ISFDB:Help desk/archives/archive 08

The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

This is an archive page for the Help Desk. Please do not edit the contents. To start a new discussion, please click here.
This archive includes discussions from March - May 2009.

Archive Quick Links
Archives of old discussions from the Help desk.

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

Expanded archive listing

Where the date seems to disappear?

Sometimes when I have submitted something new, the date field have seemed to empty itself, so that the date in approved submission has been 0000-00-00. Often I have been fairly sure that I have entered the correct date, but then thought that probably I have been mistaken after all. But now I was able to capture it. I enterd a new edition: as you can see. After I submitted it I noticed straight away, that the date field was empty. Here. And the edition has an empty field for publication year. I believe this has happened before for me at least for a few times. Tpi 07:46, 1 March 2009 (UTC)

This happens if there is a trailing space after the last digit. Unfortunately, they are hard to spot... Ahasuerus 02:56, 2 March 2009 (UTC)

One Artist is listed as two separate Artist-resulting in inaccurate credits and biography.

I am new to ISFDB and extremely unfamiliar with your system, procedures and methods.

First of all, welcome to the ISFDB! We don't bite -- at least not very often -- so feel free to ask any questions you may have. We have an extensive Help system, which explains how to edit the database, but if all you want to do is correct your own data, just let us know what needs to be done and we'll make the necessary changes. Ahasuerus 04:48, 2 March 2009 (UTC)

I have noticed that there is a mistake in your records resulting in one artist (along with his credits) being listed as two separate artist because his name was credited at various times with his middle initial and then without. It seems like an easy fix but I am not a writer or confident enough to start altering your entries.

Unfortunately, it's a very common problem. With over 54,000 author records (the term "author" also refers to artists and editors in this case) on file, it can be hard to tell whether "John Q. Public" is the same person as "John Public", so a number of our bibliographies end up being split. Once we confirm that both forms of the name are/were used by the same person, we pick one as the "canonical" name and link the other forms of the name as pseudonyms. Let us know which form you would like to be the canonical one and we can clean up your bibliography quickly. (The rest deleted after edit conflicts with Michael.) Ahasuerus 04:48, 2 March 2009 (UTC)

Also I noticed there was no Biographical information listed for him. Is there a way to submit a biography for editing or approval without actually posting it? I am the artist and do not want to upload anything that is not proper or following your normal guidelines. I do not even know if it is proper for someone to add their own personal Bio but it is another good reason for having it checked first.

Thanks, Gary —The preceding unsigned comment was added by Evilg (talk • contribs) .

First things first: Give us both names under which the artist is credited. Once it's been determined which one is the canonical name, we can create a pseudonym relationship between the names, then make variant records so that all of his credits appear under one name. And to be clear, ISFDB policy is to record credits exactly as they appear on the publication. If he's credited differently in various publications that will be how it is recorded in the database. They're only inaccurate credits if the editor enters the wrong information.

Second issue: Bios can be created by clicking on the red-linked name after Biography. You'll be led to a blank wiki page, click on "edit this page", then enter the biographical information. Before writing the bio please read this short guideline for biographical entries.

P.S. Please sign all comments with four tildes (~~~~). Thanks, Gary, and welcome to the ISFDB. MHHutchins 04:36, 2 March 2009 (UTC)

Thank you for the assistance and especially the patience. I'm not sure which name would be considered the "canonical" name. One is “Gary L. Freeman” and the other is “Gary Freeman”. It seems arbitrary to me but while “Gary L. Freeman” is more formal and complete, far more work is listed under “Gary Freeman”. BTW, I didn’t notice any thing credited to it in your list but a great deal of art work (by same artist) has been signed with my alias “Evil G”. Thanks again (Evilg 19:53, 4 March 2009 (UTC))

I think going with simply "Gary Freeman" would be the best going forth. I will make "Gary L. Freeman" the pseudonym and create variants for those titles. That will get everything onto one page. If you're aware of any book / magazine covers or interior work in which you used "Evil G", you can make changes yourself or give us a list. As a personal aside, let me say how much I admire the work you did for Asimov's and Amazing. The moment I turned the page and saw it, I knew immediately that it was a Gary Freeman work. Thanks and with much appreciaton. MHHutchins 20:20, 4 March 2009 (UTC)

Thanks again for the assistance. I'm honestly flattered by the compliment considering how much talent you must encounter through your efforts here alone. I feel like I'm getting a lot better so sometimes its painful for me personally to look at some of my older work. I admire what you and everyone here are doing and YOUR contribution to this field. It seems an overwhelming task and is much appreciated!(Evilg 22:08, 4 March 2009 (UTC))

Recording Maps as Content

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

What Has Gone Before.... xi

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

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

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

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

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

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

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

Sounds like a good idea! Ahasuerus 03:55, 6 March 2009 (UTC)

Done. Moved here. --MartyD 10:48, 6 March 2009 (UTC)

Making additions which are doubling themselves as defective clones

Apologies, but I just had a spate of duplicates created which are not complete. These two were clones. The ones before did not do this, and the ones after did not. Here. [1] top two, and [2] top two. I did the first one from title search and the other from the series list. One title, Captain, did not have problems. I never had partial clone cloning before. What is the suggested fix? Thanks, Harry. --Dragoondelight 13:46, 4 March 2009 (UTC)

There have been occasional reports of duplicate records created upon submission approval, but nothing definite has been identified so far :( Ahasuerus 03:36, 5 March 2009 (UTC)

Apparently, I am balmy as they appear to be new additions and they were done at the same time I did mine. One though is the same printing. Thanks, Harry. --Dragoondelight 17:31, 5 March 2009 (UTC)

Variant Covers

I have a copy of Stross' Saturns Children (http://www.isfdb.org/cgi-bin/pl.cgi?STRNSCHLDC2008) which, other than the cover appears to be identical to the cited entry. The cover art is even credited to the same artist (Grimando). The actual art is significantly different. I have a scan I can throw online if useful.

What to do??

First of all, welcome back! :) As far as the cover goes, a number of possible scenarios come to mind. The current scan comes from Amazon, which has been known to continue using pre-publication cover art for years even though the released book had totally different art. On rare occasions, this caused us significant problems because the credited author was changed prior to publication: in one case a co-author was added/deleted and in another case a different pseudonym was used. However, in this case the book has been verified, so I would start by asking the verifier (Al). Perhaps the book was released with two different covers, a known marketing gimmick. Ahasuerus 03:47, 5 March 2009 (UTC)

Actually, I think I remember seeing both covers on HB's of this title. I'll ping Al to see what he wants to do. It also wouldn't hurt to scan the ENTIRE cover. The artwork is continued onto the back. The painting is quite stunning compared to the original.

Regards - --Dsorgen 02:26, 11 March 2009 (UTC)----

I haven't added any cover art to ISFDB yet. The instructions seem to point to adding art to the wiki, not the ISFDB. Did I miss something?

Regards --Dsorgen 02:56, 5 March 2009 (UTC)

That's right, our Wiki software has built-in file upload capabilities, so we decided to use them instead of re-inventing the wheel. All cover scans are first uploaded to the Wiki and then linked from the ISFDB side. Was the Help page on the subject clear? Ahasuerus 03:47, 5 March 2009 (UTC)

Hmmmm... Just skimmed it (since I didn't find info on uploading to the ISFDB directly); so, I think I need to reread. Then again, if it had been REALLY clear, I probably wouldn't be asking this question. Sigh... I'll give it another shot and let you know.

--Dsorgen 20:02, 8 March 2009 (UTC)

Chapbook Editing Support?

Did I miss the announcement? It seems that chapbooks can now be edited (Chapbook has been re-added to the list of possible PUBTYPES when editing a Pub?) I was updating an ebook tonight and was halfway through the edit when I thought to myself... Huh...This isn't supposed to work... I'm supposed to clone and update... but there it was in the drop down box 'Chapbook'. I submitted the update and waited to see if it would take or if it would revert o Anthology somewhere else in the approval process... but there it is NGHBRHDWTC2006 newly updated with notes and coverart. Thanks for the upgrade! (or was I just confused?) Kevin 03:24, 7 March 2009 (UTC)

Nevermind - I realized it's editing 'TITLES' with type set as Chapbook that is unsupported... not editing PUBS with type set at CHAPBOOK. - Move along, nothing to see here. These are not the droids your looking for. Kevin 03:31, 7 March 2009 (UTC)

Problems Uploading Images

I've suddenly started getting error messages when uploading images. The images upload, but the template reports errors, and doesn't display the uploaded image.-Rkihara 07:25, 8 March 2009 (UTC)

That's because you've uploaded an image of 404 × 623 pixels - the maximum is 600 if you want to avoid the preview error. It doesn't affect final use of the image though. BLongley 11:55, 8 March 2009 (UTC)

Thanks! I'll have to readjust my Photoshop macro. I didn't know the max image size was red lined in the template. My previous scans were maxed at ~550 pixels, so when Asimov's changed size I figured 623 was close enough to the "suggested" max.-Rkihara 16:16, 8 March 2009 (UTC)

How to handle conflicting data in same pub?

In verifying HRLFKRKS1973, originally entered by someone other than me, I encountered a note that the cover art is credited to Allan Mardon on the copyright page (which it is) but "looks like Bob Pepper." While investigating, I found the cover is in fact clearly signed "Pepper". It's all one wrap-around (front, spine, back) painting, so it doesn't look like there is any other cover art to which the copyright credit could be referring. I updated the notes and changed the artist from Allan Mardon to Bob Pepper, but how should this situation of truly conflicting data in the pub properly be handled? --MartyD 16:14, 8 March 2009 (UTC)

I think you did just fine, you changed the database and documented why, and documented the incorrect information so someone else doesn't change it back. At most, I would have amended the comment to something more like "Copyright page states 'WRONG NAME', but the artwork is in the style of 'RIGHT NAME' and 'RIGHT NAME' is confirmed by the signature 'PART OF RIGHT NAME' at 'LOWER/UPPER/MIDDLE' 'LEFT/RIGHT/CENTER' of the cover art." But like I said, what you did was just fine and documents the discrepancy. Cheers! Kevin 20:14, 8 March 2009 (UTC)

Richard Powers vs. Dick Powers

What is the difference between the artists Richard Powers and Dick Powers? I want to verify this THSNFTRZNX1963 1st Ballantine printing The Son of Tarzan. It does not credit the artist, and I don't see a signature (in contrast to other books, where I've found a "Powers" signature). Googling for the F748 edition finds many references, all crediting "Richard Powers", but none particularly authoritative as far as I can tell. Later verified versions credit "Dick Powers" and have cover images with the same painting, so I don't really have a problem with the existing "Dick Powers" credit. I just want to understand the difference if they are two people. Thanks. --MartyD 10:20, 13 March 2009 (UTC)

They're the same person. The art on later printings must have been credited on to "Dick Powers", because there'd be no reason for someone to use this variant otherwise. The problem is that no one has made "Dick Powers" a pseudonym of Richard Powers. Once that is done someone will have to create variants for all of the "Dick"s (looks like about 50 credits on his summary page.) In the meantime, credit "Dick Powers" on your pub, and note that the book contains no credit and give the source as a later printing. Then merge the newly created record with that of the later printings. That will cut down on the number of variants that will have to be created. Thanks. MHHutchins 05:08, 22 March 2009 (UTC)

Many of those credits were duplicates. After merging them the page appears more manageable. MHHutchins 05:32, 22 March 2009 (UTC)

I'm game. I'll give what's left a shot and ping you on your talk page if I get stuck. Thanks. --MartyD 10:36, 22 March 2009 (UTC)

Image Linking Permission - Ace Image Library

Did anyone ever ask for a blanket permission to reuse covers from the Ace Image Library? I just came across a linked cover and checked ISFDB:Image_linking_permissions which lists the Ace Image site under the 'list of sites granting permission' BUT it also says we do NOT have blanket permission. Should Help be updated to show blanket permission yet, or should it be rearranged to show that they have not (yet) granted or been asked to grant permission? - Thanks Kevin 23:42, 21 March 2009 (UTC)

I don't think anyone has ever asked for blanket permission for the Ace Image Library, but people have asked on a case by case before (and help suggests we should still do so). But I seem to recall that the site-maintainer was known to be on hiatus and probably wouldn't respond anyway? I think all off-site links should be discouraged if you can add to the ISFDB library anyway: I know cloning will increase the problems, and our bots still link to Amazon by default, but I think "upload your own image here" is best, "link to explicit permission sites" is next, "implicit permission" last. No permission - don't do it. I seem to be the only person that's asked other sites for explicit permission recently though, judging by the state of help (the last three sites were from when I worked on van Vogt before I got ground down by "fix-ups"). :-/ BLongley 00:51, 22 March 2009 (UTC)

Recognizing that 'upload your own here' is now the preferred / long term method, what is the 'official ISFDB' email address that should be inserted into the form email on the image linking help page? Kevin 02:50, 22 March 2009 (UTC)

AFAIK, we don't have one. I don't think anybody's ever used the form letter, it was suggested by an editor that's no longer active, and it does make some unwarranted assumptions like "we are organised". ;-) . As I seem to have been the most recent person to get permissions, I'd suggest ignoring the form letter and write something less formal. I tended to start with complimenting their site, explaining how I found it and what the common interest is that led me there, pointing them at ISFDB's entries on the same subject and asking them what they think of us bibliographically, and pleaded with them to allow us to deep-link to the images we couldn't get from any other source. I couldn't promise a credit on every image we used like we do with Visco (that requires software changes), but did say we'd add some links back where possible. (I'm not a fan of HTML in pub notes, when it goes wrong it can go SEVERELY wrong, but I think most sites I used are linked to from the Wiki somewhere.) BLongley 20:41, 25 March 2009 (UTC)

Nowadays, I guess I'd add that as we have image uploading capabilities ourselves then we'd be bothering them even less (only our scanner-impaired editors need such), and would invite them to reuse the ones we've uploaded too - a sort of swap-shop deal. Whether that's by mutual deep-linking or downloading one site's images and uploading at the other is a bit open to question. I think there's benefits either way - deep-linking MIGHT lead to people discovering the other site, and is simple, but copying the image and giving credit and a link on the Image's Wiki page here would probably make such associations more visible. But people are not doing much with extra data on Images uploaded here - I can understand that, "fair use" or "out of copyright" or "creative commons" justifications on every single image are a pain and IMO unnecessary - but to maintain friendly relations with other sites it might be wise to make some effort. BLongley 20:41, 25 March 2009 (UTC)

For the record, one person recently used the form letter as a template for a message trying to get permission to link to images (not to the ACE site, but to The Trash Collector). I did edit heavily and adorn with compliments. No response, alas. --MartyD 00:55, 26 March 2009 (UTC)

An interesting site. I've rarely seen a seller go to that much effort, and it should be applauded. However, it looks like they only give images for books for sale, so once they've been sold the image is unlikely to be available? The images are also a little smaller than I like - ideally our images should make catalogue numbers and prices readable, IMO. So I suggest we don't link to their images. Whether you want to take a copy of their images for books we don't already have images for is another matter. We've been advised that "slavish copies" of covers aren't copyrightable, the copyright remains with the original book publisher or artist, and we could use the same images under the same "fair use" rules that other sites publish them under. I've never been totally comfortable with that - I don't use Amazon Images where somebody has added a "image provided by XYZ" section to the image, and there are sites where a good cover image has been "watermarked". I don't think that's the case here, but as it's a commercial site they may be able to afford lawyers. :-( BLongley 23:58, 26 March 2009 (UTC)

How to read "broken" number line?

I've stumbled across this situation a couple of times now, where a book a number line like this:

 27 26 25 24 23 22 21 20 19 18 17 16          6 7 8 9/7 01/8

How do I interpret this? Is it a 16th printing? Something else? If it is a 16th printing, is there any meaning I should infer from the trailing numbers? Thanks. --MartyD 11:02, 25 March 2009 (UTC)

It's a 16th printing. The rest seems to be an indication of year (possibly month): does expanding "6 7 8 9/7" to "1966 1967 1978 1969 1970" and reading it as "1966" make sense in this example? BLongley 20:18, 25 March 2009 (UTC)

The " 01/" might also be further year information, for "1970 1971" but that looks stilted to me. But it's for future use, not important in determining the date of this pub. BLongley 20:18, 25 March 2009 (UTC)

There's been some discussion on whether the final "/8" might mean "August" but such usage seems to go against the point of number-lines: you erase a number to indicate a later date/printing. BLongley 20:18, 25 March 2009 (UTC)

Finally, just out of personal interest - would this be a Scholastic title? As I only ever seem to get that problem with that publisher. BLongley 20:18, 25 March 2009 (UTC)

It's a Scholastic title. TK-210 Ten Great Mysteries by Edgar Allen Poe (Conklin, ed.). It's undated, except for the 1960 TAB copyright, which is why I was hoping there might be some meaning in the number line. Below the number line is "Printed in the U.S.A", centered, and on the line below that, right-justified is "06", which I suppose might be a month. I definitely got this book in the 70s. 1976 is possible. --MartyD 00:38, 26 March 2009 (UTC)

This page seems to confirm 1976. As it's not a first printing, there's no Month encoded in it. The "06" is a red herring it seems. One slight niggle - is there really no space between the final 0 and 1 in your book? BLongley 19:14, 26 March 2009 (UTC)

Really no space. You want it to be /7 = 1970 and /8 = 1980, don't you? I do! Maybe that's the right way to read it. I'm going to dig up a couple of others I have and see what their lines look like. --MartyD 00:16, 27 March 2009 (UTC)

One other thing I can tell you is it's definitely earlier than a 1980 printing, so that page's analysis that would have this be a 1980 edition of a 1976 first printing doesn't seem quite right. --MartyD 00:27, 27 March 2009 (UTC)

Thousands of books and yet pathetic reading comprehension skills. Sigh. I re-read the reference, and I can see this is a 1976 dating, with the /8 covering the 01 for the 80s and the lack of space probably a small typesetting error. Thanks for digging it up, and I will go with that dating and also check my other Scholastics. --MartyD 10:07, 27 March 2009 (UTC)

The Complete Pegāna

I seem to have gotten a stay record in my edits this title. Somehow I've ended up with a second title record here. Part of what I was doing was converting the book from a collection to an omnibus. In the course of that I was also adding the constituent collection records and I suspect that's where something went awry. Regardless, I can't figure out how to get rid of the extra title. I get all sorts of dire warnings if I start a merge of the two titles. Any help on how to proceed would be appreciated. Thanks. --Rtrace 00:25, 27 March 2009 (UTC)

Fixed. You have to "Remove title from this pub", but you can't see the Omnibus titles when they match the Publication type. I temporarily changed the pub to "Anthology", removed one of the Omnibus titles, changed the pub back to Omnibus, and merged the two Omnibus titles (which is allowable when they're not in the same pub). Please check. BLongley 17:48, 27 March 2009 (UTC)

Also, are you going to add "Beyond the Fields We Know", as one of the title notes suggested is also in there? BLongley 17:48, 27 March 2009 (UTC)

Thanks. I also adjusted the date of the title record for the omnibus. I wasn't going to include Beyond the Fields We Know since it is only a small subset of the stories from that collection that are included as opposed to The Gods of Pegāna and Time and the Gods which are completely included in the omnibus. The stories themselves are included.--Rtrace 18:16, 27 March 2009 (UTC)

Thanks, I've clarified the title notes to explain why it's an Omnibus of two books rather than three. BLongley 20:32, 27 March 2009 (UTC)

How to move pub to existing variant title?

I've read Help:How_to_unmerge_titles and am wondering if there's a better/simpler way to handle this case. The bottom-line question: Is there a way to move a pub from its current title to an existing variant of the same title, i.e., to change its title reference? The background: In re-checking my copy of Star Hunter & Voodoo Planet (ACE G-723), I did a title search for "Star Hunter" and discovered both Star Hunter & Voodoo Planet and Star Hunter / Voodoo Planet. The former is a variant title of the latter, but G-723 does not appear in its pub list because it is linked to the latter title (the other Star Hunter & Voodoo Planet pub appears in both lists, as I'd expect). The DBA in me thinks this must be a simple id/reference update, but I don't see a way to do that. And I confess to being afraid to push the "Unmerge Titles" button on Star Hunter / Voodoo Planet to see what would happen; if I should have tried that, I apologize in advance. Thanks. --MartyD 11:16, 28 March 2009 (UTC)

It should just be a simple update, yes, but unfortunately it hasn't been coded for. The Help looks a bit out of date actually - you don't have to unmerge all publications and remerge all the ones that didn't need to be unmerged. You get a set of checkboxes that allow you to select which one(s) to unmerge. You can unmerge G-723 alone, adjust the new publcation/title it creates as necessary, and remerge with the correct parent. Yes, still a bit of a pain, but not as bad as help makes it look. BLongley 12:55, 28 March 2009 (UTC)

Not scary at all! Guess I should have gone ahead and tried it. Thanks! --MartyD 13:27, 28 March 2009 (UTC)

Orphan publications (Eidolon magazine)

Hi, If I search for the title "eidolon", I get the covers for 20 or so issues of this magazine, plus a few interviews, but no titles for the actual issues. If I click on a cover at random, it gives me a publication associated with the cover. I click on the publication, I get the publication record (with a contents list so someone has labouriously entered data for the issues), but the publication seems to be an orphan - no title record associated, as far as I can tell. I haven't delved into magazines/serials much in ISFDB so I'm not sure quite what to expect. ... but I think there should be title records? How do I get them back/create them? ...clarkmci/--j_clark 05:23, 3 April 2009 (UTC)

Magazines are very different from books here. With a simple search, you're not looking for a NOVEL or COLLECTION or ANTHOLOGY in the results, you actually have to look for an EDITOR record. It's further complicated by the fact that to keep Magazine Editors author pages short, all the magazines for a year are sometimes merged under one EDITOR record for that Year of that Magazine. Or, in other cases, nobody has got round to adding the editor at all and they won't show up in the search results. In the case of Eidolon, a) the Editors seem to be unknown and b) the magazines don't even have an EDITOR record for "unknown". :-/ BLongley 18:55, 3 April 2009 (UTC)

All is not lost though - do an Advanced Publication search for "Eidolon" and you should see the desired titles, helpfully labelled "MAGAZINE". Or start from the Magazine:Eidolon wiki-page. BLongley 18:55, 3 April 2009 (UTC)

Things are getting better, though, with more and more Magazine Pubs getting proper EDITOR Titles. There are still a few thousand EDITOR-less Magazine Pubs, so perhaps we could mount a concerted effort to clean up this area. Ahasuerus 19:29, 3 April 2009 (UTC)

There's even a sort-of-project hinted at in Bibliographic_Projects_in_Progress (look at the "Other" section, first entry). But EDITOR records need a bit of explanation before we invite people to help fix them. And the standards/guidelines for unknown Magazine Editors could probably be a bit clearer. Or "Existent" - I think it's only when we got to Fanzines that we learnt to cope with such? Still, if people are interested in a particular Magazine with missing editors, and are willing to learn, some sort of Crash Course might help with the massive data-entry required. If we're not going to script a solution, documenting what needs to be done manually might get some more volunteers on board. I did some Kyril Bonfiglioli titles and some of the missing John Carnell titles but am not up to giving guidance on "unknown". BLongley 20:02, 3 April 2009 (UTC)

Well, we could slap the editors mentioned here. I think nearly all of the prozines have editor records. When I glanced at the semipromizines I realized it was going to take a lot of potentially fruitless research which would likely not result in particularly accurate results. Even those semiprozines that do have editors listed need to be thoroughly researched. Some of them have as many as a seven editors listed. --swfritter 21:27, 3 April 2009 (UTC)

Slap away, if it helps. I think the requirement for EDITOR records before the "simple" database search software works suggests something is fundamentally broken and would prefer to see MAGAZINE as a "container" title like we have with books: but that requires big software changes (far more than fixing CHAPTERBOOK even) so educating more people in how to deal with them is all we can do in the meantime. I know the magazine editors tend to get together directly and educate each other, but Help updates lag behind (same as with books - we're afraid to change stuff without consensus) and maybe we could usefully do some small updates for small parts of the process. But to a new ISFDB editor, EDITOR records are a mystery and even to an experienced ISFDB editor, spotting when the problem(s) even exist(s) is tricky. But underneath there is a small, simple fix that just needs to be applied to thousands of records: EDITOR merging and Series for regular magazine columns, etc, can wait. Even establishing what kinds of Editors get a credit can wait a bit - yes, there are often multiple editors, and I could probably find ones with more than 7, because you can get managing editor, fiction editor, nonfiction editor, poetry editor, art editor, review editor, and all those might be shared by co-editors or even subdivided (reviews of books, reviews of magazines, reviews of computer games, reviews of RPGs...) BLongley 22:05, 3 April 2009 (UTC)

I think I just made a simple task sound very complicated. :-( But there are simple tasks that can be done before you have to go down the "what are we doing this week that will make me go revisit all the magazines I've ever looked at before?" route. Maybe some guidance per currently-imperfect title is a start? BLongley 22:05, 3 April 2009 (UTC)

Checking the last backup file, I see that we have 1,277 Magazine Pubs without EDITOR records, including a number of New Worlds, Interzone, Weird Tales, Locus, Nebula Science Fiction, Omni, Science Fantasy and other well known magazines. All of them have editors defined at the Pub level, so I believe it should be possible to write a script to insert an identical EDITOR record into each pub. There are a few that clearly credit the wrong person(s), but that's a problem with our Publication level data which needs to be fixed anyway. Also, there are 146 missing "unknown" EDITOR records, which should be, in more than 50% of all cases, changed to "Editors of ...". Let me see what I can do tonight. Ahasuerus 22:18, 3 April 2009 (UTC)

P.S. Based on XML:PubUpdate, it should be a simple enough task. I'll ask Fixer to create a few submissions later tonight to see how it goes. Ahasuerus 22:24, 3 April 2009 (UTC)

It might be a good idea to work on the pubs first before invoking Fixer otherwise any errors will have to be updated twice since the editor title entries are not updated when the pub editor entries are updated and vice-versa. Most of the pubs mentioned above should be easy to verify. With Omni it might be a good idea to include both the main editor and the fiction editor. Did all of the British s-f fans throw away their mags once they read them? It would sure be nice to get some primary source data in that area.--swfritter 16:26, 4 April 2009 (UTC)

Well, doing it manually would also allow the missing months to be put in. There's one gotcha to watch out for on old New Worlds - the Pub Format or Binding of 7¼" x 9¾" (for instance) gets cut off at the first double quote when displayed for editing, and needs putting back before submission. BLongley 17:24, 4 April 2009 (UTC)

I've also noticed that several prices need regularising, e.g. "2s" to "2/-". BLongley 18:06, 4 April 2009 (UTC)

The Pub Format should probably be the closest equivalent we have with the actual size put in the notes. Of course, the stories, if they are not reprints, should have the month of the pub - and also any variant titles attached to them should have that date unless a variant was printed earlier in which case the story in the mag should have the date of the earliest variant. Fixing the editor data is not dependent upon fixing the other data although it would certainly be nice to see all this data cleaned up.--swfritter 18:40, 4 April 2009 (UTC)

I did the Science Fantasy manually as a lot didn't have Visco covers or prices and I could fix those on the way. BLongley 19:03, 5 April 2009 (UTC)

(Unindent) I have a fair number of British magazines from the 1950s and 1960s, but I haven't gotten to them yet and at the rate it is going I doubt I will until 2011. In the meantime, I think it may be better to have Fixer create EDITOR records, which will allow "normal" searching and prevent duplicate pubs from being created. Ahasuerus 03:27, 5 April 2009 (UTC)

Here's an oddity. This magazine shows no contents. It should show Lifeboat by "Will F. Jenkins" and sure enough if you look at it from the Lifeboat title this magazine is shown. I know there's a missing editor record, but is there something special about the "unknown" pub editor as well? BLongley 18:54, 5 April 2009 (UTC)

I've just fixed a load of "Cavalier" magazines with Stephen King short stories that had the same problem, using the "Editors of Cavalier" convention, but it might be wise to address these sooner than later as the stubs look useless and might attract an unwanted deletion. BLongley 18:54, 5 April 2009 (UTC)

Seen it before myself - if there is only one entry and no Editor record I don't think the one story shows up.--swfritter 22:06, 5 April 2009 (UTC)

Not good. We should fix those before the deletions get submitted. I suspect Fixer won't fix "unknown" editor? Or if he does, we'll still have to go rescue those. (There's not a lot so far though.) BLongley 22:24, 5 April 2009 (UTC)

Since there are less than 150 Magazine pubs edited by "unknown" which don't have an EDITOR record, the easiest thing to do would be to generate a list and fix them manually. Ahasuerus 23:51, 5 April 2009 (UTC)

(artist)

Somehow I missed that (artist) is being used when there is an artist of the same name as author. Is this common? Example Robert Hunt (artitst). Thanks, Harry. --Dragoondelight 15:44, 3 April 2009 (UTC)

Not very common, but when there is a collision of two names we add a suffix to distinguish them - sometimes the country, sometimes the date of birth or period of activity, sometimes the role they have in publications, sometimes just a number. Do a simple name search for "(US)", "19" or "[1]" for examples. BLongley 18:59, 3 April 2009 (UTC)

Thanks, Bill. Thanks, Harry. --Dragoondelight 11:49, 4 April 2009 (UTC)

Another unknown sig

Any ideas on this? BLongley 21:32, 3 April 2009 (UTC)

Just found a similar project. If your unknown artist ever worked on Life magazine, you might find some help there. BLongley 20:16, 4 April 2009 (UTC)

How about "P dell'osso/"? --MartyD 20:59, 4 April 2009 (UTC)

Possible. "dell'osso" seems a valid surname, but when searching for an artist like that I keep getting dragged to Italian sites. And I don't read Italian. :-/ BLongley 21:55, 4 April 2009 (UTC)

How to enter omnibus of collections of unpublished letters and stories?

I could use some guidance on how to enter Lord Halifax's Complete Ghost Book. The publication seems to contain two books, Lord Halifax's Ghost Book (1936) and Further Stories from Lord Halifax's Ghost Book (1937), each with its own introductory material.

The first book describes itself as "a collection of stories, haunted houses, apparitions and supernatural occurrences made by Charles Lindley, Viscount Halifax". The foreword describes how the man was fascinated by ghost stories, and people would send them to him. Each story has some sort of attribution, such as "Mr. X contributed these experiences" or "Related in a letter from Ms. Y", and Lindley's hand varies from pure collector -- letters from others are (apparently) reproduced -- to documentor of word-of-mouth. Lindley treats them as true stories. The second book is more of the same, although it also includes stories (according to the foreword) whose provenance is less certain or even unknown.

The first half contains some 22 major groupings (some stories, some collections of letters), comprising some 34 TOC entries; the second half is similar.

So it looks to me like the book I have in hand should be entered as an omnibus of the two other books. But how do I then do those? As collections/anthologies? As something else? If collections/anthologies, what then to do about their content? Do these types of titles have to have content? To the best of my limited knowledge, none of the stories was published separately elsewhere. Obviously, individual letters wouldn't be appropriate, but do I do the major groupings, and, if so, to whom do I attribute them, Lindley? If not collections/anthologies, then novel or nonfiction?

Thanks for any and all help. --MartyD 11:25, 4 April 2009 (UTC)

This is a gray area even though the issue comes up from time to time when entering folklore anthologies edited (sometimes quite heavily) by a professional editor. Should the stories be attributed to "unknown" or to the editor? Somewhat similarly, less savory pulp magazines were known to incorporate "stories of apparently supernatural occurrences from our readers" in their articles, which often seemed as authentic as the famous Penthouse "letters".

In this case, the book is clearly an omnibus of two collections/anthologies and should presumably be entered as such. When entering individual stories, I suppose the easiest way to handle the inherent complexity would be to attribute them to Lindley and explain any peculiarities, e.g. "related to Charles Lindley by Mrs. Y in a letter", in the Notes field. As long as this information is captured, we can always go back and change the attribution if and when we come up with a standard. Ahasuerus 02:36, 7 April 2009 (UTC)

I did enter them that way. FWIW, I stumbled on another omnibus issue of the same stories, The Ghost Book of Charles Lindley, Viscount Halifax on Locus1, listing as authors the people to which Halifax attributes the stories, even in cases where (according to my book) he wrote down a verbal telling or retelling. --MartyD 11:16, 13 April 2009 (UTC)

Are Roc printing dates legit?

I have entered numerous Roc pubs, which were published by FASA on the same date. FASA is spectacularly lax in it's documentation and I have few to cross check. Though it is suspicious that many of the 'Battletech' first printing dates match Roc. Now I am suspicious of this one which Amazon gives a May 1990 date and it shows a September 1991 date here. I match ISBN but the copyright page is showing First printing August 1987 with a first printing number line. That date would match this from Signet [3]. While it is possible Roc and FASA would print the same time, why would you print a same cover Signet and a Roc. In the FASA cases I am concerned with the 1991 datings, but this one jumps to 1987. Any sagacious comments? I really just have a gut feeling I am being misled. Thanks, Harry. --Dragoondelight 21:56, 6 April 2009 (UTC)

Well, Signet and Roc were successive imprints of New American Library (NAL) and Amazon doesn't do a very good job of keeping track of changing imprints. Checking OCLC, I see that the 1987 printing appeared as a "New American Library / Signet" book and the 1990 one as a Roc book. Ahasuerus 02:19, 7 April 2009 (UTC)

Thst's what I got, but my copyright page says Fist printing August 1987 and first printing number line for a Roc book. Add to this when Roc took over much of the Signet SF role it took over the Signet number system at top spine etc and the changed it to their own letter number sequence and then dropped it. This book has no letter numbering. In fact it is using only ISBN 13. It's number sequence should have been AE or LE 5070. My price is $6.99 and I do not think that matches 1987 either. The Signet example is priced at $3.50 and the Roc example for 1991 is $3.95. My point is this is a case of the written dates are incorrect, but people always believe the written one over something else. What do I do? Should I enter the data as given and add notes saying Roc is bunking it's data or should I guess/unknown the data field and note that what was given is erroneous? As you see, the more I look at it the more obvious it is that they are not following our rules. LOL It all reminds me of the old record scam of not changing the data and selling product under the same agreement. Millions of records sold with no way to check if they paid taxes or royalties. Again LOL. Thanks, Harry. --Dragoondelight 11:47, 7 April 2009 (UTC)

I added my copy with an unknown date and notes. Please check. Thanks, Harry --Dragoondelight 12:00, 7 April 2009 (UTC)

Author credited solely by Title?

I have run into another situation I need help with/guidance on. In both of the Lord Halifax's... collections is an introduction by his son (Edward Frederick Lindley Wood), who succeeded him as Viscount Halifax after his death. The introductions are credited to "Viscount Halifax, K.G." (and are signed "HALIFAX"). No mention of the son's name is made anywhere. I know we're supposed to follow the book's credit, but do we really want a title recorded as an author? Do I take liberties here and use the son's name instead? Do I credit the title and make that a pseudonym? --MartyD 01:08, 9 April 2009 (UTC)

I think this settled through the submission, but just in case anyone reads this: MartyD entered the piece as credited ("Viscount Halifax, K.G.") and made that into a variant of the author's true name. That would follow all the current ISFDB standards. MHHutchins 20:18, 13 April 2009 (UTC)

Yes, even though it isn't a role, it seemed analogous to "The Editor", so I went that route. --MartyD 02:14, 14 April 2009 (UTC)

Out of curiousity: out-of-print books as eBooks?

I was wondering if this idea has already been done. I was just noticing how films would lose their copyright, would thus become public domain (i.e. Night of the Living Dead) and would easily be accessible through Archive.org or distributors at a low price (in NOTLD's case: DigiView being sold for a dollar). Would this hold true then to out-of-print books and the possibility of their appearing in eBook format?--DrWho42 18:03, 17 April 2009 (UTC)

We have too different ways of documenting Project Gutenberg ebooks. See this example. Most of those that are not in the database have a tag which begins with the first letter of the author's last name followed by "pg". See this example. My page has a section which links to those titles. There are also public domain ebook entries in the system which are the product of legitimate publishers. We generally try to avoid the entry of illegally produced publications.--swfritter 18:26, 17 April 2009 (UTC)

Perhaps I misunderstand the question, but are you suggesting that any book out-of-print could be subject to e-book publication? Out-of-print is not the same as out-of-copyright. There are hundreds of thousands of books that are not currently in-print, but are still copyrighted. We're talking apples and oranges. I'm wondering which fruit you're referring to. Or are you thinking about the hybrid that Google Books is concocting in its laboratory? MHHutchins 19:06, 17 April 2009 (UTC)

Since the first sentence mentioned public domain I was assuming the question pertained to public domain ebooks? If that Google Books project goes through there may well be perhaps millions of out-of-print books of all types available for publication although those that are under copyright won't be free because authors and publishers will still need to be reimbursed.--swfritter 21:37, 17 April 2009 (UTC)

Mhhutchins answered my question. I was wondering if there was a project similar to what I was talking about, but apparently Google Books have been propounding the idea.--DrWho42 16:44, 18 April 2009 (UTC)

Infernally Intertwined Pseudonym Relationships

William H. Keith, Jr. and his brother Andrew Keith have collaborated many times, both acknowledged and unacknowledged in several books. I have two situations in three books that I would like to run by the vocal crowd here to see if I am falling into any pitfalls. My concern is that if I create the variants below, then each author will show up in the list of 'Used These Alternate Names' and in the 'Used As Alternate Name By' fields in a normal summary bibliography for each other (thus making the waters very murky). Two questions... am I understanding the results of these edits correctly? Is documenting the fact that they are brothers on the Wiki sufficient to help keep things clear? Kevin 02:30, 20 April 2009 (UTC)

March or Die and Honor and Fidelity were published with credit to both brothers. One has been verified. William's website indicates "My brother Andrew wrote these three. All I did was help design the aliens, but someone goofed and stuck my name on two of them as well." speaking about the 'Fifth Foreign Legion' series of books. I believe this is enough documentation to variant these two books and create a single author variant pointing at Andrew.
Blood of Heroes is currently indexed written by Andrew and verified, but per Williams website he indicates it was 'with J. Andrew Keith', and he claims it as a book. I believe this is enough documentation to create a dual author variant pointing to both William and Andrew.

Thanks Kevin 02:30, 20 April 2009 (UTC)

Bump! - I'm going to go with the above if no-one comments soon. Thanks Kevin 03:40, 25 April 2009 (UTC)

It's been done before (someone crediting authorship based on an author's website claims), just be careful that the pub records themselves aren't changed, only the title records. And well document the source for authorship. Go for it! MHHutchins 17:04, 25 April 2009 (UTC)

Forgot to add: I know what you mean about "incestuous pseudonyms" but when it's between two authors who actually are brothers your metaphor becomes, shall we say, "too close for comfort"? MHHutchins 17:07, 25 April 2009 (UTC)

That's what makes it actually appropriate (in an inappropriate way). Titled 'adjusted' for appropriateness. - Kevin 17:38, 25 April 2009 (UTC)

Dontcha just love the English language! What's better - the Eskimos having fifty words for snow, or the English use of the same word for such varied and nuanced meanings and interpretations? MHHutchins 18:12, 25 April 2009 (UTC)

"Eskimos have fifty words for snow" is about as valid a claim as "Europeans have fifty words for 'dinner'". And it pales into insignificance when apparently the word "set" has four hundred and sixty-four definitions in the Oxford English Dictionary. (Another internet "fact" - I don't own a copy, I lean towards the Cambridge University definitions. Modern upstarts though Cambridge are.) English-speakers can out-confuse Eskimos/Inuits/(whatever the PC term is) any day! BLongley 23:24, 25 April 2009 (UTC)

I have a copy right here, but I can't confirm the exact number since the definitions are not numbered sequentially. Still, the whole thing covers 25 pages, so I wouldn't be surprised if 464 was about right. Ahasuerus 23:48, 25 April 2009 (UTC)

Those lazy Oxford guys and numbering... ;-) I believe my father has at least one edition of a/the fuller OED, though he qualified from Cambridge. "Chambers" is the authoritative British source for "Scrabble" though, so I have a few of those. Not much point using any dictionary here though when SF tends to create the words and the dictionaries catch up later. BLongley 00:20, 26 April 2009 (UTC)

Ghost Authors, but not on the 'due to review' list?

I have been trying to 'complete' [William H. Keith, Jr.'s] bibliography. I found something in the database that I don't know how it got there. Two of his non-genre pseudonyms H. Jay Riker and Keith Douglass are already present and marked as pseudonyms, but with no titles present. The only scenario I can imagine that could create this scenario is that some of his non-genre works were previously in the database, and the pseudonym properly marked, and the works were later deleted leaving the pseudonyms defined for an orphan author. Two question... Anyone know of a different way (I already checked the Authors in only due to reviews list) that this situation was created? Any reasons not to enter this info and mark the works as non-genre?.. at 50 or so novels in plus assorted short fiction, I felt that he was above the 'uncertain' threshold for a complete novel length bibliography. Thanks Kevin 03:42, 20 April 2009 (UTC)

Bump! - I'm going to go with the above if no-one comments soon. Thanks Kevin 03:41, 25 April 2009 (UTC)

The only way I can think that these pseudonyms were created was with a publication record. But if the pub records were deleted the pseudonyms would go with them. There must be some "hidden" titles. (Don't ask me how to get to them. I don't know how to get to the records for reviews of titles by non-db authors!) I say go ahead and add the non-genre titles under these names. MHHutchins 17:02, 25 April 2009 (UTC)

Hm, let me check my local copy of the database... Ahasuerus 17:13, 25 April 2009 (UTC)

I can't find anything in the database that would reference these two Author records. I could try searching the main submissions file, but it's huge and unwieldy and probably not worth the effort in this case. My best guess is that there is a bug in the code that deletes pubs or titles and it doesn't always delete Author records marked as pseudonyms when the last pub/title is deleted. I concur with Michael re: adding the non-genre stuff for these two pseudonyms. Ahasuerus 20:10, 25 April 2009 (UTC)

As creator of the "Authors that only exist due to reviews" list, I wish I could do more than confirm that that isn't the issue here. But I wish I knew what was. I could experiment, but that would possibly only create more problem authors. For now, I agree with Michael and Ahasuerus - add the Non-Genre. If anyone deletes them later then at least we'll have some more info on the problem. BLongley 22:59, 25 April 2009 (UTC)

I could add one and then delete it to see if that removes the author? Kevin 23:02, 25 April 2009 (UTC)

Sure, it seems harmless enough. My latest theory is that the system works differently when deleting the last Title associated with an Author record and when deleting the last Publication (if it's a stray pub with no associated Title.) Probably worth a try too. Ahasuerus 23:14, 25 April 2009 (UTC)

While deleting some manga authors pages I found the strays only appeared after all the pubs with title records were gone. After deleting the stray pubs the author then disappeared. One stray author I found only by looking from the direction of the Publisher. How this would work when there's a pseudonym involved I couldn't say.Kraang 01:34, 26 April 2009 (UTC)

I can confirm that for H. Jay Riker, the record and the psuedonym connection disappeared after adding a test pub, deleting pub, and deleting title, so there were no stray publications, it was a stray author. This seems to confirm the idea that somewhere there is a bug in the author deletion process. Thanks Kevin 16:34, 26 April 2009 (UTC)

Bogus Poe pub?

Would anyone care to offer an opinion about THNNRRTVFR1856? I think it is a bogus publication. I believe the real publication is The Select Works of Edgar Allan Poe, Volume II, and "Pym" and "Eureka" are but two of the entries (or perhaps are the sole entries) in it. Here is the OCLC entry cited in the notes. This is documentation I found on the Edgar Allan Poe Society site:

The Select Works of Edgar Allan Poe, (with Griswold's memoir of Poe and an excerpt of that by N. P. Willis) 2 volumes bound as one, Leipzig: Alphons Durr, 1856. (This edition was printed as volume XIII of the Standard American Authors series, published under the superintendence of Dr. Karl Elze. It includes a selection of Poe's poems and tales, with "The Narrative of Arthur Gordon Pym" and "Eureka." A brief editorial note, introducing excerpts from the article by N. P. Willis, states "Dr. Griswold -- if these pages should ever meet his eye -- will certainly pardon the present editor for having ventured thus to tone down a highly colored picture of one of the most gifted poets of America." This comment strongly suggests that this volume was prepared without Griswold's knowledge, although the text is clearly lifted from Griswold's edition as published by J. S. Redfield except for Pym, which does not contain the various changes in Griswold's text and is instead reprinted from the American edition of 1838.)

I have not been able to find a definitive listing of either volume. I noticed this pub because it is listed as containing "Eureka", which is not the full title of the essay (it is "Eureka: a Prose Poem", as used in the pub's title). I am inclined to delete this record rather than trying to imagine how to fix it up, but I figured I can't have been the first person to run into something like this.... Thanks. --MartyD 22:06, 25 April 2009 (UTC)

I tend to leave pubs alone unless they really need deleting in favour of something we already have. And even then I'd probably add a suggestion of it being vapourware first unless I really knew better. As this has "The Narrative of Arthur Gordon Pym of Nantucket" and anything with "Nantucket" in the title will cause amusement to limerick-lovers, I wouldn't delete just yet. BLongley 00:00, 26 April 2009 (UTC)

Well, as far as we know, the publication does exist, although the only library that admits to owning a copy is the National Library of Australia, so I don't think we want to delete it. The only reference to the single word version of "Eurika" that I can find is in a secondary OCLC field, which is not supposed to be an accurate representation of what's on the title page anyway, so I would simply merge it with the main "Eurika: a Prose Poem" Title. We should also mention that the contents may be incomplete, which happens fairly often with older books and magazines. Ahasuerus 00:05, 26 April 2009 (UTC)

Their record is for The Select Works of Edgar Allan Poe, Vol. II, not The Narrative of Arthur Gordon Pym / Eureka: a prose poem as used in the pub title, although their record does include that as either a sub-title or a description (hard to tell -- do you know how to interpret the "|p" record?). Anyway, I think I will change the pub record's title to match that (making the existing title a subtitle for now, if the field is long enough), add a note, and then merge "Eureka". Coming soon to an approval queue near you. Thanks for the help. --MartyD 11:45, 26 April 2009 (UTC)

Sorry, that's what I meant to write, but I wasn't very clear. As far as interpreting bibliographic codes goes, library catalogs mostly use the MARC21 format or one of its relatives. The Library of Congress has a page which links to detailed descriptions of each field and sub-field that are allowed in MARC21 records. Field 245 is the title field and the "p" sub-field is "Name of part/section of a work". Ahasuerus 17:36, 26 April 2009 (UTC)

Actually, I misspoke about "Eureka", so no one get excited. I meant make it a variant -- it did appear published that way, so I have no way to know what's actually used in The Select Works. --MartyD 12:19, 26 April 2009 (UTC)

A follow-up on this pub: It has no title record that I can find. Before I go editing anything, how does one fix that? Thanks. --MartyD 12:19, 26 April 2009 (UTC)

You edit the publication and add the Omnibus as new content. (Same title and author as the publication.) I would guess this was entered as a NOVEL originally then half-changed to an omnibus - i.e. the Pub type was changed but not the title type. BLongley 13:24, 26 April 2009 (UTC)

Authors that won't delete

This is just on example of an author that after all titles are gone won't disappear Seth Green[4]. I've tried merging them with other author names with the same problem and they still persist. Any ideas on why they appear?Kraang 18:38, 26 April 2009 (UTC)

Maybe try updating the author.. with a fake birth year say 2009 .. to make the system 'rewrite' the author record. Then add and delete a test title again? Then again this [Seth Green] did at least have two published works - Maybe someone thought it was SF (hah) - Kevin 19:34, 26 April 2009 (UTC)

Tried that and a few other things, nothing worked.Kraang 20:48, 26 April 2009 (UTC)

I just checked the last backup (which I have installed locally) and this author is linked to Title record 774393. However, Title record 774393 doesn't exist, which suggests that there is indeed some kind of problem with Author deletion when the last associated Title/Pub record is removed. Let me write a simple script to see how many "dangling" title records we have in that table... Ahasuerus 19:39, 26 April 2009 (UTC)

And the answer is "there are 1,093 deleted Title records linked to Author records". 204 Author records exist solely due to these non-existing titles. Now to figure out how to fix the problem... Ahasuerus 20:24, 26 April 2009 (UTC)

If you find a way and it requires a one at a time approach I'll do the grunt work.Kraang 20:48, 26 April 2009 (UTC)

Oh, a volunteer! Excellent! <he said rubbing his hands> Ahasuerus 21:18, 26 April 2009 (UTC)

Foreign editions (translations) of English novels

Hi.

If I understood this page correctly foreign editions should be added as new pubs of the original (English) novel title. But the new pub's title should be the translated one - as written on the book. Did I get this right?

What about short story-collections that are complete translations of existing English collections? I'm talking especially about collections containing exactly the same stories in the same order - all by the same author.

I've found not many of such pub entries, so I'm wondering whether or not this is of any interest for other users. I do not intend to flood the DB with German pubs - my spare time is pretty limited at the moment and besides I don't own that many books at all. But I could add one from time to time if you want. Phileas 22:10, 29 April 2009 (UTC)

Add the foreign language editions under the parent title record, which is usually the English-language edition. For example, if you were entering the German-language edition of Arthur C. Clarke's The Fountains of Paradise you would go to this page and click "Add Publication to This Title". The system automatically enters the title and author, so you'll have to overwrite the title field with the title of the book as it appears on the title page of the German-language edition. Then continue to fill in as many of the fields as you can. Do the same thing for collections but with one big exception: do not create new title records for contents (such as stories or essays). When you start adding new publications, and if you have any further questions, don't hesitate to ask. MHHutchins 22:36, 29 April 2009 (UTC)

Forgot to add: the translator goes in the note field, not in the author field which is for the author of the original title. MHHutchins 22:37, 29 April 2009 (UTC)

Okay, thanks. That's about what I had in mind. Phileas 23:00, 29 April 2009 (UTC)

Stuart David Schiff and Whispers

I'm seeing some unusual behavior with Stuart David Schiff and with a particular issue of his magazine Whispers

Schiff appears to have two records here and here. It's only when I cut and paste the links that I was able to discover what the difference is between the two, i.e. the first record has a space preceding his name. Is there a preferred way to fix this, or do we need to manually edit each record removing the leading space? --Rtrace 19:59, 9 May 2009 (UTC)
I checked and his is the only author record starting with a space. I could write a script that would identify all Title records for the bad version of the name and create a submission correcting it, but since it's just one author, it's probably faster to change them manually. I have another script that finds suspicious Author records (e.g. two spaces in a row), but I haven't run it lately since I have been busy working on Fixer. Soon, though... Ahasuerus 20:48, 9 May 2009 (UTC)
"two spaces in a row" is something that we can search for here already, it doesn't need a script. I do that for authors and titles occasionally. Particularly after a lot of copy'n'paste submissions where the spaces may not have been obvious in the source. "No space after a period" is doable too, but takes 26 searches and you have to let off people like "N.A.S.A.". BLongley 23:40, 9 May 2009 (UTC)
There are a number of things that you can do with the regular search, but my script is much more sophisticated. It searches for initials without a period after them, all-uppercase names, etc. I just need to find some time to clean it up and get it to work with the latest version of my databases... Ahasuerus 01:03, 10 May 2009 (UTC)
No negative criticism intended, I just want to separate what we can do already from what we need experts to do. Hopefully I'll become one of the experts one day, if the super-duper experts don't fix it all before I get a chance . BLongley 01:30, 10 May 2009 (UTC)
Not a problem, I always assume that written communications make comments sounds harsher than they were intended :) Ahasuerus 17:42, 12 May 2009 (UTC)
I'll start off by admitting that I don't fully understand the EDITOR record yet. Especially since this help page admonishes us not to use it and says the record type is going away. However, Schiff has a title record of type Editor for Whispers #10. The title record has a pub named Whispers #11-12. What I suspect is going on here (based on what I've been able to glean about the EDITOR type from the help pages), is that every magazine has an EDITOR type, and in this case they are mismatched. Being that the title record has an award attached to it, and the pub record doesn't have any contents, it would seem that the easiest way to fix this situation would be to edit the publication to match the title record, i.e. make the publication Whispers #11-12 into Whispers #10. Whispers #11-12 could then be added as a brand new magazine creating a new pub and title (EDITOR?) record. Happily, I've got both the issues at hand (and I'll add the date while I'm in there). I just want to make sure I'm approaching this the right way.

Thanks.

--Rtrace 19:59, 9 May 2009 (UTC)

To fix 1, editing each record is safest. The completely wrong way would to be to edit the author with the leading space to remove the space - you'd end up with two identical authors and ISFDB would hide one of them. The middle way would be to do an author merge - but that screen doesn't show which one would be kept, and there's a variant of one of them. So I'd recommend record-by-record if you're going to be working on some of his publications anyway. BLongley 20:46, 9 May 2009 (UTC)

As for 2: yes, changing #11-12 to #10 will be OK. EDITOR is a sort of Title record for Magazines, but it gets used in different ways: for instance, we'll often have the same EDITOR record in multiple Magazines. (Not like here, though.) We sometimes group long runs of Magazines by merging all the EDITOR records for a given year together and retitling it with a more generic name: for instance, see 999627 - one editor record for an entire year of Interzone, which makes David Pringle's page a bit more manageable. BLongley 20:46, 9 May 2009 (UTC)

That Help page was rather misleading, so I have changed it to explain that EDITOR Title Types are (typically) not entered manually and that our software creates them automatically for new Magazine publications. Ahasuerus 20:55, 9 May 2009 (UTC)

I merged the two author records for Schiff (had not seen this discussion at the time), but when there was another submission for another issue of Whispers, all of the new records created another EXACTLY-NAMED author, which I merged again. What's the deal with this name and why do new records create a new author and not merge with the existing author? MHHutchins 22:22, 9 May 2009 (UTC)

Just created a test pub and it created a Stuart David Schiff. Freaky bug or what? MHHutchins 22:27, 9 May 2009 (UTC)

I believe the Author Merge screen doesn't display leading spaces (if any) in the author's name, so when you merged the regular "Schiff" with the space-enabled Schiff, everything got moved to the space-enabled version. I first changed the "spacey" version to "test Stuart David Schiff" and then merged it with the regular Schiff and everything seems to be OK now. Spaces -- leading, trailing, duplicate, etc -- are troublesome since the software doesn't display them well, which makes it hard to see the underlying problem. Ahasuerus 22:34, 9 May 2009 (UTC)

Yep, the fact that the author-merge wouldn't show which author would be retained is why I advised against it. I could only find one other author in my last loaded backup with the leading space problem and it turned out to be all blank. So I've changed that to literally "Blank". But now we have authors 37486, 109663 and 112137 with that name. We should probably establish why we're getting such (reviewees by pseudonymous reviewers might be a related problem) as obviously many people don't know how to fix them. And I for one don't know how to find them except from the backups - which I'm reminded I should refresh while ISFDB is up. BLongley 23:01, 9 May 2009 (UTC)

Which reminds me that I posted the latest version of the ISFDB backup file a couple of hours ago :) Ahasuerus 01:03, 10 May 2009 (UTC)

OK, I shall discard the one I downloaded three hours ago. :-/ What is the schedule these days? Given the limitations of the Web API, I'm scared to submit anything automated from the backups unless I'm pretty sure they're up-to-date. BLongley 01:21, 10 May 2009 (UTC)

I try to generate a new backup file and upload it every weekend, usually on Saturdays. I have automated much of the process, but the elapsed time is still significant due to the size of the files involved. Ahasuerus 01:26, 10 May 2009 (UTC)

Ballantine in US/UK

As many of you know, I have been trying to teach Fixer to combine data from Amazon.com and Amazon UK to create a single submission. Unfortunately, it's not a simple task since the same ISBN/ASIN can have two different:

prices (to be expected)
publishers (when another publisher distributes the book on the other side of the pond)
authors (different capitalization, different middle initials, etc)
titles (usually due to misspellings)
bindings (different coding conventions)
etc

I have made some progress, but it's a major pain.

My current conundrum is Alexis A. Gilliland's Lord of the Troll-Bats. The ISFDB record states that it was published by Del Rey/Ballantine and that the ISBN was 0-345-37467-3. Amazon UK lists what appears to be the same book -- although it mangles the author and the title -- but the ISBN is 0099192810. According to Publisher:Ballantine, Ballantine's book have been distributed in the UK as by Pan and Futura, but do we know whether they assigned new ISBNs to them? TIA! Ahasuerus 22:10, 9 May 2009 (UTC)

I see the Publisher:Ballantine page was created by DES. I suspect that was just because he was an expert in creating Wiki pages, which most of us weren't when he arrived. I see no evidence that Ballantine have actually been distributed by Pan or Futura - "distribution" doesn't tend to lead to anything noted on the publication here in the UK. Exceptions might be "World Distributors" (of London, or Manchester, depending on how you read title and copyright pages and addresses) and "Titan". But the hybrids he created don't seem to actually exist any more? BLongley 01:01, 10 May 2009 (UTC)

The Amazon.uk listing must be a mistake. OCLC has only one record, the Del Rey/Ballantine 0-345-37467-3. Abebooks.com only has two listings for the 0-09-91 ISBN (and one of them has a non-stock photo of the pub which unmistakably has catalog number 37467). That prefix (0-09) belongs to Legend/Arrow which are/were distributed by Random House (same as Ballantine). Perhaps the US printing was distributed in the UK, assigned a new ISBN, but not overprinted (maybe stickered?). MHHutchins 22:43, 9 May 2009 (UTC)

Both of the Abebooks.com dealers are UK based. MHHutchins 22:44, 9 May 2009 (UTC)

It's been a while since I bought imported new US books (there's little point now most publishers are global), but even when I did the sticker tended to have the same ISBN on the sticker as underneath. Only the price was changed. I don't have this book, but if you can give an example of a more current book I can go check what's available in stores at the moment. BLongley 23:33, 9 May 2009 (UTC)

As to other differences: "different middle initials" may NOT be an error. I've said many times, for example, that Dean Koontz and Stephen Donaldson do NOT always have a middle "R". I don't collect either, but I'm willing to go check some examples if you think the same ISBN is getting different results in different countries. BLongley 23:33, 9 May 2009 (UTC)

Unfortunately, "different middle initials etc" was shorthand for "all kinds of headaches". Here are some random example of the US/UK differences:

Arthur Conan Doyle vs. Sir Arthur Conan Doyle

P. B. Kerr vs. Philip Kerr

Val Thame vs. Valerie Thame

Harry, G. Haberman vs. Harry G. Haberman

William Shatner vs. William Shatner+Judith Reeves-Stevens+Garfield Reeves-Stevens

Betsy Cromer Byars vs. Betsy Byars

R.L. Stine vs. Carolyn Crimi (sic!)

R.l. Stine vs. R. L. Stine

Clark Ashton Smith vs. Robert M. Price (sic!)

Jack G. Voller vs. Voller

Geaorge R. R. Martin vs. Clive Barker+Lisa Tuttle+Ramsey Campbell (sic!!!)

JHHK vs. Alicia Coston

In other words, there are so many discrepancies and they are so pervasive that I can't handle them programmatically. The best I can do is to use the publisher (or some other field if we can find a better one) to guess the book's country of origin, populate all submission fields from that country's store and move the other country's discrepant data to Moderator Notes. The sooner I teach Fixer how to do this with minimal intervention from me, the more sanity I will keep for what appears to be much more interesting work on library catalogs :) Ahasuerus 01:24, 10 May 2009 (UTC)

And for bindings - well, the UK has long had a difference over "trade" versus "mass market" paperback versions. The trade size is actually more "mass-market" here (as in "more generally available") - than the standard paperback size. Bookshops will handle hc, tp, and pb - supermarkets might stock the tp size but not think the pb size with smaller margins worthwhile. Amazon UK doesn't really help unless it gives dimensions - they're just "paperback" normally. I hate it when I get a tp and expected a pb. I've bought some clothes entirely on the basis of whether I can fit a paperback in the pockets. And I'm not burly enough to wear jackets that can cope with tps. BLongley 23:33, 9 May 2009 (UTC)

I might need to clarify that "smaller margins" means "profit margins", not "area around the text that you can scrawl notes in". BLongley 23:52, 9 May 2009 (UTC)

Upon further review it looks like this Amazon UK record was in error, so it has been rejected. Ahasuerus 17:48, 12 May 2009 (UTC)

More Whispers

First, thanks to all who helped correct the Stuart David Schiff records. I was certainly willing to manually edit them, but they were all fixed by the time I got back to do more edits.

I'm not sure if this question is more approropriate for Rules and Standards, but posting it here as a continuation of the previous Whispers discussion. I have three issues of Whispers (#19-20, #21-22 and #23-24) that were issued in a limited, signed and numbered state. Essentially, the perfect bound "magazine" was placed in a cloth binding with a limitation statement and signatures on the end paper. I had intended to treat these the same way I would a book, i.e. clone the "trade" edition making a new pub under the title for the limited state. However, the software won't let me do that. My assumption is that I am prevented because of the EDITOR type of the title record. My questions are:

Should we make additional pubs for variant states of magazines?
If so, is this accomplished by editing the title to a different type, then cloning, then reverting the title record back to EDITOR when done; or some other method? Also, would a pub type of HC be appropriate in this instance (as opposed to Digest).

Thanks for your help. --Rtrace 03:33, 12 May 2009 (UTC)

You're welcome. I created those placeholder records this morning, after coming back and seeing that you'd added some of the missing issues and contents to others.

Yes, these variants should be entered as new records.
Unfortunately, magazines are not clone-able. (Because of the lack of a title reference record for magazine-typed pubs.) But you've already figured out how to get around it. You change the pub's type from magazine to anthology, but remember to change it back once it's been cloned. You'll also have to clean-up the editor records after the new pub has been cloned. This is a multi-step process that will be difficult for a non-moderator, but if you're up to it, I can lead you through it.

MHHutchins 04:09, 12 May 2009 (UTC)

I'm certainly willing to give it a try and I've submitted the first round of edits (changing the pubs to anthologies).--Rtrace 04:16, 12 May 2009 (UTC)

We'll do one together and you'll see if it's too much trouble. I'll lead you through. Next step: add a editor record to the changed to anthology pub. 1.) Edit this pub. 2.) Under "Content" click "Add Title". 3.) In the title field enter "Whispers #21-22", in the type field choose "Editor" , in the Author1 field, enter "Stuart David Schiff". 4.) Submit

MHHutchins 04:20, 12 May 2009 (UTC)

Done, though that's not what I expected the next step to be (thus the delay)--Rtrace 04:25, 12 May 2009 (UTC)

You're right. I screwed up. Instead of an editor record, I should have told you to create a title reference record. Let's do this again. 1.) Edit the record. 2.) Under contents choose "Add Title". 3.) Enter "Whispers #21-22" in the title field, choose "Anthology" in the pull-down type, and "Stuart David Schiff" in the Author1 field. MHHutchins 04:31, 12 May 2009 (UTC)

Now it's clone-able. Do that and when we've created the new pub, we'll go over cleaning up the editor records. MHHutchins 04:36, 12 May 2009 (UTC)

Clone submitted--Rtrace 04:40, 12 May 2009 (UTC)

You'll have to change the new pub to "hc", add the price if you know it, and change it's type back to "magazine". Also change the type of the original record. MHHutchins 04:47, 12 May 2009 (UTC)

Done. Alas, I don't have a price for them.

Next step: Remove the extraneous title record. Choose "Remove titles from this pub", and then check the box for "Whispers #21-22, ANTHOLOGY, Stuart David Schiff". Do this for both pubs. MHHutchins 04:55, 12 May 2009 (UTC)

That's it. If you look at the editor record on Stuart David Schiff's summary page for "Whispers #21-22", you'll only see one title listed. But if you go to one of the stories, The Bone Wizard for example, you'll see two pubs, one digest and the other hardcover. Do you feel like doing the other two issues now? MHHutchins 05:10, 12 May 2009 (UTC)

(unindent)If you don't mind approving them, I'll start submitting the next steps.--Rtrace 05:15, 12 May 2009 (UTC)

Two Titles with the Same Pub

I have been trying to clean up this title. When I started the title was listed as an Omnibus with the title Behold the Man and Other Stories. The pubs listed are primarily the novel Behold the Man. I recall that there were 3 pubs listed that were true omnibus (there may have been 2, if my actions caused a third to be created) and my thinking was that it would be easier to unmerge the omnibus pubs, edit the original title to be the novel, and merge the new titles created by the unmerge. This last step is where I'm running into a problem. When I attempt to merge this title with this other title, I get a warning because they both contain this pub. I'm not sure what the best way to fix this would be. Should I ignore the warning and proceed with the merge cleaning up any issues after it is complete? There is also a third title to be merged and the shared pub seems to be missing its eponymous novel record, but I know how to handle these and will do so after I figure out the merge.--Rtrace 05:04, 15 May 2009 (UTC)

You've run into the "one pub with two title records" situation. You'll need to remove one of the title records from the pub, and because title records for collections, omnibus and anthologies are hidden when edited you'll have to change the type of one of the title records. Doing this will cause it to show up on the list of removable records when you use the "Remove Titles from This Publication Tool". I usually change one to a POEM so it stands out from the other records (unless it's a book of poems!) Choose either one of the records and then "Edit Title Data", change the title type from OMNIBUS to POEM and submit. Once you've removed the record you'll have to delete it from the system as it will have no pubs associated with it. (It will also stick out like a sore thumb on the author's summary page!) MHHutchins 13:35, 15 May 2009 (UTC)

BTW this was caused when someone changed the content record Behold the Man to Behold the Man and Other Stories, duplicating the name of the title reference record with a content record. Now that the second one has been dropped from the pub, you'll have to edit the pub to add the novel content record. MHHutchins 14:18, 15 May 2009 (UTC)

Ahh, that explains why the novel was missing. Thanks for your help in cleaning this one up.--Rtrace 15:38, 15 May 2009 (UTC)

Variant behavior 101?

I realized there are far more pressing issues in the world, but I think I need help or an explanation about variant titles. Poe has a bunch of titles that change over time and also change the author credit over time. In the simple case, I have a title, say "Bon-Bon" with author Edgar Allan Poe. I subsequently record "Bon-Bon" by Edgar A. Poe (set up as a pseudonym), and the bibliography display shows me something like:

Bon-Bon (1832) [as by Edgar A. Poe]

I now find an earlier publication of this story has the title "The Bargain Lost". So I make a parent for the existing Edgar Allan Poe "Bon-Bon" title, and now the bibliographic display shows me something like:

The Bargain Lost (1832)
  o Variant Title: Bon-Bon (1832)
  o Variant Title: Bon-Bon (1832) [as by Edgar A. Poe]

Ok, I think I see what happened: making the new parent made it the parent of both existing titles. Fine, but not really right. So I now go and make the Edgar A. Poe "Bon-Bon" be a variant of the Edgar Allan Poe "Bon-Bon" as I had it originally. And the bibiliographic display gives me:

The Bargain Lost (1832)
  o Variant Title: Bon-Bon (1832)

with no mention of the Edgar A. Poe variant. Only if I go to the Edgar Allan Poe "Bon-Bon" title do I see the Edgar A. Poe variant. Everything seems to be inter-related correctly, and the contents of the Edgar A. Poe pub with the Bon-Bon content looks the way I think it should.

Is this set up correctly and just too bad about the bibliographic display? Or should there be just one level of variant, with everything having a common parent, despite the then-duplicated identical titles with different authors?

Thanks. --MartyD 22:37, 16 May 2009 (UTC)

Much of what you've encountered has to do with how the software displays variants, but one thing to keep in mind is that you shouldn't make a variant of a variant. As you've discovered, that will cause various display issues. You must first determine the canonical title record and have all subsequent changes in either title or author be a variant of that ONE record. MHHutchins 22:46, 16 May 2009 (UTC)

More re-reparenting is in my future, I see. Thanks for the explanation. --MartyD 22:56, 16 May 2009 (UTC)

And now a follow-up question. Since I've moved on to Variants 102, it's a multi-parter. I was under the impression that when dealing with a pseudonym, we want to record a work as published under the pseudonym, then make a variant of that same title under the canonical author even though the title might not ever have been published using the canonical author's name. Is that understanding of what should be done correct? If so, do we still want to do that in a case such as the one above, where now we end up with three title records: the parent title w/canonical author, the actual title w/pseudonym, and the actual title w/canonical author -- neither linked to any publication nor directly to the title record that led to its creation? Or should the actual title/canonical author piece be omitted (or deleted) if the actual title/pseudonym is going to be a variant of a different title? (I hope this makes sense). --MartyD 23:41, 16 May 2009 (UTC)

It's already established that "Edgar Allan Poe" will be/is the name of the canonical author record. All pieces (stories/poems/essays) will have a parent record as by "Edgar Allan Poe" even if there is no publication with that byline (unusual but possible). Next we must establish the canonical title of every piece. That will become the name of the parent title record. Every publication of that piece will be listed when a user clicks on the parent title record. If a piece is published under another authorship or another title, a variant will be created that will link it back to that one (parent) title. Is the first appearance always the parent title? Not necessarily. You must decide for the database users what the parent title will be, but common sense and a general knowledge of an author's work will be your guidelines. If "The Tell-Tale Heart" first appeared as "Revenge of the Murdered Soul" you wouldn't make that first title the parent title record, because 99% of users will look for "The Tell-Tale Heart". If the story was first published as by "Edgar A. Poe", a parent title should be created and have "Edgar Allan Poe" as the author. If we wind up with three (or ten) variants it shouldn't matter because publications of the story will be listed when clicking the parent title. To answer your last question, it depends on what you call the actual title. As the editor you decide what the parent title is, even if there has never been a publication with that story/author combination. Once you've decided what the parent record is, the system won't allow you to delete it if variants exists. You'll have to break the variant relationship first (give "0" as the variant record number.) Hope this answers your question. MHHutchins 00:46, 17 May 2009 (UTC)

Becoming a Professional Writer by Way of Southeast Asia

This is in "The Butcher's Bill" it is corrected from this "Introduction: Becoming a Professional Writer by Way of Southeast Asia" which is here [5] . Obviously, "Foreword: Becoming a Professional Writer by Way of Southeast Asia (The Complete Hammer's Slammers: Volume 1)" was the variant, but it should now go to the corrected [6] . After deleting the Introduction version, I submitted an "unmerge" for all three components, under the assumption that once they were separated all I would need do is make the others a variant. Since this was 'rejected', what is the least entangling method to right the situation? Thanks, Harry. --Dragoondelight 12:59, 19 May 2009 (UTC)

Looking at the reject list, the reason given is that the field was empty. This means you were under the wrong variant when you submitted the unmerge. (There are no pubs under this title record.) It's actually better not to unmerge contents from publications. Use the "remove titles from this pub" tool. The "unmerge" tool should only be used to separate pubs from titles. This isn't clear in the help pages and the function should be disabled for content removal. (That's an opinion.)

First thing to do is to break the original variant relationship. Go to either title record Introduction: Becoming a Professional Writer by Way of Southeast Asia or Foreword: Becoming a Professional Writer by Way of Southeast Asia (The Complete Hammer's Slammers: Volume 1) and choose "make this title a variant" and enter "0" (zero) in the parent # field. BTW, this last title's appendix is entirely superfluous. If an introduction has a title it isn't necessary to add the title of the book. A book's title is appended only if it has a generic title ("Introduction", "Foreword", "Afterword", etc.) Once the submission is accepted and the variant is broken, delete any title records that don't have pubs. Then create another variant with the older title becoming the parent title. MHHutchins 15:51, 19 May 2009 (UTC)

Thanks greatly. Harry. --Dragoondelight 20:38, 19 May 2009 (UTC)

Braided stories

I was starting to fill out the contents on Busted Flush, and I found this is a more complex work than i had expected. The copyright page lists 10 stories by 9 authors. There is no ToC. The title page calls this a "Mosaic novel". I call it an anthology with briaded stories -- It reminds me of the Merovingen Nights books.

To be more specific, most (perhaps all, i'm not done with the book yet) of the stories included have multiple sections, which are indicacted in different ways. "The Tear of Nepthys" has "The First Tear: Isis" and the 2nd and 3rd. "Double Helix" has several (many) sections of the form "Double Helix: Subtitle in several words". "Political Science" has "Political Science 101", "Political Science 201", etc. "Dirge in a Major Key" has "Dirge in a Major Key: Part I", "Dirge in a Major Key: Part II", etc. Then whe have the storie listed on the copyright page as "Shoulda, Coulda, Woulda" which never appears as an entity on an individual title page, inst4ead three are three stories entitled "Shoulda", "Coulda", and "Woulda" all by the same author.

My inclination is to enter each section as a separate shortfiction title. Note that the separate parts of a story are generally non-contiguous. That is, we have story A, part1, then story B, Part 1, etc, before getting back to the 2nd part of story A, nor is the order or number of parts the same for different stories. If I enter each "story" only once, the page numbers will be highly misleading. There is no way, aside from notes, that I can see to indicate the obviously intended linkage of the individual stories listed on the copyright page. Without notes, even the author credits don't do this, as some authors have more than one credit on the copyright page.

Comments on this situation would be welcome. -DES ^Talk 17:43, 19 May 2009 (UTC)

Looks like R.R. is at it again. Here's what I did for one of the early Wild Cards "novels". For Down & Dirty I did it this way, while someone else did the second printing of the same book another way. MHHutchins 18:07, 19 May 2009 (UTC)

I think the first method looks the best. The second method is a little too complex while the third is a little too simplistic.--swfritter 15:24, 20 May 2009 (UTC)

I'm not sure of your meaning. I only suggested two methods: 1) Each section as a separate work; 2) Each copyrighted story as a separate work, with a single entry, and possibly notes to spell out the pagination of later sections. I cant see what third method would be reasonable. I am inclining to the first of these: each story section as a separate work. I have partly entered this along those lines. -DES ^Talk 15:55, 21 May 2009 (UTC)

I was referring to MHHutchins' examples.--swfritter 17:03, 21 May 2009 (UTC)

The first two examples were really the same method. (It's just that the stories are even more interwoven in the second book.) The last example I gave was how another editor did a later edition of the second example. So I think you're saying that you preferred my method, recording exactly how the stories are broken up with the book. Right? MHHutchins 17:39, 21 May 2009 (UTC)

I must admit I pretty much went cross-eyed at the second example, hopefully worst case. If they were as simple as the first example I would say a very positive yes. The biblio pages for the authors are going to look a little odd but I can't think of any other way to accurately display the the contents.--swfritter 18:35, 21 May 2009 (UTC)

I wound up entering the various story segments as separate stories, with notes. See the results at Busted Flush. I considered using series to unify each braided story, but there are already a good many sub-series for WildCards, and i don't know this series well enough to be sure i wasn't messing things up. -DES ^Talk 17:18, 16 June 2009 (UTC)

How to record a standalone Essay?

I've come across the original Eureka: A Prose Poem by "Edgar A. Poe". In author Edgar Allan Poe we already have the title Eureka: A Prose Poem and some variants as Essay, due to its publication in various collections. Essay seems to be the appropriate type for the content (see this scan if you want to judge for yourself). But this version of the work was published standalone, and there is no pub type of Essay. What's the preferred way to record this? Make a Nonfiction pub and make the title type be Essay (if that's even allowed)? Make a non-fiction pub and title and change the existing title types? Something else? Thanks. --MartyD 10:58, 21 May 2009 (UTC)

A nonfiction work can contain one or more essays, just like a collection or anthology contains stories. look at ISOWN1967 for an example. Or you could use the CHAPTERBOOK type for this, or make it a one-item collection. Having looked at the beginning of this via your scan link essay is reasonable, although there is a lihtly fictionalized cover which would support an argument for "shortfiction". I would myself incline to go with a non-fiction publication, containing 1 essay. But other 4 methods could well be defended. -DES ^Talk 15:50, 21 May 2009 (UTC)

I agree with DES assessment that NONFICTION would be the best approach. Using this type gets you around the CHAPTERBOOK pitfall (reverting to ANTHOLOGY if a subsequent edit is made on the record.) MHHutchins 17:45, 21 May 2009 (UTC)

Poe collection pub title and type

Moderators everywhere will be thrilled to know I'm up to 1850's The Works of the Late Edgar Allan Poe. This was a four-volume set, with the first two published nearly together in early 1850, the third in the fall of 1850, and a fourth in 1856. The title page of the first volume has this structure:

                  THE WORKS
                   OF THE LATE
              EDGAR ALLAN POE:
                     WITH
         NOTICES OF HIS LIST AND GENIUS
                      BY
 N. P. WILLIS, J. R. LOWELL, AND R. W. GRISWOLD
                IN TWO VOLUMES
              ------------------
                    VOL. I
                 T A L E S.
              ------------------

The "by" credit is somewhat misleading. This is really a Poe collection. Willis and Lowell contributed introductory "notices", essays about Poe. Griswold is the editor. What do we think about a title for this pub? The most meaningful title would probably be something like The Works of the Late Edgar Allan Poe, Vol. I: Tales. FWIW, the second volume's title page is the same, with "TALES" replaced by "POEMS AND MISCELLANIES". The third volume is completely different (doesn't even refer to "Works"), and the fourth is similar but has an additional "with ... by ..." for Griswold and drops his name from the "Notices" part and just uses "IV." with no "Vol." label. I will make canonical title records in the "..., Vol. X: xxxx" format. I'm wondering to what extremes I should go in recording the pub's title as printed on the page. (Also, should this be an omnibus because of the introductory, non-editor essays?) --MartyD 11:49, 21 May 2009 (UTC)

Lots of more recent SF single-auhtor collections have introductory essays by someone other than the author, and are not changed from collections thereby. Look at TBSTRNDL198 for an example. If there were fiction by another author I would make it an anthology, but not an omnibus unless at least one of the pieces was ov novel-length, or had been previously published as a separate work.

I read the title page as having the "by" apply to "notices" anyway, si don't find it too misleading. I would standardize the titles, probably as ''The Works of the Late Edgar Allan Poe, Vol. nn: subtitle but it is up to you. You cna always add notes about the exact form of the title page if you think warrented, or even a link to a scan if one is freely available online, or if you upload one to the ISFDB. Anyway, that is my personal view. -DES ^Talk 15:40, 21 May 2009 (UTC)

Script request

I've been (randomly) going through the database looking for magazine records that don't have editor records. Is it possible that a script can be written that can find such records and a list generated for a side project for those of us who have so little to do? :) Thanks in advance. MHHutchins 17:31, 22 May 2009 (UTC)

I've done one of those already. And fixed a few years of titles, so I'll exclude those here as the backup is a bit old. And I'll exclude the unknowns so you'll have some clue as to which editor to put in.

select DISTINCT pu.pub_id, pu.pub_tag, pu.pub_title 
from pubs pu, pub_authors pa, authors a
where pu.pub_id = pa.pub_id
and pa.author_id = a.author_id
and a.author_canonical != 'unknown'
and pu.pub_title NOT like 'The New York Review of Science Fiction%'
and pu.pub_title NOT like 'Omni%'
and pu.pub_title NOT like 'Interzone%'
and pub_ctype in ( 'MAGAZINE', 'FANZINE' ) and
         not exists (select * from titles t, pub_content pc
                     where pu.pub_id = pc.pub_id and
                           pc.title_id = t.title_id and
                           t.title_ttype = 'EDITOR')
order by pub_year;

BLongley 18:10, 22 May 2009 (UTC)

Page created, with hyperlinks to make it a little easier. Feel free to break it up into smaller chunks. BLongley 18:42, 22 May 2009 (UTC)

Thanks for the fine job. I've broken it into sections in case more than one editor wants to work on it at the same time. I've also marked those pubs that I'd already created editor records for, assuming that you created the list from a version earlier than yesterday morning. Thanks. MHHutchins 23:01, 22 May 2009 (UTC)

Lost Publication

I cloned DMSDYWPN1979 in an attempt to create a record for the 4/79 printing (ISBN 0-523-40566-9). The new pub appears to be listed correctly on the title page: 10599. However, clicking through to the publication takes you to a publication for a different title by the same author DCTRWHNDT22222222223333333338197. Clearly, something has gone awry here. There is also an introductory essay 1006353 which also appears to have the correct pub on the title page, but links to the incorrect pub. Any thoughts on how to fix this? Thanks.--Rtrace 22:25, 22 May 2009 (UTC)

You have encountered a problem I was aware of and fixed a lot of (but obviously not enough). You may have noticed that Publication tags vary a lot - e.g. some are "BKTGnnnnn", some are "DCTRWHNDTxxxxxx" - at times the publication Tag is created as the beginning of the title, with vowels removed, capitalised, then a unique identifier added (involving the year, maybe). What you have encountered here is a Pub Tag at maximum length (32 characters) so when you clone it doesn't create a Unique Tag. So a lot of the links that work with publication Tags rather than Publication IDs (which are unique) are picking up the first Pub they find with that Tag. BLongley 22:51, 22 May 2009 (UTC)

The temporary solution is to edit the Pub Tags to something unique and shorter. I suspect not just in this case, but for all Pubs presently with Maximum Length Tags. I think the huge number of "Doctor Who and the" pub titles meant something broke and created some really weird long Tags (there's definitely not 22,222,222,223,333,333,338,197 of them) so the Tag generation was changed in new pubs but not in clone pubs. I feel another project coming on.... :-/ BLongley 22:51, 22 May 2009 (UTC)

Hmmm... it gets weirder. My local ISFDB can't reproduce the problem but gives big Purple Python errors instead. Definitely needs attention, but not tonight. I need to look at how the Tags are being generated, I'm not sure they're coming from the Pub Tag you clone, maybe from a pub under the title you clone or something. BLongley 23:43, 22 May 2009 (UTC)

Oh, no hurry on this. I'm actually off to a convention for most of the weekend, and physically away from the book in question. I'd just set it aside anyway until we can figure out the best way to fix it. I probably should have cloned a pub with a closer title, but I usually look for something close in date and publisher. Thanks for looking into this.--Rtrace 23:48, 22 May 2009 (UTC)

I think I see what it's doing. The "random characters" added to Tags aren't that random and for titles like this there is an extremely finite set of possibilities, a lot of which are along paths that are exhausted. In this case the default tag is DCTRWHNDTH1979 but that's taken. The software then cycles through 20 possible consonants to use as the last letter i.e. DCTRWHNDTB1979 through DCTRWHNDTZ1979 (missing out DCTRWHNDTY1979). These are all taken. Then it gives up using the letters and uses the 'try' number, for DCTRWHNDT201979. That's taken. For some reason it doesn't reset the try number so it doesn't start at 'B' again: and it doesn't recognise that it's added two characters rather than one, so it keeps the first digit and tries DCTRWHNDT2211979. That's taken. Next is DCTRWHNDT22221979, DCTRWHNDT222231979, DCTRWHNDT2222241979, etc. It rapidly runs out at DCTRWHNDT22222222223333333371979. The oddity is that the live ISFDB can then go onto to DCTRWHNDT22222222223333333338197 so there's obviously a limit applied there that isn't in the code from Sourceforge. But we're getting collisions and the tags need to be more random. I'll see what I can do. BLongley 13:49, 23 May 2009 (UTC)

Bug 2795822 recorded and partial fix submitted. BLongley 14:23, 23 May 2009 (UTC)

I seem to recall that Al improved the tag generation logic shortly before he became unavailable. I wonder if that change didn't make it to Sourceforge, which raises another question: what else may be on the live server that is not in Sourceforge? Ahasuerus 21:27, 23 May 2009 (UTC)

It might be time to recover the code from the live server and do some comparisons. Assuming the source is there? BLongley 21:46, 23 May 2009 (UTC)

We can (and should anyway) compare the installed files from the live server to an installation of the SourceForge files. We'll be able to recover any missing source changes from that, even if the source files are not present. --MartyD 01:35, 25 May 2009 (UTC)

All files in the CGI directory have been zipped up and sent to the usual suspects, er, I mean the development team. Ahasuerus 03:45, 25 May 2009 (UTC)

The publications all look good now. Thanks. --Rtrace 18:03, 25 May 2009 (UTC)

Sticky Deletes

I've just submitted a third attempt to delete contents from this pub. With each attempt more titles appear to be deleted, but not all the ones I've marked. In fact, with this latest submission, there were four titles left to delete, all of which I marked, yet it appears in the confirmation XML that only two of the four requested deletes were actually submitted. I'm not precisely asking for help here, but wanted to report that the delete titles from pub tool is behaving oddly. I'll keep submitting until they're all gone. Thanks.--Rtrace 03:36, 26 May 2009 (UTC)

I Do believe you have found some secret limit in the system for contents. Whew 400+ Contents Items. I'll see if Can delete them one at a time for you. Kevin 03:57, 26 May 2009 (UTC)

But is Rtrace trying to remove all of them, in which case Delete Pub would work better, or just a subset? Ahasuerus 04:04, 26 May 2009 (UTC)

Bug Confirmed. It appears that attempting to delete items past 400 in the alphabetical list of contents causes 'problems'. When attempting to remove contents both before and after SOME LIMIT (Presumed to be 400) some of the removals happen and others get dropped. Bug #2: When attempting to remove a single or a multiple item from above SOME LIMIT in the alphabetical list of contents, the Moderator Approval page is Blank for what we are approving. I have submitted a single edit to remove the last two titles you want gone from this Pub, and placed it on hold for the more knowledgable moderators to have a look see at. Thanks! Kevin 04:07, 26 May 2009 (UTC)

I was going to post on the moderator board, but since Ahasuerus is on the case now, I'll just leave it in his capable hands. (And I just noticed that it's not 400 items - it's just up to 400 pages and it felt like 400 items as I scrolled - A quick count yielded 217, so perhaps the problem is deleting more than one at a time above 100, and deleting anything above 200 items in a contents listing. Kevin 04:20, 26 May 2009 (UTC)

No, not trying to remove all the titles, just the ones that I marked "delete". Howard and Lovecraft are listed with their full names in the book, so this is a delete and add with the variant name. Thanks for everyone's help.--Rtrace 04:58, 26 May 2009 (UTC)

The bug/limit is in /edit/submitrm.py - "while entry < 200". There 's a 200 limit on titles, 200 on reviews, and 100 on interviews. Easy fix: just decide what the limits really should be. I guess we need over 500 to cope with things like this. BLongley 09:22, 26 May 2009 (UTC)

Can anyone think of a good reason to have a limit? --MartyD 09:50, 26 May 2009 (UTC)

3 infinite loops would take a long time to run? ;-) BLongley 12:02, 26 May 2009 (UTC)

I will eventually have over 2500 reviews in this pub.--swfritter 13:06, 26 May 2009 (UTC)

Can the code be something like "while entry < maxEntry+1", having previously read MaxEntry from the current state of the DB pub record? or doesn't that work? -DES ^Talk 15:44, 26 May 2009 (UTC)

At one point I asked Al whether there were any limits on the number of Titles in a Pub and he responded that there should be none (although it may take a while to file a pub with 10,000 Titles), so I assume that the limit should be removed. Ahasuerus 17:30, 26 May 2009 (UTC)

Can be done, but it's no longer an easy fix ("change some numbers, make them bigger"), just a small fix if you don't look at potential other (unlikely) consequences. I think DES is on the right lines, but there's three MaxEntry values to derive. Fairly simple SQL - selecting COUNT(*) for a particular title type in pub_content for the pub_id in each case. Unless somebody plans on trying to break the limits anytime soon, just changing the limits to 999 in each case has worked locally fine for me for now. Finding whether there's other limits involved rather than just in this script would take longer, and need some major stress testing. BLongley 21:57, 26 May 2009 (UTC)

Real life example: at work I removed some limits on the amount of expected data from a "calls to a particular telephone number" query. Works fine for 99.9% of all queries. Not so good when you consider how many people call customer services each day, or the emergency services. And I found that it wasn't just the slow-down from massive results that was a problem, I found limits in the Oracle database software itself - more than 64K XML nodes and it broke. (There was a 64k limit on each node too, but that didn't affect us.) But "unlimited" is dangerous. BLongley 21:57, 26 May 2009 (UTC)

Any idea what the high realistic limit is? I will not be entering the entire 2500 reviews anytime soon; it will probably take me more than a year as I am only processing two pages a day.--swfritter 22:26, 26 May 2009 (UTC)

Entering shouldn't be a problem, but I can go check if there's any obvious limit on that. (But I'm not an expert on ISFDB coding in any way yet, I'm strictly "can grab low-hanging fruit" so far.) This problem only affects removing titles. So if you get them all perfectly right, no worries AFAICS. By the time you exceed 999 titles I should know more: if people don't go for the quick fix then worry if you make a mistake after the 200th title. But there's probably workarounds for that too - the API might allow deletions that the Web interface can't deal with. And 2500+ titles in a pub is probably a concern for our bandwidth-challenged users anyway. In this case, Rtrace has a current problem - we could of course give the Tommy Cooper response "It hurts if I do this - Well, don't do that then." But there's a quick fix, an equally quick fix for your future problem (let's make it 9,999), and a future-proof fix. I know my local server doesn't worry about counting to 200, 200 and 100 much, and 999, 999 and 999 are fine. But only I use that. Can our live server cope with many users doing such? BLongley 23:06, 26 May 2009 (UTC)

I can propose a workaround that I believe will work: Temporarily change the titles to something that comes earlier in the collating sequence (prepend an underscore?). Then delete (since they now sort within the limit). Then revert the titles to what they should be. I haven't tried that yet (nor have I done the make variants that are still required) so that I don't muddy the situation while folks are still looking into it. I've got it sitting in my "more to be done" pile and it can sit there for a while longer if needed.--Rtrace 02:16, 27 May 2009 (UTC)

Poe's collected reviews -- content volume handling

The bulk of The Works of the Late Edgar Allan Poe,Vol. III is a series of short-ish passages, mostly identified by a person's name. See [7]. The first group of 3+ dozen was published as a 6-part series ("The Literati of New York City"). These are presented with no over-arching title, each having a separator and the person's name in small-relative-to-other-headers font. They are sketches of individuals. The next group of another 3 dozen or so is reviews. Some are of books or other substantial works, some are of smaller works such as poems. Except for the fact that they were written by Poe, they have nothing to do with SF. They are mostly titled using the reviewed author's name, sometimes a bit more, in normal header-sized font; starts can be in the middle of a page, however.

Rather than creating separate content entries for the sketches (which were never published separately), I took a small liberty and made a single "The Literati..." content entry to cover those, documenting it as untitled in the notes. (Here's your chance to object!)

Does anyone have an opinion about the reviews? Do we want them captured separately? Unlike the sketches, these were published separately elsewhere (although determining when and where is going to be difficult or impossible in some cases). Is there something clever and space-saving I could do that wouldn't be too much of a deviation from ISFDB Standard? Thanks. --MartyD 11:30, 26 May 2009 (UTC)

Well you could, if th4ese are reveiws of non-SF works, as I tke them to be, simply list them as "The reviews" or "Poe's Reviews" or some such as a group, much as you did for "The Literati..." again documenting that this is an invented group title in the notes, and ignore the separate pubs of what are, after all non-ge nre non-fiction. -DES ^Talk 15:48, 26 May 2009 (UTC)