ISFDB:Community Portal/Archive/Archive10

From ISFDB

Jump to: navigation, search

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: