sure, it could be done (although we would need to ask Al whether we want to use the EDITOR title_type or if something new would be needed), but keep in mind that it would require a fairly substantial upgrade of all related display logic. I suspect that it would be rather time consuming, but we could certainly create a feature request along these lines. Ahasuerus 18:39, 21 Jan 2007 (CST)
Right now EDITOR quietly means magazine editor (the software actually generates one of these records when a new magazine is entered, and it's displayed in a biblio under 'Magazine Editor'). There is a more general need for publication roles: illustrator, narrator for audio books, editor, possibly translator (that's a little thornier as a translation generates a new text). This needs some serious thought before putting something together (does a narrator get a title entry in the ISFDB? How to disambiguate between different types of editors?) Alvonruff 05:55, 22 Jan 2007 (CST)

Baffled by pseudonym-creating

The mystery writer Ellery Queen (in itself the pseudonym of two cousins) also, towards the end of his career, had various books written by prominent SF writers under the EQ byline. If you go to Ellery Queen you'll see three of them listed, Vance, Sturgeon, and Davidson, the last being entered by me. But in doing so, I also did things backwards intially, so that on the Ellery Queen page you'll see an entry saying that Queen also wrote as Davidson. I can't find any means of deleting this. Can someone help.

Also, I'd just like to say that this whole business of naming pseudonyms is absolutely opaque and impossible to understand. Even in the Help page, someone writes plaintively of "cognitive dissonance." That's fer sure! I have at least one suggestion for how the pseudonym page could be reworded so as to make it *possibly* clearer -- if anyone would like to hear it. Hayford Peirce 19:59, 20 Jan 2007 (CST)

It's my understanding that the current pseudonym logic does next to nothing which is why it's hard to understand. About all it seems to do is to trigger the "Used As Alternate Name By:" and "Used These Alternate Names:" stuff you see at the top of the author pages. It seems to have no affect at all on the title and publication displays.
As for how to delete them? I don't think you can at the moment but will add a feature request as I have an idea that should be easy to implement. BTW, my attempts to delete the existing link resulted in second link from Avram Davidson to Ellery Queen being added... Marc Kupper 00:30, 21 Jan 2007 (CST)
Well, there are two challenges here. The first one is to understand the complexities inherent in the way pseudonyms exist in the real world, including house names, self-collaborations, etc. Al has been struggling with it for many years (with some help from the peanut gallery) and the underlying structure is now quite robust and lets you mimic what's happening in the real world pretty accurately. The display logic still has a number of bugs in it, though.
The second challenge, however, is to hide these complexities from the casual user and make them look easy. I don't think we are anywhere near done in the second area. Moreover, it's an inherently more difficult task since "easy" is subjective: one person's "easy" is another person's "completely counter-intuitive". I suspect that it won't be always possible to reach a compromise, especially given out limited resources and the Web's less than sterling support for anything other than basic forms. Web 2.0 may help some if and when we migrate to it. Ahasuerus 18:35, 21 Jan 2007 (CST)

Artists and illustrators

Does anyone have any definite rule as to the correct procedure when listing books with a company only listed as the designer of the cover or cover art? And what if the bookjacket is a photograph or historical artwork ( altered or not?). Could a seperate option be included at some stage to 'Add Photographer/ Designer' or similar nomenclature? Thomas conneely 16:06, 21 Jan 2007 (CST)

(copy pasted from Thomas Conneely's user talk page as he had posted the questin in both places). I don't think there is a "definite rule" but this is what I use.
  • If it's a company that did the cover then just list the company as a cover artist. For example, yesterday I added a book where the cover art was credited to "G-Force Design."
  • If it's just the cover design, photographer, etc. then you can either add a comment in the notes or you could add an INTERIORART record ([add-title]). Let's say the title is A Way Home and so you would do
    • [Add-Title]
    • Page: cv
    • Title: A Way Home (Cover Design)
    • Type: INTERIORART
    • Author: credit whoever did the cover design
That will allow the contributors to get created and linked up both for the publication and on the contributor's page. Marc Kupper 16:19, 21 Jan 2007 (CST)
Worst so far I've seen is in Robert Rankin's books. Cover Art, Illustration, Sculpture, Photograph, Models, Technical Design and Technical Support all get credited at some point. And that's just from the first dozen I picked from the paperbacks. Sometimes it's a person, sometimes a company, sometimes it's the Author himself. He's out to make our lives miserable I think. :-( BLongley 17:17, 23 Jan 2007 (CST)

Tor ISBNs

Entering a Tor paperback, I recalled a comment of Al's that Tor mis-hyphenate their ISBNs. The ISBN on the book I have in hand is "0-812-53083-7"; it's in the ISFDB with hyphenation of "0-8125-3083-7". Do we need to make any note of this as we verify these pubs? Should we, for example, make a note in the notes field of the Tor version of the ISBN?

I'm also curious as to what the standard or definition of ISBN is that permits us to determine that Tor are doing this wrong. Can anyone enlighten me? Mike Christie (talk) 17:14, 21 Jan 2007 (CST)

Al eventually found the standards document and posted a link to it, which I don't have handy at the moment. We may want to check with Al first to make sure that after reviewing the document he still thinks that Tor is mis-hypenating their ISBNs. Ahasuerus 18:26, 21 Jan 2007 (CST)
Tor has always mis-hyphated their ISBNs and when Al added the hyphenation I suggested that he add to the tables a special case for Tor. I'm not sure if Tor's hyphenation was an error that they decided to stick with or if they knew the correct hyphenation but chose to use 0-812-5xxx so they would be more similar to established publishers that had the larger x-xxx-#####-x blocks.
See http://www.isbn-international.org/en/identifiers.html which lists the area and country codes and some formatting info. I've written a lot of ISBN code over the years and could write at length about them. :-) Marc Kupper 20:53, 21 Jan 2007 (CST)
They share 0312 with St. Martin's, and the correct formatting there IS x-xxx-xxxxx-x. 0812 puts it into another formatting group. They have the same issue with the newer 0765 ISBNs. I generally have exception handling for 0812 and 0765, but it's in different areas depending on whether it's a display, editor, or moderator routine. Which app displayed the "0-8125-3083-7" version? Alvonruff 05:44, 22 Jan 2007 (CST)
It was the pl tool. Here's the link.

Problem on the John Dickson Carr page

I just went to the JDC page to merge a couple of new entries and discovered that clicking on *any* of the novel titles brings up and error message beginning "A problem occurred in a Python script. Here is the sequence of function calls leading up to the error, in the order they occurred." Gobblydegook (to me) follows.... These errors weren't there a couple of hours ago.... Hayford Peirce 15:25, 3 Feb 2007 (CST)

Yes, there's a problem with title.cgi right now -- it broke an hour or so ago, then worked for a bit, and now it's broken again. I think I detect Al working on the code. I let him know on his talk page, just in case he hadn't realized something was broken. Mike Christie (talk) 15:40, 3 Feb 2007 (CST)

Date of First Publication vs. Copyright Date

Stories that are first published in the January issue of a magazine usually have a copyright date of the previous year. I ran across this merging two titles: one was a story's first appearance in the Jan. ish of a magazine and had a date of 1993; the other was a reprint of the story in a collection, and had a copyright date of 1992. Which date should be kept? (Of course, I didn't realize that this was the problem until I had already submitted the merge request....) Jefe 00:18, 12 Jan 2007 (CST)

Keep the 1993 date. We track official publication date, which is not the same as copyright date; see Help:Screen:NewPub (the section on "Year") for the help paragraph on this. If you happen to know the date is wildly wrong (e.g. a much-delayed publication), add a note to the note field for the pub. Mike Christie (talk) 06:01, 12 Jan 2007 (CST)
Ah... I see it now. I re-edited dates for the titles I changed, so the correct (publication) dates will be shown.Jefe 12:47, 12 Jan 2007 (CST)
Just had a query on my update of the title "Antimatter" by "Robert L. Forward". As it had a date of 0000-00-00 I updated it to the copyright date from the first publication I found it in - 1995-00-00. But as it wasn't published till 1995-09-00, have I gone too early? (Whoops - just realised I didn't add my name to this, no wonder I got no answer...) BLongley 16:37, 1 Feb 2007 (CST)
Yes, it should have been 1995-09-00 -- title dates should correspond to the date of the first pub. 1995-00-00 doesn't mean any particular month, so you haven't made it too early; you just haven't made it as precise as you could have. Mike Christie (talk) 06:35, 2 Feb 2007 (CST)

Moderator Noticeboard, help desk, other pages?

Would it be beneficial to start a new page, Moderator Noticeboard? We're again seeing a lot of traffic on this one page, and it would be good to split it up to appropriate places. I was thinking that a noticeboard page for moderators would be a place to communicate about issues related to moderation. Two current examples of things I would post there: there are a lot of GURPS deletions in the queue, which I'm inclined to leave for Ahasuerus since I know he's looked at the question of what roleplaying stuff should be in the ISFDB. I would post a note there asking him to take a look. Also there's a DAW pub delete for Invader which I would bring to Marc's attention as he's the expert on DAW. Other discussions that could take place on that page would include criteria for giving people the mod bit, and discussions about any individual candidates for it.

We've also talked about a help desk page, where questions about how the ISFDB works could be asked and answered. A lot of what is on this page really falls under that heading. Any objections if I start one up? I suggest that if we do that, then we should move active posts from here to there if they are appropriate for that page.

Any other pages that should be active that we're not using as much? Bibliographic Rules comes to mind. Mike Christie (talk) 05:19, 13 Jan 2007 (CST)

I’d rather get away from wiki messaging entirely. You probably have noticed the unread marker I’m now. That helps a lot though it’s not perfectly reliable and for things like the community portal I just spooled through 20 diffs that I realized I had already seen before I saw new traffic. A moderator-board seems fine as there have been times I’ve wanted to make a comment to all the moderators – for example, I was in the middle of a particularly messy editor submission, had to run, and wanted do document where I was at (I was mid-way through writing comments on the editor’s talk page) in case another moderator had time to pick up the thread. As for the DAW pub-delete, it’s no longer in the queue and so whoever had it on hold must have figured out how to deal with it. I suspect it was ok as there are quite a few duplicate publications. Marc Kupper 00:13, 14 Jan 2007 (CST)
I really want to get away from wiki messaging too. I'm not sure what triggers the "You have new messages" but I kept finding it didn't actually mean I had new messages - possibly my not following all links on my talk page? I also find I have pending edits that are pending for days with no reason given I can find. A more direct contact method would be good, a centralised place for discussion that doesn't need us to search for changes on other pages is good too - I'm not getting as many replies here as I expected. Given the occasional activity bursts I see where I can post a change that needs approval, and I see it approved before I can even explain what I want to do next but can't as I'm not Moderator, a real-time chat might be good too. OK, that leads to IRC/AIM/MSN/YIM arguments, but there's multi-chat clients out there. BLongley 16:50, 1 Feb 2007 (CST)
"You have new messages" is strictly tied to changes on your talk page. There does seem to be an intermittent bug as visiting your talk page should clear the message until your talk page changes again, but occasionally I've seen the message stay for a while. That's a MediaWiki bug, though, so it won't be fixed till Al finds a reason to upgrade to the latest MediaWiki software.
Pending edits -- if the edit is on hold but you haven't seen a message yet, it almost always means the moderator wants to do more research or has a question but hasn't had time to post it to you. If the edits are pending, it's just that there are no active moderators at that time. See my note below.
IRC is a good idea. I know Wikipedia admins use it extensively, and it might be useful here. However, there is also value to having a visible record of a lot of discussions that everyone can read. I certainly mine these discussions for updates/improvements to the help pages. Plus IRC only works for synchronous discussions; the wiki, and forums, are better when asynch is what's needed, which for now is more often the case.
Part of the reason for lack of replies is lack of bandwidth. I try to spend some time in the early morning catching up, and late at night; every now and then I can spend much of an evening working on ISFDB stuff. But work and family ties often prevent me from doing much for a day or two together.
I know wiki messaging has its problems, but a lot of what we do here, such as project page and help page maintenance, could not be easily done in a forum environment. My experience at Wikipedia makes me think we'll get used to it eventually. Mike Christie (talk) 06:58, 2 Feb 2007 (CST)
Based on Wikipedia's experience, admins-only IRC channels and other closed-to-the-public tools can promote cabals, political infighting, etc. Of course, Wikipedia is an inherently more political place that the ISFDB, but I think there is still value to keeping ISFDB communications open.
Having said that, although I think that the current Community Portal model is generally viable, its current implementation makes it hard to follow when the number of active sections exceeds 20 or so. I suspect it would be more user friendly if it could be organized as a proper forum with multiple threads. Perhaps the currently dormat ISFDB Google group would be a better option? Ahasuerus 08:00, 5 Feb 2007 (CST)
In sleeping on this I'd say "no" to more purpose specific talk pages. I was cleaning up my desktop as I realized I'm not regularly using pages such as Bibliographic_Projects_in_Progress, Bibliographic_Rules, etc. I sort of know they are out there but when it comes time to post a message it's often a struggle to find the correct page. For example, I know there is a conversation underway about ISBN formatting but I can't remember what page it's on and so ended up dumping some thoughts on Mike's user talk page. This is something a message board tends to take care of automatically as 1) It's a single "page," every thread has it's own subject that's visible, as interest in a thread declines it falls off the radar and does not continue to clutter the desktop, there's no need for manual maintainence to archive discussions, it's easy to see new comments, and if you use your own mail reader you can file messages, flag them for follow-up, etc. as needed. Marc Kupper 13:09, 14 Jan 2007 (CST)

Linking material

I'm entering Poul Anderson's "The Earth Book of Stormgate One" which is a collection; Anderson wrote fresh linking material between the stories. This material is fictional, in the sense that it is written from the viewpoint of someone in the fictional world. There are no titles for any of this material.

I'm having a hard time figuring out how to enter this. I think that if a Poul Anderson researcher were to riffle through the ISFDB they would want to know of the existence of this material. But it feels a bit odd to create five or ten fiction entries with a title of "untitled". Is there a better way to handle it?

For now I think I'm just going to add a note saying "Additional fictional linking material is included between each story". Mike Christie (talk) 20:34, 21 Jan 2007 (CST)

I had a very similar thing in an anthology yesterday and was thinking of for story "xyzzy" I would use Preface (xyzzy) for the "in-between" stories but then got to thinking about that to make it sort correctly I would need to have the page numbers for the stories point at the stories themselves meaning it would not match the publication's table of contents. In the end I just went with a note explaining that each story in the anthology was prefaced by a one page essay. Marc Kupper 20:58, 21 Jan 2007 (CST)
I've got a stack of Poul Anderson's here to do next, I'll add my opinions when I've discovered how hard THAT is. BLongley 17:13, 22 Jan 2007 (CST)
Update - I've added my Poul Anderson stuff, let me know if I've messed anything up. I know I had qualms over some of the "The Merman's Children" bits, but as it's the first time in ages I've picked up a book and seen real book-worm damage, I want to remove the pub from my house ASAP! :-( BLongley 18:51, 27 Jan 2007 (CST)

Unmerge Titles

What's the current status of unmerge titles? I got a copy of The Star Diaries and discovered it's under Dzienniki gwiazdowe. It looks like that title would be best dealt with via unmerge and remerging the stuff. There already were variant title though they are empty.

Also - I've forgotten - what's the ISFDB policy on minor title variations? Do we have separate title records for Star Diaries and The Star Diaries or is it ok to have a single title record for both types of publications? I'll probably change the title to The Star Diaries anyway as I started looking for The Star" and wondered why it was not on the page. Marc Kupper 19:33, 22 Jan 2007 (CST)

Al took away unmerge, largely because it had been misused more than once, the last instance being my mistake. I've asked for it to be returned, at least to moderators; I don't know if that's easy to do.
Current policy, at least the way I wrote the help files, does not define this, but does say that punctuation is enough to cause a title change: see Template:PublicationFields:Title. I think that any variation is a different title, except for capitalization and spacing.
I also just noticed that that help file doesn't say subtitles should be entered. We had some discussion about title records not using subtitles, and pub records using them (or am I misremembering?). I'd rather drop subtitles completely, and note them in the notes field if we want to, but I don't think this is the consensus position. Does this text need to be changed? Mike Christie (talk) 19:40, 22 Jan 2007 (CST)
Maybe Al will take pity on me because Dzienniki gwiazdowe has 10 publications and needs to be split into five titles.
“punctuation is enough to cause a title change” – Hmm – that would seem to add a lot of clutter to the author bibliographies. Does that mean that all of the publication titles should match the parent title directly or that the only reason the publication titles are different is because publication records sometimes include the sub-title? Any change to a publication title, no matter how minor, would trigger a new title record?
“doesn't say subtitles should be entered” – I thought the last word was that sub-titles would always get entered. I have been entering whatever is on the title page other than I sometimes usually don’t enter the series name even if it’s on the title page. If what I enter does not match what’s stated I’ll add a note explaining what’s stated plus I add a note if the title page is different than the cover for either the title or author name(s). Marc Kupper 23:03, 22 Jan 2007 (CST)
The general policy -- as I understand it -- is to record all minor title variations, including diverging articles, as Variant Titles. The only exceptions are clearly defined subtitles, which, as Mike points out above, are listed in Publication records, but not in Title records. This is done to ensure that when users look up their copies, the software can find them. Thus "Day Watch" and "The Day Watch" are both Variant Titles of Sergei_Lukyanenko's "Dnevnoi Dozor" (assuming the forthcoming Canadian title ends up being different and not an Amazon.com fluke), but if it ever gets reprinted as "Day Watch: A Novel", we would enter it as a new Publication under the Title "Day Watch". Also, please note that "Dzienniki gwiazdowe" was published in English in two volumes, "The Star Diaries" and "Memoirs of a Space Traveller" (see http://en.wikipedia.org/wiki/Dzienniki_gwiazdowe and http://contento.best.vwh.net/b20.htm#A459).
Some other thoughts on the subject. I have been thinking about Lukyanenko's bibliography ever since I read a rec.arts.sf.written discussion a few days ago. The first challenge is that the second volume in the series (which his US publisher calls a "trilogy" even though volume 4 has been out for over a year) was originally published "as by" Lukyanenko and another Russian writer, Vladimir Vasiliev. I have since confirmed that the Russian edition of volume 2 in the series (but not volumes 1, 3, and 4) was indeed a collaboratiion or at least published as such although the forthcoming US edition lists Lukyanenko as the sole author. This is easy enough to fix using our Variant Title/pseudonym support.
The second challenge is that we currently display (some) Russian language titles and author names using Latin characters. The advantage here is that most ISFDB users will be able to pronounce the titles, but the disadvantage is that there is no universally agreed upon Cyrillic-to-Latin transliteration algorithm, so we may lose some information.
In an attempt to come up with a higher level of abstraction, I have identified at least 4 different and arguably useful data elements for each foreign language Title: (1) the original spelling in Cyrillic/Chinese/whatever; (2) a Latin transliteration of the title; (3) a literal English translation; (4) the Title(s) used by the US/UK publisher(s). The question then is which ones do we want captured and displayed and how do we fit them in the current Variant Title scheme? Do we just list the original Cyrillic/Chinese title as the parent and the rest as Variant Title with an explanation of what each one means, i.e. "transliterated title", "literal translation", "US", "UK", etc in parentheses? Ahasuerus 14:05, 23 Jan 2007 (CST)


> This is done to ensure that when users look up their copies, the software can find them.
Perhaps the title search should also scan the publication records rather than us needing to duplicate all of the publication title variations at the title level. That would also allow the search to find the foreign language translations which are sometimes embedded under the English title record.
One of the reasons I have been entering notes about common misspellings or variations for author names and/or titles is so that people using Google will find the ISFDB record even if they happen to search for a variant or misspelled title or author name. I’ve thought about adding some of them as variant titles (particularly when the title and/or author on the cover does not match the title/author on the title page) to help people find the record though unfortunately they will get the VT record and then need to up to the parent to see the publications.
> The second challenge is that we currently display (some) Russian language titles and author names using Latin characters.
Where are the Russian or other foreign titles coming from? I thought we’d just be using whatever is stated in publications rather than trying to enter our own idea of how something translates particularly as there are often many ways a word or phrase can be interpreted. Marc Kupper 01:01, 24 Jan 2007 (CST)
> note that "Dzienniki gwiazdowe" was published in English in two volumes, "The Star Diaries" and "Memoirs of a Space Traveller"
I had noticed that and was wondering where to file Star Diaries: Further Reminiscences Of Ijon Tichy. The problem is "Star Diaries" is the volume 1 name and "Further Reminiscences Of Ijon Tichy" is the volume 2 subtitle meaning it's ambiguous. Sight unseen I was going to assume this was really volume 2 though that will play havoc with people looking at the bibliography. It may be vaporware as zero copies of the ISBN seem to be available. Marc Kupper 01:06, 24 Jan 2007 (CST)

Extra Info from a Primary Source

I often find a list of details from a publication that isn't represented here: e.g. when checking "Venus Plus X" I added details of the book in hand, the fourth printing. Which contains details of the second and third printings as well, down to month. Would you like those added too or are they unreliable? BLongley 17:21, 22 Jan 2007 (CST)

I think these can be added; just make sure the source of the data is noted in the notes field. I have been planning to do this with Le Guin, once I get through the Locus1 verifications, at least for the Earthsea books. Mike Christie (talk) 19:43, 22 Jan 2007 (CST)
I dealt with this a recently in that I had a 3rd printing of A Way Home that provided the publication dates for the 1st and 2nd editions. Once I had the ISFDB record for the 3rd printing nailed down I cloned it to create a record for the second printing along with a note explaining what I had done. ISFDB already had a record for the first printing otherwise I would have made one along with the record for the 2nd printing.
Note that when I do this I will blank out those fields that may be changing from printing to printing such as the price, ISBN, and cover artist and also add a note that clearly states the source of the information for the ISFDB record. The last thing I want to happen is for a future editor to think there possibly exists a real book with a particular combination of price, ISBN, and cover artist, etc. Marc Kupper 21:12, 22 Jan 2007 (CST)
Sounds reasonable. With some older (pre-WWII) books this may be the only practical way of entering rare editions and printings. It's not 100% reliable, but I think we should be OK as long as we document the source of our information. Ahasuerus 11:37, 23 Jan 2007 (CST)
Ok, good to get some consensus on this! I've been working from my reading copies mostly, but I know there's a pile of H.G. Wells hardbacks stored upstairs that I will eventually get around to - I think he counts as a pre-WWII SF author? ;-) BLongley 17:49, 23 Jan 2007 (CST)

Plans for Earthsea

Just as an FYI, in case anyone thinks this is a bad idea or has comments: when I've brought the Le Guin biblio as up to date as I can get it, I plan to use every source I can find to enter "A Wizard of Earthsea", and then post the resulting biblio to rec.arts.sf.written and ask if anyone has any edition that is not on that list. I thought it would be a good way to get people involved without having to get them using the editing interface, and I also thought it would be a genuinely useful way to validate a biblio. We might eventually pick a book a week to post the biblio and ask if anyone has a copy that's not on the list. Mike Christie (talk) 19:46, 22 Jan 2007 (CST)

The idea itself seems fine. I would set up an ISBDF “Book of the Week” page that has an explanation of exactly what data we are looking for and then has a list of the books. My experience with things like the DAW list and ISFDB is that we have pretty specific requirements for the information we are including in ISFDB and that we occasionally update the requirements as we learn of how others interpret the existing wording.
The DAW list made me realize that I’m also finding it’s easier if people enter the data directly into ISFDB rather than giving it to me and that I enter it. I don’t get a lot of “joy” out of copy/paste data entry.
Since you are looking for A Wizard of Earthsea I could be a guinea pig for what you do as I have a copy. To whet your interest, it’s a Bantam ISBN that’s not listed in ISFDB. Surprise, I thought it was a 14th printing that lists the publication dates for all 14 editions but way down at the bottom of a the page is a number line with 23 22 21 … 16 meaning it’s a 16th printing and we have the dates for printings 1 to 14. I think the thread about how to structure the title page to cluster publications by publisher had died but I suspect that would be handy in this case. Marc Kupper 21:36, 22 Jan 2007 (CST)
Just added my edition too, and it seems like a LOT of hard work ahead - good luck! BLongley 17:03, 23 Jan 2007 (CST)
Thanks! I just took a quick look at your sub and noticed you didn't specify a printing number; can you tell what printing it is? The price will help figure it out but I would like to identify the printing if I can . . . . Mike Christie (talk) 17:07, 23 Jan 2007 (CST)
No printing number. Full details are "First published in the U.S.A. by Parnassus Press 1968. Published in Puffin Books 1971. Reprinted 1972, 1973, 1974, 1975."
"Copyright Ursula K. Le Guin, 1968" and "Drawings Copyright Ruth Robbins, 1968"
I can give you the prices for Australia, New Zealand and Canada too, or page numbers for some (very small) bits of artwork too if you need them - but how detailed do you really want? I can scan bits of the book if you like.
Oh, and beware of using price to sequence editions: I'm pretty sure I've seen a later edition where the price dropped rather than inflated. I can't recall where though, but check my submissions today if you need the example. BLongley 17:22, 24 Jan 2007 (CST)
OK, that's what I needed -- it's a fifth printing; mine was a 6th and said "reprinted . . . 1975 (twice)". Thanks. Mike Christie (talk) 21:06, 23 Jan 2007 (CST)
A long time ago I was thinking of ways to document each publisher's practices so that people looking at a book could figure out how to decode it. Now that wikis exist I suspect the task would be simpler in terms of allowing for multiple contributors to the effort.
Mike - I'm not sure if you wanted me to just enter my Earthsea (and perhaps clone in 14 times for each of the publication dates mentioned) or if you want to try sending me an e-mail just like I was on rec.arts.sf.written and I'll give the book to someone like my wife and we'll see what sort of data you get back. Marc Kupper 23:46, 23 Jan 2007 (CST)

Editing or Submitting Raw XML?

Maybe it's just me, but I'd much rather be able to create and or edit an XML file like the one displayed on the post submission page then use the GUI form ... I'm a lot more comfortable typing XML then i am tabbing arround and picking itmes form pulldowns and clicking "Add Title" buttons.

is there perchance an upload page where submissions can be contributed that way? (presumably ther are also some docs on the XML format i'd want to read then as well so i'm not reverse engineering it) The previous unsigned comment was added by Hossman.

At this point there is no way for an editor to submit XML. Al may choose to make XML submissions available at some point in the future, but I think it might not be trivial. I could imagine, for example, that much of the current validation is done in the Python cgi scripts, and XML submissions would bypass that and require additional coding to revalidate and report on errors. There is a roadmap at the top of the feature list; XML submissions are not listed there right now. Mike Christie (talk) 07:25, 23 Jan 2007 (CST)
I'm by no means expert in XML, although I deal with it every working day in some way or other. I can see some appeal in creating an XSD for proposed submissions, if there's a chance we'll ever be able to submit them: it would clarify submission rules a bit for the techies. In the short term, my OWN list of publications (where I've never bothered recording publisher, pages, price or such) could fairly easily be transferred from a Pocket Excel spreadsheet on my PDA to my PC to an Oracle database and out again as an XML file with Title and Author and format, which should hopefully be easily comparable against the ISFDB to let me know where my talents are best directed. (I prefer to add new data than confirm old: just my preference though, the verification activity is valuable but seems lazier to me.) At my current editing rate, I've got a year of Primary sources to check though. :-( BLongley 17:06, 23 Jan 2007 (CST)
Rather than submitting XML to ISFDB you could issue a POST request. You will need to include a cookie (for authentication) but the interface is already public and supported by ISFDB though it is subject to change as the edit forms get updated. If a bunch of people say "I'd like to send XML" then a thought is to create an app that accepts XML and issues the POST request. Marc Kupper 00:02, 24 Jan 2007 (CST)

Can we do "Least Accurately Named Trilogy" as a Special Project?

As we're talking Le Guin's "Earthsea", and it seems there's 5 in that now... and Douglas Adams couldn't stop "HHGttG" at 5 either... and Robert Rankin has 7 or so in one trilogy, and keeps adding to previous trilogies... what's YOUR pet hate?

For most unreasonably extended series? Probably the Gor books, which should have stopped at one. Or zero. Mike Christie (talk) 20:24, 23 Jan 2007 (CST)
I was specifically thinking of series CALLED a "Trilogy" that stopped being one but kept being called one. I remember when there was a "Dune" Trilogy, but I haven't seen it called that for a while now... BLongley 17:19, 24 Jan 2007 (CST)
I think the reason for just adding more to a "trilogy" is authors and publishers don't want to call them quadrilogies and quintologies though L. Rob Hubbard used "decology" for his 10-volume set. I still call them the The Foundation Trilogy (10 volumes?) and Dune Trilogy (a whole Duniverse of books?). Marc Kupper 02:50, 26 Jan 2007 (CST)
"Dune" seems to separate into sub-trilogies fairly easily, "Foundation" less so - although the "Killer Bs" (Bear, Bentford, Brin) as a sub-series amuses me. BLongley 18:58, 27 Jan 2007 (CST)
Well, there is the Brentford Trilogy, which has contained more than 3 volumes for almost 20 years now, but is still called a "trilogy" by the author/publisher as a little in-joke. Then there is Sergei_Lukyanenko's "Watch" series, which was advertised as a "trilogy" when it was published in America in 2006-2007 even though book #4 appeared in 2005. Ahasuerus 17:26, 2 Feb 2007 (CST)

Searoad source notes question

I entered the first edition of Searoad, and verified it, a while ago. Then just now I verified it against Locus1, and added the "October" part of the date, which is not apparent from the primary. I haven't been adding notes when I do this sort of thing, since it's apparent from the verification records what sources were used to get the data. However, Locus1, while pretty stable, is web-based. Is there a CD copy of it available? If not, should we treat it as requiring a note if data is sourced only from there? Mike Christie (talk) 20:30, 23 Jan 2007 (CST)

The Locus (and some Contento) indexes are available on CD-ROM. They're actually easier to use on CD, since the CD only has one index, rather than separate indexes for each year after 1998. You can buy them from Locus Online.Jefe 14:00, 24 Jan 2007 (CST)
Yup, that's the primary reason to get it on CD-ROM and pay the compilers real money :) Ahasuerus 17:28, 2 Feb 2007 (CST)

Chapterbooks

Now that there are a couple of bugs filed against chapterbooks, this would be an appropriate time to discuss how they should really work.

I have no problem with having a CHAPTERBOOK publication type, but to me it seems dubious that there should be a CHAPTERBOOK title type. Since a chapterbook is essential a short story singleton, the situation can arise where the same text can appear twice in the same bibliography. For instance, in Le Guin's bibliography, The Water Is Wide shows up twice: once as a chapterbook, and again as a work of shortfiction. I suspect that these two titles are the same work, but the present situation leaves it ambiguous. (There are also two versions of "Nine Lives", one referring to chapterbooks and the other not, but I suspect that's because there's no current way to promote the second title to a CHAPTERBOOK).

This also happened with Buffalo Gals, Won't You Come Out Tonight.Jefe 15:13, 24 Jan 2007 (CST)
To me it seems that the chapterbook label is a publishing convention. As such, I've been making the publication=CHAPTERBOOK and the title=SHORTFICTION, and don't see the need for title=CHAPTERBOOK. What's the general feeling on this topic? Alvonruff 05:29, 24 Jan 2007 (CST)
Well, a couple of things come to mind. First is that we should change it to "CHAPBOOK"; I think there's a bug filed on that in fact; a CHAPTERBOOK is something else.
More to the point, a chapbook isn't always SHORTFICTION. Le Guin's "Dreams Must Explain Themselves" could perhaps be described that way, since there's only one fiction work in it, but I think it's really a COLLECTION. Another example: the British Novacon convention has published some chapbooks in dos-a-dos format, with two short stories in them; one from the GoH and the other from the fan GoH. I have one of these, with stories by Dave Langford and James White.
The problem seems to be that between pub_type and pub_format we may have more than two kinds of things we're addressing, or that we haven't split the options in the best way between those fields. For physical format, you could have:
  • chapbook, pb, hc, dos-a-dos, digest (and other magazine formats), tp, and perhaps others that don't occur to me offhand.
I think "dos-a-dos" is the problematic one, because the others are pretty much mutually exclusive. I had suggested using "dos" in pub_format for things like Ace doubles, but since there are dos chaps and dos hcs I don't think this is going to work.
Other conflicts: a magazine can be in chapbook format -- that is, saddle-stapled and digest-sized. NONFICTION can also be an anthology, collection or even non-genre; there is sf non-fiction and non-sf fiction after all, so something like Silverberg's "Mound Builders of Ancient America" is really non-fiction non-genre. I suppose you could have a hardcover dos-a-dos non-genre nonfiction anthology, though perhaps we shouldn't borrow trouble.
The underlying questions these fields are trying to answer for a publication seem to be:
  • Is this a magazine or a book? This has to do with marketing, distribution, and announced schedule.
  • Is this a novel, a collection by a single author (or a collaborative collection), an anthology assembled by an editor, or an omnibus?
  • What is the physical format of the source? This includes size, hc/pb, and dos & chap indicators.
  • For a title that represents a single work (i.e. one story, essay, or novel), we want to answer these questions:
  • Is this genre or not?
  • Is this fiction or not?
  • What length is it (SHORTFICTION vs. NOVEL, and ESSAY vs. NONFICTION)
For a title that matches up to a pub, we would like to match the pub's answers to all the pub questions I give above. In addition, if the title represents a single work, the title questions should be answered too. Finally, if the title questions can be answered at an aggregate level, they should be. By that I mean that an omnibus of sf novels is clearly fiction and genre, but doesn't have a length; an sf novel is a genre fiction novel; and a collection of mainstream stories by Le Guin is fiction, non-genre, and has no length.
All this seems to me to add up to the need to have a way to mark titles that relate to pubs as slightly different from titles that don't. I think Al's question about chapbooks brings up the inconsistency in the format and type fields that we've already acknowledged.
In case the above sounds too theoretical, here's a specific data scheme that would do what I'm talking about. I'm not saying this is the right answer, just trying to illustrate what I mean by the above.
pub_bkmag: new field to distinguish books from magazines.
pub_format: this field should answer the format questions about pubs -- hc, pb, tp, dos, chap, digest, pulp, bedsheet, tabloid, maybe a "special" type with "see notes" implied. (I don't know how you deal with "dos" being simultaneously present with the other types -- maybe we would have a simple "Dos?" checkbox and a separate field; or maybe "dos-a-dos" goes in the notes along with other special formats such as "element of boxed set".)
pub_type: this field goes away, eventually, because the linked title record does this work.
title_gtype: genre/non-genre/mixed
title_ftype: fiction/non-fiction/mixed
title_ptype: This field is null for titles that don't match pubs. For titles that match pubs it could be OMNIBUS/COLLECTION/ANTHOLOGY/SHORTFICTION/POEM/ESSAY; maybe others too but I doubt it.
title_type: For titles that match pubs this field is forced to match the ptype field. For other titles it could also be COVERART, INTERIORART, SERIAL or EDITOR. (EDITOR still confuses me; I hope this is the right place for it.) I would guess REVIEW and INTERVIEW also fit here.
Al has said before that he doesn't like to continue to carve up the dataspace into smaller and smaller chunks; I think that's a good instinct to have, and perhaps what I describe above is overkill. However, the requirements seem to me include the ability to make the distinctions drawn above, and the existing schema doesn't let us do that.
I also know that I have left several loose ends in the above, such as how we unambiguously know which title goes with a pub; that one has quite a few ratholes. I think it's still worth talking about how things should be ideally, though, so that we can then back off to what we're able to implement realistically. Mike Christie (talk) 08:49, 24 Jan 2007 (CST)
I agree with Mike's basic premise, although I don't understand the underlying structure to comment on his specific suggestions. Le Guin's bibliography is a good case study. She has several collections -- e.g., Searoad, Unlocking the Air, Hard Words and Other Poems -- that are collections of non-genre stories or poems, but which are currently listed as "collection" rather than "nongenre", as well as several collections of essays which are all listed as "nonfiction" rather than "collection". In both cases, the user doing the input has to decide whether structure (anthology, collection, omnibus) is more or less important than type of content (genre, non-genre, fiction, non-fiction), which will lead (probably has led) to these sorts of books being listed inconsistently for different authors. It is quite valuable to know that "Dancing at the Edge of the World" is a collection of essays, and not a single work, but you would not know this from the bibliography. (One might be able to tell that by viewing first the title record and then the publication record, but this requires two steps and also assumes that the contents were actually filled in for the collection, which is not a safe assumption.)
We also need to agree on a definition of chapbook. Mike is suggesting saddle-stapled binding, which is a fairly common usage (in-genre, anyway), I think. But note also that PS Publishing has started releasing hardcover "chapbooks" that contain only a single story. If we adopt the staple definition, we'll need an alternate way of listing things like Pulphouse's short story paperbacks which were perfect-bound (also issued in hardback, implying that they were hardcover paperbacks).Jefe 15:13, 24 Jan 2007 (CST)
Chapbooks were popular at one time and must have had their own best-seller lists as I pulled To What Green Altar? (1932) by Prescott Chaplin off the shelf and at the top of the title page is "Number One / International Chapbooks." As for defining chapbooks by binding - this one is the same as a trade paperback but is skinny enough (38 pages) that the spine is blank. There was room on the spine to print something but I think they did not want to get into registration hassles as the spine is only 1/4th of an inch. Anyway, chapbooks seem more or less like paperback books only with fewer pages. They have covers, title pages, copyright stuff on the verso, etc. I personally don't see a need for specific support for chapbooks in ISFDB, particularly as a search reveals just six titles, all of them modern and it’s possible some of those are children’s chapterbooks.
When a shortfiction gets published as a standalone work is it "bad" if a title shows up twice in a bibliography? If so, I can think of several solutions:
  • Make the shortfiction a SERIAL and use (part 1 of 1)
  • Include the shortfiction as a title in the contents of the novel/chapbook. The shortfiction would be the master title in terms of tracking down where it has been published.
  • Make the novel/chapbook a variant title of the shortfiction.
  • Merge the titles (promoting it to a novel or chapbook) and ignore the fact that it's no longer in the shortfiction list and will show up in publication contents as a novel or chapbook. Marc Kupper 03:55, 25 Jan 2007 (CST)
I don't think it's bad if the biblio makes it clear that one is a pub and the other is a title. Or in terms our users will be more comfortable with, one is a book and the other is a story. I think the options you list have drawbacks -- e.g. option 2 prevents the pub and the master title from having the same type, which I think is undesirable.
Re Jefe's point about hc chaps, format is really two things: size and binding. A magazine can be perfect bound (the usual way to do it) or saddle-stapled, and can be any size for each of those. We're using digest/pulp/bedsheet/tabloid for sizes for magazines, and pb/tp to imply size for books. Those are both standard ways to do it. However, pb also implies perfect binding, which is why we don't need a "binding" for books; and if chapbook implies saddle-stapled, then binding is fully dealt with for books, but size is not, whereas for magazines it's the reverse: you can tell size but not binding.
I think for books, the sizes that are of interest are pb, tp/hc, smaller than standard pb, and larger than tp/hc. We don't need to track sizes exactly, just "large" and "mini" or something like that would do. For magazines, we do want to track sizes. Ideally we could track bindings as well, but it's lower priority.
If "chapbook" means "saddle-stapled" then as Jefe points out we have a conflict with the way many publishers use it, which is to indicate a single short item published as a slim volume. It becomes a category like "omnibus"; difficult to define precisely, but you know it when you see it. If we want to use it in this way, I rather suspect it's going to be a flag on the publication, independent of all the other formats and types, which will govern where the publication is listed in biblios but won't imply anything else -- e.g. pb vs. hc, or single item contents vs. collection, or perfect bound vs. saddle-stapled, or genre vs. non-genre, or fiction vs. non-fiction. It might imply shortfiction vs novel, I suppose. Mike Christie (talk) 07:31, 25 Jan 2007 (CST)

Logging in

There was a discussion about this a while ago but it seems to have vanished (archived?). In any case, I've now been in and out enough times to be absolutely certain about what happens when I go to: http://www.isfdb.org/cgi-bin/index.cgi from a desktop icon. If I have rebooted the machine, or turned it off for the night, then I am *always* asked to log in again. As long as I leave the machine on, however, I remain logged it -- this is certainly an improvement on my previous status, where I had to log back in everytime I left the site (the problem, evidently, at that time was the shortcut I had first created to take me back to the site), but is still not as good as my status at Wikipedia. There I stay logged in no matter how many times I turn the computer on and off -- unless I delete all my cookies and temp files and stuff like that: then I have to log in again. But, having done that, I'll then stay logged in for a month or so or even longer. Am I still doing something wrong here at ISFDB or is this simply the nature of the beast: that log-in info is lost when the machine is turned off? Hayford Peirce 12:30, 25 Jan 2007 (CST)

As the orignal thread is gone
  • are you talking about the login to the wiki side of things at http://isfdb.tamu.edu/wiki/index.php, the ISFDB side at http://www.isfdb.org/, or both?
  • What type of machine (Max/PC and operating system) and browser are you normally using?
  • When you log in are you checking "Remember my password across sessions?"
  • What's your default cookie handling? I allow cookies for the originating site only (third party sites as for banner ads can't set cookies).
  • FWIW - Both the wikipedia and isfdb seem to use a combination of cookies and cached images to remember me. I can delete all of the cookies and the site will remember me though if I try to do something that needs a login it'll say I'm not logged in. If I delete both the cookies and cached files then the sites forget who I am. --Marc Kupper 13:10, 25 Jan 2007 (CST)
  • I'm referring to the login at:
  • http://www.isfdb.org/cgi-bin/index.cgi. This is where it asks me to log in again if I've rebooted. (I'm going there from my desktop icon, created by the Send function of Internet Explorer 7.) This is where I do editing to the database.
  • I'm using a Compaq Presario with Windows XP Service Pack 2. I normally use Avant as my browser but I've tried using Firefox and IE 7 and the results are the same.
  • When I login, I'm *never* asked to "Remember my password across sessions?". I'm just asked my name and password. When I *first* set up my account, then I *was* asked to remember and I checked Yes.
  • About my "default cookie" setting, I just don't know, things are *so* complicated when you go to Tools/Internet Options/Advanced, etc. etc. There are so many different settings -- I do know that most of my cookie stuff is enabled, otherwise lots of things won't run. You wrote "If I delete both the cookies and cached files then the sites forget who I am." Well, that's certainly the case with me, too. But also after just rebooting.
  • Hope that some of the above helps. Hayford Peirce 20:40, 25 Jan 2007 (CST)
You wrote:
> • http://isfdb.tamu.edu/wiki/index.php/Special:Watchlist I don't have any problems. I am never asked to log in.
That's an interesting clue. You are not asked to re-login on the wiki side but are asked on the database side of ISFDB. I'm not sure if it'll help but let's try logging in on the wiki side first and then the database side. First make sure you are officially logged out and then log back in.
See if that helps and in the future you may need to first log into the wiki (where there's the remember password) and then ISFDB where it's not clear how "remember password" works. Marc Kupper 02:21, 26 Jan 2007 (CST)
I did everything that you suggested, following your instructions meticulously. After doing so, yesterday, I then remained logged in to both sites for the rest of the day. I then shut down the computer for the night. When I started it up again this morning I first went to the wiki-index site, where I was (I guess) still logged in. At least nothing asked me to log in again. From there, on the Main Page, I clicked on ISFBD and went to that site. Once again, as usual, I was asked to log in. So I'm afraid that your suggestions left me exactly where I was before. It's no big deal for me to log in once a day at the ISFBD site but, since apparently other people don't have to, I'm curious as to what my own situation is. Probably it's some obscure cookie setting that I have deep within Internet Options.... Hayford Peirce 11:28, 27 Jan 2007 (CST)

Under Hill

The Gene Wolfe story "Under Hill" has two records [2] [3], which in and of itself is not that unusual, but they both link to the same issue of F&SF. Furthermore, the issue of F&SF does not list the story as a duplicate, which seems rather odd. The story titles could presumably be merged to resolve this, but I wanted to bring it up here in case there's some underlying bug that should be examined first.Jefe 20:14, 25 Jan 2007 (CST)

It looks like that issue of F&SF got entered twice and the titles were then not merged. I went ahead with merging them though had to puzzle on why the for one of them was "2002-05-06" while the other was just "2002-00-00." The magazine is from Dec-2002 and so I went with 2002-00-00. You can now see the two publication records under the remaining title. As the magazines get linked from Magazine:Fantasy_&_Science_Fiction I took a look there and for Dec-2002 it's linking to FSFDEC2002. Hmm, remarkable, the tags are the same for both magazine publication records.
I needed to get the numeric publication IDs and so from Advanced Search did a pub-search for "Science Fiction, December 2002" which includes links for
Another way to do this is from the Under Hill title where you can do diff-publications. That also shows the record numbers and the publications are slightly different meaning I'm not sure which one is the best target for deleting and what the issue are with deleting publications that have the same tag. Marc Kupper 02:00, 26 Jan 2007 (CST)
The easiest thing to do here was delete one of the magazines (which I've done), rather than merging the titles. That will whack all the title/pub linkages. Alvonruff 06:36, 26 Jan 2007 (CST)

Chapbooks, take two

Starting a new section for length reasons and also to suggest a different way of looking at it.

After some thought, I feel the right way to go about this is to look at the requirements for the author biblio display. Al has documented the way it currently works at Requirements:Author Display. That shows that the following is the sequence of display of "long works":

  • Fiction Series
  • Novels
  • Collections
  • Omnibus
  • Editor
  • Anthology Series
  • Anthologies
  • Nonfiction Series
  • Nonfiction
  • Chapterbook
  • Nongenre

These displays work from title records, not pubs.

So what do our users expect from this section of the display? Users probably fall into the following main categories:

  • Casual browsers looking for something specific, or scanning a particular author's biblio
  • Collectors checking to see what they don't have by an author

There are other reasons to use the ISFDB, but for this page, I think these are the main two reasons. In both cases the biblio's function is to present "what the author has published" in a standard way. I think separating chapbooks is useful to both these groups of users.

A publication of a short story in chapbook form is different, to most collectors, from that same text published in a collection. The collector wants a copy of that chapbook. If the chapbook exists in hardback and softcover, most collectors will only want one or the other. (It doesn't matter if the contents are a short story, an essay, or poems, or miscellanea; the collector wants to see it listed and make their decision on whether to collect it.) So for a collector the chapbook should be listed at the top level of the biblio. This agrees with the way Al has it above.

So far so good, but the problem is that chapbook, unlike the other categories above, is a judgement one makes about a publication, not about a title. "Omnibus" appears to be an exception also, but in fact it isn't: you can look at the title contents and decide if something's an omnibus or not without ever holding the publication in your hands. This is not true for chapbooks, and I think this is what makes them confusing for the biblio display.

I suggested in the previous thread that a chapbook could be tagged by having a flag on the pub record, and taking it out of the pub_format and title_format completely. The problem with that approach is that the biblio display won't ever see that flag, because it looks at titles. In thinking about that I decided that the best resolution was to flag "chapbook" at the title level, as well as at the pub level.

Here's what I think would cover it.

  • Add a new field to title to mean "chapbook". Checking this box means "This title should be listed as an entry in the "Chapbook" section of a biblio.
  • Add a new field to pub to mean "chapbook". Checking this would mean "This publication is a chapbook".
  • The biblio display of the chapbook section should list all titles that are marked as chapbooks. Those titles should not list in other sections (except series); however, all chapbooks are listed here regardless of series.
  • Title displays should show "(chapbook)" next to each chapbook pub in the list. If a title marked chapbook has no chapbook pubs, then there's no problem; none display.
  • Title display for titles not marked as chapbooks may still show chapbook pubs, for two reasons. First, The Rule of Names appears in regular collections but also in the chapbook Dreams Must Explain Themselves. This is OK; if you're looking for this story, this is a legitimate list of pubs for that story. Second, if the problem is just that the title should have been marked as a chapbook, but hasn't been, this is a consistency problem but should not cause incorrect display. We can find and fix these with a script if they become a problem.

I don't think CHAPBOOK can be correctly supported while retaining it as part of the format fields; those are too heavily overloaded already.

Here are some of the code changes that would be needed to support this.

  • Editpub and newpub obviously would have to have a checkbox for chapbook.
  • add pub would default the checkbox to the state of the parent title
  • clone pub would duplicate the checkbox but allow modification
  • edit title would allow modification of the checkbox
  • title display and author display would require modification per the above description
  • pub display would need to show if a pub was a chapbook
  • CHAPTERBOOK would be removed from the various format pulldowns

I understand there are very few existing chapbook records - Advanced search finds only six titles, and no pubs -- so conversion could probably be done by hand.

-- Mike Christie (talk) 08:58, 26 Jan 2007 (CST)

A simpler method may be to remove chapbook support entirely but to add a an "Info" (or maybe "description" is a better word) field just below the title. Whatever you put in there would get formatted after the title as title (info). While there are just six chapbooks in ISFDB there are 14,993 titles at present that contain a "(" that presumably refer to informational content that's not part of the title itself. Or, we can agree that if you see "title (something)" that "something" is not part of the title. Marc Kupper 13:47, 26 Jan 2007 (CST)
A follow-up - I have been thinking about non-genre content and a thought that's similar to the chapbook idea Mike presents is to add a "Genre" string field that would get filled in for non-genre. It would act very much like the series field but would display at a layer above everything that currently exists. Mike already listed what's on the long works page above and so we'd have
  • blank genre (speculative fiction) - that list above
  • For each unique genre among the titles, sorted alphabetically - that list above
This would allow the non-genre content to get included but displayed separately, while still being organized, at the bottom of the page. I am trying to put together and organize Maxim Jakubowski's bibliography and am running into that he has written both sf and multiple non-genre across all the categories (novels, collections, anthologies, non-fiction, shortfiction). For organizational purposes I've been appending things like "(erotica genre anthology)" to the titles but it getting messy as he has about 100 books and countless short stories. Marc Kupper 14:02, 26 Jan 2007 (CST)

Novellas That Are Also Novels

In Lucius Shepard's bibliography, there is a great deal of inconsistency about whether works are novels or novellas. The problem stems from the fact that he's published a large number of novellas as stand-alone volumes. Many of these have entries in ISFDB that appear to have been seeded from award nominations, and so appear in the short fiction section. Some of them, however, also appear in the novels section because they are also books. "The Scalehunter's Beautiful Daughter" [4] [5] and "Father of Stones" both seem to be modelled in the way that makes most sense, i.e., they are listed in both places, and the "novel" publication is included as a pub under the story's title. "Kalimantan" and "The Last Time", however, while listed in both places, are not so linked. I'm not sure how to merge them in such a way to preserve both the novel and novella status, although there is apparently a way to do this. (The same will also have to be done with "Liar's House" once somebody enters the book version.)

The way "The Scalehunter's Beautiful Daughter" is modelled, incidentally, seems like the best way to handle chapbooks -- list the book publication separately, and include the book version as a pub of the short fiction title, without the displaying the chapbook's title as part of its contents. Jefe 18:42, 27 Jan 2007 (CST)

The Scalehunter's Beautiful Daughter seems close but not perfect. The problem is that the novel version of the title has one publication while the shortfiction version of the title includes magazine, anthology, collection appearances but also has a record for a novel. It might be two publication records for the same novel, it's the same publisher, hardcover, date, but the catalogue number for one of them is a bit of a hash.
One solution would be to make one of the titles a variant title of the other. Unfortunately, ISFDB has logic where it will not show a variant title in the main part of the bibliographic display. Instead VT titles are always shown as a secondary title nested under a parent title. Thus if you made one of the The Scalehunter's Beautiful Daughter a VT of the other it will disappear from either the list of novels or shortfition.
For now there seems to be no real solution other than two disconnected titles where one would be used for novels (or chapbooks) and the other for other types of publications. It means you could not click and see all of the publications for the story. In an earlier case where I ran across a situation this I merged the titles, keeping the type as Novel. It means the title is no longer in the shortfiction list. I was comfortable doing this in that case as the story was 120 pages long meaning the standalone publications were published as books rather than chapbooks.
Note – I wrote “Thus if you made one of the The Scalehunter's Beautiful Daughter a VT of the other it will disappear from either the list of novels or shortfition” and the reason I can say that is I tried to make the VT and remove it but have succeeded in having The Scalehunter's Beautiful Daughter disappear entirely from the bibliography. I have a couple of feature requests in the queue and once those are implemented the titles, which still exist, can have the VT relationships removed and they will reappear in the bibliography. Marc Kupper 22:54, 27 Jan 2007 (CST)
As far as the vt issue is concerned, a trick I learned from Ahasuerus is to create a bogus title record via a simple new pub; delete the pub, then merge the title with the title you want to remove a vt from. The merge screen will give you the choice of selecting a blank vt, which will have the effect of blanking the vt reference.
Also: Jefe, why don't you want to show the chapbook as a pub when you click on the shortfiction? Seems to me that it's a real pub of that title; why shouldn't it show up? I would also say there should only be one title record; both represent the same text, by the same author, under the same title string. This is partly the same conversation as above, under the chapbook discussion headings, of course. Mike Christie (talk) 07:47, 28 Jan 2007 (CST)
Thank you Mike - that merge trick worked - pokes self in the eye as I should have thought of that one.
Also, I believe Jefe does want the “chapbook” novel to show up when he clicks on the title reference but also “wants to merge the two titles while also preserving the novel and novella status.” That seems like a very reasonable request as titles have been used as both shortfiction and titles for publications meaning people will be looking for the title in both the upper section (novels, collections, anthologies, chapbooks, etc.) and in the shortfiction or essay section of a bibliography.
  • Merging titles does not help as then you will loose either the novel or shortfiction status and it also will not be shown in both parts of the bibliography.
  • Using variant titles does not help as then the variant title will not show up in either the upper (publication) or lower (shortfiction) part of the bibliography.
  • One thought is to create the publication/title records as usual for the standalone publications but also to include the shortfiction title in the contents. The resulting publication contents will look odd as the title will show up twice. I just tried this with The Scalehunter's Beautiful Daughter.
  • Another thought is to create a dummy title record for the novel/chapbook that has a note “This is a dummy/placeholder record for a title that is both a shortfiction work included in magazines, collections, and/or anthologies and has also been published as a standalone work. Use the <a href="url of the shortfiction title">shortfiction</a> record to both add new publications and to view where this title has been published.” --Marc Kupper 12:51, 29 Jan 2007 (CST)
Perhaps I was unclear. I was quite pleased with the display of "The Scalehunter's Beautiful Daughter" as it was (two titles, with the novel appearing as a pub of the shortfiction). What I would like to do is link the other works that are listed separately as both novels and shortfiction, but where the novel version does not appear as a pub of the shortfiction, namely "Kalimantan" [6] [7] and "The Last Time" [8] [9]. I'd also like to know HOW "The Scalehunter's Beautiful Daughter" came to be listed this way so it can be replicated. As I said, I think the way it was originally displayed would work perfectly for the chapbooks.Jefe 13:22, 29 Jan 2007 (CST)
Well, it ought to be possible to figure out how to achieve the effect you saw. Presumably two titles, one SHORTFICTION, one NOVEL. Suppose you took the novel pub, and edited it to add the SHORTFICTION title to it; would that work? In some displays, the NOVEL title would not appear, because it's identifiably the "primary" title for the pub. The SHORTFICTION title would always appear. Each title would show that pub. Is that the effect you're looking for? One side effect is that some displays (such as Edit Pub) would show both titles together.
I don't think that would do it -- I'm pretty sure that's how I ended up with this pub.
Having suggested it, I don't particularly like it -- I think it would be better to have the pub say "CHAPBOOK" and the contained title say SHORTFICTION (if that's what it is). Then there's a single title, and every pub of it would be listed under that title. If it's long enough to be a NOVEL, it's not a CHAPBOOK and would be listed in the novels, and would be included in anthologies and collections as a novel (which works fine). Mike Christie (talk) 21:48, 29 Jan 2007 (CST)
This sounds more like what I was hoping to find.Jefe 12:05, 31 Jan 2007 (CST)

Editing James Blish Star Trek

Some things I learned today: 1) Almost every Episode that got converted into a Short Story was classified as "Essay". Is this a common problem with "novelisations" of TV shows that only end up as a short story? 2) Most of the stories I merged, but where I could only find one title I added notes as to the author of the original episode. When the Merge option was available, I left those notes out as they couldn't be added at the time. But they might be useful - some famous authors aren't getting credited. But my notes won't do that - any suggestions? 3) There's a major overlap in series. Titan Books have their own numbering: I suspect the best is to add the Titan numbering to the titles of the series currently up, but is there any chance of books belonging to multiple series under the same overall category? BLongley 19:28, 27 Jan 2007 (CST)

I saw your title updates today and wondered why the stories had been in ISFDB as essays. I believe whoever added the records thought the stories were transcripts of the television episodes and did not realize they were “adaptations” of the episodes. I agreed with you changing these from ESSAY to SHORTFICTION.
Unfortunately, there is no “correct” way to give the episode writers direct credit. In theory we could just add title records such as The City on the Edge of Forever (Star Trek television episode) by Harlan Ellison (or who?). The problem is the ISFDB:Policy#Rules_of_Acquisition allows for “published” but not “produced” works meaning television episodes don’t qualify for ISFDB. For now, adding notes exactly like what you have already done should do. You may want to link from each Blish Star Trek adaptation to the wikipedia article on that episode.
At the moment a story can’t be a member of more than one series. There is a feature request in the queue to add support for this as it’s definitely needed to deal with something like Titan books “Star Trek Adventures” series which are reprints of Pocket Books TOS titles but with their own numbering. I can’t think of an easy short term solution that does not involve shooting yourself in the foot. Marc Kupper 23:31, 27 Jan 2007 (CST)
Blish's bibliography was a complete mess when I did a pass a couple of months ago. Star Trek titles were in a particularly bad shape and I made a note to get back to them at a later point. I am not sure what had created this fiasco, but the important thing is not to take anything that you find in his biblio for granted and double check everything. Ahasuerus 23:42, 27 Jan 2007 (CST)
Hopefully it's a bit better now. I'm resisting a lot of "intuitive" edits/merges and sticking to what I have as primary sources. Which is why "Star Trek", "Star Trek 10" and "Star Trek 12", for instance, may still look messy. BLongley 17:21, 29 Jan 2007 (CST)

"Unsigned" template

I just created a template that I have found occasionally useful on WP: {{unsigned|foo}} will create a small-font note saying that the preceding unsigned text was added by user foo, with a link to foo's home page, talk page, and contribs. It's at Template:Unsigned. Mike Christie (talk) 08:05, 28 Jan 2007 (CST)

A subtitle question

A different kind of subtitle question: in verifying Le Guin's "The Birthday of the World and Other Stories I found Locus1 gives the title of the first story as "Coming of Age in Karhide by Sov Thade Tage em Ereb, of Rer, in Karhide, on Gethen". The title on the story is "Coming of Age in Karhide"; below that, between one-point lines, is "By Sov Thade Tage em Ereb, of Rer, in Karhide, on Gethen." The look is of an attribution; i.e., fictionally, Sov Thade Tage em Ereb wrote this. The question is whether this is a subtitle or not. I had entered the story title without this text, but I'm coming round to the Locus1 point of view -- whatever Le Guin puts up above the text is the title, except for her name. Sov Thade (etc.) is not a pseudonym because the book makes it clear the story is by her. If she'd put this string of text in the body of the story, it would not be a subtitle, but I think here it is. I'm going to make the change to the pub; I'd like to know if anyone thinks it should stay as the shorter title. Mike Christie (talk) 08:25, 28 Jan 2007 (CST)

Moorcock’s Corum and dealing with ISFDB title records

(this thread started on User_talk:Unapersson’s page but seems to be an issue that should have community input and so I’ve moved it here.) --Marc Kupper 13:52, 29 Jan 2007 (CST)

re: the omnibus you added - [Corum. I noticed there is already a title record for Corum: The Coming of Chaos and suspect the new title record you created can be merged with this one. Normally you would first look up the title the publication belongs under and use "add new publication for this title" which saves you from needing to do the extra merge.

The "The Coming of Chaos" part in the existing title record seems like a sub-title that can be removed as we generally do not use sub-titles in title records. We do include sub-titles in publication records though.

I'm not familiar with the "Corum" works but in looking at Moorcock's page it looks like "Corum" is

  1. a series called Corum that contains two Corum related sub-series
  2. a series called "The Tale of the Eternal Champion (White Wolf)" that contains the omnibus "Corum." This does not seem right and I believe this omnibus should be in the “The Swords of Corum” series.
  3. down in Collections is The Chronicles of Corum which probabably belongs in one of the series.
  4. down in Omnibus is Corum whish is the one you added.

Are you working on Moorcock by chance? I just did some obvious title merges and moved a couple of things into the right series but imagine the bibliography needs a bunch of TLC. Marc Kupper 17:27, 27 Jan 2007 (CST)

Yes, I have been working on the Moorcock page. I've been working on Moorcock bibliographical stuff for many years and it is a big minefield. Filled with multiple revisions, retitling, titles reused.
Yeesh! And I though my attempt on James Blish was hard work... Good luck! I'll add my Moorcock and hope it helps... BLongley 16:30, 29 Jan 2007 (CST)
OK, added mine. With some comments, maybe not enough. Corum could be simplified, Cornelius maybe not. And the Eternal Champion stuff is getting well wierd... Feel free to ask for extra info, I seem to have added a lot of new publications, many with Bibliographic comments that I didn't add but will if you like. BLongley 19:18, 30 Jan 2007 (CST)
The problem I'm trying to get around is that the Corum omnibus contain books in the corum series, but they're also part of the two definitive Tale of the Eternal Champion series. The US one (White Wolf) and the UK one (Millenium), both featured different omnibus in different orders. Both with different sets of revised texts. The Millennium ones were revised from the originals, and some of the White Wolf ones were further revisions.
So I've tried to approach it like this:
  1. the original omnibus editions which basically represented one or two trilogies go as part of the corum series with their child titles, same with hawkmoon and history of the runestaff etc.
  2. the millennium/orion editions get collected under the tales of... series with the UK ordering
  3. the white wolf editions get collected under the tales of... series with the US order
I can't see a more logical way to represent the two as titles cannot be in more than one series.
--Unapersson 12:35, 29 Jan 2007 (CST)
If a story got revised for the USA or UK that would seem to be a good case for two separate title records. You could have Story and Story (revised for UK publication) and ideally in the notes you would explain the nature of the revision.
If the story was not revised and yet is part of two definitive series then it’s a little problematic. The basic problem is that at present ISFDB does not have a way to display a title in two different places at once. Take a look at ISFDB:Community_Portal#Novellas_That_Are_Also_Novels where I just proposed two solutions to having a title appear as both a novel and shortfiction work that could also be used to allow you to have a title in two different series. A key is to leave behind plenty of notes/comments explaining any non-standard usage of title/publication records.
It seems a long term fix would be placeholder titles that would operate much like the existing variant titles but would be used to position a title in an author’s bibliography. When you view or link to the title you would see the root title’s publication list (and it could perhaps even just redirect directly to the root title though that means editing the placeholder record’s metadata would be tricky). --Marc Kupper 13:52, 29 Jan 2007 (CST)
If we put in a separate title for every revised version of one of his novels we'd wind up with a lot of title records. For example, The Ice Schooner was serialised, then published as one novel. This was then expanded in the eighties, the revision making it quite a bit longer. Then revised again for the Millennium edition. And yet again for the White Wolf edition. So you'd end up with four different records for the same work and the titling problems that entails. You also can't be sure which revision a particular minor publication is.
It's tied up with the question of how you cope with other "meta" series. What is the current intention to cope with things like Orion's Fantasy Masterworks and Science Fiction Masterworks series, the Dennis Wheatley Library of the Occult, the Corgi science fiction series etc. Series which span multiple books and multiple authors but only single editions. They should be represented somehow, but are tied to specific publications rather than titles.
It might work if there was a series for a publication as well as for a title. So the title as a whole could be in one series, but the publication in a different one. As a single publication should only appear in a single series. --Unapersson 15:43, 29 Jan 2007 (CST)
I agree that we should not try to track revisions of the text in the title; it's not possible to be sure which version one is holding in one's hands. I do think it's worth putting notes in the pub records where there are significant differences (e.g. "Moorcock regards this as a corrupt text, per later Mayflower edition, which is heavily revised." or whatever. Even if we could, there are some very fine gradations we would have to deal with -- we obviously don't want "Dune (typos corrected)" or "Dune (British spellings)", but how do we know when to stop marking distinctions? And how do people doing data entry know?
A while back, Al drew a distinction between recording publication information, which is a data entry task, and creating a bibliography for an author, which is a work of human intellect. I think the bibliographies are our goal, which is why we have supporting fields such as notes and the author biblio pages, but I think the data entry itself should not be a work of intellect. It should be where we capture information for the bibliographers to work with. Hence I think it's best not to add parenthetical comments to fields such as the title, if we can avoid it.
With regard to publication series, there is a feature request here that began as a request for an edition field, but morphed into a discussion of ways to use a publication series field for display purposes. Perhaps that feature could account for the publisher's series too? Though of course there could be multiple printings of each book in those series. Mike Christie (talk) 21:23, 29 Jan 2007 (CST)
As far as the issue of "title series" vs. "publication/publisher series" goes, it's somewhat messy. We have a number of authors whose biblio is structured around "publication series", notably Jules_Verne, and I don't think it works very well. I would venture to say that ISFDB users will typically be more interested in the fact that The Mysterious Island is a sequel to 20,000 Leagues Under the Sea as opposed to the more academic fact that these titles were books 12 and 6 in "Voyages Extraordinaires" respectively. In other cases, the same book can be a part of a bona fide "title series" as well as a part of one (or more) later reprint "Masterpieces of XYZ" series, in which case I would argue that the textual "title series" connection is the important one. Besides, it makes it easier to add other titles that are published elsewhere to the series.
On the other hand, a retrospective "Complete Stories" series that seemingly everybody in the field has had done by now (Dick, Sheckley, Sturgeon, Simak, etc) is also essentially a "publication series" since few, if any, of the stories included are linked. Still, it seems like the logical thing to do is to display these titles together in a series.
Finally, we have writers like Moorcock, who arrange and rearrange the stories in their multiverses almost at will. Eventually the relationships between the titles -- "revised", "abridged", "restored", "merged with another title", etc -- get so convoluted that a bibliographer is tempted to give up completely and do what Unapersson (and myself in my master catalog) have done, i.e. list each "publication series" as a real series instead of trying to unravel the impossible maze of Variant Titles.
So what's the bottom line? I think the primary criterion that we want to use at the policy level is "value to our users" and the principle of least surprise. What does your typical ISFDB user want to know about the books and stories that we list? I think that most users are primarily interested in the texts and only secondarily in publicatiion information. Thus "Book A is a sequel to Book B" is of primary interest while "Books A and B were printed as part of the second Ace Special series" is of secondary interest. On the other hand, when it comes to multi-volume short story reprints, the principle of least surprise suggests that "publication series" are expected by our users and should be supported at the title level. Not perfect, but seems like a reasonable compromise, at least from here.
Similarly, I think there is a great deal of value to displaying (at least major) textual variations between versions on the Summary page. It's important for casual browsers to be able to immediately discover that there have been 3 revisions of the book/story that they have either read or are planning to read instead of having to look it up in the Notes field. This can be doubly important when the author had drastically changed the text, sometimes completely altering it, e.g. Keith Laumer's notorious practice post-stroke.
Granted, it may not be easy to identify these textual variations. It's a much more time consuming process than plain data entry, as Mike points out above, but it also adds a significant amount of value to the database. As to the question of whether we want "to add parenthetical comments to fields such as the title", well, there are two title fields here, one is Title_title and the other is Pub_title. The latter is "objective" information and should be captured "as is", but the former is "subjective/derived" information and we have had numerous discussions about it in the past. Do we want to have a separate "relationship" field where we put comments like "abridged", "revised", etc, or do we use the current practice of adding this information in parentheses?
Sigh... Have to run again, but at least it's food for thought. Ahasuerus 22:41, 29 Jan 2007 (CST)
If Al implements publication series, I see no reason why they should conflict with title series. I agree that readers care more about texts, but we don't need to be restrictive in the structures we support. With regard to Moorcock specifically, yes, most people give up on trying to unravel the tangle, but if someone does unravel it, we should be able to support it. I think that's a separate issue, though; the question there is whether the series structure is robust enough to fully support Moorcock's biblio, and I suspect it's not, at least not yet.
I certainly agree that identifying textual variations adds a lot of value, and should be done. I think an additional principle to cite, though, should be that "data entry from a publication should be possible with no other data sources to hand". That would imply that information about textual revisions should go in the notes, which would be my preference; the notes would be updated by a bibliographer researching the text versions. I agree that a relationship field might be really helpful here too; perhaps that's actually a relationship between "editions", if that concept gets implemented. But it seems to me parentheses are the worst option for data entry: if we have a parenthetical addition to titles, how is a reader to know which version they have? Mike Christie (talk) 07:14, 30 Jan 2007 (CST)
I see that Al seems to be adding support for the wiki based notes where the ISFDB links now show in red if the corresponding wiki page is not present. It's working for author biographies and bibliographic notes but not yet for publication notes though I expect a "real soon now" for that one. This means that we can start documenting the more complex relationships in the wiki rather than relying on ISFDB's notes and related structures. I had been avoiding the wiki notes as it was not apparent from the link in ISFDB that the note existed meaning future editors would overlook the note but now that not the case. Marc Kupper 19:33, 30 Jan 2007 (CST)

Special Projects

OK, I've done a few that looked fast - Robert Rankin, James Blish, Terry Pratchett etc - but it seems people are doing others? E.g. Michael Moorcock, Ursula K. Le Guin. Can we declare them a bit better? I've added a few Moorcocks tonight, more tomorrow if they're useful. Or ask me for help on some subjects - I'm just picking a shelf-full of books at random sometimes, if you need a concentrated effort just say so. The worst I can say is "Who?". ;-) BLongley 17:30, 29 Jan 2007 (CST)

Maybe we should update Bibliographic Projects in Progress to take into account the new activity. Yes, I'm working on Le Guin; and it looks like Unapersson is doing Moorcock. I'll go ahead and enter all my Moorcock next, just to give Unapersson more to work with, and of course I'd be glad if folks would enter/verify any Le Guin they have. Anyone else got anything specific going on? Rudam appears to be working on Tiptree, at least some of the time. Mike Christie (talk) 21:27, 29 Jan 2007 (CST)
Adding a section to Bibliographic Projects in Progress about authors that are getting cleaned up and who is doing it sounds like a good idea. Marc Kupper 00:45, 30 Jan 2007 (CST)
Absolutely! When I created that page some time in mid-2006, we had no way of telling who may end up doing what, so the current layout reflects our tentative priorities at the time. As areas of interest get fleshed out, the Wiki Project structure can be adapted to reflect whatever we are actively working on at the moment. Ahasuerus 01:54, 30 Jan 2007 (CST)
I went ahead and added a section to track active authors; I put Le Guin, Moorcock and Pratchett in. I didn't add any content to the latter two; if the format I'm using on the Le Guin page is useful, please steal it. Mike Christie (talk) 07:35, 30 Jan 2007 (CST)
Looking at what constitutes a "Project" made me think again - I need to clarify where I'm going to sort out what I can in the next few days, as opposed to a fanatical devotion to get one topic perfect. For instance, I'm always interested in the Terry Pratchett bibliography but for the Tom Holt that's just something I can do a lot with in a short time and leave in a better state than it was before. (Similar with the James Blish Star Trek - quick fix but someone else's problem now). BLongley 14:39, 2 Feb 2007 (CST)

One publication record per printing?

When replying to an editor about something Mike had made a comment about having one publication record per printing. I personally believe it’s a good idea though it could result in many records for some publications as, for example, I recently added a 48th printing of something. The “good” part of the idea is that as publications are verified it’ll be easier to tell which printings have been verified. The downside is that ISFDB does not have fields for the First Printing Date nor Printing # meaning this data is maintained in the notes and when looking at a title’s publication list it won’t be apparent which records are for which printing unless a convention like “Avon (48th printing)” for the publisher name field is adopted.

I also believe that one record per printing will lower the chance of confusion and/or data error. For example, in the DAW list I'm looking at Strange Doings by R. A. Lafferty where it says

First Printing: Apr-1973
Published: Apr-1973
Printing: 2nd

You might think the 2nd printing was in Apr-1973 but what's happening there is that all of the printings that share order #UQ1050, price $0.95 occupy one record and we have not verified a 1st printing yet. Once we get a 1st printing the printing field will have

Printing: 1st, 2nd

If the list used one record per printing though we'd have

First Printing Printing Published Verified
Apr-1973 1st Apr-1973
Apr-1973 2nd n.d. Primary

Should we adopt one publication per printing as ISFDB “policy?”

I think the help text already states that we should have a separate pub record for each printing. Help:Getting Started isn't absolutely explicit, but is clear enough, I think. Template:PublicationFields:Year says "Note that we are interested in recording each different reprint of a publication, since there can be some significant differences between them, such as cover art, or price." Are you thinking it should be in a policy page too?
I saw your note recently that you hadn't realized this was the intention -- I wrote that help text so I tried to recall where I got the idea from, and I'm pretty sure it came up in an old Community Portal discussion and Ahasuerus or Al commented that we did track all printings separately. Ahasuerus, can you confirm this is a long-standing policy, or did I imagine it? Regardless, I agree it is a good idea, so I think we're in sync.
That's right, we have always used "one printing = one Publication record". I dont think we really have much choice in the matter since the current database scheme only has two types of records, "Titles" and "Publications", so the only alternative would be to stuff all "printing" data into the Publication Note field. There has been some talk about adding a third tier to the database scheme so that it would be "Title -> Edition -> Printing", but it wasn't clear that we could draw the lines between the tiers clearly or that it wouldn't confuse the users, so nothing has come out of it so far. Ahasuerus 10:46, 30 Jan 2007 (CST)
I had seen both of those ISFDB items Mike referenced but had interpreted it as a separate ISFDB publication record for each “different” reprint/printing. As ISFDB has fields for title, author, “year”, ISBN, publisher name, price, and cover artist the implication was that if any of those changed we would have a new ISFDB record.
Something that was (and maybe is) not clear to me, and apparently others, is what to do if some other detail of a publication changes where ISFDB is not formally collecting data. I had assumed, for example, that if the printing # changed that you would still file this data in the same record where the formal ISFDB fields match your publication. Particularly with the printing number it did not make me comfortable. I’ve been noting printing numbers in the comments though not with a consistent method and also was concerned about how other editors would interpret my notes. For example, someone removed my note about a 2nd printing and replaced it with their 1st printing. I can understand their view as they had a book in hand where the ISFDB fields matched their publication with the exception of my note which to them was “wrong” as it did not match their publication.
From what both Mike and Ahasuerus wrote I’ll change my practice of overloading publication records and instead will have record per printing. I suspect I should have a boilerplate printing note that says “This ISFDB publication record is for the #th printing. If your publication matches this record exactly except for the printing number then clone this publication record and update the note to reflect your printing’s details.” --Marc Kupper 18:34, 30 Jan 2007 (CST)

Add “First Printing” and “Printing #” fields?

(Separated from the "One publication record per printing?" thread) --Marc Kupper 18:34, 30 Jan 2007 (CST)

Should a feature request for new publication fields, “First Printing” and “Printing #” and also to rename the publication “Year” field to “Printing Date” be added? “First Printing” would be to help sort out those publishers that don’t state a date for each printing. For example.

First Printing: 1973-09-00
Printing: 9
Published: n.d. (currently called the “year” field)

--Marc Kupper 00:37, 30 Jan 2007 (CST)

Re the feature request, I think the edition feature request (which morphed into the pub series feature request) addresses part of what you're asking for. This might be an area where we should not make incremental changes via feature requests, but work with Al to identify a set of features that could be usefully added in one go with a structural change. One advantage of the series approach is that it provides an ordering without requiring an interpretation of the data -- so for example if you had a reprint and couldn't tell which printing it was, you could still place it in order in a pub series, but couldn't enter a printing number and use that for ordering. For "first printing", I am tempted to agree that this could go on a pub record, but it's not always evident and when it is it's sometimes wrong -- see Publication:ANCL00154, which is a second printing of "The Wind's Twelve Quarters"; it says "FIRST EDITION" on it, but it isn't. That's why I always say "Stated third printing" (or whatever) in the notes when I record a printing. Mike Christie (talk) 07:29, 30 Jan 2007 (CST)
I agree that implementing the edition feature request will help a lot.
Maybe we should think of it this way – should ISFDB have a record for every single printing of a book or is it enough to know that a publisher first published a book on a particular date? I believe we agreed that we want drill a little deeper and to at least know significant milestones in that publisher’s printing history for the title. ISFDB’s publication records currently have fields for various bits of data people agree are “significant” about a publication.
Is what we have satisfactory or should we drill even deeper and attempt to capture information about every known printing and/or other data? Part of the problem is that we are also simultaneously learning what may be “significant” and how to go about distinguishing one printing from the next. I was asking for “First Printing” and “Printing #” fields plus renaming the “year” field to “Printing Date” as that seems to be a set of fields that will work with how today’s larger publisher’s distinguish one printing from the next.
The other issue is if we do ask for something like “Printing #” is if editors should do translation. For example a publication may have
  • Spelled out “Third Printing”
  • Spelled out “Third paperback printing”
  • A row of dates “1981, 1982, 1984” where you count them to determine the printing number.
  • The number line “3 4 5 6 7 8 9” or “9 8 7 6 5 4 3.”
With all of these should the editor enter “3” or literally what’s in the publication?
The same is true of the First Printing Date. Many publications list multiple “first printing” dates such as for paperback vs. hardcover vs. “mass market”, in Canada vs. USA, other publishers and either the parent house or other imprints. I usually look and figure out which is the most recent date.
The “Printing Date” is sometimes stated as a code next to the number line. Should these get documented or is it sufficient to use “n.d.” or 0000-00-00 if the date is not readily identified as such in the publication?
Obviously this can all be put in the notes field but that means the format will be inconsistent and it will not display in the publication list for a title. --Marc Kupper 19:11, 30 Jan 2007 (CST)
My take: I generally add a note such as "Stated third printing" or more often "Stated third printing of the 1972 Puffin edition". I think the four forms you spell out above would all be interpreted by a reader as third printings, so I tend not to mention them; I do sometimes add "per number line" when I want to be specific. I think the key is that the information should be derivable from the publication; in this case some standard interpretation (to be described in the appropriate help files) can help.
For first printings I am more cautious for the reasons you give. I typically put "Appears to be first printing" since that is a true statement, but accounts for the not infrequent situation where there is no obvious indication on the copyright page for a reprint.
For printing date I think the same interpretation will work; there is often a date row: "78 79 80 81", which would indicate a 1978 date; again one could add "per date row/line" or simply put "1978". The verification records will show what has been looked at, which I think is enough. A subsequent bibliographic effort to organize the pubs into a coherent structure would be aware that this data isn't always reliable, but that it does represent what's apparent from the pubs. I think that's the right approach. Mike Christie (talk) 21:39, 30 Jan 2007 (CST)

Reference Works in Series

Many series eventually have reference works of various sorts written about them. Discworld has had many written and has 2 series in the ISFDB, Discworld Maps And Science of Discworld. Maps is linked to the main Discworld series, Science isn't. I'd think that they both should be since if I was using ISFDB to find what books I need to read and collect that's where I'd look. If that seems reasonable I'll start linking the reference books I have on various series to them. Dana Carson 01:48, 30 Jan 2007 (CST)

That's fine. Sometimes the reference works are classified as "nonfiction" even though they may be entirely about a fictional universe. Marc Kupper 02:41, 30 Jan 2007 (CST)
FYI, BLongley was working on the Discworld series -- you might leave a note on his talk page, if he doesn't spot your comment here, if you want to team up with him on improving the Pratchett biblio. Mike Christie (talk) 05:48, 30 Jan 2007 (CST)
Yes, feel free to talk - I only took on Pratchett as I seemed to know more than previous contributors, I won't pretend I'm an expert. The Plays could do with dividing up I'm sure, and I have (for instance) "Terry Pratchett - Guilty of Literature" to add if needed. There's a whole potential series of Discworld Diaries for instance, and the quiz books... BLongley 19:06, 30 Jan 2007 (CST)
OK, will do. Not a Pratchett expert, just have a weakness for books about books so I have reference books on quite a few series. Will add what I have for Pratchett. Dana Carson 19:42, 30 Jan 2007 (CST)

Series and variant titles

I was verifing Brain Twister http://www.isfdb.org/cgi-bin/title.cgi?23317 , which is a variant title for That Sweet Little Old Lady. It wasn't listed as being in a series so I created one. When I looked for the rest of the books in the series there is a series. But you don't see that on the Brain Twister entry apparently since its a variant. So please delete the new series and can that be added as a display feature request. Dana Carson 02:17, 30 Jan 2007 (CST)

I'm confused by "When I looked for the rest of the books in the series there is a series." and what you are asking for in terms of a feature request.
If you want to delete the new series all you need to do is to delete the series name from the Brain Twister title record. I suspect you could add Brain Twister to the "Psi-Power" series though that would mean the same story would be in the series twice.
I suspect there should be a feature request to show variant titles for series entries exactly as we do for bibliographies. Marc Kupper 02:55, 30 Jan 2007 (CST)
What I meant for feature request is that you don't see that Brain Twister is part of a series when you are looking at its title entry. —The preceding unsigned comment was added by Dcarson (talk • contribs)}.
There is a related feature request, 90007, in the Open Features section. I agree this would be useful and it doesn't seem to be handled right now. I've created a feature request for it: 90119. Mike Christie (talk) 07:49, 30 Jan 2007 (CST)
Dcarson wrote
> What I meant for feature request is that you don't see that Brain Twister is part of a series when you are looking at its title entry.
I'm confused. I went to the Brain Twister title entry, changed it's series field to "Test Series" and now see
  • Variant Title of: That Sweet Little Old Lady (by Randall Garrett and Laurence M. Janifer )
  • Type: NOVEL
  • Series: Test series
I've reverted that edit but when Brain Twister was in the Test series it was visible both in the title record and in the series list. As Mike pointed out, there is a feature request to deal with that there is no special logic at present to deal with variant titles in a series. Presumably, if both the parent and child title are in the same series then you should display them using the using "aka" and/or "[as by]" logic used for bibliographies or publication content lists. Other than that - I'm not sure what is "broken" about the current ISFDB behaviour. Marc Kupper 19:24, 30 Jan 2007 (CST)
The title it is a variant of was already in a series, I just thought that it would be also and since it wasn't listed I started one. No real problem, just seemed that if a title is in a series the variants should be also. Dana Carson 21:23, 30 Jan 2007 (CST)
Marc, I think the current situation is that both parent titles and variant titles display correctly if they are in a series, but if a parent is in a series but the vt title is not explicitly placed in a series, then the vt shows no indication that there is a series relevant to that title. If one thinks of the series as an attribute of the text, not the title, then one might expect to see a parent titles' series information inherited automatically by the vt children. The ISFDB does not do this, and I think that that's what Dana was commenting on. Mike Christie (talk) 21:30, 30 Jan 2007 (CST)
As alluded to above, currently there is a problem (Open Series Bug 30013 with the way the ISFDB code displays Series lists whenever both the parent title and the variant title are explicitly marked as belonging to the same series. For example, John_G._Hemry is currently writing The Lost Fleet series as "Jack Campbell". This results in two Title records per novel, a parent one for "John G. Henry" and a Variant Title one for "Jack Campbell". All "Henry" and "Campbell" Titles have the Series name and the Series number fields populated, which makes them appear twice in the Series list. The underlying problem is that the Series display logic is old and doesn't understand Variant Titles. Al is working on it as we type, so hopefully the problem will be addressed soon.
Also, please note that when Marc wrote "If you want to delete the new series all you need to do is to delete the series name from the Brain Twister title record", he was right as far as the apparent application behavior is concerned. However, there is a catch. When the last Title record in a Series is deleted from the database, the Series record is not deleted, it's just never (at least in theory) displayed by the software. The reason is that there are Series that contain no Title records, but do contain other, nested Series. The Series removal code was not sophisticated enough to handle all of these permutations and was eventually deactivated. Unfortunately, this causes numerous display issues with nested Series -- see Open Series Bug 30014 and related bugs. Ahasuerus 16:52, 2 Feb 2007 (CST)

GURPS

Dcarson submitted some GURPS books; I've held them for now. I was under the impression that they were outside our rules of acquisition, but that doesn't mention them one way or another. These are GURPS for specific SF worlds. Ahasuerus, as I recall you were looking at this area -- are these in or out? And should we update the rules of acquisition? Mike Christie (talk) 04:51, 31 Jan 2007 (CST)

I had approved the ones I saw as they were all part of well known series and I was also seeing title updates to add the new titles to the series. I agree they are out of scope but recalled that if a work was part of a "significant" in-scope series that we'd include it. We include the non-fiction "About this-or-that series" for much the same reason. I guess it's yet another call for genre support so that we can tag these as "Role playing game" rather than adding (parenthetical) comments to the titles. Marc Kupper 16:47, 31 Jan 2007 (CST)
This is a thorny issue and I am not quite sure where we want to draw the line. On the one hand, if there is a well established specfiction series, then I agree with Marc's approach and see no problem with adding related "association" non-fiction titles of various kinds to the database as long as they are clearly marked as non-fiction.
On the other hand, I did a rather comprehensive cleanup of the Warhammer (+40K) series during the summer and a similar problem came up repeatedly. There were tons of manuals, "adventure modules" and other types of non-fiction or "semi-fiction" set in or about the Warhammer world, but there were also dozens and dozens of bona fide novels, plus a number of thinly fictionalized books about the Warhammer world. In the end I excluded all clearly non-fiction RPG titles and included all fiction and borderline fiction titles.
My reasoning -- to use a fancy word for a lot of head-scratching -- was that any non-fiction derived from (or about) a primarily fiction world is "associational" and should be included, while for a primarily non-fiction world only the associated fiction (and semi-fiction) needs to be included. The problem with this approach is that there is a growing number of worlds (Star Wars ets) where written fiction, media items and game items are so thoroughly intertwined that it's awfully hard to tell whether the world is "primarily fiction" or not :( Ahasuerus 01:44, 1 Feb 2007 (CST)
The Gurps books are good reference books of the timeline, maps, major characters form along with the game mechanics information. Dana Carson 02:30, 1 Feb 2007 (CST)
OK -- I approved them all. Thanks. Mike Christie (talk) 21:50, 1 Feb 2007 (CST)

David R. George III

I had rejected a merge of

  • David R. George III (delete)
  • David R. George, III (keep)

with the following note:

In looking at the cover images for both author names it looks like he uses "David R. George III" and not "David R. George, III." Note - ISFDB goes by what’s on the title page and so if they are using “David R. George, III” there then yes, “David R. George, III” would be correct but I would than add a publication note to each publication that’s inconsistent noting, for example, that “The front cover states ‘David R. George III’ while the title page has ‘David R. George, III.’” Also note, if there is a publication with a slight variation in the author’s name, with the comma for example, we would set this up as a pseudonym of the canonical author name which probably is in this case “David R. George III.” –Marc Kupper

The editor replied with the following which makes sense to me but I wanted some feedback on this as it seems that we would end up with David R. George, III as the canonical name and every single one of his works would need a variant title from David R. George III to David R. George, III?

Mark, I saw that you rejected my merge of David R. George III and David R. George, III. The Help:Screen:EditTitle help page says this about suffixes:
Suffixes such as "Jr" should follow a comma and space, and be followed by a period if they are abbreviations. This should be regularized if they are not presented this way in the publication. E.g. "Sam Merwin Jr" should be entered as "Sam Merwin, Jr." Other prefixes and suffixes should follow analogous rules.
So I think the two should be merged, rather than made into variants of each other.Jefe 17:20, 31 Jan 2007 (CST)

—The preceding unsigned comment was added by Marc Kupper (talk • contribs) .

So should the same be done with pseudonyms of Arlan_Andrews differing only by a period or comma and a period, and other similar cases? I'd be in favour; it is mostly just a matter of each publisher's house style and I don't see any use in recording which one had it and which didn't, even if they could be directly verified. --JVjr 05:36, 1 Feb 2007 (CST)
I wrote the help text Jefe refers to, but it was inherited from text Al wrote for the author field. See Template:AuthorFields:CanonicalName, for example, which is more explicit and is mostly Al's original text. I agree that this should be a merge. I'll update the pub.author and title.author help text to give more examples; an advanced search for authors including a comma shows that the rule is fairly well adhered to.
Marc, I'm not clear why we would end up with variants as you describe -- the goal of regularizing punctuation in a case like this is that it permits us to avoid variants, surely? All instances of "David George III" would be transformed to "David George, III". It's the same rule that prevents "Born With the Dead" and "Born with the Dead" being variants, regardless of how each publication capitalizes the title. Mike Christie (talk) 07:32, 1 Feb 2007 (CST)
So if I understand correctly, your take is "roman numerals preceded by comma but NOT followed by period" (I think I'll go and spell it out even more foolproof, "if they are abbreviations" earlier might not be enough :-) Even ordinary search shows that this is the prevailing way, but there's much merging to do (including Jack_C._Haldeman_II canonized without comma). --JVjr 08:14, 1 Feb 2007 (CST)

(unindent as I probably should have spelled out in more detail why I had rejected the initial author mege)

Thank you Mike - Template:AuthorFields:CanonicalName seems to be pretty easy to follow and it looks like the canonical name should be “David R. George, III” and that would be “the name under which a particular author's bibliography is organized.”

For reference the help pages are:

I believe the source of my confusion is there seems to be a disconnect between the author level definition which uses canonical names and the title/publication definitions which follow what’s stated in the publications. What’s not defined is the “standard method” for connecting stated or “common” names to the canonical names for the purposes of creating an author bibliography.

How should we deal with the fact that every single publication to date for this author states "David R. George III" on the cover?

  1. The title records would have “David R. George, III” to get the bibliography filed under the canonical name while publication records would use “David R. George III” to match what’s stated in the books. It means title searches for the author’s common name, “David R. George III” would fail and that editors adding new publications would add the publication and after approval would edit the title to change the author’s name from the common name to the canonical name.
  2. The title records would have the "David R. George III" common name to match the publication and you add variant titles for all of his works so that the author’s “canonical bibliography” gets populated. Editor’s adding new publications would add the publication and after approval would make the title record a pseudonym/variant of the author’s canonical name. A benefit of this method is that title searches for the common name will work.

Maybe we can kill two birds with a stone as we have the same disconnect, and resulting confusion on my part, for titles in that the title help fields mention “canonical title” but never explains the standard method by which we get from publication titles to the canonical titles seen in author bibliographies.

--Marc Kupper 12:54, 1 Feb 2007 (CST)

OK, I now understand. I didn't manage to get the help files to say what I wanted them to say, so let me rephrase that now (whether it's right or wrong) so it's clear what I meant to put in there. Then we can try to reach consensus on what the best answer is.
What I meant was that for certain kinds of regularization, the change would be made everywhere. So if one book appears as by "John W. Campbell Jr.", another as by "John W. Campbell, Jr", and a third as by "John W. Campbell, Jr.", all three would have both title and publication author entered as "John W. Campbell, Jr." as that is the regularized form. No trace would ever appear in the ISFDB that the other forms had appeared in the publications.
The justification is that this kind of typographical variation is not interesting to the ISFDB. Of course one has to draw the line somewhere, and we could just as well have said it was interesting, but this was a rule inherited from before my involvement with the ISFDB, and I thought it was reasonable.
As far as having the author different on pubs, titles, and author records, I don't think this is possible. If I understand the schema correctly (see here and here), I think you'd have to have all forms of the name in the author table, and they couldn't remain there without at least one title or publication published under that name. So the current design requires us to be consistent -- either consistently regularize in all fields, or consistently record the name as actually printed in all fields.
For titles, the intent is the same as far as capitalization goes: regularizing capitalization does lose information that is in the original publication, but the ISFDB has decided it does not care about that information. Other changes, such as punctuation, are treated as relevant, so they get recorded as vts instead.
That summarizes what I meant the help files to say, and why. If I managed to make that clear, what do you now think? If you agree with the intent I describe above, I'll see if I can make the help files clearer, and maybe add a FAQ (or you could do the clarification -- it's often effective if the person who wants the clarification makes it). If you don't agree with the intent, let's talk some more and see if we can get a consensus. Mike Christie (talk) 22:06, 1 Feb 2007 (CST)
Hmm - that goes completely against my understanding. Isn't publication entry supposed to be raw data entry without regard to canonical author names and titles? That way if I have a copy of a publication I can look at ISFDB and see that either it matches my pub exactly or if it’s different then either someone screwed up or I should be cloning the record and making the new record match my publication exactly.
I believe the help already supports this and takes pains to italicize exactly when it comes to entering publication fields.
The question then is – how do we regularize author names and titles? I know of:
  1. The title records can have the regularize author names and titles and under that you would just list the publications. ISFDB does not care that the names don’t match as it maps titles to publications using title record numbers and publication tags (the pub_content table). The downside of this is you can’t do title searches for either the irregular title or author name. Technically there will be “stray publications” but if you make the irregular name a pseudonym of the canonical name you’ll see that the stray publication list is not shown. The title will show up in the canonical bibliography with no [as by] stuff and you don’t discover the irregular names/titles until you see the publication list for a title. I believe this is ok for small irregularities in a title and/or author name.
  2. Or, you can have title record that does match the publication author/title and instead use the variant title mechanism to link your “irregular” title record to a “regularized” record. The downside of this is you end up with a bunch of what look like spurious variant title listings to deal with slightly irregular data.
  3. Or you can use a combination of #1 and #2 above where for small irregularities you use #1 and for larger ones you use #2. Policy will need to deal with edge cases such as if adding or removing the leading “The” is a “small” irregularity.
Marc Kupper (talk) 22:45, 1 Feb 2007 (CST)
I feel that your suggested approach would be a bit cumbersome to work with. I will generally argue for entering exactly what's on the publication, but I think that some regularization is OK. After all, we do regularize capitalization: "L. Sprague De Camp" would be regularized to "L. Sprague de Camp" anywhere it occurred. I take your point about searching, but I think that's a small problem. I'll be interested to hear other opinions on this one. Mike Christie (talk) 23:08, 1 Feb 2007 (CST)

Nonsensical characters in author names

Several author names contain "&", being bad imports from multiple-author lists. It would be good if some admin corrected these, and I'd recommend adding a filter to the importing scripts to halt such cases on entry.

Also, some weird formatting bug: 38 names with "x9b", often repeated. Another with encoding accents.

Trying other numbers also gave an author name composed of Amazon image URL and some further irregularities.

I also recommend checking commas in the name: besides duplicities and mistakes any other sufficiently wide search yields, there are some cases of multiple authors merged into one without ampersand, like Arlan_Andrews,_Kris_Andrews,_Joe_Giarratano. --JVjr 05:36, 1 Feb 2007 (CST)

Similar problem even in titles; I recommend doing a more thorough search (and replace!) for all such possible characters from within the database, employing some stronger (regexp?) search terms. --JVjr 09:40, 1 Feb 2007 (CST)
  • Several author names contain "&" - All of these seem to be valid author names though some are non-canonical.
  • 38 names with "x9b" - This will be messy as every single one of those authors has stray publications and so it's more work than just renaming the author names. I fixed all of these. With all of them the pattern was that the publication record had the "9b" characters in the author's name (usually where a space would be found) but the title-reference record was fine. For the most part I was able to copy/paste the author name from the title record to the publication and in a few cases I also corrected the author names at the title/pub level as every single one of these bad publications had Amazon images.
  • Many of the "obviously wrong" commas in the name items also have stray publications and thus the title records need to be teased out using the same method as with the 9b in the name records but it's more work as the author names are often quite mangled at the title level too. It seemed to happen most often when a publication is credited "Mary & Bob Lastname" but sometimes I could not make heads or tails (literally) of the tangle of mangle.
  • Another with encoding accents. - This one had a title record which I corrected though I did not dig into what looks like extra names that are error for both the publication and title.
  • Overall - global search & replace makes me very nervous, particularly at the publication level where we are trying to match stated title/author names and also the authors/titles for publication contents where we are also often trying to match exactly what's in a publication. Thus each of these needs to be inspected and the fix worked out on a case-by-case basis.
--Marc Kupper 13:21, 1 Feb 2007 (CST)
The Dungeons & Dragons one is all gaming stuff. I'll delete them. As is the Sword & Sorcery Studios book. Dana Carson 20:44, 1 Feb 2007 (CST)
Historic City: The Settlement of Chicago seems non genre and the Chicago Dept. of Dev. & Planning probably does write a lot of fantasy but not our kind. —The preceding unsigned comment was added by Dcarson (talk • contribs) .

Just re the first point: now the author names are corrected - except "Dixon & Wenzel" in LHBBTLHBBT2004, which obviously should be deleted - but when I wrote it, there were several cases of co-author lists merged into one record.

Yes, global changes can have unpleasant unpredicted effects; but any semi-automatic help saving human work would be good. --JVjr 07:34, 2 Feb 2007 (CST)

Opening the "also wrote as" can of worms again

The great mystery writer John Dickson Carr is listed in the DB with a couple of SF items. His primary page shows that he also wrote as Carter Dickson, and gives a couple of SF items as by Carter Dickson. To be a bibliographical completist, I have been trying to enter in the fact that he also briefly wrote as "Carr Dickson" -- just one non-SF novel, it's true, but I think this info should still be noted, along with the fact that he also wrote an even earlier historical novel as by "Roger Fairbairn". But -- it appears to be impossible to enter this info. If I go to the Pseudo. page entry, where there is the gobblygook about Parent info, and try to enter Carr Dickson into the space below I get a message that he doesn't already exist. Okay, the solution then appears to make an entry for New Author. But this can't be done either. The only solution is to create a New Novel entry, along with the author (such as Carr Dickson) info. But this is ridiculous, since it means creating an entry for a NON-SF book just in order to get the author info into the DB. Surely there's gotta be a means of doing this someday! —The preceding unsigned comment was added by Hayford Peirce (talk • contribs)}.

The pseudonym author or title stuff works from the child (the pseudonym)'s record. Thus it needs to exist before you can use either of the pseudonym options. I’m going to use Roger Fairbairn as the example pseudonym here but you can substitute the other names as needed.
  • Add the Roger Fairbairn publication.
  • Once approved you will have both a title and author record you can work with. Go to the new title and
    • Select "Make This Title a Variant Title or Pseudonymous Work" from the left navbar.
    • In the lower part of the page where it says "If the parent title does not exist..." enter:
      • Title: Whatever the Roger Fairbairn title is.
      • Author1: John Dickson Carr
      • Year: Whatever the date of the Roger Fairbairn title is.
      • Title Type: I believe this already matches the desired title type.
      • [Submit Data]
    • Once approved you will see the title listed on John Dickson Carr's bibliography with [as by Roger Fairbairn]. It will also disappear from Roger Fairbairn’s bibliography.
  • You also want to set up the author pseudonym.
    • From Roger Fairbairn’s author page select "Make This Author a Pseudonym" from the left navbar.
    • Parent Name: John Dickson Carr
    • [Submit Data]
    • As a FYI - The author's record number is rarely shown on ISFDB screens I have never used the Parent Record # option. If you are interested in knowing the # for an author it's available from the Advanced Search results for an author (the middle part of the search screen).
  • Once approved then Roger Fairbairn’s page will say he's a pseudonym of John Dickson Carr and John Dickson Carr’s page will include Roger Fairbairn in the Used These Alternate Names: section. I see that Carter Dickson is already set up as an author pseudonym but that the title variant has not been set up yet. From Carter Dickson you would click on The Other Hangman and make that a variant with the new title also being called The Other Hangman and the author would be listed as John Dickson Carr.
Marc Kupper (talk) 22:08, 1 Feb 2007 (CST)
All of the above is nicely explained and works perfectly -- although I haven't quite yet grasped exactly what's going on behind the scenes. Thanks! Although I gotta say that this is most kludgey, Heath Robinson, Rube Goldburgesque way of doing an apparently simple task that I have ever seen in any earthly endeavor! In the Author Data field, couldn't there be a simple box called "Alternate Name #1" and so forth? It would sure save a lot of time in the case of someone like Carr where the goal is simply to enter in a couple of names for non-genre works....Hayford Peirce 17:43, 2 Feb 2007 (CST)
FYI, the only way I know to find an author id is to enter their name in advanced search, in the author search section; the record id comes back with the results. Mike Christie (talk) 22:11, 1 Feb 2007 (CST)
I just realized I use Author #s all the time, it's on the URL for Edit-Author, Make This Author a Pseudonym, and Titles... :-) Marc Kupper (talk) 22:52, 1 Feb 2007 (CST)

Someone please check my Kuttner Edits

I have a nasty feeling I've made "The Proud Robot" (shortfiction) a variant title of "Robots Have No Tails" (Collection). It should of course be that the Collection is a variant title of the other Collection. Also, Cloning is making me a bit lazy about Author attributions so do read my Notes about where the name is Kuttner, Kuttner and C.L. Moore, or Lewis Padgett. BLongley 14:30, 1 Feb 2007 (CST)

That used to be pretty easy to fix as you could go to the two shortfiction variants that should not be variants and via "Make This Title a Variant Title or Pseudonymous Work" you would set the parent-title to 0 (zero) to disconnect the VT relationship. If you try that now you get a "The parent title does not exist error" and the workaround is
  • Add a new dummy publication - to make it easy on you give it a title like "The Proud Robot (dummy)" (in fact, in this case it looks like you will have two dummy titles as there are two shortfiction variants called The Proud Robot.)
  • After the dummy pub is approved do an advanced-title search for them and merge the dummy title with your shortfiction. You want to keep all of the shortfiction fields except the "Parent" field at the bottom.
  • Without waiting for approval you can also submit a delete-pub for the dummy pub you added.
  • After approval of round two everything should be ok other than you add the VT for the correct title-instance of The Proud Robot.
It's not two shortfiction variants though - it's a shortfiction variant of a collection. It LOOKS as though deleting the variant title record 46738 alone would work, as the collection "The Proud Robot (1983)" isn't under that: but I'd like a Moderator to do it it really, I'm not that confident yet and the fact that the original collection is in a series too makes me nervous. BLongley 13:57, 2 Feb 2007 (CST)
OK, if Marc doesn't get to this tonight I'll take a look if I have time. Mike Christie (talk) 14:57, 2 Feb 2007 (CST)
It's fixed. It was a three step process, though the second and third steps could have been submitted without waiting for moderator approval, so really a two step process from the editor point of view. 1. Create pub "The Proud Robot (Temporary)" by Lewis Padgett, via New Novel. 2. Delete pub. 3. Merge that title with the shortfiction that should not be a vt of the collection. In the merge I chose the blank vt reference. This approach doesn't really affect the two titles except in that one field, so though it's a little clumsy it is quite safe. There's a feature request in for allowing a zero to be entered in "Make Variant"; that would be much quicker and easier, of course. Mike Christie (talk) 22:47, 2 Feb 2007 (CST)
I looked again, and the variant was still Shortfiction. That looks changeable though so if that goes through all I need to do is merge the 1983 title? BLongley 10:04, 4 Feb 2007 (CST)
Hmmm. I either didn't fix it properly, or there were two separate problems and I didn't spot the other problem. This is getting a bit long, and since it's specific to this case I'm going to move this discussion to your talk page. I'll leave a note there in a moment. Mike Christie (talk) 10:45, 4 Feb 2007 (CST)
I saw/approved your notes in the clones and they seemed clear enough to me. I think the notes are a good idea as it shows you looked for the alternate names in the pubs and did not find them rather than having people wondering why you did not attribute the credits the way they are "supposed" to be. I myself have been adding tons of notes lately and often will enter something like "The copyright page states "exactly the text that's on the copyright page about previous titles, revised stories, etc."" and sometimes also copy this to the title record(s) giving source attribution as to where the note was found. The goal being that future viewers and editors will better understand the sometimes gray, and nearly always complicated, linkage between stories. Marc Kupper (talk) 17:06, 1 Feb 2007 (CST)

Busy day

We had a lot of subs today; I looked a few times at work and realized they were piling up. I am glad someone else was able to get to them. Are we getting to the level of activity where it would be good to have another active moderator? I think we should be slow rather than quick to give out the moderator bit, because the review of inputs and the resulting discussion is educational (often for the moderator too!). However, if the submission rate keeps up (which of course we hope it does) the moderators won't be doing any submissions of their own. I was hoping to get a little further through the Le Guin and Moorcock projects I am working on tonight, but instead have spent my time on research submissions and communicating with editors. That's not a problem, but we are sure getting a lot of subs now!

If/when we do want a moderator, I think an open discussion on the Wiki of who would like to be one and who the community suggests is probably a good idea. Mike Christie (talk) 22:20, 1 Feb 2007 (CST)

I would not mind more moderators as I suspect I'll run out of free time soon. Overall, all of the current high volume editor seem to be fine. Sometimes there's an issue but I imagine someone watching/moderating what I submit would also sometimes put things on hold and ask what I was up to. People to do the right thing most of the time. Marc Kupper (talk) 23:05, 1 Feb 2007 (CST)
More moderators would certainly help editors that know what they want to achieve, but have to perform in multiple submissions. I know I've left multi-part edits for days as by the time the first submission got approved I'd put the book back on a shelf somewhere and was onto something completely different. (Which could be fixed by a feature request for Editors to be able to carry on working AS IF their changes had already been approved, but I'm an IT consultant and know such a suggestion leads to so much work it would kill the project.) So more moderators should be a long-term goal, but I have to say I'd like a bit more consistency from the current mods before they train up a new one. No criticism intended: you've all been very helpful, very supportive, but in different ways. For instance: I add a collection by an author and all the entries should be by the Canonical name - but I did it by cloning. Mod 1 corrected the few misattributed entries for me, Mod 2 just let my comments go through. I add a pub via cloning but the name has changed - Mod 3 puts it under the right variant title for me, Mod 4 just lets it through. Mod 5 appreciates added notes to a title, Mod 6 wants them moved to the Wiki page for the author. I don't know what special private conversations Moderators have, if any, but I think I'd prefer a bit more consistency as to what a Moderator SHOULD be doing before we add many more. And no, I specifically do NOT want to be a moderator - yet. I'd actually be quite happy if I had full access to the database, and will no doubt play around with a copy at some point. But as both Mods and Editors have a sub-set of privileges, it'll take me some time to learn the limitations of each. And those limitations cause me more problems than anything. BLongley 15:46, 2 Feb 2007 (CST)
If you want to play with the database, a recent backup should always be available at ISFDB Downloads. It's a MySQL database, so it's pretty straightforward, but there is more documentation here and here.
I know, I just haven't had time to look at it since I last downloaded V1 a few years ago. It's what made me look at MySQL in the first place. But conversion for my purposes is my problem, not yours - if I get the time, I'll probably play around in Oracle and learn how to do it in MySQL later. BLongley 18:22, 2 Feb 2007 (CST)
Moderator consistency is definitely a concern both in the short run and in the long run. The underlying problem is that many of the relationships are not trivial and it takes a while to internalize them. Even something as relatively straightforward as the OCLC data entry guidelines -- they used to concentrate on Publication data as opposed to Titles and Series records and also used a fairly flat database structure -- is a multihundred page monster with caveats sitting on top of caveats. What we are trying to accomplish is considerably more complicated in many ways and on top of that our guidelnes are still evolving as we discover more and more shades of gray. Add our very limited manpower to the mix and... well, you see where I am going :)
That's the problem. I'm perfectly capable of tuning a database to match requirements, but the requirements here are NOT clearly stated, and in fact are so variable I'm surprised we still have someone willing to work on any of them! That's the problem of Wikis for people of variable skills - if I had total control I'd do wonders for the database, and alienate a lot of people. I don't, so I'll take the alienation I feel and try and make helpful suggestions. But from that side, I can see how a lot of people might be put off from doing things. BLongley 18:22, 2 Feb 2007 (CST)
I think one thing that you identified above will be key in the long run: the ability to create self-contained submissions as opposed to having to go through the current multi-step approval process. It may require a fair amount of programming, though, and we are not exactly brimming with programmers :( Ahasuerus 17:12, 2 Feb 2007 (CST)
Well, being a good programmer then a good designer and a good analyst and a good communicator is how I got to be where I am at work nowadays. I still suck at being a good user, and notice I don't say I'm "Great" at any of those categories. Overall, I have enough Ego to say I could do a good job on the overall project, given time - which is exactly what most of us lack: and clear requirements, which people debate all the time. I think it's a worthwhile project, or I wouldn't put this much effort in. I also enjoy a lot of even the data-entry bits, or I'd go sleep sooner. So a couple of aims overall should still be "make it easy for newcomers without letting them mess things up too much" and "when a newcomer looks useful, have a clear training path to get them to the next level". I've no idea what a moderator sees, but a "moderator in training" level might be useful? BLongley 18:22, 2 Feb 2007 (CST)
All good points. I believe moderator consistency is best achieved by lots of debate, subsequent well-informed consensus, and good documentation on the help pages and elsewhere. I am hoping to capture some of the debated issues, initially on my user page, and use them to track those debates to completion. We have to remember this is still a beta release; we knew we would find issues for debate during the beta -- the job now is to nail down all the corners.
There aren't any special private conversations between moderators, by the way; there's no channel for such, though I do know how to contact most of the other moderators by email if I need to. Mike Christie (talk) 17:41, 2 Feb 2007 (CST)

Title corner case

Here's a test case for variant titles. Again I think I know what to do here, but am posting this note to describe a "corner case"; let me know if anyone thinks this is wrong.

In tweaking a recent submission of the first Ace edition of Leigh Brackett's "The Coming of the Terrans", I noticed that the book prepends dates to each story title. The book is a linked collection of stories, and the contents page and the story headings agree in form: "2038: The Road to Sinharat", for example. The relevant paragraph from Template:TitleFields:Title is:

"For short stories, essays and poems, take the title from the heading on the page where the work begins, rather than from the table of contents, if there is one. This distinction is not too important, and if you know that one form of the title is the usual one (e.g. the contents page has "Night Fall" but the story heading is "Nightfall") then use the one you know is standard. You can also choose to use the table of contents version where the story heading gives a non-standard presentation of the title form--e.g. if the table of contents says "Bell, Book and Candle" and the story header says "Bell, Book & Candle", you can use the former. If both the table of contents and the story title agree, though, the form given should be used, even if it is different from the standard."

The table of contents is actually titled "Chronology" and doesn't give page numbers, but it functions as a page of contents. The copyright page lists the story copyright dates without the prepended dates in the titles.

I think these dates have to go in; these are variant titles. The reason may be a chronology conceit, so that the usual titles are relegated to a subtitle of a chronology, but the net result is that the title has changed.

It's a fiddly job, since I have to dupe each title and then remove the old ones, and then make them all variant titles, so I will hold off on this just in case someone comments. I'll probably do this over the weekend. Mike Christie (talk) 22:47, 1 Feb 2007 (CST)

Maybe we should use Google for searching the site instead of MySql. That would allow us to put in notes about stuff like titles not matching between the table of contents and page header or the even more fiddly situation you have. At the moment I'm reading the short story The Sandman, the Tinman, and the BettyB which is stated as The Sandman, the Tinman, and the Betty B in the table of contents and thus Google shows both forms are common. I put the BettyB version in in ISFDB but also added a note about Betty B. I have been doing some thinking about how to do a full text index search for ISFDB so that you could just enter stuff in the main search box and not worry about which records/fields to select. The advanced search would be used for field specific stuff. Marc Kupper (talk) 23:18, 1 Feb 2007 (CST)

Are Fictional Authors Pseudonyms?

Just verified A Tourist Guide to Lancre where two pieces are by Gytha Ogg which is one of the Discworld characters. Does that get a pseudonym entry? And does anyone know which of the two authors wrote the pieces? —The preceding unsigned comment was added by Dcarson (talk • contribs)}.

I think these should get pseudonym entries; since the book overall is declared to be by Pratchett and Briggs, it is fair to assume that they both are responsible for the piece, so use both names. Mike Christie (talk) 06:23, 2 Feb 2007 (CST)
Ooh, this could be a dangerous precedent... if Mrs Ogg "exists", she's also the author of Nanny Ogg's Cookbook, in which case she's Terry Pratchett and Stephen Briggs for the essays on philosophy and etiquette, and Tina Hannan and Stephen Briggs for the actual recipes. However, some of the recipes are actually credited to other characters, such as Lord Vetinarii, Mrs Colon, Rincewind, Ponce da Quirm, Mrs Whitlow, Sergeant Angua and Lady Sybil Vimes.... how many pseudonyms do you want? BLongley 13:35, 2 Feb 2007 (CST)
You're scaring me. You're right, I hadn't considered this properly. Well, the reasons we record an alias include (a) that's what it says in the publication; (b) readers/researchers may want to know what was written under each alias; and (c) we may not know the real name. I think that if a book were published as edited by "Gytha Ogg", with stories or recipes by the various Discworld characters you list, then those would all be treated as pseudonyms. So what does the title page of Nanny Ogg's Cookbook say? If it says "by Terry Pratchett and Stephen Briggs, then it's definitely not a pseudonym. If it says "by Gytha Ogg", with no reference to Pratchett and Briggs, then it's a bit trickier. If we rigorously stick to the title page rule, then I think we'd still say "Gytha Ogg" was a pseudonym. But if Pratchett and Briggs are both credited in any way on the title page itself, e.g. "Nanny Ogg's Cookbook, by Nanny Ogg" with a footnote saying "transcribed by Terry Pratchett and Stephen Briggs" then I would say this is a coy way of crediting Pratchett and Briggs, and there is no pseudonym.
I can't even pretend to know how to identify the Title page on this book. There's eight pages of Editor's memos before anything that resembles that. Then we confirm it is Nanny Ogg writing it, next page has references to two other books (one "banned" and the other "withdrawn", neither existing in OUR world - but I wish they did.)
I should also admit I knew the original Gytha that Terry based the character on. A wonderful lady that will be missed by many. I was shocked to learn of her death. I don't think this should be linked to as an author entry here, but anyone tempted to add Nanny or Gytha Ogg anywhere should read this first: http://gythanorth.livejournal.com/ - I think she deserves a Wikipedia entry (If not an Encyclopedia Britannica one) as the Mother of Filk in the UK, but I'd be reluctant to trivialise it with an ISFDB one of less note.
For short stories, essays, recipes and so on, if Pratchett and Briggs are credited for the overall book, then so should they be for each content item. After all, the principle of least surprise should say that if we have penetrated the pseudonym at the publication level we can be said to have penetrated it for each of the content items too. If Gytha Ogg is a pseudonym, either she gets credit for the whole book, or each item is individually credited to the each pseudonym. As you say, this is the least desirable outcome -- it doesn't really add any value to the bibliography. In the case of A Tourist Guide to Lancre it does seem from the contents listing that that's exactly how that book was presented, however.
I think we have to ignore the title page and go by the copyright page that divides responsibility up better. BLongley 19:24, 2 Feb 2007 (CST)
Maybe. I'll make a note to include this as an example/test case in some additional FAQs I hope to write. Mike Christie (talk) 22:12, 2 Feb 2007 (CST)
Also, take a look at this description of a weird subtitle of a short story with a fictional author. Eventually I decided to follow Locus1 and record the subtitle and author attribution as part of the title of the story. Part of the justification was that I had attributed the book overall to Le Guin, so it made no sense to "pretend" this was a pseudonym.
I don't see an easy way to algorize this sort of situation for help-text purposes. Perhaps it's one of those cases that simply have to be discussed when they come up. Mike Christie (talk) 18:30, 2 Feb 2007 (CST)
I agree. I'd like to have a little time to see how the real Gytha's friends and family want her remembered before we record her (or the character she inspired) here in any detail. BLongley 19:24, 2 Feb 2007 (CST)
I also knew the real Gytha; it's been a long time since I saw her though. I did see the announcement of her passing and was sorry about that -- she was just an acquaintance from before I moved to the US, but I liked her a lot. Mike Christie (talk) 22:11, 2 Feb 2007 (CST)

RPG Cleanup

A title search on Advnaced brings up 81 entires. All but about 6-7 are Advanced D&D books. I can delete them if wanted but someone with privs could kill them faster. Dana Carson 03:06, 2 Feb 2007 (CST)

Al is the only one who can do this. He's always busy, and may not get to it, so another option is to list this as a task at ISFDB:RPG Cleanup. That's one of the projects listed at Bibliographic Projects in Progress. Mike Christie (talk) 06:26, 2 Feb 2007 (CST)
It seems simpler and more reliable to just delete them manually. A trick is to do the search and then to just click on the merge checkbox for all the titles you want to remove. Submit that as a mega-merge and once approved pull up the resulting title, delete all of its publications and finally the remaining title. I did a title search for RPG but that had far less than 81 entries and many are not RPG games. A search for "Advanced Dungeons" returns a returns a promissing list but I know next to nothing about RPG and thus exactly which titles should be zapped. Marc Kupper (talk) 13:09, 2 Feb 2007 (CST)
FYI, there are four target lists -- publishers, series, keywords, authors -- currently set up at ISFDB:RPG Cleanup. You have to be very careful, though, since a lot of formerly "RPG-only" designers and module writers have recenlty branched out into RPG-flavored fiction. It's not always easy to tell whether a particular Title is fiction or not just by looking at its bibliographic information at amazon.com, used.addall.com, etc. Ahasuerus 15:12, 2 Feb 2007 (CST)
OK, they've all been submitted as a big merge. Dana Carson 17:25, 2 Feb 2007 (CST)
I have put the big merge submission on hold for a few hours. One thing about RPG contributors (and comics folks) is that they tend to collaborate with each other an awful lot. If you identify one, you can then find literally hundreds of others by following their collaborators' collaborators' collaborator's etc. Six degrees of separation and all that :) Therefore I always make sure that we already have the name of the author(s) of the about-to-be-deleted title listed in the target list. With a submission this big, it will take a bit of time to cross-check. Ahasuerus 18:40, 2 Feb 2007 (CST)
OK, the target list has been updated. You can find all doomed publication in this bucket. Ahasuerus 22:10, 5 Feb 2007 (CST)
OK, deleted all of them. The merge had a lot more that that I thought. Did you kick some back because they are by writers with other SF books? I did a search and found some that the only books listed for a writer are gaming books. Dana Carson 23:56, 5 Feb 2007 (CST)
No, I approved the merge submission "as is". It seemed to contain only about a dozen Title records, though. I wonder if Al has it limited to the first 10 or 15 submitted Titles? I'll experiment tomorrow, thanks for the heads up! I guess for now you can re-run the search and submit another batch of 10 or so. Ahasuerus 01:30, 6 Feb 2007 (CST)
OK, Will terminate in small batches. Dana Carson 02:27, 6 Feb 2007 (CST)
On a more general RPG cleanup question. What about works where the writer is a general SF author. John M. Ford was a SF author first but also wrote gaming books. Do the gaming books get listed like other non genera books do? Dana Carson 17:25, 2 Feb 2007 (CST)
Well, if they are non-fiction books, then they can be listed as NONFICTION. Is that what you had in mind when you wrote "gaming books"? Ahasuerus 18:40, 2 Feb 2007 (CST)
Yes, sourcebooks and at least one adventure in Fords case. I've seen talk that one problem is that writers write game material and then when they get experience leave to write for the big money in fiction. Which says how bad the pay is in gaming. This means that the all books by a mostly SF writer, only actual SF books by a normally non SF writer will get used a lot. Ford is something like 10 SF books and collections, 2 Trek books, and 5 game books. Is 2 to 1 enough of a ratio that all books go in or just the SF books? Dana Carson 00:06, 3 Feb 2007 (CST)
re: tend to collaborate with each other an awful lot... That's a good idea. I've sometimes run across a non-genre author in ISFDB and start following the collaberation threads to find dozens of others. Though I never put in a feature request I do wish there was a way to dump out an author, title, and/or publication tree in XML so that it could get dropped onto the author's bibliographic wiki. Someone someday may want the data and having it as an XML blob would let them work with it. Marc Kupper (talk) 20:15, 2 Feb 2007 (CST)
Retrieved from "https://isfdb.org/wiki/index.php?title=ISFDB:Community_Portal/Archive/Archive05&oldid=392351"