ISFDB:Community Portal/Archive/Archive10

From ISFDB

Jump to: navigation, search

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

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


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



Contents

Lem

What should be done with Lem? I could help, but I see no Polish editions here (strangely enough, though, there are German editions of Lem like [1]). --Roglo 05:02, 17 Dec 2007 (CST)

Well, here's a pretty complete list cross-referencing English and Polish title with other languages, if you want to have a stab at it. BLongley 08:15, 17 Dec 2007 (CST)
Lem is quite popular in Germany and we have (had) a number of German editors, so it's only natural that we would have some German editions of his works.
The most immediate problem with Lem is that we have the literal English language translations of some Polish book titles listed as Variant Titles, which is highly misleading since it implies that books like Astronauci or Człowiek z Marsa have appeared in English. These variant titles need to be deleted and appropriate Notes added to the Polish language titles. The second problem is that in many cases we are missing the original Polish language Publications, which can be identified using the link that Bill posted above. Finally, we will need to sort out various permutations of the Ijon Tichy collections (IIRC, some stories were dropped and some were added in subsequent Polish editions), but that can wait. Ahasuerus 11:14, 17 Dec 2007 (CST)
OK, so the Polish titles are used as canonic titles; I will update them to reflect the full original titles; add Polish publications; remove fake Variant titles and add English translations of the titles of untranslated works as Notes. I should add Contents to collections but this will add a lot of Polish shortfiction, so the translated stories should be made Variant Titles? Do we need one Polish publication per title, or more? --Roglo 11:51, 17 Dec 2007 (CST)
That's right. As per the Help pages, when entering Titles that originally appeared in other languages, we want to make the original Title the canonical one and the English language one a VT. Sometimes there are different English translations of the same title, e.g. Sever Gansovsky's "Den' Gneva" has been translated as "The Day of Wrath", "A Day of Wrath" and "Day of Anger". Come to think of it, Gansovsky's bibliography is a good example of how these permutations can be handled, although I see that there is a misspelled version of his name lurking in the shadows. Ahasuerus 12:27, 17 Dec 2007 (CST)
Thanks! BTW, Lem's first name is Stanisław (slashed l) but I think it was never used outside Poland. --Roglo 12:47, 17 Dec 2007 (CST)
Unfortunately, this is a common problem with many European (Norwegian, Polish, Czech, etc) writers. Circumflexes, breves, acutes, diaereses/umlauts, carons, ogoneks, double acutes, cedillas, strokes, dots (in the middle or above the affected character), etc -- they all get dropped or otherwise mangled by publishers seemingly at will. And then there is Cyrillic :( Ahasuerus 13:02, 17 Dec 2007 (CST)
I understand that we leave the name as it is but give 'proper' Polish titles, not 'transliterations' like Gansovsky's titles? --Roglo 13:54, 17 Dec 2007 (CST)
Right, we use Unicode behind the scenes, so "ł" and other non-ASCII characters are supported. Cyrillic titles are supported by the underlying database, at least in theory, but they mess up searches and break the application in various "interesting" ways, so we had to go back to transliterations. We could probably set up the "l" version of Lem's name as a pseudonym of the "ł" version, but it sounds like a lot of work with relatively little to show for it. Ahasuerus 14:28, 17 Dec 2007 (CST)
Question: when there's options of which Wikipedia to use, do we want the English one or the one most suited to the title language? I've held a submission that changes "http://en.wikipedia.org/wiki/Dzienniki_gwiazdowe" to "http://pl.wikipedia.org/wiki/Dzienniki_gwiazdowe" - it's probably OK if we have the English title as well, and that linked separately to the English Wikipedia, but we don't have that yet and I'm more concerned about future titles with no English title. BLongley 11:28, 19 Dec 2007 (CST)
The English version of Wikipedia redirects "http://en.wikipedia.org/wiki/Dzienniki_gwiazdowe" to "http://en.wikipedia.org/wiki/The_Star_Diaries" which is a page about the English edition (it doesn't list all stories in this series, just those found in Star Diaries: Further Reminiscences Of Ijon Tichy, BTW Amazon has 'Look inside' with the table of contents). I will add the link to the variant title Star Diaries. --Roglo 11:38, 19 Dec 2007 (CST) [Edited to make it clearer --Roglo 05:42, 20 Dec 2007 (CST)]
Hm, I'm not sure if in such case the variant title should be Star Diaries (it is now) or Star Diaries: Further Reminiscences Of Ijon Tichy. On Amazon there seem to be at list 3 variants of the English title (the third is "The Star Diaries"), should we add them all as VTs? (and then there are multiple Polish variants...) --Roglo 05:42, 20 Dec 2007 (CST)
Re Star Diaries: I posted [2] after looking at my copies. --j_clark 19:11, 20 Dec 2007 (CST)
Lem's Polish bibliography is apparently even messier than I realized -- see recent discussions on Roglo's Talk page. I suspect that at this point the easiest thing to do would be to let Roglo finish entering various Polish collections (which he has been documenting in Notes) and then see where we are once we have more data to work with. Ahasuerus 19:34, 20 Dec 2007 (CST)
Re Star Diaries: The current practice for split books is to add both volumes to the series, e.g. Peter F. Hamilton's split novel, The Science Fiction Century split anthology. So I won't make The Star Diaries a variant title but rather leave it as a collection in the series. --Roglo 11:04, 21 Dec 2007 (CST)

(unindent) What should I do with the texts written by Lem in German (and later translated to both English and Polish)? Example: Science Fiction: A Hopeless Case--With Exceptions; the German version (related to Fantastyka i futurologia, but it is not 'an excerpt') is original (1972 or 1973, Science Fiction – ein hoffnungsloser Fall, mit Ausnahmen), later translated and published in Microworlds, and the Polish translation is even later (and no, not by Lem himself). Should I create the German title as canonical? (It had multiple reprints in German books). --Roglo 12:01, 21 Jan 2008 (CST)

The closest that I think we currently have one file are a couple of titles by Ansen Dibell and Robert F. Young. (We don't have Ken Bulmer's German books listed yet).
We seem to have handled these cases somewhat inconsistently so far. Young's La Quete de la Sainte Grille is entered using the French version of the title, but Dibell's Tidestorm Limit and The Sun of Return appear under their English language titles. I suppose the difference is that Young's novel has only been published in French and the title is unambiguous, while Dibell's books have appeared in both French and Dutch, so it's not clear which one should be used as the canonical one. Besides, the title of the Young novel is the same as the title of the novella that it was based on, so it helped avoid ambiguity.
I think the Lem situation is a little more straightforward since (unlike Dibell, Bulmer and Young) he wrote the original texts in German. After all, we don't really care what Willy Lei's, Curt Siodmak's (or, to use the latest example, Ilona Andrews') first language was/is, we just list their titles as they originally appeared. Which reminds me that we need to change the canonical title of F.P.1 Does Not Reply to F.P.1 antwortet nicht... Ahasuerus 20:49, 21 Jan 2008 (CST)

Weird Tales

There is something peculiar going on with our late 1980s-early 1990s Weird Tales records. Some are listed as Magazines (priced $3.50-$4.00), others are listed as anthologies (priced $12.99-$16.99), and some are listed as unlikely looking hybrids, e.g. Weird Tales 296 Spring 1990 is listed as an anthology co-edited by Tad Williams and Harry Turtledove (presumably because their names appeared on the cover) and price is listed as $12.99 even though you can see that the cover price was $4.00.

According to the Locus Index, some late 1980s issues also appeared in hardcover, which presumably explains the messiness of our data, but how are we going to fix it? Convert all Magazine records to Anthologies and create 2 Publication records for them? (Also, some links from the Weird Tales Wiki page to individual post-1970 issues are broken.) Ahasuerus 00:26, 15 Dec 2007 (CST)

As they first came out, and usually are thought of as, magazines, I would list them as magazines with hardcover binding plus add a note. Did the hardcover printings include the advertising or just the stories? If it's just the stories they in my mind they are different titles. Marc Kupper (talk) 02:47, 15 Dec 2007 (CST)
I have most of these magazines (unlike the original Weird Tales) in my collection, but I don't think I own any of the hardcover versions, so I can't check. Anybody have a copy handy?
As far as entering the hardcovers (assuming they are identical) as Magazines goes, the problem is that we don't allow multiple Publications per Magazine Title, so either we have to use the Notes field to document these two versions or we have to convert them to Anthologies (a la Destinies), which allow multiple Publications. The problem with using the Notes field is that it makes Verification ambiguous and the problem with converting them to Anthologies is that they will no longer show up in the Magazine Editor list.
As an aside, it's these kinds of a built-in problems with the current magazine implementation -- especially the obscure and troublesome distinction between the Publication level Editor field and the EDITOR title -- that make me wonder if we want to redesign that whole area. Ahasuerus 10:48, 17 Dec 2007 (CST)
It appears there are Trade Paperback reprints of some Weird Tales magazines, some with limited edition hardcover variants. I think the whole area needs redesign, but I've thought that for a long time anyway due to the variations by country - for some titles, it's become a "which verifier gets there first" as to whether we have the British, US or Canadian issue recorded. Fortunately, we've not had many British Magazine editors here for international titles - unfortunately, one of our major sources of covers and minutiae (Visco) often has only the British variants. I can see the edit wars coming... (BTW, anyone going to look at my second held submission?) BLongley 12:08, 17 Dec 2007 (CST)
Although I agree that the ability to have multiple Publication records for multi-country magazine issues would be very useful, especially when two versions are very close to each other the way some Canadian printings are close to their US prototypes, at least there is a workaround, namely creating a parallel Magazine series. However, when a magazine had multiple printings the way Destinies/New Destinies did or had a hardcover/paperback version the way Weird Tales did, our options are even ickier. And, as I said above, the distinction between Publication level Editors and the hidden EDITOR record is highly obscure and confusing. I wonder what Al thinks about this? Ahasuerus 21:14, 18 Dec 2007 (CST)
I might also note that the major magazines are also released in downloadable electronic format - that includes Analog, F&SF, Asimov's, and most recently Interzone. As far as I am concerned the print edition trumps the electronic edition and I don't see a particular reason at this time to clone the electronic edition - but eventually someone else will and our standards allow for it. Note: F&SF is down to a paid circulation of less than 18,000. It is my longterm guess that the only way it will survive is by going all electronic.--swfritter 22:12, 18 Dec 2007 (CST)

Hiding Contents Titles from Editors

One of the "nice things" that our editing software currently does to make our editors' life easier is hide certain types of Title records in the Contents section. For Collection pubs, the Collection Title is hidden; for Anthology pubs the Anthology Title is hidden; for Magazine/Fanzine pubs the EDITOR Title is hidden. Although well meaning, I believe that the last year worth of submissions (and subsequent torturous explanations on newbie editors' Talk pages) strongly suggests that this logic is more confusing than helpful.

The underlying problem appears to be that, our original expectations to the contrary notwithstanding, you really can't be a proficient ISFDB editor without having a working understanding of the relationships between various data elements in the database. Anything that obscures these relationships becomes an obstacle on the way to proficiency and a source of frustration for new editors.

Would you say that this is a fair summary and, if so, should we add it to the list of things to fix sooner rather than later? Ahasuerus 21:38, 18 Dec 2007 (CST)

It only took me six months to figure them out. It is particularly confusing that the Editor information in a mag is no longer tied to the Editor records after the mag has been created. Please tell me my Editor record project is still valuable. I presume you will need the Editor records to support any new methodology. I would not put it at the very top of the list primarily because it may take a fair amount of design, development, and testing time and might also involve a conversion. If we can add the missing Editor records and put them in series that might reduce some of the confusion.--swfritter 22:27, 18 Dec 2007 (CST)
I am sure our current EDITOR records will be useful regardless of how the underlying data may be eventually restructured, but "magazine redesign" is the topic of the discussion immediately above this one. This "hidden titles" proposal is much more modest in scope, I am just suggesting that we may want to display all "hidden" Contents records when editing Collection, Anthology and Magazine publications. I agree that the "magazine redesign" area is huge, complex and requires additional thinking, but I hope that this idea -- assuming that other editors agree that it is worthwhile -- can be implemented relatively easily. Low hanging fruit, you know :) Ahasuerus 23:07, 18 Dec 2007 (CST)
I'd show the special record. Maybe in its own section (as Interviews and Reviews are) to make it clear that it is a bit special? BLongley 06:47, 19 Dec 2007 (CST)
Oh, and how quick would it be to make content records' titles and authors non-editable unless they are the only entry in the database? BLongley 06:47, 19 Dec 2007 (CST)
Since there is already a place for the editor in the top portion of the magazine screen I am wondering how the display will look. Will it look like there are two editor entries? I really wish that pseudonym information were shown in the current display for both editor and cover artist. I have held off replacing the currently existing canonical Robert A. W. Lowndes entries with the more correct Robert W. Lowndes for that reason.--swfritter 15:15, 19 Dec 2007 (CST)
This would be the ideal time to clarify Editor and Editor (and maybe Editor). The book one isn't that special, the magazine one IS. And our Editors are all rather special, who else would do this work for free? We do rather overwork the word. BLongley 17:26, 20 Dec 2007 (CST)
Oh yes, and Coverart. Pub date updates are not cascaded to them. If they are attacked in a fairly systematic manner it doesn't take long to update them individually. Just spent a few hours updating more than 200 such creatures for Fantastic. I am looking forward to the January downtime.--swfritter 19:59, 20 Dec 2007 (CST)

ISBN-13s fixed?

As per this discussion on Al's Talk page, it would appear that ISBN-13s are now being processed correctly. Could somebody please do independent validation so that we could mark the bug as resolved? TIA! Ahasuerus 19:19, 19 Dec 2007 (CST)

I deleted the existing ISBN# and the submitted the ISBN-13. The ISBN-13 is fine except it doesn't create an ISBN-10.Kraang 19:54, 19 Dec 2007 (CST)
In a couple of years it will be impossible to unambiguously create an ISBN-10 from an ISBN-13. If it remained possible, then there would be no reason to go to ISBN-13 :) However, I don't think we have reached that point yet and I asked Al to add logic to create ISBN-10s from ISBN-13 (the way OCLC does it) while we can still do that. Ahasuerus 23:08, 19 Dec 2007 (CST)
I've added conversion from ISBN-13 to ISBN-10 to the publication viewer. This doesn't fix everything - Amazon doesn't accept ISBN13s in their URLs yet (they generate invalid ASIN errors), but some of the other sites linked to from a publication do accept them. I'm not going to bother converting the ISBN13s to ISBN10s for the navigation bar as it would greatly complicate the architecture, and Amazon is going to have to support these in both their URLs and Web Services sometime soon. Alvonruff 07:47, 20 Dec 2007 (CST)
Looks we are all set then. I will change the Help text to state that we support ISBN-13s now. Thanks! Ahasuerus 18:49, 23 Dec 2007 (CST)

2007-12-19 backup file uploaded

The 2007-12-19 backup file has been uploaded. Ahasuerus 23:08, 19 Dec 2007 (CST)

Dynamic Author Directory

The old out-of-date author directory has been deleted, and replaced with a new dynamic directory. The directory can be found on the front page under "Author Directory". The names are grouped into pages according the the first two letters of the author's lastname. There is a new editable field in the author table that allows you to change the sorting lastname if it is currently incorrect (while the software knows about common suffixes like "Sr.", "Jr.", "III", etc., it doesn't know *all* potential suffixes, nor foreign sorting patterns ala R. Garcia y Robertson). So if you see an author in the wrong directory, you can fix it.

It's pretty apparent by looking through some of the directory pages that there are still lots of minor author lexical variations, so there's a gold mine of fixes to be made there. Alvonruff 13:49, 21 Dec 2007 (CST)

Having just fixed "Hans Holbein the Younger" and "Jan Brueghel the Elder", I think there's two more suffixes that can be added to the automation. BLongley 07:50, 22 Dec 2007 (CST)
How about "St. X" surnames - can they be automated or is there too much risk of confusing them with John S. T. Smith? BLongley 09:29, 23 Dec 2007 (CST)
I think I've fixed all the Saints now. Most seemed logical apart from possibly Edna St. Vincent Millay BLongley 12:01, 23 Dec 2007 (CST)
I think the sorting could do with improvement, it's only sorting by last name at present? Not great when you're dealing with a common surname. Although I see the funny characters are getting in the way of surnames too - e.g. Karl Würf is coming up under "Wy" rather than "Wu", where there is a Karl Wurf. BLongley 10:02, 22 Dec 2007 (CST)
The generator seems to have problems with quotes. E. E. 'Doc' Smith comes out as http://www.isfdb.org/cgi-bin/ea.cgi?E._E._%E2%80%99Doc%E2%80%99_Smith instead of the correct http://www.isfdb.org/cgi-bin/ea.cgi?E.%20E.%20'Doc'%20Smith. Same problem with Timothy d'Arch Smith http://www.isfdb.org/cgi-bin/ea.cgi?Timothy_d%E2%80%99Arch_Smith. And E. E. "Doc" Smith gives http://www.isfdb.org/cgi-bin/ea.cgi?E._E._ . Dana Carson 02:45, 23 Dec 2007 (CST)
Underscores in names also caused problems - I've fixed two (OK, one may look worse now than before, see my Help Desk call) but as I could only do the search from an old local backup, not on ISFDB, there may be more still. BLongley 12:01, 23 Dec 2007 (CST)
How should we deal with non names. There are a couple of books with foo Society as the author and several covers by blah Studio. Leave them indexed by Society and Studio? Dana Carson 04:51, 23 Dec 2007 (CST)
I'd rather keep the Cover Artists for real artists, and leave design companies and photograph studios to notes. But I know some other editors disagree with that. However, finding publishers credited as co-authors often leads to a book that can be zapped entirely, or at least needs some more work. I've also reduced the number of authors a bit by moving translators and narrators to their proper places. BLongley 12:01, 23 Dec 2007 (CST)
It appears the directory has a problem with the identical name solution such as D.F.Jones (US). The last name is being picked up as (US)[3] and as a result of this the directory is not listing it as far as I can tell.Kraang 20:08, 23 Dec 2007 (CST)
And what's worse, you get a Python Error when you try to Approve an update to the last names. :-( The only one that went through was "Paul Chadwick (Pulp Author)" for some reason. BLongley 12:57, 24 Dec 2007 (CST)
For some reason it has 2 entries that go the same place? Dana Carson 23:14, 28 Dec 2007 (CST)
  * Edward Thompson http://www.isfdb.org/cgi-bin/ea.cgi?Edward_Thompson 
  * Edward Thompson http://www.isfdb.org/cgi-bin/ea.cgi?Edward_Thompson
Also S. M. Tenneshaw shows up twice with the same link. Dana Carson 23:26, 28 Dec 2007 (CST)
That's because we DO have two authors with each name, but only one place to go when you link by name:
	2418 	Edward Thompson 
	31783 	Edward Thompson 
	14407 	S. M. Tenneshaw
	31827 	S. M. Tenneshaw
For Tenneshaw it looks like we could merge them safely, for Thompson I suspect the Tubb pseudonym isn't really the same person as the Chesterton editor. BLongley 07:33, 29 Dec 2007 (CST)
Thompson sorted - the Tubb pseudonym shouldn't have had a 'P' in it. BLongley 07:46, 29 Dec 2007 (CST)
I've been making a listing of possible things to fix at Author index questions if anyone feels like researching. Dana Carson 23:26, 28 Dec 2007 (CST)

Another clock rolled over

I notice we just reached New Submission 900,000. Can we have a special prize for the 1,000,000th? BLongley 18:24, 21 Dec 2007 (CST)

I think at one point submission numbers were incremented by 10 instead of by 1, so the count may be off by quite a bit. I don't know if we have a single number reflecting the progress of the project, but the Title count and the publication count both seem to be useful. We have added over 12,000 Title records in the last 7-8 weeks, which is quite respectable. Ahasuerus 12:39, 23 Dec 2007 (CST)
Well, the Wiki just dated my last post as the first of 2008, so HAPPY NEW YEAR ALL! I'm going to go look at the fireworks for a few minutes. BLongley 18:03, 31 Dec 2007 (CST)
Enjoy the show and may you all (tm) have a happy, prosperous and bibliographically productive New Year! :-) Ahasuerus

I think I need another talk page

You may not have noticed, but I do tend to waffle a lot on SF publication related matters. Not necessarily ones that need a response from a bibliographer's point of view, although those are welcome too. E.g. I currently find myself wondering when and where the Dorsai Series became the Childe Cycle - bibliographically, I think we don't care so long as we get the super-series correct, but I do wonder how long Sphere kept referring to "Dorsai" on covers, for instance. Rather than me keep butting in on everyone else's talk pages (which are MOSTLY used for direct communication, until I start a discussion there that should be elsewhere) how difficult would it be to make the options to reply to someone "Username (Talk | Blog)" rather than "Username (Talk | block)"? I haven't used the Block option in ages, it seems a bit paranoid to make it so easily available... BLongley 19:00, 22 Dec 2007 (CST)

That's standard MediaWiki code, I am afraid. We have a relatively small group of editors and no known vandals (now that the bots are gone), but larger projects with thousands of participants have very different dynamics, so the "Block" option gets used much more often. However, Wiki layout may be more customizable in more recent versions of the MediWiki software, so perhaps we will have more options after the upgrade in January. Ahasuerus 19:52, 23 Dec 2007 (CST)

all I usually want to do is find out what the editor wants to reveal, comment on submissions, or chat about something. I know I can start up a separate blog, but if it's easier here that saves me a few open windows. Of course, just keeping this as a "nobody talks to you till they've got a complaint/compliment" venue is an option too. BLongley 19:00, 22 Dec 2007 (CST)

I suspect that our Project pages could be used for something like that, although they don't seem to be very popular, so a separate ISFDB blog may not hurt to raise visibility. Al used to have one, but he folded it into the What's New page once the Wiki was set up. Ahasuerus 19:52, 23 Dec 2007 (CST)
I know: "What's New" still looks a bit personal, so I've not tampered with it even if it makes all the newer moderators look a bit under-welcomed compared to what he said about me and Mhhutchins and JVjr. Thinking ahead a bit, a separate area for "blogging" ISFDB/SF/Book related matters would help, but thinking even FURTHER ahead probably gets us into the problems of image-hosting again. For instance, I'm happy to post pictures of one of my overloaded bookcases to help people advise on how to add the minimum number of extra holes for an extra shelf without endangering the stability of the entire bookcase, or its immediate neighbours... How many people DO want to ramble like I do, though? BLongley 20:24, 23 Dec 2007 (CST)

Publisher's pages

These seemed like a good idea once, and I still use the Daw Page occasionally. I've kept meaning (once sufficient data was collected and verified) to start recording some useful notes about certain British imprints/publishers, if only because when we finally get printing numbers sorted out we'll all need to know why Panther/Grafton/Granada books use each other's numbers, for instance. In the meantime, I've found some of the data useful to catch my own (and other's) errors - e.g. Star and Quantum Books recorded as Sphere or Futura or Orbit ones. The data seems good for late 1960s through to the late 1980s, maybe early 1990s, when ISBNs were pretty clear: after that it's going to get a bit messy as the big publishing houses went global. Before I get too far into that though, I wondered what YOU would all like to see from such a project. Some of my first thoughts are:

  • Converting Serial numbers to ISBNs for particular imprints
  • ISBN ranges for Publishers for particular years (nice validity check)
  • Imprints by Publishing company for particular years

And later thoughts made me think I could do things like

  • The main British publisher for Author X was imprint Y between 19xx and 19yy, then imprint Z from 19yy, or
  • British Imprint X published Authors A,B,C,D,E...

I've dumped some raw data on Sphere and Orbit for the moment. (I got tired of grabbing examples from the cold rooms in the house, and am sticking to my nice warm computer for the rest of tonight!) Have a look if you like, add comments if you want (I know it's not a talk page, but I didn't know where else to dump it for now - any suggestions?), or add examples if you have samples of publications around the transition dates. But basically, this is me offering to do some research if it's useful, and for once I'm not asking other people to check things! (Well, maybe later...) ;-) BLongley 18:59, 23 Dec 2007 (CST)

What your suggesting is a history of the publisher's and the imprints they've used and I like this idea. I think the first thing we could use would be a directory like the one we have for authors. At the moment we have different variations of the same thing ie. Del Rey, Ballantine Del Rey, Ballantine / Del Rey, etc. Is it "DAW" or "DAW Books"? On the main page we have a starter version of this already.Kraang 19:41, 23 Dec 2007 (CST)
I'm suggesting more than publisher and imprint, I'm suggesting ISBN checks, logos to look for, ANYTHING that helps us. From recent books, it seems ISBN tells us nothing about country of origin and we're duplicating books based on our local price. Back in the 1960s, 1970s, 1980s, and even early 1990s, some of these sometimes meant something though. I grew up with UK books that weren't published in the USA for copyright reasons, but that seemed to end in around 1976 for some publishers like Pan. Nowadays Pocket books are getting entered here as separate entries with same ISBN even when British-priced pubs are printed in the USA, or often Germany, or Scotland. I think I may give up on 21st Century publications altogether until that gets fixed, but I'm happy (when it's not too cold in the book rooms) to research the period we CAN make a difference for. BLongley 20:41, 23 Dec 2007 (CST)
The next major feature I've always wanted to do (after fixing numerous issues with current features) was to add direct support for publishers, in a way similar to authors. At present we have no way of merging publisher names, making variants, or making imprints that are associated with a publisher - everything is edited on a per publication basis. This would include a publisher directory as well. I think if we start experimenting with publisher pages as BLongley has done, we'll discover the details we want present in such support. Alvonruff 06:24, 24 Dec 2007 (CST)
One more spin-off project is to correlate cover artists and publishers. This would be handy to help decode ambiguous or insigned works.
> it seems ISBN tells us nothing about country of origin
The ISBN does have the country encoded in it. The explanation is long. I have the code but have not had a chance to add it to the ISBN decoder page I did a while back. Marc Kupper (talk) 02:32, 25 Dec 2007 (CST)
And Marc, I'd be very interested in the explanation of Country coding within ISBN. I can believe it for older ISBNs when they were assigned to National companies, but I frankly am sceptical about them nowadays with International Companies. For instance, today I've been spotting books by the British Broadcasting Company being categorised as US publications - same ISBN. It's been the same with Buffy the Vampire Slayer books - same ISBN, for publications printed with UK prices OR US prices, and printed in neither country. BLongley 15:53, 25 Dec 2007 (CST)
Sounds like an interesting direction to explore! I guess the first step would be to create a separate field for "Publisher's city/location". We currently have it entered as an optional "City: " prefix, but it really belongs in a separate field if we want to standardize publishers, otherwise we will have "London: Corgi" in addition to "Corgi" and so on. Unfortunately, it's not an immutable characteristic of each publisher since some of them (especially small presses and other one man operations) move around. Ahasuerus 15:11, 25 Dec 2007 (CST)
I've seen City useful for some German publications recently, but it mostly wouldn't help with the British publishers. (E.g. Sphere moved 4/5 times in the period I checked, I recorded it mostly so I could see if they moved in with another publisher, but that will take a lot more research for me to see common addresses.) Separate fields for Imprint and Publisher would help (if only one is specified, use it for both so data-entry doesn't get too hard), if we're going to add printing number support eventually: as British printing numbers can carry over imprints. (And even carry back, I think!) BLongley 15:41, 25 Dec 2007 (CST)
One of the first things that's come out of the research is Publisher vs ISBN checks: I think one of the next things that should come out is when two books with apparently different publishers are in fact the same book. E.g. if two otherwise identical pubs are by Gollancz and VGSF, they ARE the same - VGSF is an imprint of Gollancz. (Or they're both imprints of someone else now, haven't finished that one yet.) I see a lot of that duplication even from British editors. :-/ BLongley 15:41, 25 Dec 2007 (CST)
I don't have time for an explanation in English and so have posted my decoder as a text file. It's a Visual Basic for Applications code but hopefully will help people understand the process. The same code is in an Excel spreadsheet. If you have macros enabled, or know how to create a digital signature, then you can play with the spreadsheet. Marc Kupper (talk) 04:31, 5 Jan 2008 (CST)
Victor Gollancz Ltd was a publishing company, 20 years ago it wasn't even part of the Orion Group (Wikipedia has notes on its history). Only later it became an imprint... --Roglo 05:18, 5 Jan 2008 (CST)
Yes, but the Wikipedia article is crap. That's why I started researching from my own books. Unfortunately I have very few of their hardbacks, which was their main line. BLongley 07:23, 5 Jan 2008 (CST)
I have only Carr's Best SF from a library. And my remark on '20 years ago' was based on this book. But I hoped Wikipedia would be useful. --Roglo 11:51, 13 Jan 2008 (CST)

(Unindent - and very tempted to move this entry back to the active area) - we have new tools to help us now, how do we want to use them? I'm willing to update Publisher pages with all the info I've been recording so far, and people are creating "proper" Publisher pages at last - shall we just go create and update at will for a bit and sort it out afterwards? BLongley 18:18, 16 Feb 2008 (CST)

Entering Translations as separate Title records

We seem to have at least one editor who is entering non-English translations of certain popular books (e.g. Vernor Vinge's "Zones of Thought" novels) as separate Titles and then linking them to their master English Titles. I haven't checked our log files to see who the editor is, but if you are that editor, please keep in mind that at this time we do not have separate Title records for translations of English language titles. (English is privileged at this time in that English translations of foreign language titles get their own variant Titles.)

The problem with entering translations as Titles is that, comprehensively implemented, they could increase the size of popular authors' Summary pages by a factor of 10+ and would make them unmanageable, so for now we just enter them as additional Publications under the English Title record. This is a temporary fix at best, but a real solution is still being hammered out. And we really need to update our Help pages since they are at best silent on the subject and at worst misleading. Ahasuerus 17:12, 27 Dec 2007 (CST)

For an example of how bad it could get, look at one Harry Potter title. (That's going to be bad enough eventually, as there's still 40+ English printings to add.) And then multiply it by the number of other Harry Potter titles there are. This doesn't mean you can slack off in getting the foreign titles RIGHT (I've still no idea how the French editions should really be titled, for instance) - but I've had to clear up some editions where the Translator or Illustrator or Narrator is getting co-author credit too, which makes such pages even messier. Be careful out there! BLongley 18:06, 27 Dec 2007 (CST)

2007-12-27 backup uploaded

The latest backup file has been generated and uploaded. Please note that I won't be able to create the next backup until after 2008-01-06 when I am on the road again. Ahasuerus 23:51, 27 Dec 2007 (CST)

Author cleanup by checking reviews

I created a new page of fixes needed (Rough explanation of how/why they need fixing here) which seems to have attracted some attention already - it's a "One Click to see the problem" page, and each shouldn't take more than one typo-fix or regularisation, or a new pub submission, to solve. If we can clean up those we can move onto Titles that are missing, then Pubs...
Please delete the entry you've fixed, or mark it as fixed, so we don't all overlap the effort. Or just claim a range for yourself and work through it. BLongley 21:22, 29 Dec 2007 (CST)

I notice some entries are being marked as fixed. I'm simply deleting the entry, which makes me think I'm accomplishing something, if only in appearance. Mhhutchins 23:06, 29 Dec 2007 (CST)
Sure, making the list shorter is probably a better idea. Ahasuerus 23:21, 29 Dec 2007 (CST)
There's an entry that has me puzzled. The original record mistakenly credited the author as "Michael Faber" and it should have been Michel Faber. When I changed the pub ('cause we can't change the record itself, drats), it didn't change the record. I'm thinking this has something to do with the fact that there was a variant created for the author of the review (McAuley). Is it a bug, or is there some other way to fix this? Mhhutchins 23:06, 29 Dec 2007 (CST)
The problem is that there are two variants of this Review record, one for "Paul J. McAuley" and one for "Paul McAuley". Since you can only edit Reviews in the publication that they appeared in, you can only edit one of the two Review records; the second one is inaccessible. It's clearly a bug and the only workaround that I can think of for now is to Remove the review from the magazine, delete it and its VT version, then re-enter the correct version. Messy business :( Ahasuerus 23:21, 29 Dec 2007 (CST)
Another feature of this bug: the author of the book being reviewed is not shown in the contents of the publication. Mhhutchins 23:27, 29 Dec 2007 (CST)
Hm, I think something else may be going on here. We have publications where the reviewer was pseudonymous, but the review data looks OK. Probably something to escalate to Al for further investigation. Ahasuerus 23:37, 29 Dec 2007 (CST)

(unindent)I some cases, the pub where a review had author's name misspelled is verified. Can we edit it? E.g. this mag has 62 • Hardwired • Walter John Williams instead of Walter Jon Williams. --Roglo 03:18, 30 Dec 2007 (CST)

I believe the answer is "Yes, go ahead and change the author's name to the correct version and if we are sure that it was misspelled in the magazine, add a note to the publication record." Ahasuerus 11:59, 30 Dec 2007 (CST)
I can be sure about issues I have, but here I can add note that 'This listing was verified with review p.63 having author's name misspelled as Walter John Williams', or sth similar. --Roglo 15:59, 30 Dec 2007 (CST)
There are even more complicated cases, like this, where Ballard is reviewing 3 volumes from a series, so either we split it, or the titles won't match. IMHO it would be better to have non-matching reviews displayed at the bottom of author's bibliography or on a separate page. --Roglo 15:59, 30 Dec 2007 (CST)
If a reviewer reviews multiple works in one piece -- which is common in magazines -- then we enter each reviewed book/story as separate review records. The ultimate solution to reviewers misspelling the titles that they review (and sometimes I wonder if they do it on purpose, just to make our lives harder) is to allow direct linkage of review records and reviewed records the way variant titles are linked. This enhancement has been on Al's list of things to do for some time, but he hasn't gotten to it yet. Ahasuerus 16:16, 30 Dec 2007 (CST)
This would force us to add the books being reviewed to the database. (Otherwise, no parent ID. Or, perhaps, it should create the missing title automatically?). And, again, the Ballard's review is in verified publication. --Roglo 16:27, 30 Dec 2007 (CST)
I see no problem with adding books - indeed, when there's 3 or 4 reviews of it I think we want to know WHY it was reviewed so often in SF magazines. I've created a few like that already to solve the linkage problems. When it's a magazine being reviewed I've held off till we can establish whether it's an SF magazine or Manga. Videos/Films are right out. But reviews of pubs, rather than titles, will be needed: e.g. for Audiobooks where the narrator/performer/reader is probably being reviewed as much as the original author: for instance:
From the Hells Beneath the Hells 	REVIEW 	1976 	Fritz Leiber
From the Hells Beneath the Hells (audio recording) 	REVIEW 	1976 	Spider Robinson
From the Hells Beneath the Hells read by Ugo Toppo 	REVIEW 	1976 	Bill Warren
... and we haven't even figured out the original book yet! So maybe story level even... BLongley 16:35, 30 Dec 2007 (CST)

(Unindent) Before I go sleep now, let me say a big THANK YOU to all that have worked on this - I haven't seen such activity on a single project for ages! BLongley 17:44, 30 Dec 2007 (CST)

Black Gate

Where does Black Gate fit in? It's neither on the Magazines nor Fanzines pages. They have e.g. a John Justin Mallory story (Resnick) we do not have in the database. --Roglo 02:51, 31 Dec 2007 (CST)

Well, they pay 3-6 cents a word, which is a borderline pro rate, but OTOH they are not on SFWA's list of qualifying markets. When in doubt, call them a Magazine? Ahasuerus 11:25, 31 Dec 2007 (CST)
Is this an imprint of the Tangent we haven't started yet? I paused as we might need some disambiguation first, I wouldn't want to enter the WRONG zine. I've only got a couple of Tangents though, I think. Someone please start the page so I know if it's the same one. (I've started several Fanzine pages already, don't look at me for any more, I have to go back to work soon!) BLongley 19:44, 31 Dec 2007 (CST)
Tangent is a reviewzine which was published on paper in 1993-1997 and became an e-zine in 1997. I don't think it's related to Black Gate, a paper-based fiction magazine, they just review it like they review all other short fiction venues in the SF/F/H genres. Ahasuerus 20:01, 31 Dec 2007 (CST)
They just have good listing of the Black Gate issues, with reviews. Black Gate was listed in Dozois' Summation 2006 (The Year's Best...) in the semiprozine section. --Roglo 05:28, 1 Jan 2008 (CST)
I see Amazon is rather confused about them too, and has several of them in the books category, e.g. Issue 1 here. BLongley 10:57, 13 Jan 2008 (CST)
And I don't see Spectrum SF (e.g. http://www.bbr-online.com/catalogue/Items/Spectrum.shtml) neither on Magazine nor on Fanzines pages. I have 1 issue. Some of their stories are on the Locus nominees lists. Reviews of SSF are on http://www.sfsite.com and http://www.infinityplus.co.uk --Roglo 11:46, 13 Jan 2008 (CST)

Conanight?

Any ideas re: who Conanight might be? One online bookseller speculates that it may be Damon Knight, who was an active illustrator for Doc Lowndes/Columbia Publications in the early 1940s, so it's not unreasonable. OTOH, this particular painting illustrates a story by Chester B. Conant (aka Chester Cohen), so I wonder if it was some kind of collaboration between CONANt and kNIGHT? A number of folks in Doc Lowndes' circle (Knight, Kyle, Bok) were writers as well as illustrators, so it can get complicated. Ahasuerus 15:11, 2 Jan 2008 (CST)

Same artist also for this collaboration by Pohl, Kornbluth, and Lowndes. in Spring, 1942 SF Quarterly.--swfritter 21:33, 2 Jan 2008 (CST)
Knight's The Futurian - Page 90 - indeed it is Cohen who was Knight's roommate at the time.--swfritter 21:37, 2 Jan 2008 (CST)
Thanks, fixed! Ahasuerus 23:15, 2 Jan 2008 (CST)

"Something from Beyond" by "Paul Dennis Lavond"

Would anybody happen to know who was behind "Something from Beyond" by "Paul Dennis Lavond"? "Lavond" was a collective pseudonym used by a bunch of Futurians (Pohl, Kornbluth and/or Lowndes) and there may be something about it in His Share of Glory. It's not listed in its TOC, so it's not a solo Kornbluth story, but it could be a collaboration or it could be a Doc Lowndes effort. Ahasuerus 23:15, 2 Jan 2008 (CST)

Lowndes, Pohl, and Dockweiler - from Rock and Contento.--swfritter 13:27, 3 Jan 2008 (CST)
Also from Noosfere, and TomFolio says "Lowndes, Fred Pohl & Dirk Wylie" - Wylie being Dockweiler. (Not offering those as authoritative, but I'd like to know how often they concur to judge THEIR accuracy. BLongley 14:49, 3 Jan 2008 (CST)
Thanks, fixed! Ahasuerus 17:49, 3 Jan 2008 (CST)
And to follow up with another Paul Dennis Lavond question, who was responsible for "A Prince of Pluto", please? Ahasuerus 23:02, 3 Jan 2008 (CST)
According to the Day Index, Cyril Kornbluth and Frederik Pohl.--Rkihara 01:16, 4 Jan 2008 (CST)
Thanks, I'll set up a vt! (And I should have thought to check Day.) Could someone with access to the Contento/Miller CD please double check in case Day's information has been superseded? And while we are talking about pseudonyms, I wonder if D. J. Foster may be one. We have only one of his stories on file, but according to Day he published a couple more in "Tales of Wonder" (UK). Ahasuerus 08:26, 4 Jan 2008 (CST)
Still Kornbluth & Pohl in Contento/Miller CD's. No pseuodnym listed for Foster in Contento/Miller, Rock, or Robinson.--swfritter 16:29, 7 Jan 2008 (CST)

Stirring Science Stories

I have entered the first 3 issues of Stirring Science Stories in the database, but I can't seem to find the last (March 1942) issue in my collection. If anybody has it or at least has access to the Contento/Miller CD that lists the contents, could you please enter it so that we could close this gap in Magazine:Stirring Science Stories? Thanks! Ahasuerus 20:24, 5 Jan 2008 (CST)

Added. Authors are the usual incognito suspects. Now on to Kornbluth collection which will probably be only partially done tonight.--swfritter 20:30, 8 Jan 2008 (CST)
Thanks! :) Ahasuerus 22:52, 8 Jan 2008 (CST)
Thanks to the Lowndes issues you added and my minor cribbing contribution every one of the Kornbluth stories from His Share of Glory found a merge partner.--swfritter 22:57, 8 Jan 2008 (CST)

Cleanup Projects

It seems, with projects like "ISBN fields that don't start with '#', and aren't 10 digits long" and Authors that only exist due to reviews that you're all willing to work on something, however long the name, if people just set it up to be nice and easy to look at each problem! I'm quite happy to set-up and/or refresh such projects: a couple of things I've learnt so far are that they should be in smaller chunks, and it might be better if I clarified the verified pubs (for projects at pub-level). Those two projects will never quite be finished, I suspect (although each should tail-off in some standards discussions about the remainders) and I'm happy to take suggestions for the next. BLongley 18:12, 6 Jan 2008 (CST)

Happy to work on those things when I run out of books at home to enter (and have to get some from the storage unit) or am on the road and cannot take too big a pile with me. I found a nice site that takes the bad ISBN 10s, changes each digit (holding the others fixed) to make the number correct. I then look at the revised numbers in Amazon until I find the matching pub. Hasn't taken more than 4 so far to hit a valid match on the 3 or 4 I have worked on so far. Will work on this when I am in Dallas in two weeks. Would like to know how to set up a magazine series so that I can enter the 50+ Perry Rhodans I've got backed up, also like to be able to set up new series under Analog (Asimov, FSF, etc.) and add to it so I can get some of the magazine essays organized in to series. Any help would be appreciated. Thx, rbh 20:39, 7 Jan 2008 (CST)

January 8 backup uploaded

The latest backup file has been uploaded. Grab it while the system is still up -- remember the server upgrade and likely downtime/instability this week! Ahasuerus 00:18, 7 Jan 2008 (CST)

Time for some more entropy stats then:
Last Backup:
Magazine Pages
Bad	50814
Good	58049
46.6%

Pub pages
Bad	19189
Good	86626
18.1%

Pub Prices
Bad	13659
Good	99747
12.0%
This backup:
Magazine Pages
Bad	50108
Good	60482
45.3%

Pub pages
Bad	19140
Good	87695
17.9%

Pub Prices
Bad	13852
Good	100449
12.1% 
So based on four samples:
   Magazines without page numbers, 52.6% to 50.0% to 46.6% to 45.3% - all good!
   Pubs without pages, 19.2% to 18.8% to 18.1% to 17.9% - all good!
   Pubs without prices, 12.0% to 12.1% to 12.0% to 12.1% - not so good.
But understandable, given the nature of current projects where filling in a missing pub despite no price is better than no pub: and the number of good pubs has gone:
96196 to 97749 to 99747 to 100449
Looks good news overall to me! (Ahasuerus, weren't you going to create a page for these stats?) BLongley 13:12, 7 Jan 2008 (CST)
Sorry, got distracted with verifications, travel and other issues. I have massaged Bibliographic Projects in Progress and Project:Data Entropy is now prominently displayed at the top. Feel free to start populating it with your results and linking the SQL scripts that you used to generate the data. Once we have a better idea of what kind of data we have, we can organize it in spiffy looking tables :-) Ahasuerus 19:21, 7 Jan 2008 (CST)
Scripts, no problem: I think I've posted most anyway already. Tables - can do. We need nice graphs and good, positive, uplifting images - who here does Powerpoint presentations to management regularly? BLongley 19:32, 7 Jan 2008 (CST)
It's hard to beat Al when it comes to ISFDB-derived images. Have you seen his ISFDB Author Communities graphs? Ahasuerus 19:42, 7 Jan 2008 (CST)
I didn't think anyone ever looked at those. I am Powerpoint wizard, but we probably don't want to use that here (it would just be an embedded Excel spreadsheet anyway). If we can get the wiki upgraded, the new versions support embedded SVG, which means we can put graphs into the wiki without uploading any files. Alvonruff 16:15, 9 Jan 2008 (CST)
I never expected SVG! I'm about to replace a load of SVG charts at work as 1) the Corel SVG viewer is no longer supported by Oracle, and hasn't been for a while, and 2) the Adobe SVG Viewer plugin is out of support from this month too... but if it still works here I should be learning a lot about it soon, as I convert a dozen or so charts to Flash. BLongley 16:39, 9 Jan 2008 (CST)
I'd still like the option to upload SOME files here - people are asking for a library of author signatures for instance, and it should be possible to limit sizes, and make it a moderator-only option? I haven't looked ahead at Wiki upgrades though, and if there's still good reasons to NOT allow it I'll listen. (And help with off-site hosting if that's the way we need to go.) BLongley
Oh, and while you appear active on the forums, can you please tell me if my fumbling attempts at deciphering the database at this page are right, mostly right, or dangerously misleading? I'm still not confident enough to update the Database Schema itself, although some areas of that do look seriously out of date. BLongley 16:39, 9 Jan 2008 (CST)

MySQL access

I've set up a cron job to automatically keep my MySQL copy updated to the database backups Ahasuerus posts. If it would be useful, I could offer access via ssh, so that other editors could run SQL queries and so on against the database. (I'd been meaning to set up a whole ISFDb development sandbox, but that's still on my to-do list --- but the MySQL stuff is there now.) So if you're interested, just let me know and I'll look into making an account for you. It's a shell environment on my BSD box, a pretty standard UNIX system. WimLewis 15:09, 10 Jan 2008 (CST)

I'd be interested - I load the backups here fairly regularly, but as it's on a Windows box any poor SQL coding I do tends to lock up the machine totally, and I'd prefer to lock up someone else's! ;-) BLongley 12:30, 11 Jan 2008 (CST)
Seriously though - any *nix environment should be better than my local copy, and I haven't yet set up a whole ISFDB environment on any of the spare PCs yet, although Roglo seems to have done so. (OK, the newbies are doing better than I am, I'm not bitter.) BLongley 12:30, 11 Jan 2008 (CST)
Cool, send me email (address is on my user page) and I'll set up a login for you. If you can, I'd like to use public-key authentication instead of passwords (it looks like PuTTY supports it; I assume other Windows ssh clients do too). WimLewis 13:59, 12 Jan 2008 (CST)
Is the current weekly (give or take) backup schedule working for you, folks? It takes almost two hours to generate a new file, so I can't do it daily, but I could probably do it twice a week if there is sufficient demand. I don't use MySQL for my scripting any more, so I'll pass on the offer, but I would be interested in a summary of how the Python software was beaten into submission. I tried it once and didn't have much luck with the configuration files. Ahasuerus 14:28, 12 Jan 2008 (CST)
Once a week is enough for me now I'm back at work. BLongley 16:09, 12 Jan 2008 (CST)
OK with me.--swfritter 18:18, 14 Jan 2008 (CST)

Interzone 25th Anniversary

In 2007 Interzone had a special 25th Anniversary column with 'brief reflections' on Interzone's history from SF authors, editors and reviewers. I'm not sure how to record it. One one hand these include mini-essays from famous authors and are sometimes quite interesting ('my first sale to Interzone'), OTOH some are single paragraphs 'You are great and long live Interzone!'. It reminds me a letter column, with 8-10 authors on 2 pages except the authors are from the SF publishing. I would gladly record it as ESSAYs but it will clutter the contents listing. Do you have a better idea? (1 ESSAY by 8 authors would be quite misleading). --Roglo 11:30, 11 Jan 2008 (CST)

I'd say separate entries then. Multiple essays on the same page can be sorted by putting them in a small series I think - we've had some people experimenting with "serial" type but that hasn't been popular. BLongley 14:13, 11 Jan 2008 (CST)
I've used serials in New Worlds 2 (for Warwick Colvin, Jr.) but I don't see how it could help; they are not grouped together in the pub. A 'LETTER' type would be useful for hiding letters in concise mode - I would use it for this column if I could... So I'll add them as ESSAYs. --Roglo 15:35, 11 Jan 2008 (CST)
This is the conversation I'm thinking of. I think Serials are intended for bits of a (probably) Novel length work in multiple publications, that need grouping together. And I used series for something or other that needed it (lots of entries over fewer pages), but can't remember where. :-( BLongley 16:57, 11 Jan 2008 (CST)
Not that bad; for 2 issues I had to create only one new author and one new pseudonym. I was afraid it could be worse. --Roglo 16:12, 11 Jan 2008 (CST)
Not a big deal then. Ideally I'd add at least one new book each day, a new title every 3 days or so (British works are still under-represented here, IMO) and find a new author each week. Pseudonyms are controversial, but I'm sure I'll keep coming across ones that haven't been entered here yet. (I better do - as I specifically purchased a book for such!) BLongley 16:57, 11 Jan 2008 (CST)
I see that using 'series' we can display essays by various writers on the same page. This would be useful if the essays had individual titles, but here they are clearly visible and you can track them by searching for 25 or even better, 25 years of interzone. My only problem is that the contents listing doesn't look as nice as before. --Roglo 10:47, 12 Jan 2008 (CST)
Oh, and I've just noticed: an essay with 15 authors :-) --Roglo 16:28, 14 Jan 2008 (CST)
At one point our editing software had problems when dealing with more than 10 Authors per Title or Publication, but it may have been addressed. Al would know for sure. Ahasuerus 16:47, 14 Jan 2008 (CST)

January 13, 2008 backup file uploaded

The January 13, 2008 backup file has been uploaded. Ahasuerus 22:46, 13 Jan 2008 (CST)

I have rerun my scripts that find Title vs. Publication Type Consistency issues against the latest backup. The list of problem records generated for Data Consistency/Serial Dates is shorter than the one created the last time around and most of the mismatches are fairly benign (YYYY-00-00 instead of YYYY-MM-00). However, there are also a couple of improperly merged Serial records (Creep, Shadow! and The Dark World) and even one Serial record apparently merged with its associated novella record (The Dragon Masters), which somehow got past the moderators.
Data Consistency/Editor Mismatches was in better shape, with only one Essay misclassified as "Editor". Data Consistency/Serial Mismatches is also in much better shape than after the last iteration. Data Consistency/Non-fiction Mismatches has gone from 238 Kb to 211 Kb, but is still huge. Similarly, Data Consistency/Short Fiction-Novel Mismatches is down from 148 Kb to 124 Kb.
Data Consistency/Anthology Mismatches and Data Consistency/Omnibus Mismatches have a couple dozen new discrepancies each. I am not sure what's causing them. Perhaps somebody is fixing Title records but not the associated Publication records? Ahasuerus 23:06, 13 Jan 2008 (CST)
I'm sure somebody IS doing that (I keep seeing it in the Nonfiction/Novel mismatches too), but some of those seem to be due to reprints of a weird beast, the 'threaded novel'. BLongley 13:49, 15 Jan 2008 (CST)
As far as Data Consistency/Novel-Anthology Mismatches go, we are once again in better shape than we were earlier, but I am a little puzzled by some of the comments from the last pass, e.g. "OK, an anthology with 3 novels in it" seems to suggest that the publication could be profitably changed from Anthology to Omnibus. If we agree that any Anthology with 2 or more novels in it should be classified as an Omnibus, I could then update my script to ignore all Anthology records with only 1 Novel title. Ahasuerus 00:12, 14 Jan 2008 (CST)
The comments were left by me because the rules are unclear about how to classify them. Is 1 Novel & 1 Novella an omnibus? What if you have 2 Novels & 8 shortfiction titles? I felt it best to leave the note and when the rules for what classifies as an omnibus or anthology are clearer they can be relooked at and cleaned up.Kraang 07:15, 14 Jan 2008 (CST)
We can have another attempt at defining SOME clear Omnibus definitions: e.g. IMHO:
* An Anthology or Collection containing ONLY Novels should be an Omnibus,
* Introductions, Forewords, Afterwords, Interiorart shouldn't affect that.
* But "Short Novels" that were published standalone should count too - but those are being reclassified as Chapterbooks sometimes (I don't like those, they don't appear where I want BOOKS to be).
* When you start adding extra short-fiction I lean towards making it a Collection/Anthology... but not if the short-fiction was an actual book once (see above). Or at least called a (Short) Novel once.
OK, we'll never all agree: but we do seem to have made some progress on fix-ups (see here) and even a few agreements will help. BLongley 13:49, 15 Jan 2008 (CST)
I know where you are coming from, Bill, but I am afraid that any definition that depends on whether a particular Title has appeared in a standalone publication (a "chapterbook" in case of novellas) would be difficult to implement. Not only would we have to check whether a title has been reprinted separated, but we may also have to go back and change anthologies to omnibuses whenever a novella has been reprinted by itself. To make things worse, dozens of classic novellas and novelettes were reprinted as flimsy paperbacks by semipro publishers in US and Australia back in the 1940s and 1950s and we don't have most of them listed at this time. Once these publications have been added, we would have to go back and change a bunch of books from Anthologies to Omnibuses :(
I suspect that the easiest convention to implement would be to call any multi-author publication with 0 or 1 Novel Title in it an Anthology and anything with 2 or more an Omnibus, ignoring Shortfiction, Essays and artwork for the purposes of this discussion. Ahasuerus 22:06, 15 Jan 2008 (CST)
Would the above rules apply to single author's collections?Kraang 22:56, 15 Jan 2008 (CST)
Well, that would make our rules more uniform and there is much to be said for simplicity. Do you think it may result in counter-intuitive cases? For example, applying the "1 Novel rule" would make Tanya Huff's The Blood Books, Volume 3 a Collection rather than an Omnibus. Would that make it look bad (or at least significantly worse than it looks now)? Ahasuerus 23:16, 15 Jan 2008 (CST)
IMHO it depends on whether the Blood Bank collection was published earlier or not. If not, I would make it a collection. At the very least, forthcoming Blood Bank should be added, so that we really have '2 books in 1'. (And Locus describes it as Collection/omnibus). --Roglo 03:36, 16 Jan 2008 (CST)
It's the "N Books in 1" information that I want to preserve, as it's physical books I'm buying and I want to avoid duplications. An omnibus is more likely to be practically identical in content to the books that went into it - when you break it down into a list of stories you're never quite so sure that stories aren't revised/expanded/abridged. Someday they might sneak a book I've already got (e.g. this into an Omnibus, we call it a Collection or Anthology, and I end up buying duplicates. I want the same sort of visibility of prior book publications as we seem to be getting to with Fix-Ups (where I can decide whether I want the fix-up or should hunt the original stories in collections/anthologies.) BLongley 13:09, 16 Jan 2008 (CST)


(Unindent) The Data Entropy page is now populated with all the stats I produced earlier that mess up this page so much. Some backup dates are suspect. I've deliberately left the latest entry out in case anyone else wants to run the scripts and create the next table entry - I'm not claiming ownership of Stats. (Some people have a dozen backups or so and COULD create a load of missing entries for various dates, but I'm happy to stay with this and move forward.) I encourage people to add other stats and measures important to themselves so we can all feel GOOD about our work here. I think the stats I posted are generally uncontroversial, but I know I also post projects that MAY be - e.g. we should sort out whether fixing reviews by adding pubs, or questioning the sort of review we have, but Stats encourage people at times and the lists of awkward things should reduce to a few standards discussions eventually. BLongley 17:39, 14 Jan 2008 (CST)

Sounds like a good plan! I have added the data for September 2006 - August 2007 and the numbers speak for themselves. I'll wait until more "entropy measurement" scripts are posted and then run a whole bunch of them against my archives since it's somewhat time consuming. Ahasuerus 22:47, 14 Jan 2008 (CST)

January 20, 2008 backup file uploaded

...and we added over 50 Mb to the database in the last week, primarily by creating dozens of edits of large Wiki pages. It looks like we will need to completely delete some of these project pages (including their revision histories) once the affected records have been reconciled or else our backup files will become unmaintainable. Ahasuerus 23:00, 20 Jan 2008 (CST)

Variant Titles and Dates

(reposted from Rkihara's Talk page, which see for the preceding discussion)

There appears to be, or have been, some confusion over the date used for variant titles - date of variation, or date of first publication? It's easy to spot some Novels (search for title of "(rev" and see if the date in the title matches the date of the record) - quite a few could do with a quick clean-up as the variants are recorded under the wrong parent. Not so easy to spot variations of Serials, I think that would require some SQL. Are there enough to make it worth doing? BLongley 14:11, 22 Jan 2008 (CST)

This is a question that has deprived me of billions and billions of nanoseconds of sleep over the last few months :-( The Help pages currently state that "For works that have had variant titles, the date to enter is the first under any title and any pseudonym; variant titles do not have their own dates". The rationale here is to prevent misleading dates from appearing on Collection and Anthology pages. For example, William F. Nolan's "Small World" (1957) appeared as "The Small World of Lewis Stillman" in his 1963 collection Impact-20. If we had "The Small World of Lewis Stillman" listed as a 1963 Title (as opposed to a 1957 Title as per the Help pages), then a quick look at the contents of Impact-20 would suggest that the story was first published in 1963 and not in 1957.
On the other hand, the current approach makes it more difficult to find when a particular story first appeared under a variant title. To reuse the same Nolan example, you would have to pull up the list of all publications that reprint this story and review the Contents section of each one. And, of course, the original rationale is mostly moot when it comes to novels.
I think the best answer to this puzzle would be to modify the software to display the publication date (if it is different) for the parent title so that the Nolan story would be displayed as:
  • 13 • The Small World of Lewis Stillman • (1963) • shortstory by William F. Nolan (aka Small World (1957)) [as by William F. Nolan ]
Then we could change the Help text to state that we want to capture the first date when the variant title was used. If we can get 20 minutes of Al's time to make the change, I will volunteer to write a script to find all pre-existing discrepancies and we will live happily ever after :) (Ahasuerus 14:40, 22 Jan 2008 (CST)
There is another (related) problem - sometimes the canonical title is not the title of the first publication. So if we set it to the first publication of the story (as we do), we cannot use it to record the first use of this particular title. To capture the first use of the title, we would have to convert all titles so that the parent title is the title of the first publication (on the Bradbury's page you can see I'll Not Look for Wine as a variant of Ylla while it was the first title (Locus index for Bradbury stories). I think I've seen somewhere a discussion that: 1. We can usually check the date of the first publication (copyright pages) but not the first use of this particular title (as our data is incomplete); 2. We do not usually enter the newspaper where the story was first published, so if it was the only use of the original title, a note is enough. Personally, I would prefer to have the first title as the parent, so that content listings would match Locus and copyright pages (here I have 125 • February 1999: Ylla • shortstory by Ray Bradbury (aka Ylla) [as by Ray Bradbury ] which is quite useless). --Roglo 16:03, 22 Jan 2008 (CST)
Given good enough linkings, we can determine first publication date in book, and in magazine, whether it's the canonical title or not - so long as variants have the dates of variation recorded, or we work with publication dates. The "first under any title and any pseudonym" is usually only going to be accurate based on what we already have, which is often incomplete - e.g. I've chased a few titles down to first publication in magazines we wouldn't normally be entering (like Playboy, Colliers, The Strand), but not messed with title dates of later variants. Date is the comparatively easy part, assuming first book and first magazine dates are all we want. Displaying ALL the titles and ALL the pseudonyms might be desirable, but would definitely be trickier: and if we don't want them ALL, which DO we want? The title and pseudonym don't necessarily change at the same time... BLongley 17:24, 22 Jan 2008 (CST)
Roglo makes a good point. It is very important for the data on the author biblio pages to sort by the earliest appearance. In many cases the entry for the first appearance is not the canonical title. There are many cases where variant title dates do not match their parent because many of those associations were made when magazine dates were entered without month information. Since that time the month data has been updated but not necessarily to parent and variant.--swfritter 18:14, 22 Jan 2008 (CST)
True, it's an important consideration. At one point I thought about writing a script that would identify variant titles whose date differs from the canonical title's date, but figured that I should wait until we have more magazine issues validated. Perhaps a more limited script, one that would search for YYYY-MM-00/YYYY-00-00 mismatches, would be more useful in the meantime. Ahasuerus 19:34, 22 Jan 2008 (CST)

Revised versions

What do we do about situations where a variant title also indicates a significant revision or expansion of a previous work? For example When HARLIE Was One, Release 2.0 is significantly re-written from the original novel, while King David’s Spaceship is expanded (but not much if at all re-written in the early sections from A Spaceship For The King. In the latter case the variant title record includes "exp." in the title itself, but there is no indication in the main title record that the variant is also an expansion. And then there are works that are significantly revised with no change in title. For example The Time Traders where I put a mention in the Note field. What is the preferred way to handle such cases? Would explicit programmatic support for "revised version of" or "based on" be desired eventually? This might help deal with the fixup issue, also. -DES Talk 15:19, 24 Jan 2008 (CST)

There has been discussion concerning this issue and a new feature request has been added to the list in which an editor can assign relationships between pubs. This would include revisions, expansions, "based on...", maybe even handle fixups to a certain extent. This is high on my own personal wish list, and hopefully it can be implemented soon. I don't know how much reprogramming would be needed for this feature, so it's all up to Al. Until then, just place notes in the pub's note field explaining the relationship between the titles. Mhhutchins 15:58, 24 Jan 2008 (CST)
As long as we expect a software solution in the foreseeable future, we could also add "(abridged)", "(expanded)", "(restored)", etc to the affected variant titles, which have all been used at various points in time as a simple Title search would demonstrate. That way it would be easy to find and correct these title records once the software has been updated to support the proposed "relationship" field. Ahasuerus 16:41, 24 Jan 2008 (CST)

Damon Knight

I would have posted this at Author:Damon Knight but both that page and its talk page are protected

I suspect that Books: Half Loaves (1959) and Half Loaves (1967) are the same essay, or versions of the same essay. Should i merge them? How much checking is it wise to do first? in this case i know that the essay in In Search of Wonder is reprinted, possibly with changes, because Knight says so in the intro -- but he doesn't identify the specific sources from which specific chapters were taken, and in some cases it is clear that chapters represent an assemblage of content from multiple sources. -DES Talk 16:00, 25 Jan 2008 (CST)

In the past I have compared reviews in magazines to corresponding entries in the major critical essay books and found the differences to be minor to very major. Merges for fiction are generally done to combine appearances of a story that are either identical in content or have only minor differences. Expanded versions of stories, for instance, have some sort of qualifier in the title. If the book essays are merged with the magazine reviews it might leave the impression that there are no differences. There are also some pending changes to the linkage of reviews to the titles being reviewed. I don't know if anyone has mentioned this to you but there is no explicit database link between reviews and works being reviewed. They are currently matched lexically by matching the text of titles which is why you will sometimes see a review listed where you may not expect it.--swfritter 18:13, 25 Jan 2008 (CST)
I had read about the "lexical match" issue, and I can see why it is a problem. Thanks. I seems as if a simple merge is probably not a good idea in this case, at least not without a detailed textual comparison. -DES Talk 18:27, 25 Jan 2008 (CST)

Why don't we use the first printing date for reprints?

...when ten-year old kids get to do the cover art for a Robert Heinlein novel. Mhhutchins 20:36, 25 Jan 2008 (CST)

Looking at the title page for The Man Who Sold the Moon, it looks like it has a number of problems. All Signet printings are dated 1951-03-00, there is a 1951 Roc edition that purports to be a $0.95 trade paperback with an ISBN (clearly an impossibility on many levels) and some of the Baen printings are of dubious quality. Unfortunately, this not uncommon when you examine many of our Heinlein titles, which mostly go back to 1995-1996 and are often in dire need of a cleanup.
To go back to Michael's point, I agree that using the date of the edition's first printing when recording a later printing can lead to unforeseen and undesirable side effects. Hopefully, once we have a new field for print numbers, the whole issue will fade away. Ahasuerus 05:10, 26 Jan 2008 (CST)
Assuming you can tell the printing (see www.travelinlibrarian.info for a good explanation), I am not sure how much value there is to tracking each printing as a separate entry. If the only way to tell the difference is a notation on the copyright page, why don't we track printing information in the Notes? If there is a change in price/cover art it would make sense to consider it another 'publication', but not for just a printing. Holmesd 10:49, 26 Jan 2008 (CST)
We have struggled with this issue for a long time and, unfortunately, it's not always so simple :( To quote one of my recent posts:
Granted, in many cases printings can be almost indistinguishable, but consider these two printings of Alexei Panshin's Star Well. The first one appeared in October 1968, the catalog ID was G-756, the price was $0.50, the page count was 157 and the cover was done by Frank Kelly Freas. The second one was published in August 1978, it used an ISBN (0-441-78405-4) instead of an old style catalog ID, the price was $1.75, the page count was viii+211 and the cover was done by Vincent Di Fate. So here we have two books that are as different as any two paperbacks can be, yet according to the publisher they are two printings of the same edition!
When dealing with a later printing of an edition that we don't have on file, we usually create two separate Publication records. The first record is for the actual printing that you are verifying and it should be dated 0000-00-00 if you don't know that printing's publication date. The second record is for the first printing of the edition and is created based on the publication information found in the later printing, with the source of information clearly stated in the Notes field and the record left unverified. That way we capture as much information as possible and our users can get a pretty good idea of the history of that edition. Unfortunately, this approach doesn't work too well when the imprint was (deviously) changed in between printings, so it's not a rule but rather a guideline of limited applicability :) Ahasuerus 13:35, 26 Jan 2008 (CST)
P.S. Since this question has been asked 3 times in the last week, I have added it to the FAQ. Ahasuerus 13:46, 26 Jan 2008 (CST)
Of course 10-year-old kids can do cover artwork! How else do you explain this rubbish? :-/ BLongley 14:26, 26 Jan 2008 (CST)
I agree, this wasn't thought through fully - people forget Coverart records (well, until they try a normal search and find they can't find the novel among the covers and reviews) so this might be partly lazy cloning to create stub records: people should always take off the details that probably or possibly did vary - I know taking out number of pages does cause bibliographic warnings, but usually these stubs are created from later editions that only list the DATE of a previous printing in (probably) the same format. You have to know the publisher well before you can guess anything else. I'm not sure why the Artist is still listed for prior editions to the verified one: and the coverart date should be Unknown (0000-00-00) or at worst for the date of the first verified, dated pub with that art. You can see possible problems if and when we canonicise publishers too - this looks like multiple attempts were made to improve on the '0000-00-00' ugliness for each printing, whereas one or other would have done. BLongley 14:26, 26 Jan 2008 (CST)
Holmesd - after a year of dealing with these, I've come to the conclusion that keeping things at the lowest level of detail is necessary for now UNTIL we can roll-up the data meaningfully. We have a lot of duplicates that we don't KNOW are duplicates: e.g. someone enters dated editions from a respected catalogue of publications, someone else enters them from books which don't show the date on the physical publication, someTHING has entered data from Amazon - until someone provides more detail on why there's, say, an Ace 1973-04-00 edition with no printing number, and someone looks at all the 0000-00-00 publications to see if any have a note that matches, we can't match the records together. Making the printing number more visible, against the days when we might have to read the notes of dozens of 0000-00-00 publications, has been an experiment that certainly helps at times: and cloning FORWARDS rather than backwards doesn't tend to cause the problems Mike points out. But roll-on Printing number support anyway - it'll be a start, if not a solution! BLongley 14:26, 26 Jan 2008 (CST)
Holmesd - Keep track of the printings in notes seems like a good idea. However, I tried it for a while in my personal book database and discovered that in the long run it tended to introduce error and confusion. Unless the notes were incredibly detailed it was sometimes unclear what comments may apply to a particular printing as publishers often do introduce subtle changes from printing to printing in the layout and formatting of things such as the price codes, copyright page, etc. One record per printing turned out to be vastly more reliable as then I could add notes about a particular physical publication. Later, if I looked at the same book again, or another copy of the same printing, I could compare it against the notes and know that if a discrepancy came up that it was about *this* printing and that the research to locate the source of the error/discrepancy is much easier. It also simplifies verification in that we can verify with confidence a single printing but can only make assumptions that the same information is true for other printings. Marc Kupper (talk) 01:21, 27 Jan 2008 (CST)

Entropy - Prices

I have added Entropy data for prices to the ISFDB:Data Entropy page. Only 0.89% of the publications don't follow the standard price convention, but it still adds up to 1,027 records that need to be reviewed manually. What would be the best way to handle these problem records? I could create a Wiki page for them, but it would be huge and unwieldy. Should I break it up in chunks, 100 per page? Ahasuerus 22:07, 26 Jan 2008 (CST)

I guess that's the only way for now, until Publication Searches are fixed. When those work, it's often better to describe how to find the duff pubs dynamically using ISFDB, rather than using an ever-increasingly out of date static list. For example, if a significant proportion are British prices ending in 'p' then I'd prefer to search for '%p' and fix from that list rather than a Wiki page. BLongley 06:04, 27 Jan 2008 (CST)
Still, if a Wiki is the only solution for now, can you put the verifier id against the pubs that need looking at? That either lets us clean up our own mistakes, or allows people to choose to avoid pubs that need consultation with someone else. BLongley 06:04, 27 Jan 2008 (CST)
Sure, I can add a column for the pub's Verifier. As far as British prices go, at this time I have things like "20p" lumped together with "6/3", "3/-", etc under "Old British prices", but I can separate them if necessary. Most of the "bad" prices are "$3.75 (Can $4.50") or "None" or "pp" or "$A5.00" (instead of "A$5.00") or "3.75" (instead of "$3.75") or "DM 12.00" (instead of "DM12.00") plus some general purpose garbage. It's probably easier to start by posting a list of "Nones" and then, once they have been fixed, move on to other problem records. Ahasuerus 06:00, 28 Jan 2008 (CST)
And let's not forget about prices like "12 zł" (perfectly valid Polish price). --Roglo 11:49, 28 Jan 2008 (CST)
I have the Italian lira, the Japanese yen, the German mark and a few other things accounted for, but I am sure I will keep adding currencies to the list :) Ahasuerus 12:50, 28 Jan 2008 (CST)
What, no Altarian Dollar, Flainian Pobble Bead or Triganic Pu yet? We must be missing several editions of the Hitchhikers Guide to the Galaxy still. ;-) And I'm sure I've entered some publications with Galleons, Sickles, and Knuts. BLongley 16:54, 28 Jan 2008 (CST)
I checked: prices in Galleons are even rarer than prices in Guineas. We do have a couple of books verified as priced at "14 Sickles 3 Knuts" though. As usual, feel free to ask me to explain British prices: I also understand shillings, tanners, bob, ha'pennies, thrupenny bits, sixpences, crowns, half-crowns, etc. Guineas and farthings and groats are a bit before my time, but I'll have a stab at those too if needed. BLongley 17:05, 28 Jan 2008 (CST)

URL permissions

I have created an ISFDB:Data Consistency/Disallowed URLs page to keep track of how many URLs we point to on a domain by domain basis. We may want to check with the owners of the domains that we haven't secured permission to point to yet and, if they refuse, zap the URLs. (Whether the URLs are still valid is a different question, one that I will explore in a later script.) Ahasuerus

I have added links to the problem publications whenever the number of pubs didn't exceed 10 per Web site. Hopefully, it will make it easier to clean them up. Oh, and it turns out that we do link to 1 MediaWiki image after all. Ahasuerus 23:50, 30 Jan 2008 (CST)

Duplicate Series Names

I have created an ISFDB:Data Consistency/Duplicate Series Names page to monitor and correct duplicate series names. As many of you know, duplicate series names result in our software displaying individual titles multiple times and causing other kinds of mischief. Sometimes these duplicate series are really just one series, e.g. David Brin's Out of Time, and all titles need to be consolidated under one of the series. Other times, one series is a sub-series of the other, e.g. Duel Of Sorcery, and the superseries needs to be renamed to something like Duel Of Sorcery Universe.

Please keep two things in mind. First, Series record can't be deleted at this time. Once you empty a series by moving all of its titles to another series, you will want to rename it to something like "<old series name> - To be deleted" so that we can get rid of it once we have the ability to delete series. Second, apostrophes have been known to cause problems with Series names, so it may be safer to rename the series that you are changing to something that doesn't have an apostrophe and then change it back when you are done. Third, due to yet another software limitation, you can't remove a series from its parent super-series at this time. Due to these limitations, I would recommend than only moderators work on this project since it requires a lot of intermediate steps and it's hard to explain what you are doing when you can't approve your own submissions. Ahasuerus 00:24, 27 Jan 2008 (CST)

Award Displays

IIRC one of the significant displays under the old compiled ISFDB was to display all titles that won (or were nominated for?) a given award in a given year. This info is in our database, but there does not seem any interface to run the relevant query and display the results. Do others think that this would be a desirable feature? If so, should it be put on the feature request list? Or am I missing the proper way to access this info? Ideally a more complex search, allowing for a range of years and/or awards might be nice, but that is probably feature creep at this point.-DES Talk 03:22, 27 Jan 2008 (CST)

We have the pages, but you're right, no search. So I knocked up an Awards Wiki page for now. Feel free to pretty it up a bit. BLongley 07:52, 27 Jan 2008 (CST)
That is good, but of course it will not be visible to most users (who do not get onto the wiki). Still, it provides a way for anyone to get this data who wants it and asks, and it costs no coding. -DES Talk 10:20, 27 Jan 2008 (CST)
It so happens that Al is in the middle of redoing Award support :) Ahasuerus 23:55, 27 Jan 2008 (CST)

January 27 backup uploaded

The latest backup has been uploaded. The full backup has grown by another 74 Mb over the last week, a sure sign that we need to start deleting some pages' histories while I can still get this old and tired notebook import the 1.7Gb file in the MySQL database. In addition, Al posted the latest version of the Python scripts earlier today. Ahasuerus 23:55, 27 Jan 2008 (CST)

P.S. I went ahead and archived all August-December 2007 discussions. Take a look at ISFDB:Community Portal/Archive/Archive09 if you think we need to resurrect any of them. Ahasuerus 00:02, 28 Jan 2008 (CST)

Changing the misspelling of the title in an award

I added this pub to a publess award and then merged it with the correctly spelled title record to link the two. The award was attributed to the Dillons, but the work is actually a chapbook of Nancy Willard's poem which they illustrated. And the title attributed in the award is misspelled as well ("Heironymus", instead of "Hieronymus") I know we can't edit awards, but maybe Al can? Thanks. Mhhutchins 19:03, 29 Jan 2008 (CST)

Al can change any field in any database record directly using SQL commands, but it's always somewhat dangerous and sometimes time consuming. Since he is currently working on an award editor, he tends to concentrate on that as opposed to on fixing individual problems. Once he is done, we will have to do a lot of cleanup in the award area, so it would be probably easier to make a note of this problem and attack it once we have the editor in place. After almost two years of tiptoeing around award issues, revenge will sweet! :) Ahasuerus 20:04, 29 Jan 2008 (CST)

Synopsis

I notice that very few titles seem to use the Synopsis field. Is it deprecated, or is it just that few people have entered one. Is entering Synopses for works I have read encouraged? Are there any style standards for a Synopsis? Should spoilers be avoided? -DES Talk 19:35, 29 Jan 2008 (CST)

The Synopsis field hasn't been deprecated, it just hasn't been at the top of the list of priorities. There are two main sources of Synopsis information, editor-submitted summaries and brief publisher descriptions that are usually printed on library cards (the Library of Congress has an online database).
Spoilers can be tricky. Our Help pages currently state that a "short non-spoiler synopsis can be entered here", but it's easier said than done -- in certain cases even Tags can be spoiliferous. Heck, even titles can include spoilers, e.g. Flight into Yesterday (aka The Paradox Men).
Ideally, we would have two fields for Synopsis data: one would avoid spoilers and be used by prospective readers and the second one would include spoilers and be used by scholars. I doubt we will go that far, though, so I guess we should just try to follow the current guidelines, use common sense etc. Emphasis on "try" :) Ahasuerus 19:58, 29 Jan 2008 (CST)
I'm from the scholar school. With your Wikipedia experience you may have some ideas. One of my thoughts was that we could link to Wikipedia reviews which allow spoilers. The only problem with Wikipedia is that articles seem to be removed or truncated with remarkable frequency. It seems like editors have to fight to preserve nearly everything they write. I'm wondering how long a synopsis of a two page vignette would last there.--swfritter 20:21, 29 Jan 2008 (CST)
I'm one of those wikipedia editors who objects to excessivly long plot summeries there, and has fought to trim quite a few. I was on the losing side of the argument about removing spoiler warnings from wikipedia articles -- I felt that some of them, at elasr, should stay. The issue on wikipedia would be not how long the piece was, but how "notable" it was, and also whether the synopsis dominated the articel, or was balanced by "real-world immapct" with proepr citastions. On the other hand, most editors have given up trying to trim plot summeries and such in any article about a work of fiction with a sizable dedicated group of fans, such as "Buffy" or "Dragonlance". It jsut isn't worth the hassle. -DES Talk 20:37, 29 Jan 2008 (CST)
Overall, I don't think we have major problems with Wikipedia's policies, but they are a general purpose encyclopedia and can't support the level of detail that we are after. There are ISFDB titles that we will happily provide Synopsis data for that Wikipedia would immediately (and understandably) delete. For example, I am sure we will eventually have a Synopsis for that article in Future announcing the election of John B. Michel to succeed Doc Lowndes as the director of Futurians after the latter turned pro. It was a major event in the SF world at the time, not only indicative of the Futurians' rise in the pulp arena, but also suggesting that Michel's faction was still dominant among the Futurians. OTOH, it would mean little to Wikipedia unless, perhaps, used as a footnote.
It so happens that a number of us were active Wikipedia editors at various points in time, e.g. at one point I contributed to hundreds of articles and wrote or rewrote dozens of them ranging from Gerry Turner to Leon Trotsky, but it's a very different animal in many ways. Ahasuerus 00:25, 30 Jan 2008 (CST)
Absolutely. I did not mean to imply that policies that are right for Wikipedia ought to govern the ISFDB, their purposes are different as you say. And I don't for a moment think that I am the only person with relevant Wikipedia experience (insofar as such experience is in fact relevant). I was reacting to swfritter's comments which seemed to be criticizing Wikipedia's policies in ways with which I disagree. Not that I think Wikipedia's polices are perfect, indeed I was rather known as a critic of them to some extent. In any case, The question here is what the ISFDB's policies are or ought to be. I think i now understand the current consensus on synopses: feel free to enter them, but they are not required and may editors have other things as higher priority; when entering, try to avoid blatant spoilers; if entered, they should be reasonably factual, and be neither blurbs nor pans. Have I understood correctly? -DES Talk 10:45, 30 Jan 2008 (CST)
That about sums it up :) One minor irritation is that you can't submit Synopsis and other Title-level data when creating a New Novel/Collection. You can't even submit Interior Art or Essays when submitting a new Novel, which necessitates 2 or 3 passes per book and that can be a turn-off for new editors who can't approve their own submissions. Oh well, something else to add the the list of requested features... Ahasuerus 12:26, 30 Jan 2008 (CST)
But essentially you answered my question about synopsis on Wikipedia. They need to be of relatively significant works and it cannot be assumed that they will stay there or that their contents will not be changed as a result of changes in policy. That's just the nature of the beast. Democracy can be a little messy. I might also note about synopsis data in The ISFDb that variant titles also have a synopsis field. I guess the obvious logical place to put the synopsis is in the synopsis field for the canonical title but that when viewing a variant title from a pub.--swfritter 15:58, 30 Jan 2008 (CST)
Well, almost. The first thing is that anyone who depends on any article on Wikipedia having contents remain substantially unchanged is gambling at best. It is not so much that a "less significant" work will automatically have a plot summery deleted. But Wikipedia policy says that an article should not consist solely or primarily of a plot summery, nor should such a summary be detailed enough to be a derivative work of the original. Plot summaries are supposed to be balanced by content that shows the effect or influence or importance of the work in the world at large, and if they cannot be, should be cut down. These policies are enforced highly inconsistently, and are more likely to be enforced on a "less significant" work. However, if a fictional work or series, has a hard-core group of fans active on Wikipedia, it is likely to have more in the way of "in-universe" content than policy suggests, and most editors won't bother to try to overcome this. -DES Talk 18:33, 30 Jan 2008 (CST)
I've noticed the options to add a separate Synopsis/Note on the Variant title, but I don't recall using such though. I do add title NOTES at times - recently to add the variant number within a Series, when the series number changed between UK/US publishers. That was for the same title though, and gets usefully displayed: entering such on a Variant title would be pretty lost on our general displays. We seem to have plenty of extra places to record data if we want, but the display logic won't help make the intention clear. I think we need to look ahead a bit and guess Al's intentions, or demand he tells us his intentions before he "fixes" duplicate entries that we've misused to NOT be duplicates. BLongley 17:16, 30 Jan 2008 (CST)
It would make some sense for certain data to be shared by the various titles in a variant title setup.--swfritter 20:07, 30 Jan 2008 (CST)

(unindent) This issue has come up before, but I don't think an elegant solution has been found yet. When two+ variant titles represent the same text, it doesn't make sense to have separate Tags, Votes, Notes, Synopsis etc fields for each Title record. If anything, it can be misleading and confusing, resulting in data duplication and dispersion. On the other hand, if we allow variant titles that represent related but different texts ("abridged", "restored", "revised", etc), then it's entirely possible and even likely that we will need to populate these fields with different information. I suspect that until this underlying issue is addressed, there will be ambiguity about the use of Title level fields in the context of variant Titles. Ahasuerus 20:59, 30 Jan 2008 (CST)

It is my view that the ideal design would be to have some sort of relationship field, so that when one title is an abridgment, expansion, revision, etc of another title, or is "based on" another title in some way (such as a fixup), two separate non-variant title records can be linked via the relationship. (See ISFDB:Proposed Design Changes#Based on for more detail. Of course, this requires code changes, and may not be feasible. Expanding the use of variant titles to cover this would be possible, but it would not handle the fixup situation, nor the case where a work is significantly expanded with no change of title. I agree that we can't settle the use of fields in variant title records until we decide finally what they are and are not to be used for. I also agree that if variant titles are only to be used for essentially identical texts, then there is no reason for a separate synopsis field, or separate tags. However, even then a separate notes field might help to document the circumstances in which the variant title appeared. -DES Talk 10:52, 31 Jan 2008 (CST)
We have been discussing adding a new "based on" one-to-many relationship field to link fixup novels with the stories that they were based on. We have also considered adding another one-to-one relationship field to describe the nature of the relationship between Variant Titles, e.g. "abridged", "revised", etc. What you seem to be proposing is adding only one new one-to-many relationship field that would cover the "based on" type of relationships as well the "abridged"/"revised"/etc type of relationships (and breaking the currently existing variant title relationships between all non-identical titles).
Sounds quite promising, but we'll need to run it by Al and consider all structural and display implications of the design change. <wanders off to play with data elements> Ahasuerus 04:27, 1 Feb 2008 (CST)
Yes that is my suggestion, made in ignorance of the detailed code and structural implications It seems to em that functionally they are versions of the same thing. in particular, a fixup may also be an expansion, as it often adds new content, or a revision. However, if it seems better to implement this as two distinct relationships, and use variant title for revisions, expansions, abridgments that do not involve multiple parts, so be it. Note that if variant titles are used for such cases we need to account for situations in which the actual title is changes, and other situations in which it is not. I am going to copy this exchange over to ISFDB talk:Proposed Design Changes and hope for more discussion there. -DES Talk 09:19, 1 Feb 2008 (CST)

E-book pubs

I "own" (have paid for and downloaded) copies of a number of the ebook versions of SF works put out by Bane books through their Webscriptions service. Is it a good idea to enter those as additional publications of the titles involved? Obviously, the text is usually identical or near identical to the hb or pb version published at the same time -- Baen does not usually publish an ebook unless it is also publishing a non-ebook version at the same time. But in some cases I have the ebook and not the "dead tree" version, and no one else has entered a very similar printing. What is our general policy on ebook printings of titles also published physically? The Baen versions seem to be stabler than some ebook sources are, but there is never a guarantee, of course. -DES Talk 11:50, 30 Jan 2008 (CST)

As per ISFDB:Policy, we include "e-books with ISBNs", so if we are missing any (and I am sure we are missing a lot), it's not by design. I am afraid you will discover that there are big holes in our coverage of certain areas, especially if that area has been unsupported until relatively recently, e.g. fanzines. Ahasuerus 12:32, 30 Jan 2008 (CST)
Fine, no one siad this is the final stage. ("Resistance is futile, you will be catalogued!") Just didn't want to waste time and server space putting in content that was considered pointless. -DES Talk 12:46, 30 Jan 2008 (CST)
Jim Baen's Universe is in the system because it meets the standard of having an ISSN. In addition it is available in more permanent downloadable versions. I subscribe to Analog and Asimov's through Fictionwise and read them on my palm T/X but it is no priority for me to get them into the system. I have also been using tags to document Project Gutenberg versions. See my User Page for a list of the links.--swfritter 16:17, 30 Jan 2008 (CST)
Thanks. The websubscription versions all, i think, share the ISBN of the print version issued at the same time. -DES Talk 16:38, 30 Jan 2008 (CST)
Yes they do. I find it a bit annoying since that makes then look like the book entry. For the couple I've done I think I put a # in front since it isn't really a ISBN. Dana Carson 19:19, 31 Jan 2008 (CST)
Jim Baen's Universe strangely enough have both ISBN's and ISSN's but the other mags have only ISSN's.

Display of short fiction in series

When a work of short fiction is part of a series of fiction works, it displays on the author's bibliography page with the notation [SF] after it. I presume this stands for "ShortFiction". But given the name and focus of this site, it might be confused for "Science Fiction" -- a user might mistakenly assume that we visibly distinguish between science fiction and fantasy, say. Perhaps the display could use [short] or something similar, instead. Obviously this is not a top priority item, but OTOH it ought to be a very quick change, if people agree that it should be done. -DES Talk 11:54, 30 Jan 2008 (CST)

That's right, "[SF]" stands for "short fiction" when displayed next to a title in a series. I think I mentioned this issue a while back, but it didn't seem to elicit strong opinions one way or the other. Perhaps we could create a new feature request? Ahasuerus 12:21, 30 Jan 2008 (CST)
I just went to do that, on the wiki principle that one does for oneself when possible, rather than asking someone else to do it for you, but ISFDB Feature List is protected, so I can't. Perhaps semi-protection would be sufficient for such pages. <pause> I Just checked, and semi-protection is only available with wikimedia version 1.9 and higher. I understand why we can't upgrade right now. Sorry. -DES Talk 12:44, 30 Jan 2008 (CST)
Sorry, I should have checked first. It's unprotected now, featurify away! Ahasuerus 13:25, 30 Jan 2008 (CST)

Should i merge?

This essay and at least some publications of this essay are the same text. The titles don't match (the first includes "Introduction:") and one of the publications of the second has been rewritten slightly (see note). Should these two works be merged? Any advice? -DES Talk 17:00, 30 Jan 2008 (CST)

If there's a difference worth noting, then (IMNSHO) we shouldn't usually merge them. To do so would effectively over-ride what the Noter intended. "Worth" is of course a subjective issue and if all the notes say are "this was published later there" and "this was published earlier here" then it might be worth overriding the Noter. If in doubt, leave more notes as to why they should or shouldn't be merged. E.g. "This version is longer" is good enough a warning to stop me merging it. Explaining that Heinlein or Clarke kept revising a similarly titled essay every few years to see how predictions matched up with reality would be even better. Of course, few of us are lucky (or unlucky?) enough to have multiple versions of the same work to check, so merging may or may not be controversial. BLongley 17:37, 30 Jan 2008 (CST)
Actually, the only "noter" here is myself. The essay in question first appeared as the introduction to a collection (Hammer's Slammers by David Drake). It was then reprinted as an essay in an anthology (There Will be War), with a few sentences changed so as to remove the explicit references to the Drake work. We have two title records for it. One is attached to a single publication of the Drake book, and includes the works "Introduction" as part of the title. The other drops the word "Introduction", and is attached to two other publications of the Drake collection, and to the Pournelle anthology. I added a note to the later, documenting the re-write -- and only then found the second title record. I have the first printing of the Drake, and an early printing of the Pournelle, and can verify that the versions of the essay in those two pubs are the same except for the minor rewrite I have mentioned. I have no reason to believe that there was any rewrite in other versions of the Drake collection. Is this a mess, or what. -DES Talk 18:24, 30 Jan 2008 (CST)

Mass Notification of change to Silverberg's "Road to Nightfall"

I have changed the length to novelette and it appears in four other verified pubs.--swfritter 21:07, 30 Jan 2008 (CST)

It's my understanding that the "length" field is very subjective and so I don't have a problem with people changing it, even if it's included in verified publications. There may be a few stories where someone has the exact word count. Hopefully that gets into the title-notes and then it's a matter of what definitions of the various lengths you use. Marc Kupper (talk) 00:54, 2 Feb 2008 (CST)

Authors that exist only due to reviews

The project seems to be progressing nicely, if fitfully, and I think has been pretty useful in clearing up some typos in reviews, adding books we'd previously overlooked (which I think was one of the main reasons we recorded reviews in the first place), sorting out some variant names, and is getting down to the odd titles that deserve some discussion. And maybe some discussion over HOW we've fixed some of the "problem" authors. I suggest people read The Rules of Acquisition again before commenting. But we're already bending some of the Rules of Acquisition to fix things. Here are my thoughts for now:

  1. Adding missing SF titles is fine, and to be encouraged.
  2. We need to clean up some reviews where a translator of narrator or artist is credited as a co-author.
  3. Adding missing Nongenre titles is fine, if reviewed in an SF pub. Make sure they're marked as such though so we don't go grab everything by the author/editor.
  4. Adding missing Nonfiction titles is fine, if reviewed in an SF pub, OR is SF relevant. I'm thinking of Bibliographies, collections of SF artwork, even books recording second-hand/collector book-shops that deal in SF.
  5. We need to STOP recording reviews of things we don't actually want as "reviews". This is the big controversial one: if we don't want Manga or Graphic Novels, then we shouldn't create review entries that link to an "author" that only exists because of reviews of such. It's also where the borders of the Rules of Acquisition need to be reviewed - is a review of a book of the making of a film that is probably SF in or out? Currently we wouldn't include the film, we would include a novelization of the film, we'd exclude a graphic novel of the film, a book of the making of the film is in the vague area in between. We currently even have reviews of the soundtrack of an SF film recorded!

Some of us are coming to this as a database problem (links should work), some from a Web basis (links should work) and some from a "just record the data, F**K the links" basis. It's easy to stop bad links - the question is where? I'd prefer to have a dead-end from a "link" that needs to be recorded: e.g. a review of a Star Trek Calendar in a magazine needs to be recorded to keep the magazine contents intact, but does NOT need to be recorded with an author: a Reviewer link is enough. A link to a Science title can be recorded, but if it's marked as a NON-Genre title we can stop further links at that point. Recording artists is fine, but I'd suggest we stop at their genre works, books ABOUT their genre works, and definitely not add extra. BLongley 19:36, 2 Feb 2008 (CST)

Here's an example of stopping scope-creep. Good move? Bad move? Would it help if I moved this to Rules and Standards? I normally only move things there when I have a concrete proposal to make, and sort of hoped people would look at the project and come up with some ideas of their own first. There was a little discussion here before this action. BLongley 16:26, 3 Feb 2008 (CST)
An interstign question -- I'm not sure where I think we should cut such things off. Did the reveiw treat the boook as SF or SpecFic in any sense? I recall a review, in one of the standard genre mags (I think it was Asimov's) of Tom Clancy's The Sum of All Fears, which argued that the book was in fact science fiction. Would such a review justify listing that book in the DB? Clancy's other books? Many of them involve fictional technology in advance of any that really exists, but they are not what i usually mean when i say "Science Fiction". -DES Talk 16:34, 3 Feb 2008 (CST)
I personally think that we need to record a reviewed book somehow: I hate non-functional links, and suspect that demanding a book review be OMITTED from a Magazine will not be popular here! Changing to an Essay should satisfy both camps, but there'll always be an argument over whether it's SF or not. When there IS an argument, I'd suggest inclusion - e.g. Analog Science Fiction and FACT will lead us to reviews of many Science-Fact books that we can then encourage to be the end of it. Asking everyone to check a review to see what it's actually a review OF is going to be unpopular too. Anyway, this is a discussion - I have a pile of Tom Clancy related books I've never read that LOOK SF, but I can't say for sure. If I have to READ them before we can decide I'll be gone for a while. ;-) BLongley 17:53, 3 Feb 2008 (CST)
Fair enough, some common sense is needed. As for Clancy, i have read most of his novels, including the one at issue. They are set in an indeterminate near-future/present time (as of publication), with a few tech developments that are not RL, but fewer than the James Bond books have. Eventually, as later works assume the reality of the fictional events in the earlier books, you might sort of call them "Alternate history" but no more so than the seqels to Advise and Consent (or for the matter of that West Wing) IMO. I don't personaly shelve them among my SF. But the point could be argued. -DES Talk 18:05, 3 Feb 2008 (CST)
I feel Ian Fleming is under-represented here, even if the James Bond books are considered non-genre. We put a lot of effort into The Man From U.N.C.L.E. even though his exact contribution is debatable. I'm happy if James Bond is IN - I can do some more primary verifications. And the novelisations of some films probably SHOULD be in, e.g. Moonraker the movie is SF IMO, but Moonraker the original Novel is just a thriller. What I want are rigidly defined areas of Doubt and Uncertainty. :-) BLongley 18:27, 3 Feb 2008 (CST)
Bill's original dividing line, i.e. "books are in, the rest is debatable", seems like a good starting point. If we can list all categories that should be listed as "essays" rather than "book reviews", i.e. movies, TV shows, games (tabletop, computer and online ones), calendars, comics, manga, audio recordings, etc, then the rest can be made more manageable when we examine them on a case by case basis. Hopefully :) Ahasuerus 22:53, 3 Feb 2008 (CST)

SF onthology?

A new editor has just posted an interesting article/essay on Chris J's Talk page. I am not sure how/whether we could take advantage of it, but I thought I'd link it here for those who do not check Recent Changes every 10 minutes :) Ahasuerus 22:06, 2 Feb 2008 (CST)

DOIs

Several of my downloaded ebooks from Baen books show a "DOI" number instead of an ISBN, which according to doi.org is a "Digital Object Identifier". It appears to be a system like the ISBN system, but for any and every digital object. It appears that all DOIs start with the text “10.” Followed by a mumeric string (the “prefix”) which identifies a group or system of identifiers, not unlike the publisher section of an ISBN. Following that is s slash (/), and following that is the “suffix”, which can be any character string at all, including an ISBN. Different prefixs may support different suffix coding schemes. DOIs do not have the same kind of invariable checksum that ISBN's do, although a particular suffix scheme might include a checksum. All DOIs can be uiniquely and persistantly "resolved” to either a single URL, or to a URL plus additional metadata. The URL may point to a copy of the digital object, or just to information about the object. To resolve a DOI, one can construct a UTL starting with “http://dx.doi.org/”. For example “doi:10.1000/123” would be resolved from the address: "http://dx.doi.org/10.1000/123". There is a lot in the doi.org FAQ comparing DOIs to ISBNs -- it might be worth reading, particularly if other ebooks start using DOIs. (I gather that a major system of providing online copies of academic journal articles is already using DOIs for individual articles.) A problem is that at least some of the Baen books DOIs appear not to be validly registered DOIs. -DES Talk 23:56, 2 Feb 2008 (CST)

Authors that have little or no spec. fiction

As I've been working my way through The Data Consistency non-fiction mismatches I'm finding authors that have little or no spec. fiction on their summary page. Some of these authors have a large body of work but most of it is non-genre novels or non-genre non-fiction. Should we retain only the speculative fiction and delete the rest or do we want to keep all their works? The help page suggest only the spec. fiction should be retained and the rest should be deleted. Here's a list of the authors I've found so far.

  1. -Andrew M. Greeley[4]?
  2. -Colleen McCullough[5] Appears to be romance/historical romance no spec. fiction
  3. -Mary Hoffman[6] mostly pre-school books but has some spec. fiction for teens and pre-teens.
  4. -Robert Louis Stevenson[7] most of his work is non-genre or non-fiction
    • I agree, most of what is listed for RLS does not belong here -- a number of items listed under "novels" are not even fiction. However, IMO The Strange Case of Dr. Jekyll and Mr. Hyde does belong here. -DES Talk 14:02, 3 Feb 2008 (CST)
  5. -Chinua Achebe[8] Except for one shortfiction story the rest appears to be all non-genre & non-genre/non fiction.

If anyone knows more about these authors let me know.Kraang 11:07, 3 Feb 2008 (CST)

Stevenson has a half page in Tuck, though I'm not familiar with most of the stories listed. The Stevenson page could use a bit of cleanup.--Rkihara 15:07, 3 Feb 2008 (CST)
I've been whacking the Mary Hoffman page about a bit for a few hours today - it doesn't look much better now than when I started, but I've learnt some more Spanish in the meantime. ;-) I'm sure there IS some SF in there, but till we can sort the wheat from the chaff I'd keep it, categorise it like we normally do, and learn in the meantime. I'd be in no hurry to delete it - for instance, some of the "co-authors" are actually illustrators, and keeping such illustrators can help us find OTHER authors worth deleting eventually. There's also translators we might want to watch. We're not immune to Sturgeon's Law - we DO have a lot of crud here, it's just not easily found when you come here with a good idea of what you want. I encourage people to pick authors at random, now that we have an Active Author Directory, and try and clean up their bibliography. Yes, we might end up with a perfect Bibliography of someone we don't actually think BELONGS here - but we'll have checked it thoroughly first, and we don't have any OFFICIAL target practice. Let's just Round 'em up, put 'em in a field, and BOMB THE BASTARDS! (Eventually.) BLongley 16:54, 3 Feb 2008 (CST)
It's true that Robert Louis Stevenson's bibliography is our poster child for the old children's ditty updated for the 21st century: "Don't ever trust a bot AI, no matter friendly how!" At one point Dissembler saw a couple of legitimate SF pieces by Stevenson already in the database and went and grabbed every Stevenson book that it could find on Amazon.com... which made his bibliography completely unusable. It will take a lot of work, mostly of the slash-and-burn variety, to get it in shape. At the very least, non-genre works need to be clearly marked as such, which is not always easy since at this time only Novels can be profitably converted to "Nongenre"; any other title type (e.g. Collection, Anthology or, heaven forbid, Short fiction) will break the display logic in various unspeakable ways. Certainly something to work with Al on when he gets to it.
Then there are bibliographies of non-genre writers who were manually entered by their fans years ago. Some of them are very comprehensive and it would be a shame to destroy something that other people could potentially find useful. We have discussed finding a loving home for these bibliographies, but if it proves impossible, then there is always Wikipedia. Unfortunately, Wikipedia, for all its strengths, is a very fluid place and any article may be drastically altered or even disappear at any time. For example, at one point I compiled a reasonably comprehensive secondary bibliography for the [http://en.wikipedia.org/wiki/John_Randolph_of_Roanoke John Randolph of Roanoke article, only to see it cut in half.
Finally, there are author records in the ISFDB that may appear to be of marginal SF interest based on what we currently have on file, but a further investigation may reveal that the author in question was indeed responsible for a significant number of SF works. Earlier today I stumbled on Martin Noble's bibliography, which at the time was limited to his novelization of Who Framed Roger Rabbit. After poking around some, I found and entered 4 more SF novelizations (and there are also some borderline SF juveniles that I haven't entered yet), so at this point I would be comfortable entering his Nongenre work as well, something that I wouldn't have considered useful 3 hours ago based solely on the data that we had on file at the time.
Hmmm, I have "Bloodbath at the House of Death" but never really considered it relevant: still, I guess it's technically "Horror" although the "Comedy" aspect is more important overall. BLongley 14:15, 4 Feb 2008 (CST)
Checking the plot summaries at IMDB and Wikipedia, it looks like it's supernatural horror with a modicum of an SFnal background :) Ahasuerus 14:36, 4 Feb 2008 (CST)
The film was mainly a vehicle for Kenny Everett (originator of the "Round 'em up, put 'em in a field, and BOMB THE BASTARDS!" quote I posted above - how's that for coincidence?) to move up from TV. It took several characters, and even an actress, from the TV show. I've never actually seen it, just read a bit of the book and concluded that it probably wasn't worth continuing. BLongley 14:52, 4 Feb 2008 (CST)
It wasn't exactly a coincidence :) I looked him up on Wikipedia and IMDB after you posted and when I saw that one of his vaguely SFnal movies had a novelization, I checked whether we had the book on file. The rest is history! Ahasuerus 15:06, 4 Feb 2008 (CST)
I guess the bottom line is that each case is unique and many require a careful investigation and preferably a discussion before we turn on the blowtorch. Of course, we also have many manga, comics and RPG records, where things are much more cut and dry and we typically don't need a lengthy discussion before exterminating them. But even there we have to keep in mind that more and more publishers and other intellectual property owners commission bona fide SF novels based on superhero comics, RPGs, etc, so it's not as simple as it was 40 years ago. But then again, what is? :) Ahasuerus 21:21, 3 Feb 2008 (CST)
Is series support for non-genre on Al's list of things to do? Kraang 21:42, 3 Feb 2008 (CST)
Well, it has been requested, but I am not sure what Al's plans in this area are. I assume we will discuss new/requested features when Al gets to that section of his list the way we discussed new features the last time. Ahasuerus 22:01, 3 Feb 2008 (CST)
Should we turn a blowtorch on this author Colleen McCullough [9] (appears to be romance / historical romance no spec. fiction)? Kraang 21:42, 3 Feb 2008 (CST)
I am unaware of any SF written by Colleen McCullough, but I may well be off. Anyone? Ahasuerus 22:01, 3 Feb 2008 (CST)
Not read any myself. A quick scan suggests that "A Creed for the Third Millennium" published in 1985 is SLIGHTLY futuristic, but probably not enough. Not sure about the Troy book - how speculative-fiction are Greek myths considered to be? BLongley 15:21, 4 Feb 2008 (CST)
One of the criteria we could use would be if the author has won or been nominated for any SF awards. Second, have any of there published works been reviewed in any SF/Fantasy/ Horror magazine and third have any of their works been verified by anyone. McCullough's works fail all of the above, so if anyone thinks we should keep here page speak now. I'll wait till the weekend and then zap her! Kraang 19:13, 4 Feb 2008 (CST)
That would catch all notable SF writers, but they generally are not the ones in doubt since they are, well, notable :) In McCullough's case, one reviewer mentions that "McCullough seems to vacillate between the no-nonsense realistic approach and the magical realism one. Oracles and prophecies abound in her book and are always truthful, but the characters act motivated by purely human and worldy causes, and many mythical themes receive explicit mundane rationalizations (e.g. it is said that someone who claims to have a God as father is usually a bastard). This makes the mythical components left in the story seem somehow unreal and fake." This seems to suggest that there is a supernatural component to at least one of her novels, although apparently a weak one. It sounds like the rest of her novels are non-SF, but it may be worth checking further before we zap them. Ahasuerus 19:25, 4 Feb 2008 (CST)

(unindented)Many of the historical romance novels written about this time period (and later periods) will incorporate these elements of the supernatural. I know my better half reads this rubbish, not like the high quality SF I read! :-) Being a numismatist I'm familiar with this time period and the gods were part of everyday life. They were as normal to the Greek's and Roman's as the current crop of religions are to us. If a current editor submitted any of her work it would probably be sent strait to the reject bin. I've also read all the Amazon blurbs and comments, all would suggest non-genre. Her web page would also support this opinion. :-)Kraang 20:02, 4 Feb 2008 (CST)

I have read her series centering on the Life of Julius Ceaser (The First Man in Rome. The Grass Crown, etc.) The tone is not unlike that is some works of "Alternate History", and it might well appeal to a good many SF fans. I wouldn't be suprised to learn it had been reviewed in an SF magazine, although I don't know of any such review. But the only "supernatural" elements are that the characters (usuprisingly) mostly belive in the Roman gods, as far as i recall. I wouldn't class these as SF. I havn't read any of her other work, and it might be that some of them are borderline SF. As to the criteria above, as a verifier supposed to verify that the work is in fact SF? The help pages don't tell a new verifier that this is part of the task. -DES Talk 13:31, 5 Feb 2008 (CST)

February 3, 2008 backup uploaded

Enjoy! :) Ahasuerus 22:45, 3 Feb 2008 (CST)

Thanks! (...but the date on the ISFDB_Downloads page is still January 27, 2008) --Roglo 04:20, 4 Feb 2008 (CST)
Sorry, fixed now. Ahasuerus 08:43, 4 Feb 2008 (CST)

What is the most frequently used title in SF?

I happened to come across this question in a Usenet discussion earlier tonight and whipped up a script to see if I could answer it. Here are the results if anybody is interested. Ahasuerus 00:11, 4 Feb 2008 (CST)

Grantville Gazette essays

I have been entering issues of the Grantville Gazette recently, including contents. Each issue has several nonfiction eassya included. I know that fiction and non-fiction don't display in the same series, so i haven't entered any sereies indication on any of these. But there are enough of them, with sufficiently common themes, that I think that an essay series, perhaps "Grantville Gazette Essays" would be a good idea. Does anyone else have any thoughts on thsi matter? -DES Talk 14:07, 5 Feb 2008 (CST)

Intentionally incomplete collection

I just added data on a pub of the collection Forty-Two Poems by James Elroy Flecker. I did so because it contains "To a poet a thousand years hence" which we already have listed and have only one pub for. However, this collection also contains 41 other poems, and i have no particular reason to think that any of them belong in the ISFDB. (On scanning, the first few do not seem to be SF in any reasonable sense). Should I none the less complete the contents of the collection, adding 41 title entries? Or what? -DES Talk 17:15, 5 Feb 2008 (CST)

Personally, I'm quite happy to enter incomplete Collections and/or Anthologies, so long as it's clear they are incomplete. E.g. This book took me ages to get all the links right for, and I have no intention of adding the "Detection" half. More recently (well, today in fact) I entered this magazine - I doubt I'll enter many/any more as the data is already available on the web, and even if it wasn't the NON-fiction reviews are often of books of reviews or critical studies of SF - not the Speculative Fiction I want to find here. Two many degrees of separation for my liking, but anyone can enter the rest if they like. BLongley 17:59, 5 Feb 2008 (CST)
I've also entered first magazine publications for SF works when they're in a Non-Genre magazine - some are obviously non-controversial, like The Strand Magazine, so I could enter the non-sf too: but when the first publication is in Playboy, do people really want me to go add all the other contents and search for a Cover-Image as well? (OK, I'm sure some WOULD, but I'm not going to!) BLongley 17:59, 5 Feb 2008 (CST)
Ok I'll add a note as soon as the epublication edit is approved. -DES Talk 18:15, 5 Feb 2008 (CST)
This issue comes up fairly often. On the one hand, it's nice to be complete and include the complete contents of each magazine issue/anthology where SF appeared. On the other hand, you would end up with a lot of essays and fiction that are completely unrelated to SF, which would make it harder to find the good stuff in the ISFDB. Besides, our software support for non-genre works outside of novels is pretty weak, so it would be an even bigger mess than one might first think. After going back and forth in mid-2006, we decided not to list non-genre works in primarily non-genre publications which we are only listing because they contain some incidental SF (New Yorker, Times, Playboy, Norton's Book of American Litereture, etc). We still want to create separate Publication records for these books/magazines, though, so that they could be physically Verified. And please add a note about the omitted non-genre contents to the pub :) Ahasuerus 00:39, 6 Feb 2008 (CST)
Note added. -DES Talk 10:21, 6 Feb 2008 (CST)

More on Ernest Bramah

I note that for Ernest Bramah we have a fair amount of detail on the "Kai Lung" series, I guess because these are considered to be fantasy, but rather less on the "Max Carrados" series. i am not sure why we have the latter listed at all. Well I suppose "The Ingenious Mr Spaniola" (title from memory, may be mis-spelled) might be called SF, but it is a stretch. Are we considering that Bramah is a "primarily Genre" author, and thus all his work should be included? If we are, I have a copy of the "Best Max Carrados Stories" and can add the contents and verify it. Otherwise, perhaps we ought to simply delete all the items in this series, and also his autobiography English Farming and Why I Turned it Up and his other non-fiction? (I think that The Secret Of The League can reasonably be classed as SF, based on the comments in the intro to a Max Carrados collection -- I have not read it.) I also note that Author:Ernest Bramah does not yet exist. -DES Talk 10:30, 6 Feb 2008 (CST)

That's right, Bramah wrote a significant amount of "pseudo-oriental" fantasy, which was popular and influential in his time. He also wrote a less influential novel about a future war/revolution, a very British thing to do in the last decades of the 19th century (which, as we all know, ended in 1914), and "future war" stories are explicitly included in the scope of the project as part of "proto-SF".
The problem with the Max Carrados books is that they are non-genre and our software doesn't support non-genre series at this time. Thus if we were to change these titles' type from Novel/Collection to Nongenre, they wouldn't display as a series -- see Lloyd Biggle, Jr.'s bibliography for an example of how his Sherlock Holmes books are displayed. It's a no win situation, I am afraid, which is why we are all waiting for Al to fix it when he gets to the display issues on his list. Ahasuerus 17:19, 6 Feb 2008 (CST)
Ok, but the question is, why are these non-genre titles included at all? The Rules of Acquisition say "IN: Works that are not related to speculative fiction by authors who have not published works either of or about speculative fiction over a certain threshold." and the next rule excludes such works by authors who do not meet this threshold. Do you consider that Bramah meets this threshold? If so, it seems to me that the detailed contents of the Max Carrodos collections should probably be entered. If not, they should probably be deleted. -DES Talk 17:55, 6 Feb 2008 (CST)
Ah, I see! Yes, I definitely think that Bramah, one of the 5 or 6 Big Names in pre-genre-pulp fantasy, meets the requirements for inclusion of non-genre works -- unlike, say, R. L. Stevenson, Honoré de Balzac, E. Phillips Oppenheim, etc, whose SF was peripheral to their oeuvre. I have been hesitant to include Bramah's non-genre stories because we have no easy way of handling them: the software assumes that anything labeled NONGENRE is a novel :( On the other hand, if we were to enter the Max Carrados stories as "short fiction", they would look like they are SF, which would be misleading. Another no-win situation and the reason why we have an outstanding request to beef up our support for non-genre works. Unfortunately, Al has higher priority things on his plate (e.g. bugs) that he will be taking care of first, so it will have to wait a bit longer. Ahasuerus 18:59, 6 Feb 2008 (CST)
I See. I thought of Bramah as a mystery writer who also wrote some fantasy -- now that I think of it that was mostly because that was the work of his that I have read and liked. How foolish. In terms of word count his fantasy is probably more than his mystery writing, but not by that much. I see why you fell that inclusion of the detail of his non-genre work would be a bad idea until software changes are made. Very well. I will not move on the matter pending such changes. (I must say, I think that there are quite a few names in pre-genre-pulp fantasy I think of as bigger: E.R. Eddison; Lord Dunsany; William Morris just off the top of my head. But that doesn't really matter, if he is "big enough" it doesn't matter if he is in the top 5 or the top 20.) -DES Talk 19:10, 6 Feb 2008 (CST)
Oh, I agree that at the very least Eddison (who didn't publish his major works until the 1930s), Dunsany, Morris and Haggard were more important (and generally more prolific), but then it gets murkier. Chambers was arguably more important and F. Anstey did much to popularize light fantasy, although only his first novel was really major. E. Nesbit was certainly important in her niche. After that I can't think of any major competitors, but then I am sure I'll think of something as soon as I click "Save Page". So let's compromise and say "top 10"? :) Ahasuerus 19:27, 6 Feb 2008 (CST)
Um, yes, George MacDonald and L. Frank Baum. As I was saying... Ahasuerus 19:30, 6 Feb 2008 (CST)
I've made a minor change to the title in one of McCullough's series[10]. This should make them easy to find in the future when we can change the title record to "non-genre" and still have the series record display.Kraang 19:34, 6 Feb 2008 (CST)
Looks very good - we can always change the series name once "the fix is in" :) Ahasuerus 19:39, 6 Feb 2008 (CST)
I'm going to attach this to collection titles, series and non-fiction titles. I think this will help organize and find the non-genre in the future. Tonight I'm going to pack it in early, all your tornadoes have turned into snow up here and I've spent my day and probably a bit more of tonight shoveling the white stuff!Kraang 20:06, 6 Feb 2008 (CST)
Unfortunately, modifying Collection and Non-fiction titles would break various Review relationships that have been lovingly maintained by the magazine folks :( The good thing about Series names is that nothing relies on them, so they can be changed in any way we want. Ahasuerus 20:33, 6 Feb 2008 (CST)
P.S. Thinking back, another solution was proposed a few months ago. Instead of changing the titles of non-genre short fiction, we could use the "non-genre" tag in conjunction with the regular "short fiction" type. That way we will know which short fiction titles need to be changed once non-genre support has been improved. At the time this approach was proposed, I thought that it was not a viable long term solution since we would still be displaying non-genre short fiction in the same section with SF short fiction and our users would need to click each story to see whether it was SF or not. As a short term measure, however, it may be a reasonable compromise that will enable us to enter non-genre short fiction without creating display problems down the road. (And sorry about that snow, Kraang, that wasn't the plan at all!) Ahasuerus 22:57, 6 Feb 2008 (CST)

Project Gutenberg

I have documented the practices I have been following in entering PG etexts as publications at Publisher:Project Gutenberg. Anyone else interested please take a look, and if your practices have differed, or if you disagree with mine, perhaps a consensus can develop. -DES Talk 10:45, 6 Feb 2008 (CST)

It looks reasonable, but I don't think we have documented a standard for the price field when entering free books yet. The last time a related issue came up, the consensus seemed to be that we should be leaving the field blank as opposed to entering "npp" for books with no printed price, but free books are different. Should we use "free" or "$0.00"? "$0.00" seems to be too US-centric since Gutenberg is accessible worldwide. Ahasuerus 12:08, 6 Feb 2008 (CST)
We could point at the Australian version instead and use Zero Australian Dollars? ;-)
Actually, that's probably another can of copyright worms - would our US hosts get in trouble for linking to something legal in Australia but not in the USA? Can we list the Australian etexts without linking? Or should we just refuse to admit they exist? BLongley 13:31, 6 Feb 2008 (CST)
And if the etext number is useful for generating a URL, maybe a feature request to automatically create such would be in order? Or would that annoy the Commercial sites we link to? (I doubt it, as we link to those by ISBN and anything with an ISBN isn't on Gutenberg yet... but give it time...) BLongley 13:31, 6 Feb 2008 (CST)
(after edit conflict) I'll follow whatever the consensus is, of course. I think that leaving the field blank is a mistake, because that is what we do for items with unknown prices, and here the prices is known -- also if anyone runs a query on price it would be better to be able distinguish free books. I would prefer some form of 0 to free so that if we ever develop stats on average prices or the like these will work properly. I was using ($0.00) because we use a currency with all numeric prices, and while the Project Gutenberg is accessible worldwide, they make a significant point of being a US-based project -- specifically they look only to US copyright law in determining what is in the public domain, and have posted works where someone has claimed that a non-US copyright is still in force. Note that there is a separate Project Gutenberg Australia (which carries a number of works that are not PD in the US, but are in Oz), and a separate project Gutenberg EU, and I think that a separate Project Gutenberg Canada is in existence or being formed. I would mark works published by each of those as zero in their respective native currencies (euros for the EU PG). -DES Talk 13:46, 6 Feb 2008 (CST)
Those last five words decided it for me - I prefer zero without a currency symbol. No way would I want hard-working British contributors to have their works revalued in foreign money. ;-) And zero converts to zero worldwide - well, for currency anyway, it's not like Centigrade to Fahrenheit to Kelvin. BLongley 13:57, 6 Feb 2008 (CST)
Actualy the PG-Europe site has been concentrating on works not in English, particularly works where full unicode representation is desired to handle accented characters. British works are mostly being done by either PG-Aus or PG-US, because of the ways in which the copyright laws interact. But that is jsut a tendency, not an invariable rule (to misquote Prof. Parkinson).
(In resposne to Bill) Reprints of stuff that is on PG could well carry ISBNs. In general, linking to a site that might carry copyright infringing content is not itself infringement (and so not illegal) unless it is shown that the linking site intended to induce people to unlawfully access the infringing content. So far this has only been claimed in cases where the linking site got a commission, or in some other way based their business on people following the links. Where a link is merely a form of reference, is is not likely to be considered "contributory infringement", but this is the kind of thing that might be well to get a proper legal opinion on. -DES Talk 13:46, 6 Feb 2008 (CST)
Sure, reprints will carry ISBNs - I should have been clearer. I think anything that was ORIGINALLY published with an ISBN is still out of any of the Gutenberg's scope though? (Probably less than a decade before that's not the case though.) The reprints are often using the same "out of copyright" rules as Gutenberg. BLongley 13:57, 6 Feb 2008 (CST)
No, i'm afraid not so. Items published in the US before 1923 are now PD, pretty much no matter what. Items published before 1964 are now PD if the copyright was not renewed. Items published after 1923 are copyrighted for 95 years after the date of publication if the copyrights were renewed, and items publsihed after 1964 are copyrighted for 95 years if there was a copyright notice on the origianl publication. So items from, say 1969 will not be PD in the US until 2064, unless they were published originally with no copyright notice, which is unlikely for a commercial book with an ISBN. See this site. -DES Talk 14:11, 6 Feb 2008 (CST)
Oh, and some few texts still in copyright are on PG by explict permisison of the author. Some of those were initally published with ISBNs. I don't think there is any SF in that category, however.-DES Talk 14:18, 6 Feb 2008 (CST)
I'm thinking of it the other way round - SBNs were invented circa 1966 and morphed into ISBNs. Given the shortest copyright periods on the various sites, which I think is 50 years after death(?) SOME of the sites will be free to publish works from authors unlucky enough to have died just after then. So in 2016 we should see the first originally-published-with-SBN legally appearing on a Gutenberg site. "Explicit permission" already excepted - but commercial hosts can't argue about that either. BLongley 14:27, 6 Feb 2008 (CST)
In US copyright law (which is what the original Project Gutenberg uses) the 70 years after death (The US never adopted the Life+50 standard) only applies to works both written and published in 1978 or later. Works "created" (aka written) before 1978 will not be PD before 2047 no matter what. Australia (and thus PG Australia) uses Life plus 50 for authors up to a certain death date, and life plus 70 after that, I think, but I don't recall what the changeover date is. So it is just possible to have a book initially published with an ISBN PD under Australian law in about 2017, yes -- but it won't be PD in either the US or the UK, unless the author was a native of Australia, and the books was published inside AUS before (at least 30 days I think) it was published anywhere in the EU or US or anywhere else that uses a Life+70 term. -DES Talk 14:50, 6 Feb 2008 (CST)
I think you're missing my point. No commercial site we link to can complain we're undermining their sales with a rival link to a free version that may be illegal in some countries: if we only link to them by ISBN, and nothing with an ISBN is yet guaranteed to be out of copyright, we wouldn't have an etext version linked. OK, Copyright may have been waived, but if so they still have no grounds for complaint. At the moment, we can safely say that anything originally published with an (I)SBN will not cause any problems with conflict of interest in our links. If we have no ISBN - no link to commercial site. If there's an ISBN - there shouldn't be an etext version, or they're allowable for some reason like being placed in the public domain or under a creative commons license or being an ISBNed REprint or suchlike. BLongley 16:21, 6 Feb 2008 (CST)
Oh I see. Yes, you are quite correct on that point. Of course, unless a commercial site is paying us, I don't see that they have much grounds to complain about links to an alternate version anyway. Are we set up to be paid for links to commercial sites? -DES Talk 16:45, 6 Feb 2008 (CST)
Well, I don't make anything from them! Some of the links do look like they could be part of a referral program - if Al is making a few pennies from it I don't really begrudge him it after all the work setting this site up in the first place. BLongley 13:33, 7 Feb 2008 (CST)
I think we are Amazon.com "partners", but I am not sure what that entails. Al would know for sure, of course. Ahasuerus 13:56, 7 Feb 2008 (CST)
Oh, and on a related note swfritter has been using Tags for Gutenberg titles, it might be good to merge the effort - I'd like to have ONE solution. Do Gutenberg have only one edition of a book, or do they have multiple versions of some titles? If the latter, multiple publications would be better than single tags. BLongley 13:41, 6 Feb 2008 (CST)
PG can, and does, have multiple versions of some titles. First, they may correct errors or add missing features (such as adding an HTML version to a text that was previously ascii only). These changes will create a new revision of the same etext number (the bare url for a given etext number always gets the most recent revision). Second there may be multiple "editions" (etext numbers) of a given work. This most often happens when a work is presnted in multiple languages. But it can also happen when the text is based on different printed editions with significant textual differences. This is rare, but does happen.
On looking at User:Swfritter#Project Gutenberg Science Fiction It appears that he is using only 26 distinct tags: all PG pubs by authors whose last name begins with A get the tag "pga", all PG pubs by authors whose last name begins with B get the tag "pgb", and so on. Thsi is in effect a hack to provide a search by publisher/author for PG titles only. I have no objection to adding these tags, and they in no way conflict with what I have been doing with PG pubs.
I am going to copy some of this discussion over to the Publisher:Project Gutenberg page. -DES Talk 13:59, 6 Feb 2008 (CST)
The last time I got involved in the Project Gutenberg discussion was one of the few times there were actually some rude comments. I instituted the tags because they are by definition user-defined and only in rare situations are they forbidden. If I remember correctly I even got some grief for using the tags. My own inclination was to treat novels, collections (rare), and anthologies (none so far as I know) as regular books with Project Gutenberg as publisher and the etext number. Price seemed irrelevant I thought it might be interesting to collect standalone stories into virtual expanding collections by author. As far as multiple formats - for Jim Baen's Universe I use Electronic: MultiFormat and use the downloadable HTML version for doing data entry - copy and paste to The ISFDb. I can also determine word count by copying and pasting borderline cases into a word processor and using the word count function. There are actually some other sites [Manybooks http://manybooks.net/] and [Blackmask http://www.blackmask.com/] that have additional non-Gutenberg content that may, from a legal standpoint, be more questionable. The previous Blackmask got nailed by Street & Smith for it's Shadow and Doc Savage content. I prefer getting my Gutenberg content from Manybooks because they have the mobipocket versions.--swfritter 18:39, 6 Feb 2008 (CST)
I must have missed it but was there an actual recent discussion somewhre about entering Project Gutenberg titles before the methods were actually implemented?--swfritter 18:03, 7 Feb 2008 (CST)
I had discussed, in a couple o places, entering ebooks from Baen. Based on what I was doing with Baen, I entered a few PG works. I created Publisher:Project Gutenberg by documenting the practices I had been implementing, and an invitation to others to advise if what i was doing could be improved, and to hope that a consensus on the best way or ways of handling PG publications emerged. I said that if a consensus emerged that was different from what I Had been doing, i would, of course, follow it. This thread arose from my announcement. Possibly it would be better continued on Publisher:Project Gutenberg or Publisher talk:Project Gutenberg. -DES Talk 09:09, 8 Feb 2008 (CST)
I totally support the basic concepts and would like to see even more electronic editions accepted. Aeon Magazine, for instance, has been successful enough as an ezine that they have managed to double their author word rates. But they don't have an ISSN. It seems odd that we can include fan fiction from fanzines but not stories by professionally noted authors in ezines. I think there is a misunderstanding about the difference between ezines and webzines. Ezines are every bit as permanent as ebooks and Project Gutenberg entries.--swfritter 13:07, 8 Feb 2008 (CST)
I'm quite happy to have stable ezines, and even entries from sites like PG, if the entries are stable enough. The first time I looked at it (quite a while back admittedly) it seemed likely that a lot of entries would get deleted for copyright reasons (either somebody submitting something in error, or from a different legal perspective, or just the US changing copyright law again to put some books BACK into copyright). Once there's a certain mass of standard, identified, catalogued and preferably not-easily-modifiable editions of a work out there with the public I don't particularly care if you can't get it from a certain site any more - there's plenty of physical books and magazines we record that you can't buy anywhere now, but we record exactly what you WOULD expect from it if you COULD get it. So still no web-zines for me, but if people are downloading something of SF interest that can be preserved intact and unmodified, I lean towards inclusion. Yes, we may lose all physical copies of it eventually (memory cards wear out, DAT tapes become obsolete, CDs will be replaced with DVDs, etc) but I'm told Pulp Magazines are a bit vulnerable to age as well. BLongley 14:38, 8 Feb 2008 (CST)
It might be time to point Dave at the Baen discussions then restart a Rules of Acquisition discussion again - I think we're going to need one anyway for the Reviews problem and maybe the Too Little Speculative Fiction discussions. BLongley 14:38, 8 Feb 2008 (CST)
As to copyright causing a deletion of entries posted to PG, you might be surprised at the level of research PG does to verify the public domain status of a submission, and nothing is posted until that research is done, submitted, and verified by a second person. (The level of research is, I believe, comparable to or greater than what a reprint publisher would insist on.) To the best of my knowledge, PG has yet (in over 25,000 etexts) to withdraw an etext due to a copyright claim. As to the US restoring copyright to stuff now in the public domain, no one can predict what the US Congress will do. But current policy has been stable since 1998, and stable except for the effects of the GATT treaty since 1988, and there seems to be no pressure for significant changes in the direction of further restoration. PG texts have been stable (with consistent etext numbers) since 1993, and there are copies at multiple serves and multiple off-line backups; the URLs have been stable for several years and the organization has indicated an intent that its canonical URLs remain stable. Thus, unless people object, and pending further discussion, I am treating PG texts of works that would be IN if published on paper commercially as being IN. -DES Talk 15:17, 8 Feb 2008 (CST)

Special projects for Polyglots?

I had a look at this author today (not sure why, but approving a load of Dissembler submissions does take one to strange places at times, that might have been involved. Or maybe I just followed some strange links most editors don't follow...) Anyway, it's a page of titles that I THINK should be a quarter of the size, if people know enough about foreign translations. You don't really want ME trying to fix it - does anyone else want to try? And if so, can we find a place for such projects that will attract those capable and leave me OUT of such? BLongley 18:40, 9 Feb 2008 (CST)

She has been on my radar screen for some time and I should probably add her to my "To Do" list, but if anybody knows Spanish, please jump right in.
BTW, looking at her Summary page reminded me about something else that I have been pondering for over a year. We have a fledgling Entropy page, but we don't have any measurements of completeness even at the Author level, much less at the Title level or at the Publication level. How do we know which SF writers are currently not covered by the ISFDB? How do we know what Titles we are missing for the covered Authors? How do we know which Publication records we are missing for the covered Titles? For now, we are using primary sources like our personal collections (and other libraries) and secondary sources like Tuck (which Michael has been working on), Clute/Nicholls, Reginald, Rock, etc. We also have a verification matrix for each Publication, but we don't have a way to tie all of these sources together.
Every time I run an OCLC/Bookwhere/Amazon/etc search and reconcile the records that I find there with what we have, I say, "You know, it would be nice to document what I have just done so that the next editor to look at these Titles/Authors wouldn't have to run the same searches". I used to do author level reconciliations and add notes on the Author's Wiki page, e.g. Author:Mercedes Lackey, but I haven't had time to do this regularly ever since we opened the ISFDB up for outside editing and, besides, a hundred scattered Auhtor-specific notes in the Wiki do not a completeness matrix make. Perhaps we could add something to the Author Directory or start a new Wiki page? And if we do, how do we make it maintainable since a matrix with 40,000 author records would be useless?
We don't tend to use the Wiki pages for an author much, and even less for a title or pub. We probably SHOULD, even if it's just a "this author doesn't belong here judging by a Google Search" comment or an "Amazon shows another dozen titles that might be relevant" comment. I prefer to look at such rather than squabble over a spelling mistake somewhere. (BTW, somebody might want to look at "Ed Emshweller" entries...)
OK, now we've lost those nit-pickers we can get back on subject. ;-) It SHOULD be fairly easy to flag an author as needing more attention for inclusion or exclusion. Starting a new project seems to encourage activity for a bit, but it always peters out uncompleted. Maybe a feature request for "flag this author as needing more research whenever someone visits the page"? BLongley 20:16, 9 Feb 2008 (CST)


(BTW, someone needs to archive bits of this page, judging by the huge warnings I get posting this..) BLongley 18:40, 9 Feb 2008 (CST)

I archived most of 2007 (through mid-December) just a couple of weeks ago. I thought it was safer to keep at least two months worth of discussions on the main Community Portal page, but if we keep growing at the current rate, we may have to start archiving more frequently or perhaps spawn off more sub-pages. For now the way to avoid "size" warnings is to edit the last section of the page and add the new section at the bottom. Ahasuerus 19:43, 9 Feb 2008 (CST)

Organizing SF stories into an ontology of SF concepts.

I'm posting this here instead of the where it was at Chris J's Talk page in the hopes that more people will respond to it here.

For a little over a year now I've been developing scifidb.org. It's a site that attempts to organize stories from the magazine Analog into an ontology of SF concepts. To start with I just want to be clear that I'm not trying to advertise the site. I've given up on being able to make the site popular all by myself. Furthermore, it takes a lot of time just to categorize Analog stories. I'd never be able to categorize all SF works and I really think that'd be a useful thing to have. I'm trying to make use of the work I put into the site so that all my efforts don't go to waste. I've tried to attract attention to it from Analog readers by advertising in the classified ads in the magazine, but I've never built up more than a trivial amount of traffic.

Yet, I've put a lot of work into the ontology and I really want it to be put to some use. So, I was hoping that perhaps ISFDB would either make use of the ontology directly or at least add the infrastructure to your site and database which would allow for organizing SF works into an ontology of SF concepts. I seeded my ontology with the basic information about Analog stories (from ISFDB for the older issues), but then added topic ontology information as I read each Analog story.

I'd be happy to contribute the work I've done. I'd also be happy to help teach people what the heck an ontology actually is. When people try to categorize things they usually either create a simple tagging system (something very popular in social sites these days) or they go a little further and create a hierarchy like the Dewey Decimal System.

An ontology is a much more thorough way of describing something. For instance, if you go to scifidb.org and take a look at the "Topics" section near the bottom on the left side, you'll see a list of top level topics. If you click on "sci fi topic" it'll open up a tree of subtopics and display a page that shows the topic, all the stories that use that topic, along with a forum messages that apply to just that topic and some images that apply to that topic (the smaller the image, the farther up the hierarchy the image was inherited).

Note that in the tree, some topics have a larger font. The larger the font, the more stories there are which use that topic. It's a way of showing the popularity of various nodes in the topic tree.

If you click down deeper it'll seem like you're just going deeper into a simple topic tree. But nodes in the tree can have multiple parents. For instance:

   http://scifidb.org/know/sview?topic=http%3A%2F%2Fscifidb.org%2Fscifi%2Fscifi.owl%23AliensDoNotValueIndividualLife

Look at the top of the page to see the breadcrumbs (which mirror the topic lineage section): The first breadcrumbs list get to the topic via the topic of mundane society which leads to alien society.

   mundane topic « society « alien society « alien psychology « aliens do not value individual life

The second breadcrumbs list gets to it via sci fi topic then alien then alien society and then alien psychology.

   sci fi topic « alien « alien society « alien psychology « aliens do not value individual life

That example had just two parents and both parents were simply separate trees that the topic was present in. Here is a really good complex example:

   http://scifidb.org/know/sview?topic=http%3A%2F%2Fscifidb.org%2Fscifi%2Fscifi.owl%23TimeTravelRoundTrip

If a story is categorized as having the topic "time travel round trip" then that implies that the topic involves time travel forward and time travel backward. That's the great thing about this system. It's the reason why ontologies are so great. If one person tags a story as "time travel round trip" then they don't need to remember to also tag things as "time travel forward" and "time travel backward". It happens automatically so that anyone who looks at "time travel backward" will always see any story that's tagged with "time travel round trip".

An even cooler capability would be that anything tagged with both "time travel backward" and "time travel forward" would automatically have a suggested tag of "time travel round trip". But, that would probably be more difficult to set up and many such matches would not be valid, so I never implemented that sort of capability. But, it could definitely be done and would be an excellent way to suggest to editors other possible tags that should apply to a story.

My favorite tag is "impossible". If there are any science fiction concepts which suddenly are proven impossible by physicists then it's possible for me to go in and change that tag to also be under the impossible hierarchy. As a result, all stories that were tagged with that tag or any other tag under it will end up also having the tag "impossible". This allows an editor to quickly change the hierarchy and, as a result, modify the tags of all stories without having to go to each story one at a time and make a determination for each one if it should be updated or not. It works the other way too. If cold fusion is deemed impossible and then some guy proves it's possible then we can move that tag out of the impossible hierarchy and all stories would be updated to no longer be marked as impossible.

I've implemented my system using an RDF data store using the OWL language. But, I'd be happy to help you develop a standard relational schema which approximated this sort of system. It wouldn't be as good as a true RDF store, but it would require the least change to your systems.

I can definitely see a difficulty in maintaining the hierarchy of SF concepts. My suggestion would be to allow users to tag stories in any way they want using a standard social site tagging system except that for any tags that didn't exist before, the user is encouraged to search the system for an existing tag that matches their concept. If they can't find one then they would be required to add a description of the tag (this is a departure from a standard tagging system where users just add tags without descriptions) and moderators would look at the tag and either change it or add it to the hierarchy. People could submit requests to change the hierarchy or add additional nodes, but the moderators would maintain control over it. I know that the ontology of SF concepts that I've created could do with some outside suggestions for cleaning it up. It makes sense to me, but I'm sure that some parts could be made more succinct. Some nodes should probably be moved around, etc.

There are two parts of this system that are important. One part is the ontology of SF concepts all by itself. It's just interesting to look at various nodes in the topics tree and see all the associations between topics. The second part is the relationship between stories and the topics. The topics allow you to look at very specific or very general topics and see all stories that use those topics. The way the tree is constructed allows for very easily manipulation of topics so that editors can organize large numbers of stories without needing to understand every story. They just need to understand the topics they are manipulating and the organization of the stories just falls into line.

Authors can use this sort of hierarchy to research concepts for their stories. They can find other stories that apply very specific story concepts and then update their own stories after reading those other stories.

I'll try to come back to this site in the near future to see if anyone has any interest in this at all. When you're used to standard relational databases, it can be difficult to wrap your mind around this much more fluid type of organization. But, I hope someone is interested because I think this is a great type of thing to have. You can also email me directly. Just go to scifidb.org and click on the "send me email" link at the bottom of the page.

I'm a java programmer, so I can't help with the Python programming required to enhance your site to include this information (although I am open to learning Python for this if my help is necessary), but I'd be happy to help with the ontology and consult on the design of the schema. If you want the SF topic ontology that I have, I'd be happy to contribute the rdf database I have. I can send it in either OWL or N3 file format.

HappyEngineer 21:56, 2 Feb 2008 (CST)

Oddly enough I just read another review in the January 1953 issue of Imagination about Alastair Cameron's Fantasy Classification System which was an attempt to do much the same.--swfritter 14:55, 3 Feb 2008 (CST)
Hm, I didn't realize that Alastair G. W. Cameron was interested in SF early in his life. I have updated his Author Data. Ahasuerus 18:24, 3 Feb 2008 (CST)
The site has a lot of interesting features. The hierarchical system appears to be an especially good way to display a lot of information. A search for authors lists stories in the following tree structure (magazines/year/month/story). This gives a nice visual listing of stories chronologically sorted. This might also be a good way to organize our pseudonymous authors.--Rkihara 00:52, 4 Feb 2008 (CST)
So, how should I proceed? I'm glad that Rkihara likes some of the other organizational features of my site, but what about the topics? Is anyone interested in this? What can I do to get the ball rolling? Is there a discussion forum where you discuss additions to the database schema?
Swfritter, I searched around for information on the Fantasy Classification System and found a page in German that actually had a part of the classification system listed on it in English. I didn't see anything particularly interesting (it looks like a traditional hierarchy which only tries to categorize an entire book into one leaf of a tree instead of tagging a book with topics which are themselves organized in a tree) and given that the book is over 50 years old, I suspect it may be of limited use in trying to classify SF by topic today. But, I'd obviously need to actually get a copy of the actual book before making any real judgments. Sadly, the cheapest copy I found was $80. I might buy it if I could find a google book preview which showed me what was in it.
HappyEngineer 17:11, 6 Feb 2008 (CST)
Ok... I guess there is no interest in this. If anyone decides this is worth pursuing or wants me to contribute in some way then feel free to contact me at: scifidb at g42.org. HappyEngineer 19:08, 14 Feb 2008 (CST)
I am not sure what the rest of the editors are thinking, but my reaction was "Hm, looks intriguing, but how exactly do we fit it into the current project?" And since I didn't have a good answer, I waited for other folks to come up with ideas. Ahasuerus 20:14, 14 Feb 2008 (CST)
I personally liked the idea quite a bit, since it allows you to search for stories that you vaguely remember, or to find stories of a similar type, and I could see other ways that it could be useful. I was also waiting for more responses. Part of the problem is that we have only one programmer for this database, the rest of us are editors/moderators, and with a few exceptions, are unfamiliar with the complexities of adding a new new layer of programming. The data entry is also daunting, since you have to read the story, then mentally split out the classifications. I skimmed over your site and didn't look very deeply, but since I'm interested, I think I'll go back and spend a bit more time there.--Rkihara 12:24, 15 Feb 2008 (CST)
By the way, I wonder if the ontology project may intersect with the Technovelgy project (which I discovered 5 minutes ago) at some point? Ahasuerus 10:47, 18 Feb 2008 (CST)
I'm fascinated by ontologies, so of course it's of interest. This reminds me of a couple of other projects I've seen:
  • Everett F. Bleiler, Science-Fiction: The Early Years (1991) and Science-Fiction: The Gernsback Years. These works carefully dissect early works of SF, listing an extremely detailed synopsis. In the back of each work there is a concept index, making it easy to find the first stories with ray guns, for example. Bleiler does not attempt to gather the concepts into hierarchies.
  • Eric S. Rabkin, at the Michigan State University uses the ISFDB for his Genre Evolution Project. Students read stories for this project and enter results in an Access database. While the project uses concepts, the purpose of this project is to track how those concepts evolve over time into standardized tropes. As such there are standardized "Genre Content" fields, but again they have not established a hierarchy.
In general, your work is interesting, but I'm not sure how to fold it into the main database as yet. My concern is that most people are currently focused on hard data like page numbers and ISBNs, and have not expressed much interest in soft or derived data yet. The tagging concept has been around for about a year and only 1% of the titles have tagging data associated with them. Only 0.5% of the titles have voting information attached to them. Between resource constraints for programming a solution, and the current interests of the ISFDB editing population, it's probably too early to try to integrate your work with the ISFDB. I'll continue to poke around your stuff to see if any ideas pop out. Alvonruff 12:45, 18 Feb 2008 (CST)
I think there is a fair amount of interest in derived information out there. Occasionally, when somebody posts a question about "the most frequently used titles", "the average age at first sale over the decades", etc on rec.arts.sf.written, I write a script, run it against the database and post the results there. The response seems to be positive, if not overwhelming, which is encouraging.
Having said that, we need to collect a lot more hard data before we can have a reasonable degree of confidence that whatever derived data we come up with is accurate. First we need to have 99% of the stories entered, then we need to have them tagged and then we will be in a position to start looking for trends, discontinuities, etc. Ahasuerus 13:19, 18 Feb 2008 (CST)

2008-02-10 backup file uploaded

... and I will be rerunning my title-pub mismatch scripts shortly. Ahasuerus 22:59, 10 Feb 2008 (CST)

New mismatches posted. I have improved the script that finds problem anthologies. It no longer selects Anthology Titles that appear in Omnibus Publications as long as the pub contains one Omnibus record and at least one Anthology record. I will work on improving other collection-anthology-omnibus scripts tomorrow, travel gods (more like demons, really) willing. Ahasuerus 00:13, 11 Feb 2008 (CST)
Almost all title-pub mismatches have been updated using a new and improved algorithm that reports fewer false positives. Some pages have been reduced to just a handful of suspect records, but we still have a few pages with dozens of records that need work. Ahasuerus 22:17, 12 Feb 2008 (CST)

Data Consistency report for duplicate title publication records

Can someone run this script with the latest backup file to create a new list of duplicate title publication records? One of these days, I'll learn how to do it myself. Until then... Thanks. Mhhutchins 17:18, 11 Feb 2008 (CST)

Done. I have also run a bunch of other MySQL scripts, which were mostly clean. We probably want to create separate sub-pages for these mini-project to make the page more manageable. Ahasuerus 23:09, 11 Feb 2008 (CST)
Thanks. Yeah, that page is getting rather cluttered. If you can clean it up, my hat's off to you. Mhhutchins 22:47, 12 Feb 2008 (CST)

Placeholder numbers for eBooks?

I'm holding a submission from DESiegel160 adding contents to this eBook, and he's used numbers to maintain the story order. I know this shouldn't be used in printed books, but since this eBook is unpaginated, I can see the advantage of placeholder numbers. Anyone have an opinion, pro or con, on the matter? Mhhutchins 10:52, 13 Feb 2008 (CST)

Maybe capitalized letters or Roman numerals could be used to distinguish ordering of ebooks vs real books?--Rkihara 11:16, 13 Feb 2008 (CST)
My understanding is that the page field supports Roman numerals (as it must for front matter which is often so numbered), but not letters in an ABCD sequence. Besides, what is the proper way of handling more than 26 items in such a sequence? Consider Tales of War from the same author. Note that I have carefully indicated that the page numbers are placeholders in the publication notes.-DES Talk 13:22, 13 Feb 2008 (CST)
I'm OK with dummy page-numbers for ebooks, although I would recommend spacing them more (e.g. 10,20,30,40 rather than 1,2,3,4) just in case someone wants to add some detail you don't feel worth recording (e.g. Interiorart). I'm not sure how far Roman Numeral support goes - I can't recall seeing a number over XXIII, would it break when we introduce L or C or D or M? BLongley 13:46, 13 Feb 2008 (CST)
That makes some sense, although in this particular case I have included all items present -- there is no interior art, for example. On the other hand, a sequence like 10, 20, 30, etc looks more like "real" page numbers and might be more likely to fool a user, perhaps? -DES Talk 14:11, 13 Feb 2008 (CST)
That's been an argument against using fake numbers at all, true. But if the note is clear it shouldn't harm. However, notes aren't always clear and sometimes we've wanted to record orders of real contents in real books on a sub-page level, or in completely unpaginated books, so the whole area could do with addressing. A sequence number to be used instead of or in addition to real page numbers maybe? BLongley 15:44, 13 Feb 2008 (CST)
If you are talking about direct software support for such a thing, that might be a good ultimate solution. But there are a lot of requested features still to implement, and this may well not be at the top of the lsit. Furthermore, thre is the risk of making the already soemwhat complex publication editing screeen overly compicated and hard to use. In any case, we need a decision for what is to be done in the current state of the software, as well as an idea of where we want to go in the future.
My feeling that if page numbers are 1, 2, 3, 4,..., given how rare it is for multiple items to be only a single page long, this may well help to give the reader a heads-up and call attention to the note (which IMO should always be present when placeholder page #s are used), while 10, 20, 30, 40,... is less likely to draw that attention. But I will go along with whatever the consensus is, of course. -DES Talk 16:02, 13 Feb 2008 (CST)
Might I suggest that the best place for this discussion is at Publisher:Project Gutenberg#Pages fields or at Publisher talk:Project Gutenberg? -DES Talk 13:22, 13 Feb 2008 (CST)
Too specific - Project Gutenberg is not the only etext we have to consider. Here's fine for now, or Rules & Standards if there's a definite proposal. BLongley 13:46, 13 Feb 2008 (CST)
Fine. However, if we come to any consensus, the Publisher:Project Gutenberg#Pages fields section should be updated to document practices to be used in future. -DES Talk 14:11, 13 Feb 2008 (CST)
Agreed, any general policy should be clear as to when it does or doesn't apply to special cases. BLongley 15:44, 13 Feb 2008 (CST)

(unindent) my feeling is that when an ebook has no page numbers, and the order of items in the contents is or may well be (in the judgment of the relevant editor) significant, the use of placeholder page numbers should be the accepted solution. I don't see that using roman numerals for this purpose adds much, but I don't object to it that strongly provided that there is good software support for, say up to 99 items. This needs to be checked or confirmed by someone who knows.-DES Talk 16:02, 13 Feb 2008 (CST)

Ordinal numbers work.--Rkihara 16:46, 13 Feb 2008 (CST)
Wonderful. I think this is the best solution that I have yet seen that does not involve a software change. I hereby propose that ordinals be the standard for all placeholder item/"page" numbers in ebooks, until and unless a software change is made, and I hope this will make such a change unneeded. -DES Talk 17:10, 13 Feb 2008 (CST)
Maybe it doesn't work? I put another entry in out of order, after the original submission and it didn't sort. I'll have to experiment with it a bit to be sure.--Rkihara 17:43, 13 Feb 2008 (CST)
Ah well. If this proves to work, it seems ideal to me. If it doesn't (and I wouldn't have expected it to) then we are back to where we were before, and my comments above are still my view. -DES Talk 17:46, 13 Feb 2008 (CST)
I ran a couple more experiments with numbers in the title field, and found that the sort gave first priority to numbers at the end of the field over letters at the beginning, but letters in the first position priority over numbers. I then tried leading zeros to denote place holders in the page field, which seemed to sort properly (the title gives the entry order), but I'll have to try it a bit more to make sure it works all the time.--Rkihara 18:46, 13 Feb 2008 (CST)
Ordinals are fine with me. They may have to be reworked on rare occasions and they only need to be used if there is a reason - keeping related stories in sequence or tying artwork to stories being examples. Hopefully whatever solution we use is a temporary fix until an invisible sort sequence can be implemented.--swfritter 18:58, 13 Feb 2008 (CST)
I am suspicious of putting numbers in the title field, as opposed to the page number. i don't trust that the sort will remain accurate. See ISFDB:Community Portal/Archive/Archive03#Order of authors in a collaboration where the kind of reordering of DB records that may happen is discussed. (in particular, I refer to the statement: "The same is true of title order in a collection - there is no way for the ISFDB to know what order the titles appeared in unless you tell it - with page numbers. You can't depend on the order that they appear in the database - because that is EXACTLY what the ISFDB is doing now. If you first have collection A, with stories in order 1, 2, 3, and 4; and then have anthology B with stories in order 5, 3, 2, 6 - then the order in the title table will be: 1, 2, 3, 4, 5, 6. Displaying the anthology without page numbers would then give an order of 2, 3, 5, 6. There's simply no way to preserve the title order without specifically stating that order, and we're using page numbers to do that.") Even if it did work, consider the problem when a short work appears in multiple publications, and not in the same ordinal place in each? This does indeed happen in the works of Dunsany which are what set off this thread. Consider Idle days on the Yann, for example. So I don't think numbers in the title field are a good idea, although leading zeros in the page number field might be. -DES Talk 19:39, 13 Feb 2008 (CST)
There is no sort order for the title field. They sometimes appear in the order in which they are entered but often have a seemingly random order. The needs of the USER should take precedence over the efforts of the editors. Ordinal numbers make much more sense to the user than place holder numbers with gaps.--swfritter 19:52, 13 Feb 2008 (CST)
I agree, but it appears that the page number field does not sort on ordinal numbers (that is numbers of the form "1st, 2nd, 3rd, ...") in the current state of the software. (If anyone can show that this does work, I would be very happy.) It looks like our currently available choices are:
  1. Use ordinary cardinal numbers (1, 2, 3, 4, ...)
  2. Use cardinal numbers with gaps (10, 20, 30, 40, ...)
  3. Use roman numerals (i, ii, iii, iv, ...)
  4. Do not use placeholders in the page number field, and accept that the contents in an ebook will not display in any fixed order
I incline to the first choice, where the order is perceived to be significant, and to the last choice otherwise. i don't strongly object to the roman choice if it will work for large enough values. -DES Talk 20:35, 13 Feb 2008 (CST)
I think leading zeros with option 2 would be a clearly denote a placeholder, otherwise, a pub such as "Magill's Reviews . . ." with 500, essays of nearly identical length, would appear be numbered with placeholders. If you use option 1, then if you misplace a title, you'll have to renumber everything that comes after.--Rkihara 20:53, 13 Feb 2008 (CST)
Revised list of options
  1. Use ordinary cardinal numbers (1, 2, 3, 4, ...)
  2. Use cardinal numbers with leading zeros (01, 02, 03, 04, ...)
  3. Use cardinal numbers with gaps (10, 20, 30, 40, ...)
  4. Use cardinal numbers with leading zeros and gaps (010, 020, 030, 040, ...)
  5. Use roman numerals (i, ii, iii, iv, ...)
  6. Do not use placeholders in the page number field.
Any other choices? -DES Talk 21:01, 13 Feb 2008 (CST)
Use ordinary cardinal numbers if it is absolutely necessary to logically group entries. Use series if possible - at least the data will sort correctly on the author bibliography (as soon short fiction series are displayed anyway). If only a few entries are affected document with publication notes.--swfritter 21:16, 13 Feb 2008 (CST)

(unindent)I would also ask Al to make sure that we are not abusing some unintended functionality that he is about to eliminate/change :) Ahasuerus 23:27, 13 Feb 2008 (CST)

It appears that there are still multiple options on the table, so I'll just describe the sorting rules as they exist, and should remain:
  • Without page information, titles are not sorted in any way. They appear in the order they reside in the title table.
  • Titles without page information have a temporary page number of 0. If a page number is 0, the page number is not displayed.
  • Titles with page information that is parseable as roman numerals are printed first.
  • Titles without any page information is printed second (in no particular order). This section also includes titles with page information, but the page is not a valid roman numeral and is not a valid integer.
  • Titles with page information that are not roman numerals, and are valid integers are printed third, in numerical order.
So the one area which could be easily played with, without affecting the current algorithm, are page numbers that are not valid roman numerals nor valid integers. For instance, if the page numbers were: E1, E2, E3, E4 (for eBook numbers) or O1, O2, O3, O4 (for ordinal sorting numbers), then it would be simple enough to sort as numbers (by dropping the first character). Some small amount of additional code would be needed to suppress the printing of the bogus page numbers. Alvonruff 07:18, 14 Feb 2008 (CST)
Another option: Manybooks.Net converts the Project Gutenberg files to PDF's which have page numbers.--swfritter 09:27, 14 Feb 2008 (CST)
Thanks for clarifying the way in which the current algorithm works. I gather that your idea of E1, E2, E3, etc would require some (small) code change to sort properly? I presume that in the current code such entries would be classed as simply "invalid" and not sorted at all? Given that, would it be any harder to drop a suffix instead of a prefix, and support ordinals of the form "1st, 2nd, 3rd, 4th, 5th, ... 278th, ..." This is more humanly obvious than E1, E2, E3, etc, IMO, and if a code change is to be made at all I would think little or no harder than the leading letter. (It could be made more specific as "drop trailing string if and only if it exactly matches 'st', 'nd', 'rd', or 'th' and the remaining string is a valid integer"). In any case, if code changes are to be made to support special placeholder forms, i would suggest that the display logic display them as entered, to at the same time make the intentional ordering clear, and also make it clear that these are not page numbers in the usual sense.
As to Swfritter's comments above, i would not limit placeholders to where it is "absolutely necessary to logically group entries". There are a number of cases where the sequence of related but separate items in a collection is obviously part of the author's intent, where the effect is obtained at least in part through the sequence. This includes, but is not limited to, cases where shortfiction items function almost as chapters in a larger work. In my view, in any case where the order of the items seems a significant part of the design of the overall work, placeholders might be justified. The problem with using series is that, at least at present, series entries have no effect at all on the display order within a publication listing. As to manybooks.net, that would reduce but not eliminate the problem, as they do not, as I understand it, reproduce ALL pg ebooks, and in any case there are other ebooks where this issue arises. Besides, part of the argument for including PG works as separate editions is their stability, which includes their multiple mirror sites, and their accessibility, which includes not using proprietary formats like PDF. -DES Talk 11:16, 14 Feb 2008 (CST)
Like many other things, we may then have to wait until there is software support - the simplest mechanism would be a flag that determines whether the page numbers are displayed. Until then perhaps the only information we should use is that which can be found in the publications. In the case of Jim Baen's Universe the page numbers are derived from index numbers in the table of contents which is why both the stories and art share the same page number.--swfritter 12:02, 14 Feb 2008 (CST)
And to complete the circle, the software support will have to wait until it's decided how you guys would like to do the data entry and how it's displayed. It doesn't really make a difference to me whether or not we use prefixes or suffixes (or something else entirely), we just need to come to consensus and I'll implement the changes. So far, they seem pretty innocuous. Alvonruff 12:07, 14 Feb 2008 (CST)
I'm a big fan of gaps given the number of things people have found to add recently. They're probably easiest done with cardinal numbers too, lest we get into edit wars over XXXX versus XL. BLongley 13:24, 14 Feb 2008 (CST)
I agree on that; but I think some kind of prefix would also be a good idea. It would make it plain that these are not actually page numbers, while allowing for insertions. (And they'll come. I've seen several magazines lately where someone, sometimes me, has tried to enter everything but missed one illustration or a book review. With page numbers, this isn't too much of a problem (except where there are illustrations called "Somestory", Somestory [2]", etc. and an early one was missed).) -- Dave (davecat) 15:02, 14 Feb 2008 (CST)
With programmer support I think we can simplify things. There are two issues - the visual aspect and the editor aspect. Visual aspects - is there any reason for the user to see the placeholder numbers as long as the entries are sorted correctly? The programming changes that would have to be made are to allow the editor to indicate that the page numbers should be suppressed and then suppress them upon display. Editor aspects - enter the page numbers as usual but allow gaps for inserting further entries?--swfritter 15:23, 14 Feb 2008 (CST)
No doubt this discussion will go on for weeks, and the workaround of "Add Notes, Use Cardinal Numbers, Leave Gaps" is fine by me for now and needs no programming. I'll leave that as my final view on the workaround. BLongley 15:49, 14 Feb 2008 (CST)
As to what we want in the end - I want some indication as to when Interiorart relates to a story. When they're on the same page, that's usually a good guess. If it's the page before, maybe. If it's 6th place between a 5th and 7th short story I'm even less sure - it might be a "showcase" piece unrelated to anything else. But that's "Naming Standards for Interiorart (and other stuff)" though, and doesn't have to involve page numbers, real or otherwise, at all. BLongley
Another issue that probably only concerns book editors is that when we edit a title to correct page numbers, it doesn't appear in page number order. This is a major problem when we clone one fully-entered pub to another similar one - I actually have to look for a BADLY entered title with contents but no pagination to make it easy. In the long run, when all contents are good, this will be a major problem. But that's probably a "I want to clone a book but lose page numbers" or "I want to copy contents from one pub to another" feature request. (Or at least be able to delete page numbers from a pub easily after cloning - that might have been fixed recently though, I think Ahasuerus complained?) BLongley 15:49, 14 Feb 2008 (CST)
Yes, fixed as of a couple of weeks ago -- see Al's Talk page for details. Ahasuerus 18:11, 14 Feb 2008 (CST)
In response to swfritter, it is my view that it is desirable for the user to see the placeholder numbers, provided that thy can be displayed in a way that makes it clear that they are item numbers or ordinals. That informs the user that the order is meaningful, but does not represent page numbers per se. I suppose that my ideal system would be if placeholder could be entered as, say "E10, E20, E30, E35, E40" but display to the user as "1st, 2nd, 3rd, 4th, 5th". But that may be tricky to code, and I would rather not insist. I do agree that the user display ought to take priority over the editor display. I am willing to renumber things, if need be, when items are inserted, and IMO for ebooks it is easier and more reasonable to insert everything at first in most cases, given that entry in generally by copy&paste, so insertions should be less common for ebooks than it has been for printed works. It would be nice indeed if the edit display appeared sorted in order of any page numbers present when the edit started. I think the interior art matching ultimately needs to be done by some sort of title/label, rather than just by page or item number, but that is another discussion. For the moment, i am willing to do whatever we agree to, pending software changes, I could agree to Bill's "Use cardinal numbers, with gaps" or to "use cardinal numbers, without gaps". In any case, i think any use of placeholder numbers demands a publication note until we have a clear and unvarying standard, and probably software support. -DES Talk 17:19, 14 Feb 2008 (CST)
Since it seems that any method will require an explanation I do not see the need to display placeholder numbers, the purpose of which will require even more explanation and probably further programming to designate artwork as opposed to story content. Without the placeholder number the explanation needs only be "Data listed in published order" or something similar and interior artwork will only need have the title of the story being illustrated otherwise 'untitled' or any independent title the piece may have.--swfritter 19:20, 14 Feb 2008 (CST)
One other option: Enter numbers with gaps and display ordinals. 10, 20, 25, 30 on the data entry page would display as 1, 2, 3, 4 on the user display. Visually this would be precisely what DESiegel60's initially wanted. Logically it would be what Blongley wants. It might make Davecat happy too. And it would probably make the most sense to the user. Basically everybody gets what they wanted in the first place. The only question would be the method for indicating to the program that this is the desired display option and Al might have a better idea than we do.--swfritter 00:28, 15 Feb 2008 (CST)
That would do the job nicely, but if we are doing that much coding, then perhaps an "insert item" button would be no harder, and would seerve the same purpose as the gaps. Either would be ok, IMO.
As to what should be displayed, I disagree that there is no reason to display placeholder numbers, since a note should always be present and will do the work. Indeed a note should always be present, but notes are entered manually, and sooner or later one will be ommitted. More importantly, notes are not obtrusive, and will sometimes be overlooked by users. IMO the ideal placeholders should display, so that it is clear that the display order is significant and intentional; but the ideal placeholder should display in a way that is obviously different form normal page numbers, to call attention to the situation, and prompt people to actually read the note. That is why, for display, I like 1, 2, 3, ... better than 10, 20, 30, ... -- it looks less like commonly occuring page numbers (sure items all 1 page long are possible, but not common). But a display that is absolutely unique would IMO be better yet. Dispaly ordinals would serve this purpose (1st, 2nd, 3rd, ...), so would display of placeholders in a different color, or some such indication. Obviously any such special display logic depends on a way to flag the numbers as placeholders for the software. This could involve a "this pub uses placeholders" checkbox, or a special way of entering the numbers (E1, E2, E3, ..) or some other method. Note that I have been speaking about ehat would be, in my view, ideal. Less than ideal solutions would still be quite acceptable, IMO. -DES Talk 07:44, 15 Feb 2008 (CST)
I can think of a series where each book has one page per story, but at least they don't start at 1... but any placeholder-numbers could mislead in some cases, so in the long-term should not be displayed. BLongley 13:44, 15 Feb 2008 (CST)
Anything at all could mislead someone -- the only way to be absolutely sure no one ever gets a wrong idea from the db is not to have a db at all. No displayed numbers can mislead a user into thinking that no sorting info at all is present. That is why IMO the best long term solution is a display significantly different from the display that appears for any non-placeholder pub, including different from pubs with no page numbers entered. No ordinary pub displays page numbers as ordinals, so such a display can't plausibly mislead anyone into thinking that these are page numbers, for example. -DES Talk 14:13, 15 Feb 2008 (CST)

Going foreward

In my view we need to make two decisions.

  1. How do we want to handle placeholder numbers in the long term?
    The above discussion seems clear that we want some form of placeholders or other ordering method for contents items when page numbers are not available, and that software changes are quite possibly desired for the entry and display of ordering information, whether through placeholders "page" numbers or some other mechanism. We need to decide on what we want, and get confirmation that the coding required is not excessive, nor will it interfere with the rest of the system.
  2. How shall placeholders be handled in until software changes are made?
    What shall be done with ebooks such as this one until the final design is implemented? It would be good if we all agreed on a standard for this, and that standard must use only the existing software. One choice would be not to enter such publications at all, or to enter them without any placeholder page numbers. Another is to use placeholders, but only in restricted circumstances. Yet another would be to use them fairly freely, whenever it seems useful. I think we would all agree that any use of them should include a pub-level note, as should any pub with contents but no page numbers.

I am confident that we can come to reasonable decisions on these two points. -DES Talk 09:39, 15 Feb 2008 (CST)

One - I'd leave it to Al, he knows the problem and desired effect, and can determine the quickest solution. Two - it doesn't really matter to me if we all use different conventions so long as the convention is explained in the notes. However, if I find myself in the position of having to renumber an entire pub to insert extra contents in the right place I won't be happy - unlikely though, I don't own many books that would require this sort of workaround. Al would presumably like something that can be converted programmatically later though, so feel free to come to a decision that makes it easy for a computer program to identify the affected pubs. (Yes, I know that probably excludes my preferences for current practices as they look too "normal".) BLongley 13:44, 15 Feb 2008 (CST)
Al's comments above seem to imply that he is waiting for the rest of us to come up with a desired design, and then he will implement it or indicate why it would be a problem.
While it is true that as long as things are explained in the notes different conventions can be used, i suspect that it improve consistency and make conversion easier if we are use the same or similar conventions. But if our decision is 'do want every you think best provided you document it" then so be it.
I think, from what AL said above about the current sorting rules, there there is no possible scheme which can be converted programatically without at least manual creation of a list of pubs for the conversion to work on that will also sort correctly with the software as it now exists. -DES Talk 14:07, 15 Feb 2008 (CST)
Al suggested: For instance, if the page numbers were: E1, E2, E3, E4 (for eBook numbers) or O1, O2, O3, O4 (for ordinal sorting numbers), then it would be simple enough to sort as numbers (by dropping the first character). Some small amount of additional code would be needed to suppress the printing of the bogus page numbers. That would be easy to pick up later. Using a new standard with current software that cannot be confused with any current system makes it perfectly possible to mass-convert a short-term solution into a long-term one without any effort on the part of current editors, apart from agreement. If Al can work on this soon - pick one of those options. If not - well, editors will have to rework their own standards if they want them picked up in the conversion, but in the meantime they can pick one that's easier for editors and moderators of such intermediate-standard pubs. Which is actually quite a low number - I can count the number of editors affected by this problem on one hand, the number of moderators on two. (OK, I know binary and could count both categories on one hand, but basically there's five or less people submitting place-holder numbers and ten or less Mods wondering about them.) I can read notes and am happy with most of the short-term fix proposals. IF I had to deal with an additional entry in a short-term-solution "1,2,3,4" paginated pub I'd be very tempted to tell the person that created it "I told you it wasn't a good idea!" - but we don't have that level of detail that I could chew out the right person. :-/
I just KNOW, from (mumble) years of experience that NOT leaving gaps in sequences causes trouble eventually - I first caused that problem for myself around 1980, and still have to edit more sequence numbers from other people than I should in 2008. If you go "1,2,3,4..." there will almost ALWAYS be a future exception that needs a 2.5, or a 1.3 and 1.5 and 1.7 - of course, we could always demand Al supports such, but it's easier NOT to introduce such problems in the first place. So I'd recommend that even if Al can deal with "E" or "O" prefixes easily, that you get the NUMBER bit right. BLongley 16:50, 15 Feb 2008 (CST)

"Tags" and "popular tags"

A recent garbled submission alerted me to the fact that we use the term "tag" to refer to two different fields in the database. For example, "MOMA1993" is the tag used by Publication record 23030 (Greg Bear's Moving Mars) while "nanotechnology" is a "popular tag" used on the Moving Mars title. Do we want to rename the Publication level tag field? Renaming Title level tags would probably cause more disruption since they are displayed on each title page. Ahasuerus 13:17, 15 Feb 2008 (CST)

I don't use "Popular Tags" (well, not any more - I think I have 3 still): and only deal with "Publication Tags" when they go wrong and we have missing ones and/or duplicates (I've wasted too much time fixing Doctor Who Pubs with tags like 'DCTRWHNDTH%') - I'd prefer those to be replaced with Pub IDs sometime. Actually, "Popular Tags" are most memorable for some Python Errors they caused too... I'm happy if we rename both, they're always bad news! BLongley 13:55, 15 Feb 2008 (CST)
Really? I like the idea of the title-level tags, and have been starting to enter them with some frequency. I agree that the publication level unique ID tags should not routinely be edited, and probably should not be available for editing without taking an extra step, to avoid messing them up. I also agree that calling both "tags" could be confusing, although I'm not sure if that was the reason for the confusion in the submission that started this thread. -DES Talk 14:18, 15 Feb 2008 (CST)
Tags actually mean three different things, all taken from common usage when introduced. We have the publication tags; "tag" in this case is the old computer science meaning of a unique identifier (like cache tags, and others); we've deleted all the other tags that we used to have, but have kept this one. We have XML tags which are the things between left and right angle brackets. And we have user tags, which in the Web 2.0 world is the first thing that comes to mind these days when someone says "tag".
The word "tag" with respect to publication tags doesn't show up in the database unless someone edits a publication. It would be easy enough to change the label to something else in the editing tools (like "identifier") if that would make things clearer to someone. Alvonruff 19:40, 16 Feb 2008 (CST)
The reason I started this discussion was that a new editor had submitted a Magazine record and entered something like "postmodern counter-culture" in the Tag field. Clearly, the editor had used the database before, was accustomed to the way we use "user tags" and had no way of knowing that the term meant something else when submitting a new Magazine pub. If we can change the name of the "Tag" field in the "New Magazine" form, we will likely avoid all kinds of "interesting" tags down the line :) Ahasuerus 19:49, 16 Feb 2008 (CST)

Magazine:Fantasy & Science Fiction FSFFEB2008

Could someone add Feb 2008 to the Magazine:Fantasy & Science_Fiction? Thanks! --Roglo 03:36, 16 Feb 2008 (CST)

Done. BLongley 07:56, 16 Feb 2008 (CST)

Al's new present to us

OK, in case you haven't noticed, we now have Publisher Search and a quick link to Publisher pages. What do we do with these new super-abilities? Regularise Publisher names? Go find misuse of the field? Dump all our detailed Publisher knowledge on the publisher pages? BLongley 17:18, 16 Feb 2008 (CST)

As usual, I'll ignore asking the creator until we've found ways to abuse or misuse the new powers. ;-) Some immediate thoughts are that we can go search for publishers with "printing" in the name and undo a lot of Marc Kupper's experiments: or "reprint" and undo someone else's. Or find "DC Comics" or "Viz" or "TokyoPop" and obliterate a lot of titles. (We've still no super-powers to make the deletion of Manga and Graphic Novels and RPG accessories easier though... :-( ) Do we use this to ADD information (I've been recording Publisher info elsewhere until we get Publishers properly started here) or "Cull the Imperfect"? BLongley 17:18, 16 Feb 2008 (CST)

Looking at some recent activity, maybe we should work on clearing up the fact that there is more than ONE "SFBC"? BLongley 18:06, 16 Feb 2008 (CST)
I have already used this to elimiante alternate names for Project Gutenberg, and i am now working though all the PG items we have on file, and verifing them, or correcting them if need be. -DES Talk 18:54, 16 Feb 2008 (CST)
One step ahead of you, Bill! I'm changing all of the UK editions to "Science Fiction Book Club [UK]" because their editions actually credit themselves and not the original publisher as the book's publisher. For the US club (Doubleday/Bookspan), I'm creating the following standard: "Original publisher / SFBC". The original publisher is what is printed on the title page and/or spine. The slash will indicate that it's a book club edition, but that "SFBC" is printed no where in the book itself, even though it's the printer of the edition, not necessarily the publisher. Note that I'm placing spaces before and after the slash, so that it doesn't get confused as an imprint, e.g. Berkley/Putnam, Arbor House/Morrow, etc. Many of the club's editions are listed as simply "SFBC" which will eventually disappear as a publisher (give me a couple of months!) Mhhutchins 18:58, 16 Feb 2008 (CST)
Forgot to add: there are rare occasions where Doubleday/Bookspan did print "Science Fiction Book Club" on the title page. That's the reason I'm adding the "[UK]" to the British bookclub (created by Sidgwick & Jackson, later Readers Union). Mhhutchins 19:02, 16 Feb 2008 (CST)
Some publisher and imprint titles have a space(Ballantine Del Rey) and some a slash(NAL/signet). Do we need to have some formal rules now that we can standardize publishers and imprint names?Kraang 15:16, 18 Feb 2008 (CST)
Feel free to abuse as you see fit. This is a good opportunity to start thinking about the relationships between publisher names. Take Baen for instance (I chose Baen because the sample size is small). There are six different publisher names in the database that contain 'Baen':
  • Baen
  • Baen (Canada)
  • Baen / SFBC
  • Baen Books
  • Baen Starline
  • Baen/Starline
From this I see:
  • Imprints - Baen Starline is obviously an imprint. Baen (Canada) might be one as well.
  • Variants - Baen Starline may be an imprint, but which spelling is correct? Is the other version a "variant" that needs to be tracked?
  • Canonical - Which version, "Baen" or "Baen Books", is the canonical version? Is the other a variant?
  • Licensing - Baen / SFBC is not an imprint, but represents a licensing situation between Baen and SFBC. Or should these all just be SFBC? (Seems like some information is lost in that case, as when Guild America does a SFBC exclusive omnibus).
I'm going to add support to the publishers section of the database for things like webpages and wiki links, as well as relationships like imprints and variants. These should be displayed on the publisher's bibliography much like pseudonyms are displayed on a canonical author's bibliography.
As you play around, keep these things in mind, as well as other aspects of publisher data that you grow to find annoying. I'll then start prototyping some changes. Alvonruff 19:28, 16 Feb 2008 (CST)
Where is this stuff? The links points to "what's new" and a static HTML page from years ago. Marc Kupper (talk) 19:58, 16 Feb 2008 (CST)
There will be many interesting regularization issues to debate (we have over 50 records for Putnam alone, including half a dozen permutations of "G P Putnam", "G.P. Putnam", etc), but the first thing that comes to mind is "place of publication". There are over 350 publishers with the words "New York" in their name and I think the first thing that we need to do is to decide whether we want to create a separate field for "place of publication". And if we do, do we create it in the publisher table or in the publication table since publishers, especially small publishers, have been known to wander all over the place? Ahasuerus 20:29, 16 Feb 2008 (CST)
Where is this stuff you are writing about? Marc Kupper (talk) 21:07, 16 Feb 2008 (CST)
Oh, now I see what you are asking! The drop down box in the "Search" section on the main ISFDB page now lets you search by publisher :)
Thank you - that was hidden in plain sight. :-) I did not see it as I never use the drop down but enter something to search for, tab and the first letter of the type of search (T for title), tab again to [Go] and then space bar. Marc Kupper (talk) 21:20, 18 Feb 2008 (CST)
P.S. And now that I am thinking about it, would it make sense to add "ISBN" to the drop down list if it's quick? Ahasuerus 21:22, 16 Feb 2008 (CST)
Done. Alvonruff 06:56, 17 Feb 2008 (CST)
Very nice, thanks! Would it be easy to change the ISBN column header from "ISBN" to "ISBN/ID" and perhaps format 10/13 digit ISBNs with dashes? Ahasuerus 11:22, 17 Feb 2008 (CST)
The label is easy enough; the formatting requires moving some large chunks of code around, so probably later. Alvonruff 12:12, 18 Feb 2008 (CST)
Hey, we'll take what we can get, thanks! :) Ahasuerus 13:10, 18 Feb 2008 (CST)

Print series?

#Al's new present to us says "and a quick link to Publisher pages" but that page has been around since 2005. I recently copied it to ISFDB Publishers and added more things but that does bring to light that the Phython code links to the old http://www.isfdb.org/printseries.html. Marc Kupper (talk) 19:55, 16 Feb 2008 (CST)

Good point, the new page would be a better link. And it would be nice if the publisher search also worked from the "New Submissions" page, as I find myself there more often than the true "Home" page. BLongley 04:03, 17 Feb 2008 (CST)

Covers with separate titles

In the discussion about linking cover art with specific works of short fiction, i mentioned the possibility of a work of cover art having its own, distinct, title. Well it has happened, and I'm not sure how to handle it. Specifically, Taking Flight has a credit that says "Cover art (Flight by Dalmazio Frau) Copyright 2001 by Dalmazio Frau", and Blood of a Dragon lists "Cover art (Dragon by Dalmazio Frau) Copyright 2001 by Dalmazio Frau". It seems to me that the specifically stated titles of these works of cover art ought to be recorded -- failing anything else I would put them in the publication notes, but perhaps there is a better place?

There is another issue with each of these publications, also. Each mentions the previous del Rey publication, and gives a copyright notice for "that edition" and a separate copyright for "this edition", which suggests, but does not prove, that the work was revised for the wildside edition. The author's web site mentions the wildside edition, but refers to it only as a "reprinting" with the addition of a bonus story. I guess I'll email the author and ask -- he is fairly active online. -DES Talk 11:56, 20 Feb 2008 (CST)

Lawrence (and don't make the mistake of calling him 'Larry' if you want to live :-) has been active on rec.arts.sf.written for the last 14 years and even wrote an r.a.sf.w-sponsored novel a couple of years ago, so I am sure he will answer. It's always best to go to the source while the parties involved are still available as any number of bibliographers and genre historians will tell you.Ahasuerus 01:07, 21 Feb 2008 (CST)
Thanks, I know, I was one of the contributors to the first of his two "Street performer" novels, and I pointed out an inconsistancy in an Ethshar book to him a couple of years ago. I've been intermitantly active on rasfw myself for soem years. I will email and ask him as soon as I have a chance. -DES Talk 08:29, 21 Feb 2008 (CST)
To address the original question, using the Notes field seems to be the simplest/default way of recording this information for now. It's probably safer to record it both in the Publication record and in the Cover Art record. Ahasuerus 01:07, 21 Feb 2008 (CST)
OK, that is what ai will do. I think we might want to consider an additional field in the coverartt record in the future, but notes will do for the time being. -DES Talk 08:29, 21 Feb 2008 (CST)
Does this mean you both don't like swfritter's solution to alternate names for artwork? That looks usable for more than just Cover art. And will be easier to find when things do get sorted out. (Overloading fields causes problems, overloading Notes less so but only because we'll NEVER be able to do much with them programatically, abusing standard functionality is actually the easiest to sort out later, so long as Al doesn't do something to obliterate such in the meantime!) BLongley 14:50, 21 Feb 2008 (CST)
I posted a comment on (very busy) Alvonruff's talk page. There may be technical issues involved in either solution. I think there are some potential, but I don't think radical changes, to the way both Editor and Coverart records will be processed.--swfritter 16:08, 21 Feb 2008 (CST)

2008-02-20 backup posted

Sorry about various travel-related delays. The next backup is tentatively scheduled to be posted on Sunday 2008-03-10. Ahasuerus 00:59, 21 Feb 2008 (CST)

Publisher changes - Project(s)?

We seem to be more active and concentrated when we have a fairly-clearly defined goal and lots of links to "problem" titles. They always trail off in indecision, but there's a lot of good work done in the meantime. I think some directed activity at certain publishers/imprints might help with eventual decisions on "Canonical Publishers" (which seems to be the topic of the day). Al's given us some nice pointers, but before we go crazy "correcting" everything in sight, maybe we can claim a few specialities, or spread them around those that know them quite well? For instance, I'd quite like to take over "Panther" titles in all forms and create a canonical imprint, some publishers, and a final resting place. "Orbit" and "Sphere" next maybe, but I'd like some help on recent titles. I don't mean I'm going to over-ride verifiers, but I'd use them as sources of info and might spark them into activity again, where they've got less interested. But separating some Publishers and Imprints into projects might solve some of the problems we are looking at - I think many can be resolved locally, and then we can look at the big problems later. It would keep me from complaining about Bantam/Corgi for a bit for instance! Does anyone else want to claim a publisher and/or imprint to work on? BLongley 17:55, 21 Feb 2008 (CST)

I'm working on Wildside Press and its imprints, Cosmos Books, Juno Books, Borgo Press and Prime Books. Cosmos, Prime and Borgo started as publishers and where eventually taken over. I'll add some history and as must address info. as I can into the Wiki. I'll also link all the publisher and imprint names to each other.Kraang 18:25, 21 Feb 2008 (CST)
I'd just dive right in - I think you'll find out soon enough if your work happens to overlap with someone else's. One area that I see could use work is someone with time and the willingness to work with Python and MediaWiki. There's lots to do on the Feature Request lists plus many of the recent lists (DAW, SFBC, and now Panther) could use integration into the main ISFDB database. The wiki is a great quick-prototyping tool but is not as useful for large lists/tables. Marc Kupper (talk) 22:07, 21 Feb 2008 (CST)
Are you, in fact, set up for anyone but Al to work on actual coding? -DES Talk 09:36, 22 Feb 2008 (CST)
No - I tried to get set up but needed some help from Al who hit a rough spot in his availability and then I was unavailable for a long time. I up the ISFDB parts (Apache, Python, etc.) without a problem but ran into a snag trying to locate a copy of MediaWiki that matched the one ISFDB used. I believe Al found a copy recently but I need to coordinate getting that, and a full db backup, from him as the public backups are missing MediaWiki tables. The MediaWiki stuff is an issue as ISFDB's login mechanism uses MW tables. That said, my development system had a crippled version of ISFDB. It used the production Python code, and the current at the time version of MediaWiki. On the ISFDB side of things I could view publications, etc. but could not log in to do edits without modifying the code to either bypass MediaWiki or to use the current MW table layout. I'm referring to the development machine in past tense as while it still runs something strange happened to it and I can't log in remotely meaning my access is now via a remote physical console thing that's a pain to use. The machine is a vmware virtual box. The current installation is Windows 2003 but what I should do is to set it up as a Linux box and then if someone wants a copy I'd then be able to set up the vmware VMD/VMX files on a server. Marc Kupper (talk) 16:34, 23 Feb 2008 (CST)
TAMU was supposed to upgrade our servers and the Wiki software the second week of January, but they ran into administrative problems and I don't think anything has happened yet. If and when they are able to get us to a recent version of the MediaWiki software, things may get easier. Ahasuerus 19:00, 23 Feb 2008 (CST)
I hope the wiki upgrade gets coordinated with Al as it'll need changes to the Python code. Marc Kupper (talk) 16:09, 26 Feb 2008 (CST)
Al is the only conduit between the ISFDB team and TAMU, so if anybody is going to be notified, it will be him. Hopefully ;-) Ahasuerus 16:34, 26 Feb 2008 (CST)

Self published works

I recently verified Stewards of the Flame By Sylvia Engdahl. The previous entry, which was pretty much correct (all i added was notes) listed the publisher as "BookSurge" It appears that BookSurge is a PoD Printing and Distribution service (not unlike LuLu) owned by Amazon, and that this book and any others by BookSurge are really self-published. No "publisher" is listed, and the back matter makes a point that it has been self-published in order to avoid genre-focused marketing limits (the book is science fiction by any reasonable definition -- it is set on another planet, the VP character is a starship captain, and it is an "if this goes on" cautionary tale of a kind very common in SF). In view of our recent intent to capute more and more useful publisher information, what should be done with a book such as this? (Since the author has published several noted works of SF through traditional publishers, this is clearly IN, where a book that was purely self-published might not be.) -DES Talk 17:10, 23 Feb 2008 (CST)

Actually, we have been including all POD and vanity published books for a while now. As ISFDB:Policy states:
  1. In - Works of speculative fiction originally published in English, including works published within and outside the genre. "Published" is defined as published by/in:
  • professional publishers
  • small presses
  • prozines
  • semi-prozines
  • paper-based fanzines (note: software support added in early 2007)
  • print on demand (POD) publications
  • vanity publishers (note: policy liberalized with the rise of Amazon etc after 2000)
  • e-books with ISBNs (note: software support added in mid-2006)
  • audio books (note: software support added in mid-2006)
Obviously, it's not something that most ISFDB editors tend to spend a lot of time on, but Dissembler captures this information and eventually somebody cleans it up. Emphasis on eventually :)
As far as the publisher-vs.-printer/distributor distinction goes, it's not always a clearcut one whenever "non-traditional" publishers are involved. Some claim that they "reject 70-80% of the manuscripts submitted to them", but will actually publish any submitted manuscript -- see the (in)famous Atlanta Nights sting by the SFWA crowd. At least the BookSurge guys are honest about the services that they provide and give the authors credit for all the extra work that they have to do when they go that route.
It occurs to me that the line between a "publisher" and a "printer/distributor" could be conceivably drawn in various places and it can be hard to tell what the distribution of responsibilities was in any given case. Moreover, our publisher information may be used by ISFDB users to look for books in stores and online, so it's probably safer to record the name of the publisher as it appears on the title page or wherever it may be printed.
In this particular case, there is no publisher name given: The title page has a blank line in the usual spot for the publisher, and the copyright page says "printed and distributed by BookSurge". I have a copy in front of me as I type. Given that, should BookSurge still be listed in the publisher field, or should it be left blank, or should the author be listed as the publisher? -DES Talk 00:15, 24 Feb 2008 (CST)
The "Publication" field in this book's OCLC record reads "Eugene, Ore. : Author ; printed and distributed by BookSurge", which seems to be a reasonable compromise except that we will end up with a lot of "publisher" records if and when this business model becomes more popular. Also, quite a few BookSurge books are listed in OCLC as by "[Charleston, S.C.] : BookSurge" or "[Charleston, S.C. : Book Surge LLC]", although perhaps their MO was changed after they were purchased by Amazon.com. Ahasuerus 15:36, 26 Feb 2008 (CST)
Just to be clear, are you suggesting that the publisher in this case be listed as "Author" (which could give us lots of listings under that name in time) or as "Sylvia Engdahl" (which might give us lots of publisher records with few books in time). -DES Talk 16:55, 26 Feb 2008 (CST)
I was thinking "Sylvia Engdahl", but you raise a good point. Anybody else have suggestions? Ahasuerus 17:56, 26 Feb 2008 (CST)
If we go that route, then I guess the next question is whether we want to clarify the nature of the relationship between the author and the publisher/distributor at the Publication level or at the Publisher level. The Publisher record seems like the logical place to record this information, but what if some Publishers serve as traditional publishers in some cases and printers/distributors in others? Ahasuerus 20:08, 23 Feb 2008 (CST)
In that case we would need notes or other data at the publication level, i should think, but in most cases a given firm will be one or the other. -DES Talk 00:15, 24 Feb 2008 (CST)

Publisher Issues

Very impressive work on the new Publisher pages, folks! :) Now that it has been a few days and we have some experience with the new functionality, let's see if we can compile a list of publisher-related issues that we can take to Al. Here are a few:

  1. The new "Publisher" and "ISBN" drop down choices appear on some pages, but not others. Reported a few days ago, but appears to be fixed now. Anybody know of any pages where the new choices are still unavailable?
  2. Links from Publisher search results to Author pages are broken if the author's name contains apostrophes, e.g. if you look up iUniverse - Books Published in 2004 and click on Robert 'Dean' D. Russell, you will get an "Author not found: Robert ’Dean’ D. Russell" error.
  3. Wiki pages for publishers start with the "ISFDB" prefix, which the Wiki software reserves for "ISFDB Projects" like Project:Data Consistency. Do we want to create a separate namespace for publishers? If we make it follow the same convention that Authors and Series use, it will be "Publisher:". It will be easy to move all publisher-specific pages to the new namespace. In fact, I will volunteer :)

Anything else? Ahasuerus 21:11, 23 Feb 2008 (CST)

The new options are missing on the submission queue screen, which is a nuisance when you're approving edits of publisher details. And yes, new Namespace please while we have a volunteer around! BLongley 06:59, 24 Feb 2008 (CST)
A note field in the publishers bibliography would be very helpful.Kraang 07:51, 24 Feb 2008 (CST)
Moves to new namespace done, except:
  1. Publisher:DAW is protected, so an admin must either move it, or remove the protection so that a non-mod can.
  2. There were two DAW/Titles pages, and by mistake I moved the older one first. That has now been moved to Publisher:DAW/Titles/old (and possibly it should be deleted), but the redirect left by that move prevents me from moving Publisher:DAW/Titles to Publisher:DAW/Titles -- an admin must delete the redirect first.
  3. There are now a lot of redirects from the moves in the ISFDB namespace. It may be a good idea to delete these, but a moderator would need to do that if it is to be done.
I hope this was helpful. -DES Talk 09:28, 24 Feb 2008 (CST)
Thanks! I'll post the list of issues on Al's Talk page pronto so that he could change the Python code to point to the right namespace. We'll need to keep the redirects around at least until then. Ahasuerus 10:22, 24 Feb 2008 (CST)
In addition to the note field, a Start date for the Imprint and/or Publisher would provide a nice sanity check. For instance, Wizards of the Coast are credited as the publisher of many D&D novelisations up to TEN YEARS before they actually took over TSR. (I think the blame lies with Amazon's policies, but we'll have to clean them up some time.) BLongley 14:02, 25 Feb 2008 (CST)
I'm not sure what this thread is about but have unprotected Publisher:DAW. DES - so you don't need to be the wiki-admin to add name spaces? I thought you did. Marc Kupper (talk) 16:02, 25 Feb 2008 (CST)
To do a true namespace ad, you need to change the wiki installation tables, which most admins can't do either -- I think in our current setup it eould take Al, but i'm not sure. the curent "publisher:" pages are a pesudo-namespace, which is to say they are simply pages whose names all begin "Publisher:". When and if a true publisher namespace is created, they can be mopved there easily, and we get most of the benefits in the meantime. Adding custom namespaces is a lot easier in newer versions of teh wiki software, adn i suspect a true publisher namespace may not be added until we are able to upgrade to a newer version. -DES Talk 09:32, 26 Feb 2008 (CST)
Thank you David. Yes, with the current ISFDB setup it would take Al to add name spaces. I've asked for them from time to time and had held off on doing page moves as I was not sure what happens if a page is named Publisher:DAW for example and a namespace named Publisher then gets added. You indicated that the pages can be moved into the namespace easily. Does this mean for example, that we'd need to locate all of the Author: pages for example and move them to the Author namespace? Marc Kupper (talk) 16:07, 26 Feb 2008 (CST)
According to the mediaWiki help page custom namespaces may be added simply by editing the LocalSettings.php file. If pages already exist in a pesudo namespace that matches an about to be created custom namespace, those pages should be first moved to a new name, then moved back after the namespace hass been created. For example Publisher:DAW might be moved to Publisher2:DAW, anf later, after the Publisher custom namespace has been added to the LocalSettings.php file, it would be moved back to Publisher:DAW. According to the MediaWiki help page, there is a script (maintance/namespaceDupes.php) avialable to do such moves as a batch operation.
In light of this, I may have jumped the gun by making the moves I did, but it doesn't appear that any major harm was done. Also, contrary to my comments above, it looks to me as if the only requirement for havign a custom namespace is a simple 1-line edit (per namespace) to a configuaration file, which should take no ti,e at all for all or anyone else who has access to the server wiki config files (if there is anyone else who has access). But if there are existing pages in a pesudo-namespace, they must be moved before the namespace is added (or while it is commented out), or they will be effectivly hidden and inaccesable. See the MW help page linked above for more detaiul. -DES Talk 16:47, 26 Feb 2008 (CST)

Wiki help

I got confused by the moves, but David left the following advice on my talk page: reposting here in case it helps anyone else.

To create a redirect page manually, the first and only line of the page should be 
#REDIRECT [[link here]]
with the words "link here" replaced with the name of the target of the redirection. 

(i.e put "Publisher:XYZ" inside double square brackets as the target.) Note that a PREVIEW of the page will take the '#' symbol as a numbered link, which scared me off trying it at first, but persevere and it does indeed work. (An old Mediawiki bug, maybe?) And I don't know how to show the Double Square brackets required without them being interpreted as special characters... :-/ BLongley 16:41, 24 Feb 2008 (CST)

Actually if you display a redirect page itself, you may see the # displayed as a numvbered list. This is how mediawiki has worked for a long time, it is at least arguably a bug, but since it only affects dispaly of pages only editors normally see, it has never had priority for fixing I think. To display wiki-code without executing it, put it after a nowiki tag, and before a /nowiki tag, as I have done above. -DES Talk 16:57, 24 Feb 2008 (CST)
Thanks again! I don't suppose there's any chance of you just fixing the "Help:Editing" page we always get offered when we post here but that never leads anywhere? :-/ I keep thinking someone just forgot to install part of the mediawiki software, but for all I know it's always been this confusing. BLongley 17:24, 24 Feb 2008 (CST)
My pleasure. It appears to me that the wiki software is fully installed, albiet not on the most recent version. But there are a number of pages, such as the one you refer to, that are empty by default. Each local installation is supposed to write its own version, or else borrow an appropriate one from an active wiki, such as wikipedia. I'll see what I can do, unless the page is protected. -DES Talk 18:06, 24 Feb 2008 (CST)
Help:Editing now exists, as well as a number of the pages that link to it, although far from all. The text is copied, with changes (mostly cuts) from the MediaWiki help pages, that are general/default help for all mediaWiki wikis. Note that some stuff refered to in these pages may only work when/if we upgrade to a newer version of the wiki software. -DES Talk 01:02, 25 Feb 2008 (CST)
  • any questions about editing, with or without answers, can be added to Help:Editing FAQ . I will try to answer questions if i can. -DES Talk 11:50, 25 Feb 2008 (CST)
A very good start, thanks! I already see a solution to a problem I've noticed - "If multiple sections have the same title, add a number". This has occurred several times (e.g. "Nomination statement") and I've never known the way to fix it before. BLongley 13:07, 25 Feb 2008 (CST)
Thank you DES - BTW, if you run across protected pages then drop a note somewhere (ISFDB:Moderator noticeboard will do) and we can unprotect them. The protection was only in place to deal with a spam attack. In the past the only page that was protected was ISFDB:General disclaimer as that seemed to be a frequent spam target. Last year some robots showed up and I ended up protecting much of ISFDB as Special:Recentchanges was becoming next to useless. Marc Kupper (talk) 15:22, 25 Feb 2008 (CST)

All the wiki-specific help pages are now listed in Category:Wiki Help. Take a look if you are interested. -DES Talk 12:47, 27 Feb 2008 (CST)

"Unknown, but after X" dates

It seems to me that it would be a good idea if we had a way to enter a date as "unknown, but not earlier than nnnn". When the best evidence in a book is the copyright date, we do not know the actual publication date with certainty. But we can be tolerably sure that it is not prior to the copyright date, and it is unfortunate that entering the date as 'unknown" makes it sort before publications that may well be known to be earlier, just not how much earlier. Therefore I think it would be a good idea if there were some way to enter a date so that it would code for, and display as "on or after nnnn-nn-nn" Perhaps if the day were coded as "99" that could be taken as meaning "not earlier than" the specified year and month. Does anyone else think this would be worth putting in the list of requested features? Obviously it would not be a top priority. -DES Talk 10:59, 25 Feb 2008 (CST)

There is often circumstantial evidence that allows us to come up with a fairly narrow date range, e.g. I own an undated ex-library copy of Red Planet and the first "checked out" stamp is from August 1974, so given the use of an SBN, the $5.95 price and this stamp, we can estimate 1970-1974. This is why many library systems let you enter a date range for books whose publication date is uncertain, e.g. 1950-1959 for a book published in the 1950s. I am not sure how much work it would take to add this functionality to our software, but I suspect the answer would be "quite a bit". No harm in asking Al, of course :) Ahasuerus 12:11, 25 Feb 2008 (CST)
I thought that Year->infinity ranges would be both the most common range we would have, and easier than a more general range feature to implement, as only the display logic need change, no additional editing fields would be needed. But a fully general date range facility would be wonderful, if it could be implemented. -DES Talk 12:38, 25 Feb 2008 (CST)
An UPPER limit is just as useful in the long-term - e.g. for instance ALL our '0000-00-00' titles currently have an upper limit of 2008-02-25 and could have had the date of entry automatically stored as the upper limit when the record was created. (I'm assuming nobody enters "forthcoming books" as "0000-00-00", we tend to get those from Dissembler.) But yes, it has been frustrating when I have an undated second printing and can't put it between dated 1st and 3rd printings. We've never agreed on a way to misuse the date field: but you will see past experiments with printing number in the Day part of the date, or added to the publisher name, based on First printing date (the lower limit you're proposing, probably). I think we'll have to bite the bullet and propose a couple of extra fields for upper and lower dates to be used when we can't put an exact (or nearly exact) date in the publication or title. Do we want to use these for unknown Month as well though? It can also be annoying when a magazine story with an exact Month gets sorted AFTER its later publication in a no-Month Anthology or Collection the same year. BLongley 13:00, 25 Feb 2008 (CST)
I don't disagree that an end-point is useful, I was just assuming that it was less common, and that if we had to focus on a single special case the start-point was the most obvious. Additional fields are obviously the most full-featured way to do this, abusing the existing field is only a hack, but it might be acceptable. (BTW, I would enter a forthcoming book if I saw it somewhere fairly reliable, and Dissembler hadn't gotten it and i suspected that it wouldn't. Would that be a mistake?)-DES Talk 13:10, 25 Feb 2008 (CST)
If you entered it with enough details and a definite date I guess it would get past most Moderators: then when Dissembler submitted it we'd at least have something else to check against. I can't say I've seen many such submissions though, and the only ones I remember actually managed to get the TITLE wrong, so probably didn't help. :-/ I don't know if Dissembler checks to see if we have records already, you'd have to ask Al if prior Human Input affects Dissembler's actions. And as to where Dissembler gets data from - well, having approved a few runs of Dissembler Submissions I'm pretty sure Amazon.com is one of them (even when Amazon.co.uk would have been better) but there's no official list of "where Dissembler looks, so you don't have to". BLongley 13:26, 25 Feb 2008 (CST)
I agree with the concept. The publication date is stored as a SQL date field rather than a string meaning that an extension in this area would require additional db fields. An idea that comes to mind that sounds easy to implement would be to add a single boolean field, pub.DateEstimated which if true would cause "(estimated)" displayed after the date in bibliographic displays. Within the notes you would document your logic for how you determined the date. I've frequently estimated dates but tend to hold off on entering them into ISFDB's date field as it would inject garbage/noise into later queries that are date related. If an estimated boolean field was added then the queries could use this.
While I believe it's a good idea I'm less comfortable with a date range as proposed by DES as I don't see a clear path on how to use this in the code. The pub_year field date is used in in both the title and publication display and in the title display it's also used to order the results. If we allowed ranges we'd have to develop rules for how this changes the ordering along with how it should be displayed not to mention adjustments to the help and user input rules.
A related area I've thought about is a pub.DateSource identity field plus a pub_date_source table which would allow programmatic entry of the source. This is similar to pub.DateEstimated but would allow us to add a list of sources that could be selected from. Right now the only source implemented (and implied) is "Publication" but there are others such as Amazon.com, "estimated", etc. The advantage of this over the boolean pub.DateEstimated is the queries could become more interesting. From a coding standpoint this is a little more work than the boolean but is also something that would not need a lot of "thinking." Marc Kupper (talk) 15:15, 25 Feb 2008 (CST)
To address Dissembler-related questions, Dissembler's internal logic is not public and it changes often, but as far as I can tell Dissembler doesn't create a submission if the ISBN is already present in the ISFDB. As far as Dissembler's reach goes, it trawls Amazon.* and perhaps other sites for data (IIRC there was some work done on Z39.50 a while back), but it doesn't know about obscure publisher's plans if they don't notify Amazon.* prior to publication. In a way, this is a self-selecting process since small presses that don't notify Amazon are known for delayed publications and canceled books, so their lists of forthcoming books are dubious at best and entering them in the database may do more harm than good. Forthcoming books are generally a tricky area since there is no easy way of telling if a projected book has been canceled and then we get stuck with phantom records that we can't easily find and delete. It doesn't happen very often when major publishers are involved, but small presses are a different story. Ahasuerus 18:08, 25 Feb 2008 (CST)

Publishers' Wiki pages

I've noticed a link at the top of the page when searching by publisher, e.g. Wiki Link: ISFDB:Doubleday. Since the namespace for publishers is now Publisher:XXX, I believe it would be nice if this linked to Publisher: Doubleday allowing us to then create a page for this publisher. If we create a page to the one that's linked now, the page name would have to be changed or redirected. I'm rather new to Wiki procedures so don't flame me if this sounds naive (or just plain stupid.) Thanks. MHHutchins 19:48, 26 Feb 2008 (CST)

It's a perfectly reasonable request and was recently posted on Al's page as per the results of this discussion. Al responded to the effect that he will make the change once his Internet connection is once again stable. Ahasuerus 19:57, 26 Feb 2008 (CST)
Thanks for pointing out the previous discussion. These pages have been so busy lately it's hard to keep up with the changes! MHHutchins 20:11, 26 Feb 2008 (CST)
And now the change has been implemented along with the the other two requested fixes! (And yes, the Wiki has been quite busy and sometimes hard to follow lately, probably due to all the software changes and the new editors joining, but that's a good thing :-) Ahasuerus 21:01, 26 Feb 2008 (CST)

Red Planet restored edition?

I just entered (not yet approved, as I type) some details for this publication. I mentioned in the notes that this is the restored edition, with the Dalglish edits undone. But should I create a variant title? This has been discussed in connection with RAH's works, but has not yet been done for this book, nor for Podkayne of Mars , or Stranger in a Strange Land. There might be later reprintings of the unrestored versions of any of these titles -- do we want to clearly indicate which printings contain the restored text? -DES Talk 15:05, 28 Feb 2008 (CST)

If I recall correctly, Heinlein restored some of the Red Planet changes for the first round of Ballantine reprints in 1977 (?), but it would be worth asking Bill Patterson or some other RAH expert about the details. For now, we may want to add notes to the Title as well as to the affected Publication records. Certainly some of the changes, e.g. in Stranger, were major and deserve their own Title record. Ahasuerus 16:47, 28 Feb 2008 (CST)
Ideally, IMO, we would record such significant changes. But we haven't in the past, and so as soon as we create the variant we might misdirect people into thinking only that one version is the variant. Heinlein is a pain for me anyway as I have to cope with No-A variants already and have to go back and check what people have done with those every few months, but I do try and keep the title notes up-to-date a bit. (I've currently got THREE "Stranger in a Strange Land"s at the moment if people want to try and separate those though. Or a brace of Farnham's Freehold. Probably a couple of Podkaynes. Only one Red Planet I think. Who's up for some actual reading?) BLongley 16:53, 28 Feb 2008 (CST)
Patterson wrote the intro to the 2006 Del Rey TP ed of Re Planet in which he says that "Finally, in 1992, Del Rey Books published a restored version..." -DES Talk 17:24, 28 Feb 2008 (CST)
I am sure the 1992 version was restored, but I wonder if my memory that Heinlein made some minor changes to the late 1970s Ballantine reprints is accurate. He was in pretty bad shape at the time (the famous comment about sleeping 16 hours and being worthless the rest of the day was made about that period), so even if I am right, I doubt it was anything major. Thankfully, Bill reads r.a.sf.w and it should be easy to check. Ahasuerus 21:09, 28 Feb 2008 (CST)
There are two questions here. One is the question of fact: which editions/printings were restored or changed, and how large are the changes from the original editions? That we can presumably find a factual answer to, and in this case without large problems, I think. The other is the question of process. Should a revised edition, known to be revised, but marketed under the original title, with no "revised edition" or "restored version" (or the like) subtitle on the title page be recorded as a variant title at all, or should this simply be a matter of publication-level notes? In particular, should this be done for Red Planet and those other RAH titles that were similarly restored late in his life or after his death? If not, the question of which publications were restored becomes more or less moot.
By the way, there is also the technical question, if a variant title is created, can existing publications be assigned to it, and if so, how? -DES Talk 11:41, 29 Feb 2008 (CST)
Yes, they can - unmerge the pubs you want to move from their existing title, then merge the titles that that creates with the one you want to move them under. The difficulty is establishing which pubs need to move. BLongley 12:24, 29 Feb 2008 (CST)
I see, thank you. This leaves the question of whether to create variants at all. I am getting the impression that people are in favor of doing so, but I would like to have this more clearly stated. i also have the impression that I should perhaps not start this until I have a reasonably clear idea of which pubs I would move into the "restored" variant, at least at first. Am I correct? -DES Talk 12:33, 29 Feb 2008 (CST)
There is a request to add support for "relationships" to the database. Once implemented, we will be able to link "expanded", "restored", "abridged", etc titles. If the new "relationship" field allows "one-to-many" links, we will also be able to use it to associate fixups with the stories that they are based on. Until then, I am inclined to continue our (sometime) convention of creating new Title records and adding " (abridged)", " (restored)", etc in parenthesis. The advantage here is that it will be easy to find all of these occurrences once the new field has been added. The downside is that in the meantime it will break the lexical match logic that is still used by reviews and serials. Hopefully, the lexical match logic will be replaced with real links sooner rather than later (it's already gone in the case of Awards), but you never know. Ahasuerus 16:52, 29 Feb 2008 (CST)
You can see from a quick search for titles with the words "expanded" or "revised" or "abridged" that people have recorded such variations, but looking closer usually suggests that these have been entered from secondary sources. I've actually removed a few variations that were labelled "Revised (UK)" when all that had apparently happened was that the publisher had run it past a British spell-checker. I'm quite happy to have significant variations recorded, although as usual they do cause other problems: people won't necessarily enter future publications under the right title, and someone will have to go check which variant they belong under. We've seen that with Authors with the same name already - new publications get attributed to the one we didn't adjust to [US} or [UK], often incorrectly. So it's probably best if ALL versions get an "Abridged" or "Unabridged", or "Standard edition" or "Expanded" suffix so that when a 'normal' edition arrives we know we need to push it one way or the other. We've often chickened out by leaving notes in the title record instead - e.g. I think I might have done that with "Stranger in a Strange Land", but am happy to give it another go. There the size difference is significant enough that I'd have a good guess at the versions.
And yes, we should look at keeping the review links in the meantime - again, I look at reviews as pointers to publications rather than vice versa, so as long as we keep the reviews matched we should improve the quality of review records (recording which version was being reviewed) and not send people in search of a a not-abridged but not-expanded-either version. BLongley 17:12, 29 Feb 2008 (CST)
I see Bill's point. In this case, i take it that there would be two titles: Red Planet (original) or perhaps Red Planet (as first published) or perhaps even Red Planet (Dalglish edition) for the version initially published , and Red Planet (restored) for the more recently published version. All publications would go under one or the other of these variants, at least if we can find sufficient data to show for a given publication which variant it belongs under. A similar approach would be taken with Stranger Puppet Masters and perhaps Podkayne. This would have the benefit of making it explicit which state any given publication was in, and if a new pub was entered without putting it under one or the other state, it would stand out as needing attention.
Does anyone else agree with this method of handling these works? I would like to come to a decision on this issue, but I suppose there is no particular urgency. -DES Talk 18:34, 29 Feb 2008 (CST)
I think we should be OK as long as we identify and adjust all related Review records to reflect the new titles of the original and revised books. However, keep in mind that in some cases this kind of treatment can involve a significant amount of work and result in a very busy page. For example, the Vance Integral Edition Project revised most stories/novels published by Jack Vance, so you would have to go through their notes (thankfully available online) and adjust dozens and dozens of records. Having said that, it's probably worth doing as long as it makes our readers aware that a restored/corrected edition of their favorite book exists.
Also, as Bill pointed out, sometimes the changes between editions can be very minor, from the typical UK/US spelling alterations to substitutions of certain terms like "the philosopher's/sorcerer's stone". Do we want to create separate Title records for that level of detail or do we want to relegate them to Notes? And where do we draw the line, e.g. would the changes between the US and the UK versions of Tim Powers' The Anubis Gates, which, IIRC, amounted to restoring about a dozen pages worth of very minor abridgments justify a separate title record? Ahasuerus 23:44, 29 Feb 2008 (CST)
"Stranger in a Strange Land" might be worth a look for another reason, the "audio" versions. I can tell just from the price that one is an abridged version and the other is probably the full one, but I also get annoyed that people still think "audio" is a good enough binding description. I don't buy audio books myself, but I know I'd be annoyed if I used this site as a link towards getting the audio book I wanted and discovered I'd bought a cassette when I wanted a CD. Or bought an MP3 CD when I can only play normal CDs. There are audio books on DVD as well now. "Abridged" and "Unabridged" are probably far more important for audio books than dead-tree ones, but PLEASE, folks, don't dismiss all audio books as "'audio' for Binding and '0' for Pages removes the bibliographic warnings, that'll do". BLongley 17:29, 29 Feb 2008 (CST)

(unindet) I agree this needs judgement. In some cases it may be too much work to be worth while. I don't think a few spelling changes justify a varient title, nor does corretion of typographic and copyediting errors in the original. An interesting (and questionable) case is the Eric Flint edits of the reissues of James Schmitz's reissued work that causeed so much furor over on r.a.sf.w. a few years ago. I think that the heinlein restorations (certianly in Red Planet' and Stranger) are significant enough to justify going ahead, carefuly preservcing all review recordss as mentioned above.

As to Audio, that is quite another issue. I am no so worried about people purchasing the wrong format -- anyone who clicks on a purchase link and fails to chaeck that the merchant is selling, on that page, the exact product desired deserves little sympathy, IMO. There is no way we can check that mercahnt links always lead where thy ought to, even if all our info is correct. However, documentign the differencve between a cassette, a CD, and a DVD seems to me just as proper and importan as documentign the difference between a hardback, a trade paperback, and a mass market paperback, and for pretty much the same reasons. I am going to rais this over on Rules and standards discussions. -DES Talk 06:47, 1 Mar 2008 (CST)

What happened to the magazine pages?

The magazine pages for Amazing Stories, F&SF, If, and maybe others are now blank. Is this temporary, or does something need to be fixed?--Rkihara 10:43, 1 Mar 2008 (CST)

That is the result of the true "Magazine" namespace beign installed, see User talk:Alvonruff#Publisher Namespace. The pages created in the pesudo-namespace need to be moved out of their old names and then back into the true namespace. Al will have to do this or cooperate with whoever does it. -DES Talk 10:53, 1 Mar 2008 (CST)

Namespace problem

We have a problem. Al installed the requested namespaces. However, the various pages which formerly lived in pesudo namespace (i.e. any page whose name pagan, as of yseterday, with "Publisher:", "Author:", "Publication:", "Magazine:" or any of the other new namespace prefixes) are now all inaccessible, and will be until they are properly moved to their new locations. This process will require cooperation with Al. Until it is done, when people create/recreate pages like Publisher:DAW, which have the same name as one of the temporarily hidden pages, a problem is caused. The hidden page can only be put back in its proper place if the new page is moved elsewhere (and the resulting redirect deleted), or itself deleted: either of these will require moderator intervention. Creation of any more Publisher:, Author:, Magazine:, or Publication: pages before the old pages are moved back into place will only make the amount of ultimate manual intervention needed larger. -DES Talk 17:01, 1 Mar 2008 (CST)

Oops and thank you for the heads up David. I was just working on Author:James Gunn, Publisher:The Easton Press, and Publisher:Easton Press (just a redirect) but have moved them to Author-Temp:James Gunn, Publisher-Temp:The Easton Press, and Publisher-Temp:Easton Press for now so that there will not be a conflict. I have no idea if the pages already existed. Marc Kupper (talk) 17:13, 1 Mar 2008 (CST)
See prior discussion of the namespace issue in #Publisher Issues and User talk:Alvonruff#Publisher Namespace. -DES Talk 17:20, 1 Mar 2008 (CST)
To check if a page existed previously but is "in hiding" for the moment, go to Special:Allpages and enter the name of the page you are looking for, or anythign that comes before it alphabetically. if it is listed, it existed and may be "in hiding", this is a list of all the "in hiding" Publisher pages, and this is a list of the "in hiding" author pages. I hope and trust that this problem will be fully dealt with fairly shortly.-DES Talk 17:29, 1 Mar 2008 (CST)
Please note that this "allpages" solution will, as far as I can tell, not work if it is tried after a new page has been created, whether or not it has then been moved or deleted. The new creation overwrites the old record whose presence or absence is the test. -DES Talk 17:41, 1 Mar 2008 (CST)
I e-mailed Al but don't know if he's working on this. All of the Author pages vanished (and then reappeared) and so I assume he's working on it. Roglo created a new DAW page, I see that it was moved to Publisher-Temp:DAW and so I have deleted the original Publisher:DAW was that had a redirect to the new page. Publisher:Books on Tape is new but that should be ok as it's not in the root namespace. I'm headed off line for a while. Good luck. Marc Kupper (talk) 17:52, 1 Mar 2008 (CST)
Publisher:Books on Tape was mine but can be recreated if it causes problems. I'm not Wiki-Knowledgeable enough to know, so I'll just hold off with the rest till someone gives the all-clear, OK? BLongley 18:50, 1 Mar 2008 (CST)
True new pages, ones that did not exist on 29 feb 2008, will cause no problems to the best of my understanding. -DES Talk 19:02, 1 Mar 2008 (CST)
Short explanation. The wiki has an internal table listing the "namespaces" that are active -- each has a number and a name/tag. any page whose name starts with the tag of an active namespace is stored in that namespace, and wiki-links to it include the namespace number in their internal form. Pages whose name starts with a tag and a colon where the tag does not match one of the currently active namespaces are stored in namespace 0, the "main" or "article" namespace. Such pages are said to from a pseudo-namespace, because in many ways they can be treated as a namespace, but the internal wiki code does not recognize them. If a namespace is later created that matches existing pages in a pseudo-namespace, when a user tries to display or edit the old page, the wiki software now looks for a page with the internal number of the now-active namespace, doesn't find it, and reports the page as not existing. It is in limbo, with no way to get to it though the normal wiki interface. To fix this, either a) a script has to be run that updates things behind the screen (said script is provided by mediawiki) or b) the new namespace must be made made inactive (commented out), all pages in the pseudo-namespace(s) moved to new names that don't conflict with the new namespace(s), then the namespace is reactivated, and the pages moved back, which will put them into the new, true, namespace so that the wiki software will find them. I hope that is clearer. -DES Talk 19:02, 1 Mar 2008 (CST)
Commenting out the namespaces in LocalSettings.php has no effect on the current situation (individual magazine pages are still in limbo). The script in question (namespaceDupes.php) is not part of our current distribution, and since I don't have shell access to the system, we can only execute php scripts through the web interface. If someone can post exactly what needs to be done to repair things (as I am not a namespace expert), then I can look into what can be done here. The namespace lines in LocalSettings.php are currently NOT commented out. Alvonruff 21:20, 1 Mar 2008 (CST)

(unindented)A quick check of all the publishers pages that I put notes into are all empty. How do I get this info back or is it lost? Even stuff that I put in last night is gone. Kraang 22:00, 1 Mar 2008 (CST)

It's still there, just not visible at the moment. To quote DES, "Don't panic!" :) Ahasuerus 00:16, 2 Mar 2008 (CST)

The MediaWiki Help page gave the following instructions:

If the number of pages affected is small (e.g. "Bar" held five pages created before the namespace was defined in the site configuration), then the following approach might be suitable:

  1. Comment out the namespace definition in the configuration file
  2. Access each affected page, and move it out of the pseudo-namespace, e.g. move "Bar:Some page" to "Bar2:Some page"
  3. Un-comment the namespace definition
  4. Move the affected pages back into the new namespace

or

Within the maintenance directory, there is a maintenance script which performs the above operation more effectively for a large number of pages; namespaceDupes.php. It is simple to use, but as with all MediaWiki maintenance scripts, consult the available usage information first (use --help) as an option.

I have downloaded namespaceDupes.php and the include files it calls, and placed a copy at User:DESiegel60/Namespacedups -- i'm not sure if this will be hlpful or not. It does include SQL code that might be extracted.


When you (Al) commented out the lines in LocalSettings.php did you restart the wiki software, or anything esle that might be needed to get them to take effect? Did you try to acces any of the preempted pages by going throuhg special:allpages? -DES Talk 01:07, 2 Mar 2008 (CST) Oh, i assume you forced a browser cache refresh? -DES Talk 01:31, 2 Mar 2008 (CST)

Just to confirm the "Don't Panic" advice, you can still see the old DAW page here (but please don't edit it). --Roglo 07:31, 2 Mar 2008 (CST)
Hi Al - I see you have made good progress. Is there anything you need help with? For example, I see
  • In the Author namespace the pages need to be moved from "Author:~name~" to "~name~". Someone has moved Marly Youmans up to Z already.
  • The Talk pages are still in limbo land for Author, Magazine, Publisher, etc.
The DAW pages need some cleanup as I see they are spread among AllPages, AllPages Talk, ISFDB, ISFDB Talk, and Publisher namespaces. Marc Kupper (talk) 13:26, 2 Mar 2008 (CST)

Trying to get Started

I came upon ISFDB while doing a (so far unsuccessful) search for a SF book containing a very specific term. I don't know yet if this is even a valid use of this site. But here goes (if I'm out of line just tell me): a SF book in roughly the 1960 time frame used the term (best guess)"magellanium" which was a previously unknown super dense metal which was a key element in the story. I'm trying to find the title and/or author of the book. Thanks.

The book was identified on r.a.sf.written by Mike Stone back in 2002. It's David_Duncan's (not to be confused with the much younger Canadian writer Dave_Duncan) Dark Dominion. See http://groups.google.com/group/rec.arts.sf.written/browse_frm/thread/93668917a3a2140c/6a122c6f1c271d3c?lnk=st&q=&rnum=1#6a122c6f1c271d3c for details. Ahasuerus 22:04, 2 Aug 2006 (CDT)

Request to Speak with Admin. about Editing Author Content

Perhaps someone , hopefully one of the founding admins ,can contact me regarding the specific request to get an authors/artist information updated on the index ? After much searching thru this site i was unable to find any contact information / email , and hope i am not inappropriately postly here , but i need to speak with someone connected to the database / index and there did not seem any other place to post this. thanks much for understanding. my email is : chezzwick@optonline.net - if anyone could respond it would be much appreciated. thank you (message added by Chez)

The following stories need to be added to the short fiction of Steven Popkes: "Hellcatcher", Night Cry Spr 1986, "Winters Are Hard", Sci Fiction website Nov 13 2002, "Holding Pattern", Fantasy & Science Fiction, July 2006, "Stegosaurus Boy", "Realms of Fantasy", February, 2003, "Boulder Country", June, 2002 Realms of Fantasy -- Spopkes

At this point user submissions, including data corrections, are disabled as we are completing a long and painful migration from one database format to another and getting the bugs worked out. With luck, they should be re-enabled later this summer -- well, it could be a reeeeeeeally long summer :) -- and then you can submit the correction. For now, feel free to document the correction in this Wiki or even on this page and I'll move it to the appropriate Wiki page. Thanks! :-) Ahasuerus 15:48, 8 Aug 2006 (CDT)

New Editor Question/Adding a Collection

I just attempted to add the Fritz Leiber collection The Black Gondolier to ISFDB. I could have sworn I added the author name but when I checked over the data and hit enter, I got an error saying I had not added the author and then lost all the data I had entered when I hit back. Any suggestions so that this doesn't happen again? I'll probably try it from a different computer for one... Sigil23 14:09, 8 Jan 2007 (CST)

Did you enter Leiber's name for each story in the "Content" section or did you also enter it at the top of the form? The latter is always required for all types of books, be they collections, anthologies or novels.
As far as error handling goes, unfortunately it's somewhat rudimentary at this point. Ideally, the software should have put you back in edit mode without any loss of information, but we are not quite there yet :( Ahasuerus 14:27, 8 Jan 2007 (CST)
Thanks for the feedback, it looks like it happened because I didn't add the author name for each story. Doubt I'll make the same mistake again! Understand about the software, will work around. :)
Feature Request 90083 has been updated by Al von Ruff so that it now calls for submission re-editing capabilities instead of the current behavior, which is limited to displaying an error message and losing the submission data. Ahasuerus 18:53, 8 Jan 2007 (CST)
You can also work around this by using Firefox, it doesn't blank form fields when you go back in situations like this (and I'm sure there are other browsers that work the same, I use Epiphany). I've encountered this when forgetting to enter an author name when entering an anthology, but never suffered any data loss because of it. I just press back and enter the missing details. --Unapersson 17:50, 16 Feb 2007 (CST)

Dating Question

I can see that the date for Great Wall of Mars by Alastair Reynolds is wrong - it says 1990 and it should be 2000 or 2001. How can that be corrected?

[11] Lawiedc 07:09, 8 Mar 2007 (CST)

From 99521 click on "Edit Title Data" in the left navigation bar. The date format is YYYY-MM-DD meaning you would enter 2001-00-00. If you have the story and the copyright is 2000 you may want to add a comment to the notes about this to help explain why some sites will list this as a year 2000 story even though there were no publications of that title published in 2000. Marc Kupper (talk) 12:35, 8 Mar 2007 (CST)

Accidental duplicates resulted in rejected submissions

I accidentally added a "new novel" twice (the first submissions gave me an error message, so I submitted a second time). My submission was rejected due to "duplicate submissions". How can I still get the novel listed?

Thanks —The preceding unsigned comment was added by Firefang (talkcontribs) .

I am happy to report that the first submission made it to the submission queue and was approved at 4:01pm ISFDB time -- see this record for details. Are there any other editions of this book you would like to add? Ahasuerus 00:17, 6 Apr 2007 (CDT)

Jim Baen's Universe Revisited

This paragraph was originally posted with minor differences on the Ahasuerus talk page. I addressed it specifically to him partly because of his knowledge of technical issues but now I take it to the Community Portal.

Magazines with ISSN numbers are supported: Publication Fields. There is support for ebooks with ISBN numbers: ISFDB:Policy. So doesn't it follow that ezines with ISSN numbers should also be supported? The distributed downloadable forms of Jim Baen's Universe, in my case the mobipocket reader edition, have ISSN numbers. The main reason this issue bugs me is that it gives the printed mags an unfair competitive advantage. Note the advertising info for F&SF.--swfritter 11:27, 18 Sep 2007 (CDT)

There was an unnecessarily long convoluted discussion of this issue under the heading of ezines.

The software supports publications with ISSN's. The software supports electronic publications. It does not require a script to add the data. I own all the issues and will volunteer to add and maintain them.--swfritter 08:28, 20 Sep 2007 (CDT)

For more informations on ISSN numbers goto What is an ISSN?. Take special note that any change in the content of a magazine with an ISSN requires that a new ISSN be issued.--swfritter 14:26, 20 Sep 2007 (CDT)

There was also a long thread on Dcarson's Talk where there is a discussion of the original submission being for the website. My proposal is that only the downloadable version with ISSN numbers be added to the database and that a notation be made on the magazine summary page that the stories are also available online. The website could vanish at any time - we still have magazine entries for SF Site which document stories that are no longer available on the web.--swfritter 11:04, 21 Sep 2007 (CDT)

Once I discovered that e-zines with ISSNs can't change their contents after the fact (duh!), I agreed that we want to catalog them. The biggest headache that we had with Web publications in the past was their mutability and as long as that issue is addressed, I think we should be in good shape. Ahasuerus 16:56, 23 Sep 2007 (CDT)
I vote FOR inclusion, yet again. (Can we please come up with some rules for making a Voting Issue clear? Steal the Moderator Nomination process or something...) BLongley 17:16, 23 Sep 2007 (CDT)
Marc Kupper has turned this one over to me. He also seems to agree that they should be included. The original submission was for the website so it is quite justifiable that the submission was questioned. I am assuming that we have a consensus and will approve the submission in a couple of days. I will also update the appropriate Help screens.--swfritter 12:03, 25 Sep 2007 (CDT)
And as it turns out the point of whether or not to include them as magazines is totally moot. The issues have ISBN numbers and are therefore technically ebooks which have been acceptable for quite awhile. The ISSN number appear to apply to the series and are incorporated into the ISBN number. But it was fun getting to know all the people who participated in the discussion.--swfritter 18:55, 29 Sep 2007 (CDT)

Feature request: Search Pubs by Note field

On Advanced Search -> ISFDB Publication Search Form there is no Notes field on the dropdown list. And we have both issue numbers and book series ('publisher's series' like SF Masterworks) recorded. BTW the field is called Note in the editor, but Notes on the Publication Listing page. --Roglo 05:10, 17 Dec 2007 (CST)

Hm... "Notes" are actually a separate table in the database and I am not sure how computationally intensive an exhaustive search would be. Something to ask Al, I guess. Ahasuerus 10:37, 17 Dec 2007 (CST)
It's not a huge table though: e.g. I just searched for notes with "Sf Masterwork" in them and it took three-eighths of a second. One second more to get the publication ID and title to go with them. OK, sometimes I suspect my PC is more powerful than the ISFDB server, but it doesn't look too intensive. BLongley 11:48, 17 Dec 2007 (CST)
By the way Roglo, if you're actually interested in such searches before Al implements them, I can run them against the last backup for you if you like. Just drop me a note on my talk page. BLongley 11:48, 17 Dec 2007 (CST)

Split novels - theory and practice

The Help:Screen:NewPub (Pub Type -> "Split" novels) states that the original book is still treated as a novel; it does not become an OMNIBUS because it contains two works later republished as novels. However it links to example The Reality Dysfunction, which is recorded as an OMNIBUS. Is there a reason for this, or should I try to fix it? --Roglo 16:27, 18 Dec 2007 (CST)

Good spotting! We often have some really poor examples of "rules", sometimes because we've changed the rules since the help was written. (We all seem to prefer following the rules we THINK we have rather than what the help actually SAYS, and we're often too lazy to change the help.) An example of the opposite problem might be Mick Farren's "Phaid the Gambler" which still needs sorting out. I can't think of a GOOD example of a split novel where we've got Verified pubs on both sides and have all agreed that's what we want, can anyone else think of an example offhand? BLongley 17:01, 18 Dec 2007 (CST)
(Oh, and in the case of Peter F. Hamilton: I've picked up two of his books THINKING I've bought an Omnibus, only to find it's not only a single 1000+ page Novel, but part of a Trilogy! They should come with warning labels... ) BLongley 17:01, 18 Dec 2007 (CST)
Hm, I have to edit the Title and all publications to change the type to NOVEL or maybe they are somehow connected? --Roglo 17:26, 18 Dec 2007 (CST)
I'm afraid all the pubs need doing separately - you can change the Title type on one of the edits too, but that's about the only short-cut I can offer. BLongley 06:29, 19 Dec 2007 (CST)

Problems with REVIEWs

I have a strange problem with REVIEW record. Here is a review of Black Man, and here is the Black Man record (the canonical title) which doesn't display the review. What's wrong with this REVIEW? I have run out of ideas... (BTW Reviews of books are displayed on bibliographies of eponymous stories, e.g. here. But perhaps sometimes stories are being reviewed, too?) --Roglo 05:11, 19 Dec 2007 (CST)

This was a bug in the title.cgi code (which would occur anytime a title had a review and a variant author existed for that title). Fixed now. Alvonruff 06:04, 19 Dec 2007 (CST)

Thanks! I have even more problems but perhaps it's browser's fault? (Firefox 2.0.0.10): See Reaper’s Gale. The title has printer's apostrophe. The publications have plain ' (I suppose). After I entered a review, the titles didn't match. I was trying to fix it, but while the text in edit mode matches the NOVELs title, the text on the publication's listing here doesn't match. I mean, when I copy the reviewed book's title from magazine's Publ. Listing to the Search field I get REVIEW, when I copy the text from input box (while editing the magazine's contents) to the search field, I get NOVEL. Strange, but repeatable. And the NOVELs bibliography doesn't list the review. --Roglo 06:23, 19 Dec 2007 (CST)

Apostrophes are one of the more common annoyances I find here - certain screens helpfully "correct" non-standard apostrophes to regular ones on entry, but then they don't match the non-standard ones already stored here. It would be nice to standardise all the underlying apostrophes with a big SQL clean-up, but I fear it would destroy certain foreign characters. In the meantime, the workaround seems to be to remove all apostrophes from the titles you want to match, then put them back afterwards. BLongley 16:00, 9 Feb 2008 (CST)
As for OTHER problems with reviews - how did we get this and this? Another case for Detective Roglo, maybe? You could call it the "Alan Parsons Project project". ;-) BLongley 16:00, 9 Feb 2008 (CST)
Well, well, the Case of Publess Reviews ;-) --Roglo 16:16, 9 Feb 2008 (CST)
The first one was 'replaced' by 17 • Review of Tales of Mystery and Imagination (audio recording) by the Alan Parsons Project • essay by Harlan Ellison in this pub. The essay is quite recent. The Alan Parsons Project was removed from Authors that only exist due to reviews on 29-30 December 2007. So now I'm learning how to break things :-) --Roglo 06:05, 10 Feb 2008 (CST) (edited --Roglo 07:02, 10 Feb 2008 (CST))
And this is the essay replacing the second Alan Parsons review. --Roglo 07:31, 10 Feb 2008 (CST)
For now, we have to delete reviews and interviews by hand after using Remove Titles From This Pub because, just like for other titles, only the pub_content records are removed by Remove Titles. Probably reviews and interviews should be removed automatically if there are no more pub_content records for them (but this has to be checked, as the pub might be cloned or reviews might be reprinted in books and records merged perhaps?). This is something to discuss. --Roglo 07:27, 10 Feb 2008 (CST)
And then the Mystery of Publess EDITOR? --Roglo 03:41, 10 Feb 2008 (CST)
The New Yorker seems to be a source of other strange records of no apparent use. BLongley 05:59, 10 Feb 2008 (CST)
I can explain the stray Alan Parsons Project reviews by Harlan Ellison. When I first entered the issue of Delap's containing these reviews, I naively recorded them as REVIEWs, making the band the AUTHOR and the album title the BOOK. After Bill created his list of reviewed authors without pubs (where "Alan Parsons Project" appeared along with other creators of non-books), I made a decision to make these reviews ESSAYs. I did this to several Delap's that I'd verified, and deleting the original review title records. Unfortunately, these two Alan Parsons reviews escaped me. I'll go ahead and delete those two stray titles. In the future, I plan to enter all reviews of non-books as essays to avoid getting on Bill's List. Hasn't there been a discussion on changing the standards about recording non-book reviews as an ESSAY instead of a REVIEW? Mhhutchins 10:16, 10 Feb 2008 (CST)
There was some preliminary discussion here - I didn't put a solid proposal forward for standards discussion as we might want to extend it to some nongenre or nonfiction books and magazines and fanzines as well. Essays for reviews of soundtracks and films and TV shows should be fairly uncontroversial: I'm not sure how many types of review we've actually got yet, but we're getting close to the end of the original list. And should we standardise the name of the essay? BLongley 11:59, 10 Feb 2008 (CST)
It looks like there is consensus re: recording "non-book reviews" as Essays rather than Reviews, where "non-books" are defined to include films, TV shows, games, and other performance arts (songs, theater, etc) and defined to exclude all printed material and audio books. If we use this definition in Help, it should cover about 98% of our problem cases and then we can spend the next three years arguing about the other 2% :) Ahasuerus 13:22, 10 Feb 2008 (CST)
OK, I've updated the 'Stray Authors from Reviews' page a bit, someone go update the help and we'll nail this down eventually. Of course the next step is 'Missing Titles from Reviews' - which will be a MUCH bigger project. BLongley 16:33, 10 Feb 2008 (CST)
Should we record all such reviews as individual essays? (E.g. I have movie review column in Interzone, should I add an essay for every movie reviewed?). --Roglo 13:33, 10 Feb 2008 (CST)
I wouldn't - but I'd leave it up to the individual editor's choice. I think I'll take the opportunity to actually get myself listed as an author though. (It saves me from writing a proper BOOK!) :-) BLongley 16:33, 10 Feb 2008 (CST)
BTW I added publess reviews/interviews to ISFDB:Data Consistency. --Roglo 13:33, 10 Feb 2008 (CST)
And there are 233 publess EDITOR records, including 34 of Yandro and 27 of Locus. What to do with them? E.g. publess Locus - 1995 (Semi-Prozine) with an award, while we have 11 Locus issues from 1995. --Roglo 13:54, 10 Feb 2008 (CST)
I see Marc Kupper is working on them and empty EDITOR records seem to be used to mark that the magazine was published that year but we do not have the content listings yet. --Roglo 14:02, 10 Feb 2008 (CST)

Page count

How exactly is interpreted the word text in It is fairly common for the last page of text in a book to have a different graphic layout which may not include a page number. The "last printed page number" rule would then use a page number before the end of the work. In these cases, count forward to the end of the text and use that as the last page number? (Help:Screen:EditPub). Is it all the text printed in the book, including contents list, publisher's notes etc., or the text of the main works(s), i.e. the text of the novel, stories etc.? The part on the last page of text in a book [having] a different graphic layout suggests the last page of the work (often if the text doesn't reach the bottom of a page, the page number is not printed) and later the end of the work is mentioned but I usually count till the last printed page. Should I use the last page of the work?

For books, I usually use last page of the items I bother to enter, and the items I bother to enter are usually mostly numbered. This means that I'll often not enter an unnumbered "About the Author" page immediately after the last page, although I'll read it to see if it improves our author Bios. Sometimes it's something I'd want to find here, like a full bibliography so far for the author - that could be worth entering even if unnumbered. Or a list of Awards - I think we've got most of the standard ones recorded though, I'm not sure I'd repeat Nebula or Hugo listings unless numbered or someone else has already entered that section. I definitely don't enter advertising for other books by the same publisher (or even author) - and recently I've stopped bothering with "excerpts from forthcoming books". BLongley 14:54, 22 Dec 2007 (CST)
Mostly all I need to do is look back one page and see that the last but one page was numbered, so add 1 to that for the last page for the count. BLongley
Thanks! So the rule is to record the last page number of the last text recorded in the Contents listing (a novel is there, too, just not displayed, usually). In the Contents listing we record Afterwords but not small publisher's notes nor printed contents listings. --Roglo 05:52, 23 Dec 2007 (CST)

The case of Mixed Series

I've created a new series The Apocryphs. Space included. Initially I couldn't find it on the author's page. Now, I've found it. One ESSAY was enough to move it to the Essay Series section. Not quite what I expected. (I'm not sure where these texts do belong but most of them are fiction). The SHORTFICTION in this series is still displayed in the Shortfiction section, the ESSAY is displayed only as part of the series. And the Bibliographic Comments link doesn't work due to the space in the name. Should I rename the series?

Also, I've noticed that while on the author's page the variants are grouped with their parents in the series, on the series page the variants are included only if they were explicitly added to the series. Should I edit all the missing variants to add the series names, or this will be handled automatically at some time later? --Roglo 10:50, 23 Dec 2007 (CST)

The way the software currently works, a fiction series consisting solely of short fiction Titles will not be grouped together on the Summary page. If you have at least one book length entry (a Collection or a Novel) in the series, then all Titles will appear together. It's a known bug, I am afraid :(
As far as the space in the Series name goes, it should have been converted to an underscore when linking to the Wiki. I have just tested it and it worked fine for me. Are you sure you didn't get redirected to isfdb.tamu.edu? If you did, just change the URL to www.isfdb.org as per the instructions on the Main Page of the Wiki. Hopefully, this issue will go away in mid-January once TAMU has upgraded its software.
Yes, sorry, it was underscored and I was redirected. My mistake. --Roglo 14:36, 23 Dec 2007 (CST)
Finally, the Series display page is far behind the Summary page in terms of functionality and variant titles are not displayed at all. The original idea was to get the Summary page to work the way we want it to work and then make all the other pages work the same, but then Al, our programmer, had to take an extended leave of absence due to RL issues and is only now getting back into the swing of things. It will get fixed -- eventually :) Ahasuerus 14:16, 23 Dec 2007 (CST)
So I don't have to worry about how the works look on the series page. Thanks! --Roglo 14:36, 23 Dec 2007 (CST)

Brian Aldiss cloned?

It doesn't seem to be a serious problem but Brian W. Aldiss has alternate name Brian Aldiss displayed twice, and on the pseudonym's page the canonical name is displayed twice. Also, on the pseudonym's page, 4 essays are displayed and they seem to be missing from the author's canonical page. --Roglo 11:05, 23 Dec 2007 (CST)

Essays sorted - they just needed a variant creating under the canonical name. Unfortunately the pseudonym has been set up twice so I can't sort the double display. BLongley 11:09, 23 Dec 2007 (CST)
Thanks! --Roglo 11:24, 23 Dec 2007 (CST)

Good Wiki

Oh, and by the way, thanks for fixing the Wiki.--swfritter 16:54, 3 Feb 2008 (CST)

The Incredible Disappearing Ed Emshwiller

Emsh vs. Emshwiller. I just checked the covers that were credited to Ed Emshwiller during the first six years of Galaxy and all of the about twenty covers that were credited to Emshwiller in The ISFDb were credited to Emsh in the magazines. I changed those but did not go any further in order to avoid having somebody make them pseudonymous. I strongly suspect that the vast majority of remaining Ed Emshwiller attributions are wrong. Anybody thing of a way to avoid somebody making mass changes to the Emsh entries that would make them pseudonymous works of Emshwiller until this can be sorted out? This might be one of the few cases where a script designed to wipe out a few pieces of data, the alternate names assigned to Emshwiller might be justified.--swfritter 21:02, 4 Feb 2008 (CST)

My F&SF November 1972 copy clearly lists Ed Emshwiller as the cover artist. rbh (Bob) 17:07, 5 Feb 2008 (CST)
That appears to be only one of two F&SF issues that were credited to Ed Emshwiller. All the rest are credited to either Emsh or Ed Emsh.--swfritter 22:27, 5 Feb 2008 (CST)
Well, I guess the best thing to do is go through my collection, modify any cover art attributions that are wrong, and hope that nobody does a mass attribution to Emshwiller until we can determine the most valid canonical name.--swfritter 18:25, 5 Feb 2008 (CST)
I can't think of any way to PREVENT a certain sort of edit being made without Al changing the programming. Discouraging Moderators from APPROVING such edits should be as simple as changing the Moderator Help - but to be honest, how many Moderators keep up to date with it? I suspect we new Mods read it a few times until we think we understand it, and then we think we know enough and don't notice any changes. To be sure everybody was aware of something you feel really could be disruptive, I'd look at the Moderator Availability list and message ALL of them individually. BLongley 18:55, 5 Feb 2008 (CST)
Isn't it standard practice that when a canonical name has been established that variants are created of all pseudonymously credited title records? I'm not talking about changing the credits within the pub record, just the creation of a variant so that it will appear on the canon's summary page. Dave has been creating variants for the interiorart Analog credited to "Kelly Freas" for some time now. And I've been approving them! Was I wrong? Mhhutchins 19:54, 5 Feb 2008 (CST)
Oh, I get it. You feel a canonical name has not been established. It appears that "Emsh" has been made a pseudonym of Ed Emshwiller, so Emshwiller is currently the canonical name. Is the determination of which name becomes the canonical one based solely on the number of title records? If so, I can point toward several authors in the database where that is not the case. (In some cases, the entire page is filled with [as by..]'s.) I'd love to try my hand at changing some of those, but so far have let the temptation slide. Mhhutchins 20:04, 5 Feb 2008 (CST)
I don't think a strict number count should necessarily be the decider - especially for artists. You could factor in the later credits, name recognition, and possibly an emphasis on the person's real name if there are a significant number of pieces by that name. Another factor here in favor of Emshwiller - if you search for emsh you will find emshwiller but if you search for emshwiller you won't find emsh. Emshwiller would be fine with me but I want to make sure we have the correct data to make a decision.--swfritter 20:15, 5 Feb 2008 (CST)
This is an interesting discussion. Galaxy was one of the first magazines that I started editing. I noticed that the Ed Emshwiller attributions for the covers did not match the credit, which was Emsh, but I was new and uncomfortable about changing something that someone else had entered (I still am), so I left them alone. I did however, credit the interior illustrations to Emsh, as signed or credited. The majority of the illustrations were credited to Emsh, so I entered a few as Emsh, when they were signed Ed Emshwiller, or Emshwiller, or Ed Emsh, especially if the particular magazine usually credited him as Emsh. The Strauss index shows only Emsh for the covers of both Galaxy and F&SF. I should probably go back through with swfritter and correct some of these, since I've contributed to the situation both passively, and actively.--Rkihara 20:43, 5 Feb 2008 (CST)
Don't worry about the covers. I will get them all. I have no doubt that some faithful editors will at some point in time will be making a comprehensive pass through all the mags just to make sure the artwork attributions are consistent and well documented.--swfritter 21:16, 5 Feb 2008 (CST)
A sweep through Galaxy, Amazing, Astounding, If, and F&SF and more than 95% of the Emshwillers were Emsh or Ed Emsh. Most of the remaining records are probably a little suspect.--swfritter 22:27, 5 Feb 2008 (CST)
Yet another pseudonym for Emsh - Harry Gars which is actually miscredited in the table of contents of the March 1953 Amazing to Ed Garo - the editor misinterpreted the signature. The Emshwiller page is looking pretty slim but there are still plenty of arguments both ways. I would strongly suggest not changing anything credited to Ed Emshwiller until there has been a physical verification of the pubs. Since the Gars pseudonym is so obscure I would like to get the information into the system before I forget about. Either way there will need to be some cleanup because some titles have been made pseudonymous entries for Emsh and some for Emshwiller.--swfritter 18:58, 6 Feb 2008 (CST)

Unprotection

Is it safe to enable editing for all the pages that were nailed down because of the robots? The list of protected pages should be at list of pages that link to Template:Protected Marc Kupper (talk) 23:12, 6 Feb 2008 (CST)

Keep in mind that TAMU is in the process of deciding when/how they will upgrade the ISFDB software, including the Wiki. If we wait another week or two and they upgrade the Wiki to version 1.9 or above, we should be able to improve our anti-bot defenses significantly, so it may be the prudent thing to do. Ahasuerus 19:08, 9 Feb 2008 (CST)

Robots are blocked?

Why is ISFDB blocking all robots? I used to use Google all the time to search the ISFDB sites but now Google only has

  • 39 pages for www.isfdb.org
  • The same 39 pages for isfdb.org
  • 336 pages for isfdb.tamu.edu (wiki pages)

Even with the few pages that are left Google is not showing cached results meaning I can't see how old Google's copies are.

robots.txt is dated April 25, 2007 and says

User-agent: *
Disallow: /

Marc Kupper (talk) 23:23, 6 Feb 2008 (CST)

Hm, that's a very good question, but I suspect that only Al knows the answer. Unless, of course, it's something that TAMU did unilaterally, but that seems unlikely. Ahasuerus 23:54, 6 Feb 2008 (CST)
I'd like at least the Wiki to be searchable at Google: we tend to only add to that, whereas you could argue that people should always query the LATEST version of the database. BLongley 13:42, 7 Feb 2008 (CST)
I use Google regularly to search the entire site. For example, to track down where Orsin_Scott_Card comes from I'd use site:www.isfdb.org "Orsin Scott Card". Marc Kupper (talk) 15:28, 9 Feb 2008 (CST)
Well, if you don't have a good set of backups then the Wayback Machine should be even better than Google, but that's banned too. :-/ To answer your example though, "Orsin Scott Card" comes from this record. I see we don't have that at Stray Authors caused by reviews, but I'm reluctant to refresh that till we can agree on what we want to record about non-book reviews, or reviews of books that are at least two degrees removed from SF. BLongley 15:41, 9 Feb 2008 (CST)
Thank you - I dealt with that review by fixing the author name and adding a note to the publication asking a future reviewer to check for "Orsin." Marc Kupper (talk) 00:20, 10 Feb 2008 (CST)

Kuttner and C. L. Moore

There are a fairly large number of Henry Kuttner stories that have been credited to Henry Kuttner and C. L. Moore despite the fact that Kuttner is listed as the only author in the pubs. Is there are any bibliographic source that justifies co-authorship credit for specific stories?--swfritter 16:11, 8 Feb 2008 (CST)

I have read several sources that state that published author credits to either Kuttner or Moore do not reliably exclude the work actually being, to some degree, a collaboration. I recall at least one source asserting that after their marriage, pretty much everything they wrote was in fact co-authored, whatever was on the title page. How authoritative this is, i can't say -- i would need to do some digging to even find where I read this. I don't recall reading any source that claimed to give detailed author attributions by work for these authors. -DES Talk 16:19, 8 Feb 2008 (CST)
That was also my understanding but I was hoping that somebody might have knowledge of a true bibliographical source. Considering the number of modifications H. L. Gold made to the stories printed in Galaxy there are some people who suggest he should have been credited as co-author for every story that appeared there.--swfritter 16:53, 8 Feb 2008 (CST)
I'm not sure where some of the collaboration credits come from, but after a major bad merge in my early days that got through ("who is this Lewis Padgett bloke?") I've avoided Kuttner mostly. This title is one of the few where I've let the pseudonym credits stand as it's fairly clearly written in the publication. The First volume looks debatable. A year later, I'm still not confident with edits in this area - but am happy to explain the credits on titles I own. BLongley 17:03, 8 Feb 2008 (CST)
I am unaware of One True Source for Kuttner/Moore attributions, but there are various tidbits of information that are strewn across various publications. For example, as I wrote in the Note field for Fury, "according to Moore's introduction in the Magnum/Prestige Books edition of the book, Moore wrote about one eighth of the text", even though the novel is usually attributed to Kuttner alone. Worse, publishers have been occasionally lax with attributions, e.g. the Gallegher stories, which are generally attributed to (and published as by) Kuttner alone, were published in 1983 as by Kuttner and Moore. Well, sort of. To quote our Notes field: "Spine has only Henry Kuttner's name. Title page adds "with C.L. Moore". Copyright page mentions originally published and copyright 1952 "Lewis Padgett (Henry Kuttner & C.L. Moore)"."
The bottom line is that the Kuttner/Moore bibliography is one big minefield and extreme caution should be exercised when trying to "correct" our data. Not that it's currently perfect, of course, but there is often a method to our madness :) Ahasuerus 17:20, 8 Feb 2008 (CST)
I am going to leave things as I find them for now, leaving only notes in the pubs indicating how the authorship is actually listed.--swfritter 17:52, 8 Feb 2008 (CST)
I'd agree with that - When doing sight-unseen cleanups I tend to leave notes on the possibly affected publications asking future verifiers to pay extra attention to specific elements. The main thing this system is missing is a way for the verifier to respond back "yes, I checked and "this" is exactly what's stated." A way to do this would be to start a wiki page for each cleanup project and the publication notes asking for extra inspection would link to this page. That would allow someone down the road to see that there was (or is) a project and what the results were. The goal is to create a list of citable sources for the spelling of an author name, story title, publication date, etc. Marc Kupper (talk) 14:33, 10 Feb 2008 (CST)
Good idea. This would also help spot cases where a title or attribution has been incorrectly changed. I have a feeling that in the case of Kuttner it might be a perpetual project as people keep adding bits and pieces from various sources - many of which will contradict each other. This may be a case where the real truth is NOT out there.--swfritter 17:53, 10 Feb 2008 (CST)

Magazine Reconciliation Project

I have created a page for the Magazine Reconciliation Project. The project page has links to a published Google Spreadsheet created with Google Docs. The spreadsheet covers American mags of the 1950's and provides a good statistical analysis of data that has been entered and tasks that have been accomplished along with documenting some titles that should probably be in the database. Google docs are shareable by multiple editors and can automatically be published to the internet whenever changes are made. As soon as I get them cleaned up a bit I also have spreadsheets for pre-1950's and the 1970's. I intend also to create data for non-American titles although I haven't decided yet whether to combine all the magazines on one page or split them up.--swfritter 19:08, 8 Feb 2008 (CST)

Quick links to Magazine Lists

Now that the magazine Wiki pages are filling up wonderfully with essays, series links, and other data the magazine list is not instantly accessible. Look at Amazing Stories. I have placed an {Issue List} link there that takes the user directly to the magazine list. Planned changes will probably make this method obsolete in the future but would like to run the idea by everyone.--swfritter 16:22, 10 Feb 2008 (CST)

Personal tools