ISFDB:Community Portal/Archive/Archive41


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 July - December 2016

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


LC Online Catalog is down 49 hours

LC Online Catalog is now down for 49 hours per notice atop "Search Results". --Pwendt|talk 20:59, 29 July 2016 (UTC)

Related to the earlier DNS attacks they were experiencing a few weeks ago?
Library of Congress wracked by DNS attack
Turkish hackers claim credit for Library of Congress attack
Albinoflea 19:24, 1 August 2016 (UTC)

Working languages

Someone wants to make a guess at what the language for these should be?

  • Elena Dudina - born in Russia, living in Spain, hailed as a Spanish artist. Will the language for her be Spanish? Or English (most of her relevant work seems to be in English language publications)
  • Don Landwehrle - 2 French works, 1 Dutch, 2 English in our DB and I cannot find where he does live actually (same named photographer shows up in VT,USA but I am not sure I can make the connection).

Anniemod 22:54, 31 July 2016 (UTC)

Help:Screen:AuthorData says:
  • For bilingual authors enter the more frequently used language
so I would go with "English" for now. We can always change it later if we discover that the bulk of these two artists' work is in other languages. Ahasuerus 01:03, 1 August 2016 (UTC)
That's the problem - these are not bilingual - because it is art, artists can work in a lot of languages and for most it is fairly easy to find the main language. Bilinguals I am untangling just fine - it is when there is 3 or more involved with almost the same frequency or power that get me :) English it is then. Anniemod 01:20, 1 August 2016 (UTC)
It sounds like we should change "bilingual" to "multilingual" then :) Ahasuerus 01:40, 1 August 2016 (UTC)
Or clarify the rules for artists compared to authors - art does not need translations in most cases so it make sense to go with "most published works in a language" as opposed to what language the artist uses themselves. Anniemod 02:13, 1 August 2016 (UTC)
Well, the rules already tell us to use "the more frequently used language", which can be applied to writers as well as to artists.
The most common problem with bilingual and multilingual authors is that we don't always know which language is used more frequently. For example, if we were to go by what we list for Sam J. Lundwall, we would have to say that his working language is English. However, the Swedish Wikipedia lists many more Swedish titles. Or consider the case of Alexander Lomm aka Václav Klička: we list only one Russian title, but the Czech Wikipedia lists a fair number of Czech works in addition to his Russian titles.
In the grand scheme of things, it's not terribly important. The "working language" field is not a statement about the author and his or her language preferences. It's just a flag that the software uses to determine which titles have their language displayed next to them on the author's Summary page. If we were to change Lundwall's "working language" from Swedish to English, the "[English]" tags would disappear and "[Swedish]" tags would appear. Ahasuerus 04:50, 1 August 2016 (UTC)
Those two examples are clear though -- we do not have all the works but a quick search finds more information - I did that for a few yesterday. My problem with these two above was that there wasn't enough information to be found :) They were just annoying me. Thanks! Anniemod 19:10, 1 August 2016 (UTC)

Through the Earth serial, novel, chapbook

Through the Earth by Clement Fezandié A1767 is

  1. early 1898 serial novelette, 4 parts perhaps 4000 words ea. -- 4 parts in the database as a serialization of the novel
  2. late 1898 novel, much expanded, perhaps 40,000 words; also issued 1971 -- both publications in the database as novels
  3. 1972 chapbook facsimile edition of the serial novelette; also issued 1980 -- both publs in the database as novels; the latter also in the database as a chapbook novelette

The latter chapbook novelette P580458 is new today, with chapbook and shortfiction titles both new today. Having added the [32]-page count and Notes to the 1972 "novel", I found this 1980 publ in library records and added it as the first step in separating the [32]-page from the 238-page novel.

Only this hour I noticed that duplication (two publications ISBN 0-89370-028-2 from Borgo Press). And also realized that the 4-part serial is in the database as a serialization of the novel, which it is not, so that to separate the [32]-page books from the 238-page books will not clean up the records.

It occurs to me that the novel, the 4 serial parts of the novelette, and the novelette might all be grouped together in a new series named Through the Earth. But I decline to take another step alone. --Pwendt|talk 01:53, 1 August 2016 (UTC)

P250763 and P580458 are the two records of the 1980 facsimile ed. of the novelette, both now essentially complete, as NOVEL and CHAPBOOK respectively.
Emphasis added above.
There must be numerous serializations of novels, so-called, where the novel is actually a substantial expansion of the serial story. If this co-ordination of all database records under a single NOVEL title is appropriate, T20771 (with Note extended moments ago), then P#580458 and its two title records should be deleted. --Pwendt|talk 17:56, 3 August 2016 (UTC)

Regularizing Publishers by ISBN

I've been having a discussion about regularizing publishers by the publisher code embedded in the ISBN. Although, this would only apply to books with ISBNs, this would go a long way to regularizing the publishers, and make it easier for editors. There is probably a master list of publishers, and this could be applied automatically in the database, assuming no programming difficulties.--Rkihara 17:54, 1 August 2016 (UTC)

"Regularizing Publishers" became policy .... when? As noted below, this is something that simply can't be automated. I did a little search of the above-mentioned 'code' page. One publisher only so the sample is limited but very enlightening: BAEN. So far there have been at least four separate prefixes. The first was 671, widely used for Timescape, then Pocket though the parent remained Simon & Schuster. In fact for that particular prefix BAEN doesn't even appear! In fact BAEN never appears on that page! There were a series of e-books released by BAEN of the Chandler Grimes novels, that, for whatever reason, carried the original DAW ISBNs. By the logic above all the current records with BAEN or the several sub-BAEN categories would ALL revert to either Simon & Schuster or DAW or whomever ..... even Locus didn't do that. At what point do we cease to be a bibliographic site and become simply an Index? Bibliography is in the details. It's what the vast majority of the software is set up to accommodate. I would like to see the merging of publishers [for the nonce, at least] unavailable. I'm really tired of having days/weeks/sometimes months of work nullified by "Regularization" for its own sake. What's on the title page is supposed to be the 'mantra', at least it was. --~ Bill, Bluesman 03:53, 16 August 2016 (UTC)
I concur with Bill, that what I thought at first was our policy ("What's on the title page is supposed to be the 'mantra'") rapidly proved false as I was called to order for entering data as it was. As for entering publishers strictly as on title page, that will need a lot of regulations and explanations. Take this book for instance, there is "Mayflower" on bottom of the title page but "Granada Publishing Limited" and "Published ... by Mayflower Books Ltd" on copyright page, which may cause some anguish in our contributors. I spare you this not uncommon occurrence in french publishing where the paperback publisher just photo-reproduced the whole tp here including the title page so we have "Laffont" on the title page of the Livre de Poche printing, which will means a mandatory regularization. Hauck 09:57, 16 August 2016 (UTC)
Unfortunately, ISBNs are not always publisher-specific. To quote
  • Question: If one company buys another, can the buyer use the ISBNs or do they have to get new ISBNs?
  • Answer: The company being bought can have all its ISBNs transferred to the new owner. If the company is a subdivision or subsidiary of the new company, it can maintain its own unique ISBN prefix. If the company being bought is being demoted to an imprint (a brand) with no legal standing of its own, then the new parent company can use the ISBNs at will for any title it publishes.
For this reason the process could not be completely automated. The best we could do would be to create a table of ISBN blocks for publishers and then add a cleanup report to identify ISBNs that do not match their "assigned" publisher. However, maintaining an ISBN/publisher list may be time-consuming since publishers come and go. Moreover, as we discussed a couple of months ago, some European publishers assign multiple ISBNs to the same book, which makes things even more complicated.
My current plan is to wrap up transliteration and translator support in the next few months and then revisit ISBNs. The first thing to do will be to change the current ISBN field to an infinitely repeating pair of fields. The first field will be a drop-down list of supported "identifier types" (catalog ID, ISBN, corrected ISBN, ASIN, LCCN, OCLC, BnF ID, etc) and the second field will be a "free text" value. This will help us process multi-ISBN books, self-published books without ISBNs (which Fixer can't handle at the moment), books which have a catalog ID as well as an ISBN (e.g. recent SFBC books) and so on.
With that out of the way, we can revisit publishers and see what we can do about them. The first thing that comes to mind is that we will want to allow entering multiple publishers -- think of all the SFBC/[some other publisher] collaborations -- but there will be a number of things that will need to be sorted out before we have a complete design. Ahasuerus 18:49, 1 August 2016 (UTC)

Dotted/Dotless I transliteration

I had been looking at the last few lagging non-transliterated Turkish titles - all of them have a non-standard element of one of the pairs: dotless (I ı)/dotted (İ i). The small dotless one and the capital dotted one are not in the standard Latin script. Any objections to just converting them to the Latin equivalent (İ -> I and ı -> i)? Technically the dotless pair is pronounced differently but if we go for pronunciation, we need to catch everything that starts with or contains I as well (as this is the dotless capital letter) and that gets into transcribing and not transliterating. Thoughts? Anniemod 00:46, 3 August 2016 (UTC)

I am not familiar with the Turkish alphabet, so let's try an example to make sure that we are on the same page. Kızıl Mars is a Turkish translation of Kim Stanley Robinson's Red Mars. The cover title uses all capital letters, but various Turkish sites as well as OCLC use dot-less "ı"s when referring to the book's title. If that's how it appears on the book's title page, then I think we want to keep the dot-less "ı"s in the title field and add "Kizil Mars" as a transliterated value. Unless I am misunderstanding something, of course. Ahasuerus 01:22, 3 August 2016 (UTC)
You got it. If you look at the title it looks like a regular "I" but because it is Turkish, it is the capital letter of ı and not the capital of i (That would have had a dot even on the capital). Pronunciation of the dotless one is like/close to the vowel in the English fuss for example - so kizil and kızıl are read differently... But we are transliterating, not transcribing so it makes sense to standartize all "i"'s (dots or not) to "i/I" in the transliteration fields similar to how we collapse the 3 Romanian a's into 'a' (although there at least the capital of one of the non-standard one does not look like a standard letter) Anniemod 01:39, 3 August 2016 (UTC)
Ah, I see, we are talking about using "Kizil" as a transliterated value. It all makes sense now, I was a bit confused. Ahasuerus 03:31, 3 August 2016 (UTC)
Was I overexplaining again? :) Anniemod 03:38, 3 August 2016 (UTC)
I am probably just tired. I am working on entering new pubs for September, the busiest month in the publishing industry. I hope to be done in another 3-4 days and then I can go back to enhancing the software. Ahasuerus 04:25, 3 August 2016 (UTC)
On a separate note - when you are done with the translation stuff you are working on, will non-Latin titles be searchable based on their original alphabets as well (if you have the proper keyboard, you can type it; if not - use transliteration - in a perfect search both will work) Anniemod 01:39, 3 August 2016 (UTC)
The regular search logic for titles/publishers/pub. series already checks the main field as well as transliterated values. Try a Fiction search on "Човекъ" vs. a Fiction search "chovek" and you'll see :) Ahasuerus 03:31, 3 August 2016 (UTC)
But not for authors yet, right? I know I tried a few of the Bulgarian guys (I admit I never checked the titles) Anniemod 03:38, 3 August 2016 (UTC)
Not yet. Once the "transliterated values" field has been added to author names, it will work the same way. Ahasuerus 04:25, 3 August 2016 (UTC)

Fixing a publisher name

Is there a way to fix a publisher name (wrong alphabet used) without updating all the publications? I can go through the publications of course but thought I would ask first :) Thanks Anniemod 23:39, 3 August 2016 (UTC)

Publisher names can be edited in "Edit Publisher", but, as per Help:Screen:EditPublisher, "Note that only moderators can edit this field once the publisher record has been created." If you post a request on the Moderator Noticeboard, I am sure it will be taken care of promptly. Ahasuerus 00:07, 4 August 2016 (UTC)
Thanks! Anniemod 00:15, 4 August 2016 (UTC)

An Amazon problem with the 'SF Masterworks' series

A heads-up for editors with verified pubs in the first Millennium/Gollancz SF Masterworks series: Amazon may be replacing correct cover images with incorrect images of the books' appearance in the 'SF Masterworks (II)' series. Have a look at John Sladek's Roderick at Random to see what I mean: 2001 pub date, now with a post-2010 cover – this was not the image in use when I verified this record, and I have seen several more incorrect Masterworks (II) covers on my verified records in the last few days.

I don't know how or why this is happening but we have around 200 verified pubs (including later impressions) that may need checking and I don't have the time or the inclination to do all of them myself! Editors will need to find another Amazon link to a correct Masterworks cover and use it instead, or failing that, find the cover elsewhere or scan our own books then upload to the server. Cheers. PeteYoung 18:23, 7 August 2016 (UTC)

Magazines outside of the major few

Just making sure I am not stepping on someone's toes... I had been looking at some of the non-major (for some value of major) magazines (Strange Horizons, Nature, Cosmos: The Science of Everything, the web edition of Clarkesworld, Beneath Ceaseless Skies) and they had not been updated since late 2015/early 2016. I do not mind getting them uptodate and keeping them uptodate (and yes, I know - only the story from Cosmos/Nature) but do not want to step on someone's toes that had already prepared the data. So what is the policy? Should I just start working on them or is there a place to notify someone? Anniemod 21:46, 9 August 2016 (UTC)

I don't think there is a policy re: entering magazine issues. Some of them are submitted by our hard-working robot, but mostly it's done by whoever happens to be interested in a particular magazine. Editor availability tends to fluctuate, which is why some magazines occasionally fall behind.
That said, if there are two or more editors working in the same area, they usually try to coordinate their efforts to make the data entry process smoother and more efficient. Ahasuerus 22:32, 9 August 2016 (UTC)
So what is the best way to find out who had been updating a magazine? I don't mind leaving them a message on their page and working out a coordination but I cannot find a way to figure out who had entered them (except when there is someone that had been verifying as well - for those it is an easy answer)? Post a message here and hope that someone will identify themselves does not seem very efficient :) Anniemod 22:41, 9 August 2016 (UTC)
Some magazines have matching Wiki pages whose Wiki "history" can provide clues. However, Wiki pages are optional and often very dated. We are also in the process of migrating them to the database proper.
Other than that, there is no easy way of telling who has touched a publication. At one point we attempted to capture "history" information, but it turned out to be a huge can of worms. Of course, if a pub has been verified, the verifier's name is available, but that's pretty much it. Ahasuerus 23:06, 9 August 2016 (UTC)
OK then - I guess I will just start adding the missing ones. If someone is ready to resume working on something, we will find a way to work out a schedule. Thanks! Anniemod 23:19, 9 August 2016 (UTC)
Look to see who, if anyone, has verified some recent issues. Those would be the natural candidates for who might have entered the data. Chavey 02:51, 11 August 2016 (UTC)
Yep, that seems to really be the only way :) Thanks Anniemod 22:14, 11 August 2016 (UTC)

"Uncredited" author links nixed

As per this discussion, "uncredited" authors are no longer linked to the "uncredited" Summary page. Ahasuerus 22:36, 9 August 2016 (UTC)

Changes to Author Editor

The ability to edit canonical names has been restricted to moderators. It has always been a potential problem area and it will become more of an issue as we get closer to the software change that will let us enter non-Latin author names. Simply changing "Reki Kawahara" to "川原 礫" or "Boris Strugatsky" to "Борис Стругацкий" will invalidate many translated titles, so we will need to tread carefully.

Once we modify the bulk of non-Latin author names, we will revisit this issue. Ahasuerus 22:35, 10 August 2016 (UTC)

Is there a way to do an automated edit which updates only those pubs in the same language as the author? Maybe create a CSV file with the current canonical name and the new one? Just trying to figure out a way to make the process faster, but avoid the issue you mentioned. It would then ignore any pubs in a different language than that of the author. ···日本穣 · 投稿 · Talk to Nihonjoe 00:31, 12 August 2016 (UTC)
I have been thinking about this issue. There is another, closely related, problem. Let's use Donko Naidenov, a Bulgarian author who has only one SHORTFICTION title on file, as an example. If a moderator changes the canonical name in the author record from "Donko Naidenov" to "Донко Найденов", there should be no problem. However, if a non-moderator, who can't change canonical names, changes the author in the title record, the software will delete the author record for "Donko Naidenov" and create a new one for "Донко Найденов". And that's a problem because we will lose all author-specific data, including Web pages, biographical data, working language, etc. I'm thinking of creating a cleanup report to help identify single-title authors whose working language is non-Latin, but I haven't quite gotten to that point in the process yet. Ahasuerus 01:06, 12 August 2016 (UTC)
When someone submits something like that, can the system flag the submission if there is only one entry under that author name so the moderator can check to make sure nothing gets lost ···日本穣 · 投稿 · Talk to Nihonjoe 01:20, 12 August 2016 (UTC)
Some moderator pages have "moderator warnings" for certain scenarios, e.g. new series, new publishers, new authors, etc. The current system isn't sophisticated enough to determine that approving a submission would result in author deletion, but it may be tweakable. It's one of the options that I am currently looking into. Ahasuerus 02:13, 12 August 2016 (UTC)
(or just change the author canonical name instead of accepting the submission)? ···日本穣 · 投稿 · Talk to Nihonjoe 01:20, 12 August 2016 (UTC)
That would be a much bigger change. Ahasuerus 02:13, 12 August 2016 (UTC)
I was trying to say that if the moderator saw this suggested flag, they could change the canonical name instead of accepting the submission if that's all that was being submitted. ···日本穣 · 投稿 · Talk to Nihonjoe 03:01, 12 August 2016 (UTC)
Oh, I see... Ahasuerus 03:06, 12 August 2016 (UTC)
Maybe have it tied to the cleanup report, so any author on that report gets the flag? ···日本穣 · 投稿 · Talk to Nihonjoe 01:20, 12 August 2016 (UTC)
An interesting thought. Thanks! Ahasuerus 02:13, 12 August 2016 (UTC)
I ended up developing a new cleanup report which finds authors with:
  • at least some author data; and
  • one title which happens to be in a non-Latin language; and
  • no other titles
There are 108 affected authors, so we should be able to clean them up easily once our software supports transliterated author names. I plan to deploy the new cleanup report later today once I finish fine-tuning its performance. Ahasuerus 21:12, 18 August 2016 (UTC)
Sounds good. ···日本穣 · 投稿 · Talk to Nihonjoe 17:50, 19 August 2016 (UTC)

"Interviews with" moved to a separate section

As per FR 252, the list of author interviews has been moved from the top of Author pages to a new section, "Interviews with This Author". The section appears immediately after the "Interviews by This Author" section. The sorting within the section is chronological or alphabetical depending on the type of the author page. Each line displays the following values: title, year, language (if different from the author's working language), the name(s) of the interviewer(s) and the name(s) of the co-interviewee(s), if any. At this time the section does not display variant titles. Ahasuerus 00:31, 12 August 2016 (UTC)

That looks good. Chavey 03:58, 12 August 2016 (UTC)

Too many Gerbers cluttering up the landscape.

I will be listing a book with the artist’s name of Gerber. However, the signature has some suspicious resemblances to the signature of Mark Gerber who is listed on the internet as being the man behind Gerber Studios.

Funny how none of these have any real portfolios on the internet to look at.

To make things even more confusing, listed on this site is Stephanie Gerber and Stefanie Gerber, both of whom have not only done solo work, but also collaborative artwork with Mark Gerber. I cannot even guess if Stephanie is Stefanie, or vice versa; if (just plain) Gerber is Mark, Stephanie, or Stefanie, or a combination of both, or all, or none. All the Gerbers were active in the paperback field at the same time, and for the same publishers. Can anybody clear any of this up? MLB 07:53, 12 August 2016 (UTC)

I now have a headache. Time for a nap. MLB 08:03, 12 August 2016 (UTC)
Here's the link to Gerber Studios. If any of the Gerber art has visible signatures, compare them to the one on the main page of that website. Based on a quick perusal of the covers, the art style matches Mark Gerber's on the Gerber Studio website. Therefore, I think "Gerber" is a pseudonym of Mark Gerber. I think Gerber Studio should be a pseudonym jointly for Mark and Stephanie. Based on comparing the art from Stephanie and Stefanie, I think they are the same person. Since Stephanie has more entries, I'd go with that one as the canonical, and Stefanie as the pseudonym. ···日本穣 · 投稿 · Talk to Nihonjoe 17:24, 12 August 2016 (UTC)
When somebody takes the credit, they must take the blame. While I have seen three different signatures on the various signed work that I've seen, all the work on the Gerber Studios website is copyrighted by Mark Gerber, so both credit and blame is his. The merging of Stephanie and Stefanie's records seems a no-brainer. When I got time I'll start on this. MLB 20:51, 12 August 2016 (UTC)
The Web page that Nihonjoe linked lists "mark [at]" as the point of contact. The copyright date is 2008, so I am not sure the e-mail address is current, but there should be no harm in e-mailing him and asking for clarification. Ahasuerus 21:52, 12 August 2016 (UTC)

Self-review fix

Self-reviews like this one have been changed to appear correctly on Author pages. Ahasuerus 19:27, 12 August 2016 (UTC)

Co-interviews fixed

Author pages have been adjusted to distinguish between co-interviewers and co-interviewees. Ahasuerus 21:44, 12 August 2016 (UTC)

Pseudonymous self-interviews fixed

The VTs of pseudonymous self-interviews (yes, we have a few!) no longer appear twice on Summary pages. Ahasuerus 22:13, 12 August 2016 (UTC)

Ongoing changes to Author and Series pages

Please note that I am in the process of streamlining and improving the software that drives our Author and Series pages. I expect that I will be deploying a number of patches over the next few days. I don't plan to post patch notes because all of the changes will be behind the scenes, but if you see anything unusual, please report your findings here. Ahasuerus 23:05, 14 August 2016 (UTC)

The last performance-related change was installed a few minutes ago. Long pages like Robert Silverberg and Star Wars Universe should take significantly less time to load and display. Next: implementing support for transliterated author names. Ahasuerus 16:41, 18 August 2016 (UTC)

New language reports

A new section, "Language Statistics", has been added to the ISFDB Statistics and Top Lists page. At this time it contains two reports, Authors by Language and Titles by Language. The latter revealed that we have 247 titles with an "undefined" language, presumably due to title merges. I'll need to take a closer look. Ahasuerus 18:55, 18 August 2016 (UTC)

So we really have 77194 authors without an assigned language? ···日本穣 · 投稿 · Talk to Nihonjoe 19:13, 18 August 2016 (UTC)
We do. However, considering the fact that approximately 80% of our authors had no assigned language a couple of years ago, 54% is not too bad! Ahasuerus 19:20, 18 August 2016 (UTC)
Progress! ···日本穣 · 投稿 · Talk to Nihonjoe 21:03, 18 August 2016 (UTC)

Analog volume numbering errors

I have been working on the Wikipedia article on Astounding/Analog, and have been using the ISFDB magazine grid to help me organize the data. While I was at it I noticed some errors in the volume numbering -- they're either incorrect, or it's an error in the magazine that isn't noted as such. For example, this issue is listed as volume 108, but the surrounding issues are 107; ditto for this issue; and this issue and the next are volume 107 when I suspect they should be 108. I think there were a couple more errors after that, but I stopped relying on the issue pages for volume numbering after that so didn't look too hard. I would fix these myself but my copies are in storage; if anyone else has access to copies it would be great to either verify or correct these. Mike Christie (talk) 22:14, 19 August 2016 (UTC)

I have fixed the years 1987 + 1988 according to my copies --Chris J 22:48, 19 August 2016 (UTC)
Thanks. Mike Christie (talk) 11:14, 20 August 2016 (UTC)

Patch and downtime -- Transliterated author names

The long-awaited "transliterated author names" patch will be installed at 6:30pm server (North American Eastern Daylight Time) time. There will be a brief downtime, around 5 minutes. The changes will be similar to the other "transliterated names/values" changes, so there should be few surprises. The affected areas include:

  • Author editing
  • Author merges (moderator only)
  • Regular Search, which will let you search using original and transliterated names
  • Advanced Search, which will have a "Transliterated Name" selector added
  • New cleanup reports

Please note that, for the time being, we will continue using transliterated names in the "Family Name" field. They are used to build our Author Directory, which is basically a 30 by 28 table. Going forward, there will be a new cleanup report identifying authors with non-Latin "Family Names" (and a display change from "Family Name" to "Sorting Name".)

As far as the process of updating author records goes, we have 2 types of cases.

First, we have authors all of whose records use the same form of the author's canonical name, e.g. Rumen Vuchkov or Kosta Sivov. They are easy to deal with -- all you need to do is move the current canonical name to the "Transliterated Author Name" field and enter the original non-Latin form of the name in the "Canonical Name" field. (As we discussed earlier, only moderators can change canonical names at this time, so non-moderators will need to post a request on the Moderator Noticeboard to perform this step. )

Second, there are cases where simply changing the author's canonical name may not be the right things to do. Typically, it has to do with translated titles. Let's use Reki Kawahara's Summary page as an example. At this time we have 17 Japanese titles and 17 English VTs on file. All of them use "Reki Kawahara" as the author's name. If we change the author record to use "川原礫" as the canonical name, it will affect all 34 titles, including the English VTs, which is clearly wrong. What we want to do instead is change the author name associated with the 17 Japanese titles while keeping the current author name associated with the 17 English VTs. We can make the requisite changes in one of two ways:

  1. Change the canonical name to "川原礫", then update all English titles to use "Reki Kawahara"
  2. Change the author of all English titles to something like "Reki Kawahara (Latin"), then change the main canonical author name to "川原礫", then change the canonical author name of "Reki Kawahara (Latin") to ""Reki Kawahara"

It doesn't really matter which method we use in this case, although the second method may be marginally safer since it forces you to change each English title's author first. (I may need to add another cleanup report to look for any discrepancies.)

The total number of affected titles is in the 4,200 range, but the required number of submissions will be much lower since a single canonical name change can take care of multiple titles. Ahasuerus 22:00, 21 August 2016 (UTC)

P.S. Serbian authors will require a separate cleanup report because the Serbian language uses two different alphabets. I will work on it tomorrow. Ahasuerus 22:07, 21 August 2016 (UTC)

The patch has been deployed. Ahasuerus 22:38, 21 August 2016 (UTC)
I've submitted one to test how it works. ···日本穣 · 投稿 · Talk to Nihonjoe 03:31, 22 August 2016 (UTC)
And here are the results. Ahasuerus 03:35, 22 August 2016 (UTC)
Now I'm confused. J. G. Ballard is an English author, so shouldn't his Canonical name remain in English?--Rkihara 06:58, 22 August 2016 (UTC)
The canonical name will remain "J. G. Ballard" while "J・G・バラード" will become a pseudonym just like "Jim G. Ballard" or "Dž. G. Balard".
Come to think of it, the way Edit Author works is somewhat misleading. It says "Canonical Name" even when you are editing a pseudonym. Ahasuerus 13:52, 22 August 2016 (UTC)
Does this mean (if I understand correctly) that we'll have to replace canonical names by their non-latin transcriptions, whenever they exist ? For instance, Jean Ray by "Жан Рэй", or Francis Carsac by "Франсис Карсак", etc., in Russian publications ? Linguist 14:31, 22 August 2016 (UTC)
That's right. The basic rule, which applies to all titles, publications, publishers, publication series -- and now authors -- is the same: We enter all data "as stated" in the book and then we create variants or add transliterated names/titles as needed. If a Serbian publication credits J. G. Ballard as "Dž. G. Balard", we enter "Dž. G. Balard" and set up a pseudonym/VT. If a Japanese publication credits him as "J・G・バラード", we use "J・G・バラード. If a Russian publication calls him "Дж. Г. Баллард", then that's what we'll use. And if another Russian publication spells out his first name and his middle name ("Джеймс Грэм Баллард"), then we'll use that version of his name. Once the raw data has been captured, we can use pseudonyms and VTs to link everything together. Ahasuerus 14:52, 22 August 2016 (UTC)
OK. Thanks for making this clear ! Linguist 15:09, 22 August 2016 (UTC).

(unindent) BTW, here is a trick that I find to be useful. We have a contributor, User:Zlan52, who has been entering Russian translations of Philip José Farmer's works. Many of them credit "Филип Фармер", i.e. "Philip Farmer", which is currently set up as pseudonym. It is tempting to go to Philip Farmer's Summary page and change the name to "Филип Фармер", but how do you know that PJF has never been credited as "Philip Farmer" in English (or other non-Cyrillic) publications? You can review PJF's Summary page, but it contains hundreds of titles and a mess of pseudonyms/VTs.

Well, it turns out that the "Philip Farmer" page has a "view all titles by this pseudonym" link. (For non-pseudonyms the same data is available under "Show All Titles" in the navigation bar.) If you click it, it will list all of the titles that are currently attributed to "Philip Farmer", including their language. There is only one English language title, so the easiest way to fix this author is to:

  • Change the author of the English title from "Philip Farmer" to "Philip Farmer (Latin)"
  • Change the canonical name of "Philip Farmer" to "Филип Фармер" and add "Philip Farmer" as a transliterated name
  • Change "Philip Farmer (Latin)" back to "Philip Farmer"
  • [Edit: Forgot a step] Set up "Philip Farmer" as a Philip José Farmer pseudonym

Four edits and you have cleaned up 35 titles! :-) Ahasuerus 15:13, 22 August 2016 (UTC)

Except you changed it so only moderators can edit canonical names. So most of us have no way to even submit the change. ···日本穣 · 投稿 · Talk to Nihonjoe 18:39, 22 August 2016 (UTC)
That's right. As I mentioned earlier, "only moderators can change canonical names at this time, so non-moderators will need to post a request on the Moderator Noticeboard to perform this step."
I am currently working on Cyrillic authors and I am finding that it's very easy to make a mistake and accidentally corrupt a bunch of records. For this reason I'd like to keep this restriction for the time being. Let's try the current process for now and see where it will get us in the next few days. So far I have cleaned up 200+ titles, so we are making steady progress. Ahasuerus 18:51, 22 August 2016 (UTC)
IIRC there was a moderator-only edit that non-moderators having a specified number of approved edits were allowed to make changes to. Or was that a proposal?--Rkihara 18:51, 22 August 2016 (UTC)
The software checks the number of Wiki edits when deciding whether to let you edit certain Wiki pages. In addition, the moderator queue highlights the submitter's name if s/he has edited the Wiki fewer than 20 times. I don't recall any other limits based on the number of edits. Ahasuerus 18:56, 22 August 2016 (UTC)

On a related topic, please note that I've shifted to a partial exploitation of the cleanup reports, I'm going to leave aside everything that seems (to me) pertaining to transliteration problems in order to avoid interfering with the work being done. Hauck 08:02, 23 August 2016 (UTC)

Thanks for the heads up! Ahasuerus 15:12, 23 August 2016 (UTC)

Transliterated author names - canonical names and legal names

Now that our software supports multiple transliterated canonical/legal names per author, some page headers look very busy. For example, consider 小松左京 (Sakyo Komatsu)'s Summary page. One third of the "real estate" at the top of the page is occupied by two forms of his transliterated canonical name and three forms of his transliterated legal name. Nihojoe mentioned this issue a few months ago when all we had was transliterated legal names, and it has become more serious with the addition of transliterated canonical names.

I'd like to propose changing the way these transliterated names/titles are displayed at the top of author, title, publication, publisher and publication series page. Instead of displaying one transliterated name per line, we could use the same mouse-over help bubbles that we use elsewhere. In certain cases it may make finding useful information a little harder, but on balance I think it will be an improvement. Would that work for everybody? Ahasuerus 19:09, 23 August 2016 (UTC)

I would favor the mouse-over like with transliterated titles. ···日本穣 · 投稿 · Talk to Nihonjoe 19:49, 23 August 2016 (UTC)
Related: Should we add a "Same as canonical name" (or whatever wording) to make it so that is used for the legal name if they are the same? It would save a lot of duplication in some cases. ···日本穣 · 投稿 · Talk to Nihonjoe 23:13, 23 August 2016 (UTC)
I am not sure it would work since canonical names and legal names use different formats. In addition, there is no way to get from one to the other without additional information. For example, how would one know that the canonical name "Patrick Nielsen Hayden" corresponds to "Nielsen Hayden, Patrick" and not "Hayden, Patrick Nielsen"? And then we have Hungarian and Japanese authors whose family name is listed first. Ahasuerus 01:29, 24 August 2016 (UTC)
I guess to make it work, we would have to have separate fields for FamilyName, FamilyName2 (for those with more than one), GivenName, GivenName2, GivenName3 (however many we wanted). It could get really complicated. ···日本穣 · 投稿 · Talk to Nihonjoe 02:18, 24 August 2016 (UTC)
That's right. I have seen the software used by certain organizations to parse names. It was not a sight for the faint of heart.
For now I have created FR 925 to address the original issue. Ahasuerus 22:02, 27 August 2016 (UTC)

Mouse-over bubbles for transliterated names and titles

At this time the transliterated names and titles that we enter serve 2 purposes. First, they facilitate searches, e.g. a search on "Gordevsky" finds Дмитрий Гордевский's page because his transliterated name is "Dmitry Gordevsky". Second, transliterated values appear in mouse-over bubbles when non-Latin names and titles are displayed, e.g. on this page.

This works well for Cyrillic, Japanese, Chinese and other non-Latin scripts and alphabets. On the other hand, Latin-based alphabets that are not limited to Latin-1 characters (Polish, Czech, Romanian, etc) are different. For example, consider the case of Karel Capek, a Czech writer. We have added transliterations to titles like "Továrna na absolutno" and "Povětroň" in order to facilitate searching. And that's fine, but do we really need to display mouse-over transliterations on his Summary page? Are our users likely to benefit from these mouse-over bubbles? Or are titles like "Továrna na absolutno" pretty much self-explanatory and the bubbles only serve to make the page look busy?

What I'd like to propose is a modification to the display logic for mouse-over bubbles. It can be updated to examine the transliterated values associated with each record and then decide whether the bubble should be displayed. If the record's transliterated names/titles contain only Latin and Latin-derived characters like "ě", "ň", "á", etc, then the bubble won't be displayed. How does it sound? Ahasuerus 20:28, 23 August 2016 (UTC)

At first glance, this seems like a good idea. I can't currently think of a case where it would be a bad thing. ···日本穣 · 投稿 · Talk to Nihonjoe 23:15, 23 August 2016 (UTC)
OK, FR 924 has been created. Ahasuerus 21:57, 27 August 2016 (UTC)

Post-patch tweaks

A minor patch has been installed. It includes the following changes:

  • The cleanup report "SERIALs without a Parent Title" is now available to non-moderators. At the moment it shows 32 Japanese SERIALs that will need to be varianted to NOVEL or SHORTFICTION titles.
  • All "X Titles with a Latin Author Name" have been moved to the "Authors: Languages" section.
  • The logic behind the cleanup report "Authors with Author Data and One Non-Latin Title" has been adjusted to include all authors with a single non-Latin title, not just a subset.

Ahasuerus 18:31, 23 August 2016 (UTC)

Default author languages?

At this time newly entered authors do not have a "working language" assigned automatically. You have to go back and enter one manually. Would it be OK to modify "New Publication" submissions to use the publication's language as its new authors' working language? Ahasuerus 21:07, 23 August 2016 (UTC)

Yes. Hauck 21:12, 23 August 2016 (UTC)
This would be awesome. ···日本穣 · 投稿 · Talk to Nihonjoe 23:13, 23 August 2016 (UTC)
It would be great! Right now there seem to be more than a handful of new authors per day that need that extra treatment. Stonecreek 06:40, 24 August 2016 (UTC)
Why not do all of the authors in that manner, not just the new ones? That would also take care of one of the cleanup reports.--Rkihara 16:33, 24 August 2016 (UTC)
Our current medium-term goal is to identify all non-English authors/titles and assign appropriate language values to them. Once that has been done, we will be able to run a database-wide script which will assign "English" to all authors and titles without a language.
We aren't quite there yet, but we are getting close. As I work on converting Romanized author names to their original values, I fix missing and incorrectly assigned languages, but it's getting to be increasingly rare. Once we are done with the current cleanup operation, I will develop additional database queries to look for anything unusual, which may lead to new cleanup reports.
One way or another, I expect that we will be in a position to declare victory in a few months at the most. And with the proposed software change to NewPubs, there won't be any way to create titles/authors without a language going forward. Ahasuerus 22:08, 24 August 2016 (UTC)
I think this would be a big help but I can see a few cases where it might work at cross purposes... certain world SF anthologies for instance (like this one).
Maybe there could be a flag or highlight on the moderator's approval screen to show when a new author is having their language set, sort of like now when it highlights new authors, pseudonyms, disambiguated authors, etc. so that in cases where something different than the automatic assignment is desired it's easier to see and respond to? Albinoflea 17:34, 24 August 2016 (UTC)
(after sleeping on it) Yes, I can see how a translated anthology can mess things up if it adds new authors or pseudonyms. Let me think about it... Ahasuerus 16:34, 25 August 2016 (UTC)
Maybe only do it for individual works? ···日本穣 · 投稿 · Talk to Nihonjoe 17:09, 25 August 2016 (UTC)
Well, if it's an established name of a certain author the existing language should be used. When the name has to be varianted the problem of possible different languages has to be addressed in or after this step, and if it's an entirely new author it may actually happen that he slips under the radar and has to be assigned his or her canonical language later (just as is the case now). Stonecreek 17:49, 25 August 2016 (UTC)

(unindent after additional brainstorming) The underlying problem with the originally proposed solution, i.e. using the publication's language to assign working languages to authors, appears to be the fact that publications can (and frequently do) contain titles in language A while their authors' working languages may be B, C or D.

For example, a Japanese edition of a foreign language anthology would contain Japanese titles. Each title would have one or more foreign (non-Japanese) language authors associated with it. Since we now enter authors "as stated", author names would be in Japanese, e.g. "クリフォード・D・シマック" for "Clifford D. Simak" or "イタロ・カルヴィーノ" for "Italo Calvino". If this is the first time that these forms of Simak's and Calvino's names have been submitted, they will have no author record on file. If the software were to use the language of the publication, it would incorrectly assign "Japanese" to these author records.

I think the most straightforward way to handle this mismatch would be to use a "two page solution". The first Web page would be the standard NewPub/EditPub/ClonePub/AddPub page. The second page, displayed after the submitting editor clicks "Submit", would only appear if the submission contained new authors. It would display a list of new authors along with a drop-down list of supported languages next to each author's name. Each drop-down list would default to the language of the publication, but let the editor choose a different language if necessary. In most cases the editor would simply click "Submit" again, which shouldn't take more than a second.

How does it sound? Ahasuerus 19:01, 25 August 2016 (UTC)

I like it but that will complicate a lot the addition of anthologies, magazines and other items with multiple authors from new contributors (and even the now straightforward novels...). The system is already hard to understand in places - an additional page when you are entering a new work (especially when you are working on a non-latin alphabet language and your authors come in as first entries will be really confusing. Just my 2 cents Anniemod 19:11, 25 August 2016 (UTC)
I like this idea, too. It may also be useful to have a page where transliterations can be entered for the content and authors (especially with anthologies and magazines) if they are not already in the system. Then the system could automatically create the variants. ···日本穣 · 投稿 · Talk to Nihonjoe 19:21, 25 August 2016 (UTC)
Transliteration or translations for the works? It sounds like you mean translations (variants use that, transliterations are kinda irrelevant for variants)? Anniemod 19:25, 25 August 2016 (UTC)
It sounds like Nihonjoe would like to use the proposed second page to enter transliterated values for new records -- author names, titles, perhaps publisher names and publication series names as well -- at submission time instead of waiting for the submission to be approved and then creating a bunch of additional submissions. Ahasuerus 19:31, 25 August 2016 (UTC)
Yeah but the last sentence talks about creating variants automatically - that is why I asked. I'd support being able to add transliterations directly and not waiting for an approval first :) Anniemod
Yes, that's what I meant. Be able to enter variants/pseudonyms for the entered content. I think it would remove a submission step, though the transliterations could be saved by the system as separate submissions, perhaps, to make it easier to handle them more precisely. This would be most useful on anthologies, collections, and magazines, where there is a big list of content titles. Being able to enter the "Variant of" title and "Pseudonym of" for the author for each of those content titles at the initial submission would save a lot of time. ···日本穣 · 投稿 · Talk to Nihonjoe 19:42, 25 August 2016 (UTC)
I wonder if this feature should be delayed until translator support is built in; it seems like once we're adding translators, there would be a sense that the the title included in a pub is not in its original language, so the logic of whether to apply the title-level language to the included works would be a lot more straightforward.
I also like the idea of being able to enter name variants when entering the pub itself (and I suspect it's been discussed before); the site Discogs does this with their data entry screen where you enter the canonical name and then can open a separate field (ANV: Artist Name Variant) to list the name as it appears on the item itself. Albinoflea 19:44, 25 August 2016 (UTC)
A couple of additional considerations:
  • We have an FR to add the ability to 'Edit' submissions. If we add a second page to New/Edit/Clone/Add Publication submissions, it will make it harder to implement this FR because it will be hard to load existing submissions into an edit form.
  • NewPub submissions, unlike many other submission types, are always valid. Edit Author, Make Pseudonym, Edit Title, Edit Pub, etc can become invalid if one or more of the records that they reference are no longer on file when the approving moderator looks at them. When this happens -- which it does fairly often, especially if submissions are approved out of order -- the only thing that the moderator can do is reject the submission. NewPub submissions don't have this problem, which is really nice because redoing long submissions is a pain. If we were to add the ability to link new Contents title to existing titles and new authors to existing authors, it would become possible for NewPub submissions to be invalid.
Ahasuerus 22:58, 25 August 2016 (UTC)

New title for Harry Potter Universe ?

As J. K. Rowling's Wizarding World is used at the author's web site and as logo on Harry Potter and the Cursed Child: Parts One & Two, it seems sensible to rename the series according to that guideline. Are there any objections (I'd prefer just Wizarding World) ? Stonecreek 07:14, 24 August 2016 (UTC)

I have the "Cursed Child" book, and I don't see anywhere on it that uses a "Wizarding World" logo or wording (at least not one that says "Wizarding World"). I don't see any such logo on her official site or on Pottermore, either. Pottermore mentions "the Wizarding World" on the very first page, but not anywhere else on a quick glance. I don't know that changing it would be beneficial. ···日本穣 · 投稿 · Talk to Nihonjoe 19:44, 24 August 2016 (UTC)
It might only show on the back of the British first publication by Little, Brown, but it is obviously the term used now by Rowling on her page (even if the logo doesn't show). As we try to mirror the terms as preferred by the creators, the benefit lies within that circumstance. Stonecreek 17:53, 25 August 2016 (UTC)
Ah, I see it now. So we currently have this setup:
  • Harry Potter Universe
  • Harry Potter
  • all the Harry Potter books
  • Harry Potter (Associational)
  • in-universe books, promotional books (post card, poster, etc.)
  • Harry Potter: Nonfiction
  • Reference works about the series and world
I have a few suggestions:
  1. Rename "Harry Potter Universe" to "Wizarding World" per Stonecreek
  2. Rename "Harry Potter (Associational)" to "Harry Potter (promotional)" and include post card books, poster books, and similar titles in it.
  3. Move the Beedle book, the Hogwarts Library books, and the Cursed Child book so they are directly under "Wizarding World" (since they are only mentioned in the main Harry Potter series, and are not directly part of that series)
  4. Rename "Harry Potter: Nonfiction" to "Harry Potter (nonfiction)" to standardize the series titles
Thoughts? ···日本穣 · 投稿 · Talk to Nihonjoe 19:37, 25 August 2016 (UTC)
Overall that's what I had in mind, too. I just thin that it'd be better to stay with '(associational)' as the series really didn't need promotion and the published titles were a gimmick for the readers / fans. Stonecreek 09:32, 26 August 2016 (UTC)
That's fine. For some reason, I don't like "associational". It sounds too vague to me. Maybe "memorabilia"? ···日本穣 · 投稿 · Talk to Nihonjoe 15:32, 27 August 2016 (UTC)

Deutscher Phantastik Preis and Deutscher Fantasy Preis?

Would it make sense to add Deutscher Phantastik Preis and Deutscher Fantasy Preis (Wikipedia article) to the list of supported awards? Ahasuerus 16:31, 25 August 2016 (UTC)

Sure, but I'd think that we'd need someone to add the awarded works first. Stonecreek 17:56, 25 August 2016 (UTC)
Deutscher Fantasy Preis seems to be limited to authors, so it wouldn't require adding new titles. On the other hand, Deutscher Phantastik Preis is a mix of author-based and title-based awards and we seem to be missing many/most of the titles. Calling for volunteers! L) Ahasuerus 18:33, 25 August 2016 (UTC)
While we're at it, the Kurd-Laßwitz-Preis could also be added as well. Records for most works in the categories Best German novel, Best foreign work and Best translation should already exists here. As for the SHORTFICTION categories... they might be missing mostly. I could help a bit adding missing titles. Jens Hitspacebar 19:55, 25 August 2016 (UTC)
Oh, a volunteer! Grab him before he has a chance to reconsider! :)
The new "award type" for Kurd-Laßwitz-Preis has been created. I am not sure I got everything right, but it should be easy enough to tweak as needed. Now we need to create award categories, which may be a but tricky since they have apparently changed over the years. If you want to post a preliminary list here, any moderator will be able to add them. TIA! Ahasuerus 22:20, 25 August 2016 (UTC)
Here's the list of award categories for the Kurd-Laßwitz-Preis and how some changed over the years. Please reword/adjust the wording if necessary:
  • Bester deutschsprachiger Roman (Best German-language SF novel)
    • Beste deutschsprachige Kurzgeschichte (Best German-language short story, until 1996)
    • Beste deutschsprachige Erzählung (Best German-language novelette/novella, until 1996)
    • Beste deutschsprachige Erzählung (Best German-language short fiction, since 1997)
    • Bester Übersetzer ins Deutsche (Best translator into German, awarded to a translator in general, until 1992)
    • Beste Übersetzung ins Deutsche (Best translation into German, awarded to a translator for the translation of a certain work, since 1993)
    • Bester Grafiker (Best artist, until 1992)
    • Beste Grafik einer deutschsprachigen Ausgabe (Best artwork of a German-language publication, since 1993)
  • Bestes ausländisches Werk (Best foreign-language work with first German-language publication, was called "Bester ausländischer SF-Roman" until 1996)
  • Sonderpreis für herausragende Leistungen im Bereich deutschsprachiger SF (Special price for extraordinary accomplishments in German-language SF)
  • Bestes deutschsprachiges Hörspiel (Best radio play)
  • Bester Film (Best movie, only from 1987 to 1993)
Hitspacebar 14:37, 26 August 2016 (UTC)
I have added a bunch of categories and set up a tentative display order. We may need to check the cut-off years in the Note fields to account for the differences between the years the awards were given and the years the eligible works appeared. Also, the award site suggests that the name of the "combined" post-1996 short fiction category is "Beste deutschsprachige Kurzgeschichte" rather than "Beste deutschsprachige Erzählung". Is that correct? Ahasuerus 16:01, 26 August 2016 (UTC)
Partly correct. I missed that when I compiled the above list: the category changed again in 2008, or was probably just renamed, from "Beste deutschsprachige Kurzgeschichte" to "Beste deutschsprachige Erzählung". Hitspacebar 16:23, 26 August 2016 (UTC)
Let me make sure that I got it right. In 1981-1996 there were two German language short fiction categories:
  • "BD Erzählung", which covered novelettes and novellas
  • "BD Kurzgeschichte", which covered short stories
In 1997 the two categories were combined. The combined category was known as "BD Kurzgeschichte" in 1997-2007. It has been known as "BD Erzählung" since 2008.
That's correct. Hitspacebar 14:07, 27 August 2016 (UTC)
The question then is how we want to structure these categories within the database. Because of the potential for confusion, I suspect that it may be best to have 4 separate categories with disambiguators added to their names, i.e.:
  • Beste deutschsprachige Erzählung (novelette/novella)
  • Beste deutschsprachige Kurzgeschichte (short story)
  • Beste deutschsprachige Kurzgeschichte (all short fiction)
  • Beste deutschsprachige Erzählung (all short fiction)
What do you think? Ahasuerus 17:07, 26 August 2016 (UTC)
That's one possibility. Another option is to combine the shortfiction category names from 1997 onwards, like the German Wikipedia article does:
  • Beste deutschsprachige Erzählung (novelette/novella only)
  • Beste deutschsprachige Kurzgeschichte (short story only)
  • Beste deutschsprachige Erzählung/Kurzgeschichte (all short fiction)
If we want to get as close as possible to the "real" name of the awards then your solution is better. But I think users rather look at the type of work an award is for than expecting an accurate reproductin of the award name. That would make my solution the better one.
BTW: you got bullet point three right in your suggestion. Though it says "Kurzgeschichte" it's indeed for all shortfiction types. I just checked, and the change of the shortfiction category name in 2008 is obviously just a renaming and it included novellas and novelettes since 1997 as well already, despite its misleading name. Example: Wolfgang Jeschke's Das Geschmeide is a novelette from 2004 and was awarded "Beste deutschsprachige Kurzgeschichte" in 2005.
Hitspacebar 14:07, 27 August 2016 (UTC)
Indeed, it was the main reason why I thought that using category names verbatim and adding disambiguators might help clarify the situation. If we were to combine categories, then a user looking at a single award page may not have the context needed to understand why we are using a category name that is different from the one that was used by the awarding organization. Ahasuerus 17:01, 27 August 2016 (UTC)
Yes, the missing context might be a problem for some users. Either way, both solutions have their pros and cons for users. Let's go for your 4 categories. That's also more consistent with the other ISFDB rules where all kinds of names and titles have to be used as stated. Hitspacebar 19:28, 27 August 2016 (UTC)
OK, I think I got it. Now I need to change the software to display the categories in some semblance of logical order. Ahasuerus 19:57, 27 August 2016 (UTC)
Thanks a lot! One minor thing: there's an error at the end of the note for Beste deutschsprachige Erzählung (all short fiction). It reads "..., which see for details, in 2008.". Hitspacebar 20:40, 27 August 2016 (UTC)
"Which see for details" is technically correct, but it's dictionary-speak :-) I have removed it to avoid confusion. I have also added a "Display Order" column to the award type page. I am not sure the current category display order for this award type is ideal, but it's easy enough to change. Ahasuerus 21:04, 27 August 2016 (UTC)
One more minor note error: the last sentence 'Superseded by "Bestes ausländisches Werk" in 1998.' must be removed from Bester deutschsprachiger Roman. Hitspacebar 23:02, 27 August 2016 (UTC)
Oops! Sorry, I must have gotten lost in a maze of categories, all seemingly alike... Ahasuerus 23:27, 27 August 2016 (UTC)

Language mismatch

While we are fixing languages all over the site, nvented words are causing some funny side effects :)

The Faded Sun and I mondi del sole morente contain the same 3 novels - in different languages but with the same names. Someone changed the language of the 3 to Italian and messed up all the English ones. So before I start doing something, can someone verify that my plan is correct (and help me figure out how to finish it - I seem to be missing the last step...):

  • Change the language back to English
  • Create a new Italian variants for each of the 3 novels.
  • In the Italian Publication, remove the 3 now English variants via Remove Titles From This Pub

How do I add the 3 Italian titles into the publication? Or am I missing a lot easier way to get this untangled? Thanks! Anniemod 23:36, 26 August 2016 (UTC)

If you create the new variants, you can get them into the pub via Import Content and then providing the title IDs. An alternative to that approach is to skip your second step (don't create variants), add the three titles as new content in the publication, then, make those new titles variants of the English titles. I would do it the latter way because it's easier to tell what is going on. In fact, if you add the new titles first, you can make the page "del" (or some such) on the titles you want removed, and the moderator handling the submission might notice and kindly do the removals for you.  :-) --MartyD 00:28, 27 August 2016 (UTC)
Import content... somehow managed to forget that I can do that - I think I had used it twice only. I was planning to write a note to the handling moderator on what I am doing and why when I am submitting. I think both ways are the same number of steps (3 language changes, 3 variants are always needed, 3 removals from the publication, import vs editing the publication). I think it comes down to where one wants to start untangling from :) Thanks! I will go try to fix this. Anniemod 01:03, 27 August 2016 (UTC)

Clarissa Ross

According to us, we have two authors named "Clarissa Ross", this one and this one. According to Wikipedia, these two are the same pseudonym. Does anyone know whether we are wrong or Wikipedia is wrong? Chavey 05:33, 2 September 2016 (UTC)

I don't know, but the wikipedia article lists the last novel as by "Clarissa Ross" for the year 1982, which speaks for different authors. Stonecreek 05:52, 2 September 2016 (UTC)

Family Names in Japanese Author Pages

I've been transliterating some Japanese author names and I've noticed that the programming puts the entire name in kanji into the field for family name. There's also no ability to transliterate the family name. Should I leave this alone, enter family name only in kanji, or enter the transliterated family name? I'm hesitant to make changes in this field, since pending possible programming changes anything I do here could make more cleanup work later.--Rkihara 18:09, 2 September 2016 (UTC)

I have updated the relevant Help template to cover this issue. How does it look?
Also, there is a Feature Request to "Create a cleanup report to look for authors with non-Latin Family Names". I hope to get to it in a few days once I finish the current Fixer run. Ahasuerus 19:21, 2 September 2016 (UTC)
The template looks good, and I'll go back and enter the Japanese family names for the authors I edited. Under the third bullet, the caution about Japanese names entered as canonical in Latin text may cause some problems, since it assumes that "all" Latin forms of the Japanese name has been entered in Japanese order.--Rkihara 19:46, 2 September 2016 (UTC)
I am not 100% sure I am reading your comment correctly, but I have eliminated "canonical name" and made a number of other wording and formatting changes. Does it address the issue that you had in mind? Ahasuerus 21:32, 2 September 2016 (UTC)
Yes, the changes took care of my concerns. I realized after posting that my comments were a little confusing.--Rkihara 21:50, 2 September 2016 (UTC)
Great! Ahasuerus 22:55, 2 September 2016 (UTC)
Maybe the software can be altered to use the last part of the transliterated name?--Rkihara 19:46, 2 September 2016 (UTC)
At this time there is no way to enter transliterated author names when entering new authors in Edit Title, NewPub, EditPub, etc. We would need some kind of mechanism -- a pop-up window or a second Web page -- that would checked the entered author names against what's already in the database, identify any new ones and ask the editor to provide additional data about the author. It's doable, but quite time-consuming. It also raises the same issues that popped up when we discussed entering author languages in NewPub. Ahasuerus 21:20, 2 September 2016 (UTC)
We probably need a cleanup script for this too.--Rkihara 19:56, 2 September 2016 (UTC)
Would it be in addition to the one that I mentioned above? What would it look for? Ahasuerus 21:20, 2 September 2016 (UTC)
Sorry, speed reading got me again.--Rkihara 21:44, 2 September 2016 (UTC)
P.S. We also have another Feature Request, "Change Family name to Sorting Name", but it's just a change to the field label. Ahasuerus 21:35, 2 September 2016 (UTC)

The Oxford Book of Adventure Stories

I would have thought that The Oxford Book of Adventure Stories was a non-genre anthology. But it turns out that 13 of the 23 stories in it are already in this database (and one that isn't, is a ghost story). Do you think I should just index the whole thing? Adventure fiction is kinda-sorta related to spec fic, especially since authors often took their exotic locales as reasons to include the supernatural (when you go to the farthest places, you will find strange things). But still, it's a stretch. Some of the stories that are already in this DB have essentially nothing speculative about them, and I'm reluctant to index them in this book. --Vasha77 21:03, 3 September 2016 (UTC)

Nicholas Meyer's Holmes Pastiches

What is the guideline for adding Holmes pastiches? While this pub by Caleb Carr should be deemed speculative in nature (ghosts) and should be marked as in genre, Nicholas Meyers's pastiches do not appear to have any speculative content, specifically this one and this one. There is another Holmes pastiche written by Meyers. It seems to me than rather than adding the third book, it would make more sense to remove the other two. (I have not read those two novels. I'm simply working from the Wikipedia entries. and 21:19, 3 September 2016 (UTC)

Done. Hauck 07:18, 4 September 2016 (UTC)

Phonetic Entry of Japanese Authors in HIragana

I'm working my way through the cleanup script, "Japanese Titles with Latin Author Names," and I cannot find the proper rendering of some of the author's names in Kanji. The only way to clear the script is to enter the names phonetically in Hiragana. For example, あずみゆきのぶ for Yukinobu Azumi, instead of the Kanji. I think that should be alright, as a native speaker would probably search phonetically first, unless he knew for certain which Kanji to use. I'll proceed in this manner if no one objects.--Rkihara 03:54, 4 September 2016 (UTC)

My knowledge of Japanese is close to non-existent, but the idea is to enter author names as they appear in publications. As far as I know, authors' name as they appear in Japanese books and magazines use kanji. My concern is that if we use phonetic (hiragana) forms of their names, they will disappear from the cleanup reports, which will make it difficult to find and correct them in the future. Ahasuerus 04:44, 4 September 2016 (UTC)
Okay, I'll leave them as is.--Rkihara 05:26, 4 September 2016 (UTC)
If you make me a list of them, I will find the kanji for them. ···日本穣 · 投稿 · Talk to Nihonjoe 16:36, 4 September 2016 (UTC)
I'll do that. It may be a while, since I have to fix one of lapses.--Rkihara 19:15, 4 September 2016 (UTC)

What about entering the Hiragana (with an interpunct) as the transliterated name? I've been thinking of doing that for all Japanese authors anyway.--Rkihara 23:48, 9 September 2016 (UTC)

Japanese Authors' English Translations

When Japanese author names were converted from transliterations to the Japanese language, for some the canonical name was simply changed with no regard for translations. This means the English translations (and perhaps other languages), including primary verified publications, now have the wrong credit. These were published as by the transliterated name and should have remained that way. Impacted authors include 菊地秀行 (Hideyuki Kikuchi), 天野 喜孝 (Yoshitaka Amano), 瀬名 秀明 (Hideaki Sena), 安部 公房 (Kobo Abe), 村上 春樹 (Haruki Murakami) and probably others (I stopped looking, but we have quite a few others that were converted so each will need to be checked). There is a significant amount of clean-up required here. -- JLaTondre (talk) 13:16, 4 September 2016 (UTC)

Looking through the edit history, for some of these, the transliterated name was adjusted at the same time the canonical name was changed. This means the current transliterated name cannot be simply used to update the translations. Each one will need to double checked. -- JLaTondre (talk) 13:37, 4 September 2016 (UTC)
Thanks for identifying the issue! It was discussed as a potential problem when the patch that added support for transliterated names was installed, but I guess some people missed it. I will scan the submission history and identify the affected author names. We have the original names in the backup files, so it will be relatively easy to undo the damage. Ahasuerus 13:49, 4 September 2016 (UTC)
Actually, it turns out that there is an easier way to do it. As JLaTondre pointed out, in some cases the transliterated name was adjusted at the same time the canonical name was changed, so we can't just move the transliterated name to the canonical name field. However, the "Subject" field of the submission, which is displayed on the Recent Edits page, contains the author's canonical name as it existed at the time the submission was created. So all we have to do is check "Recent Edits" for AuthorUpdate submissions and use the values in the "Subject" column to restore the data. Ahasuerus 14:06, 4 September 2016 (UTC)
However, digging back through recent edits is tedious (to the point of being painful and easy to miss when needing to go back days). Hence the feature request to be able to search it. ;-) -- JLaTondre (talk) 14:18, 4 September 2016 (UTC)
True. For now, it may be best to create a list of affected authors and post it here so that multiple eyeballs could double check than nothing was missed. Ahasuerus 15:10, 4 September 2016 (UTC)
Since I caused the transliteration problem, I'll fix it. It's easier for me to track them down, since I can work through "My Recent Edits," rather than have others try to find these in "Recent Edits." I'm making the changes in the chronological order of my edits. Some of the authors cited by JLaTondre have already been fixed.--Rkihara 18:33, 4 September 2016 (UTC)
Great, thanks! Ahasuerus 18:48, 4 September 2016 (UTC)
Yes, I've corrected (via our cleanup reports) some such cases where books in English by a Russian author had their credits changed (I suppose) to the name in Cyrilic, which was evidently wrong. Hauck 13:45, 4 September 2016 (UTC)

All done, other than varianting titles in English where the language is undefined.--Rkihara 23:26, 4 September 2016 (UTC)

It looks like publication authors (as opposed to title authors) are still outstanding as well. For example, the six English pubs under Inter Ice Age 4 credit "安部 公房" instead of "Kōbō Abe". They will show up on the cleanup report that finds pub/title author mismatches, so it's not a big deal. Ahasuerus 00:41, 5 September 2016 (UTC)
I'll go back to the start and fix those too.--Rkihara 02:07, 5 September 2016 (UTC)

There may also be problems with REVIEWS, if the review credits the tranliterated name and (IMHO) is the sole to do this, it will generate a line in the "Authors That Exist Only Due to Reviews". I've corrected one here by setting the author's name to the canonical. Hauck 07:35, 5 September 2016 (UTC)

Spacing in Japanese names

In Japanese, names do not have spaces in them. I have seen many Japanese names here that follow that pattern when using the kanji version of the name. However, I have also seen some with spaces added (細田 守, for example). I think we need to establish a standard for this so all of them are one way or the other ("細田 守" or "細田守"). Thoughts? ···日本穣 · 投稿 · Talk to Nihonjoe 21:05, 5 September 2016 (UTC)

It's not customary, but the Ja-dot-Wikipedia shows all names that way. For example, Hayao Miyazaki. When I cut and paste a name from the Wiki, I leave that space. If the name is already entered into the ISFDB with/without a space, I leave it the way I found it. I've also noticed that some legal names have been entered with a comma and space to indicate that the family name comes first.
Well, it's not all of them, but a large number. I don't think we should use commas in the kanji name, however. The space is good enough for that, and we can use the comma in the Romanized version (the transliteration) of the legal name. ···日本穣 · 投稿 · Talk to Nihonjoe 16:52, 6 September 2016 (UTC)
I agree that we should standardize. I would vote for a space, since it removes the ambiguity in parsing the kanji, but for someone who can't read Japanese it makes no difference.--Rkihara 08:25, 6 September 2016 (UTC)
I can read it either way, so it makes no difference to me. I agree that standardizing is good, however, to make a consistent experience for those using the site. ···日本穣 · 投稿 · Talk to Nihonjoe 16:52, 6 September 2016 (UTC)
There's a related problem with the transliteration of English authors into Japanese. The names are often separated with an interpunct for disambiguation. If we drop the interpunct that makes it consistent with spacing the kanji in names. It might cause a search problem for native speakers of Japanese. There's a fuzzy search feature request that may take care of this if it is implemented.--Rkihara 18:19, 7 September 2016 (UTC)
Yeah, I would leave the interpunct if that's how it was on the credit since that is common in Japanese (even in some Japanese names that don't use kanji). Also, regarding spaces in kanji names, the article titles on the Japanese Wikipedia never have the spaces. It is only in the body of the article that the spaces sometimes appear. In writing credits, there are rarely spaces unless there are 3 or fewer kanji in the name. ···日本穣 · 投稿 · Talk to Nihonjoe 22:42, 7 September 2016 (UTC)
Names look more natural to me without the space, so I'll enter names without the spaces.--Rkihara 23:50, 8 September 2016 (UTC)
I agree. Should we document this somewhere so people can easily find it? ···日本穣 · 投稿 · Talk to Nihonjoe 20:04, 9 September 2016 (UTC)
Probably under Help:How to enter foreign language editions, which also needs a section on transliteration. We could write it up and submit the change here for general approval.--Rkihara 21:42, 9 September 2016 (UTC)
That works. Want to take a first stab at it? ···日本穣 · 投稿 · Talk to Nihonjoe 22:23, 9 September 2016 (UTC)
I think it would be better for you to write it up, since you have a better sense of the issues involved.--Rkihara 23:52, 9 September 2016 (UTC)

Sylvia Townsend Warner autobiography

Would the autobiography of noted speculative writer Sylvia Townsend Warner (Kingdoms of Elfin etc.) be of sufficient interest to include in the DB? It is called Scenes of Childhood (but covers her adulthood alo). She says almost nothing about writing in it, though. --Vasha77 04:38, 9 September 2016 (UTC)

I would be in favor of including it. Autobiographies of SF authors are always good to have on file. Ahasuerus 16:11, 11 September 2016 (UTC)

Standardizing automated record deletion

Earlier today we had a discussion of automated record deletion on the Moderator Noticeboard. Based on the discussion, I have created the following (very tentative) FR:

At this time the rules for record deletion are as follows:

  1. Publications can only be deleted manually. There is a cleanup report that finds publications without titles.
  2. Titles, including titles with no publications, can only be deleted manually. There is a cleanup report that finds titles (except ESSAYs, POEMs and SHORTFICTION) with no publications.
  3. Series, including empty series, can only be deleted manually. There is a cleanup report that finds empty series. We have an FR to "Enable series auto-deletion".
  4. Authors are deleted automatically when the last title associated with the author has been deleted.
  5. Publishers are deleted automatically when the last publication associated with the publisher has been deleted.
  6. Publication series are deleted automatically when the last publication associated with the publication series has been deleted.
  7. Awards, award categories and award types can only be deleted manually. Empty award types and award categories are flagged by cleanup reports.

It would be desirable to standardize the rules for automatic record deletion. The following record types would be affected: authors, series, publishers, publication series. The updated algorithm would check if the about-to-be-deleted record has additional information associated with it: notes, Web pages, biographical data (for authors.) If it does not, then the record would be deleted. If it does, then the record would not be deleted; instead it would appear on a record type-specific cleanup report. The new cleanup reports would be periodically reviewed and any relevant data would be migrated to existing records prior to manual deletion.

The proposed algorithm would require three new options:

  1. Delete Author
  2. Delete Publisher
  3. Delete Publication Series

It would also require the creation of three cleanup reports for authors, publishers and publication series.

Is this the direction that we want to take? Ahasuerus 16:09, 11 September 2016 (UTC)

I'm not against the move (although I'm quite comfortable with the automatic deletion process) but it will add other reports to handle. I'm not sure that our manpower is sufficient as the interest in cleanup reports is perhaps not that huge. On a related topic, I'll be very interested by an extension of the cleanup report that find publess titles to other categories (perhaps gradually starting with one of the three types). I know that some of our gafiated colleagues used to create such publess records in special circumstances (e.g. when asked by an author) and think that all publess titles should be eliminated. Hauck 16:21, 11 September 2016 (UTC)
I agree that expanding the cleanup report that finds publess titles would be useful. In many cases publess titles are merely debris left behind after our data was reorganized. They are just sitting there waiting to be deleted. In certain other cases I believe the preferred course of action would be to enter the (often non-genre) pub in which the pub-less title appeared.
However, there is yet another type of cases that we need to consider before we make any changes. Let's use K. J. Hannah Greenberg's story "Betting on the Outfield" as an example. It first appeared online in Bewildering Stories #362, November 23, 2009, as by "Chennie Greenberg". In 2012 it was reprinted in Don't Pet the Sweaty Things as by "K. J. Hannah Greenberg". We have two title records on file for this story, one for "Chennie Greenberg" and one for "K. J. Hannah Greenberg". Bewildering Stories is a non-genre webzine that, as far as I know, is not eligible for inclusion under our rules. If we delete the 2009 title associated with "Chennie Greenberg", the author name will disappear from the database. I don't think that would be desirable, so I would be in favor of "ignoring" this title record. Does this sound about right? Ahasuerus 17:27, 11 September 2016 (UTC)
IMHO such data about publications outside our scope should go in the note field, but it will have to be decided on a case by case basis. Hauck 17:58, 11 September 2016 (UTC)
My larger concern is that deleting data that was painstakingly researched and added to the database by other people can cause friction. When editors discover that their work has been destroyed, they are less likely to contribute and may drop out altogether. It's happened to me as well and it was very unpleasant. Ahasuerus 19:50, 11 September 2016 (UTC)

At the moment, by the way, we have Author "Channie", with one variant Title so credited (two child records in the database if i understand correctly), and "Chennie" in the Note field of the canonical Title (one parent record iiuc). The linked webpage gives "Channie". Feel free to delete this notice when appropriate.

Transferring information from child Title records with no publications to their parent Title records, I suppose, is one thing moderators now do manually before deleting child Titles without publications.
I would have guessed that Author, Publisher, and Publication Series records are retained when they contain "any" information [including part of an URL, if that is what generates a bluelink in contrast to a redlink; probably not]. Offhand I think they should be retained, except when "any" information is nothing but Language in an Author record, say. --Pwendt|talk 21:31, 11 September 2016 (UTC)
[edited immediately above]. On further thought, I would prefer such a record be retained for manual deletion, even when empty in the relevant sense, whenever it is the target of a link from another record in the database (not to say a page in the wiki). Perhaps the database can be searched for the full URL of the record --in view such as "Summary Bibliography" for an Author record-- and held for manual treatment rather than automatic deletion if there is a search hit.
In other words, very much yes, make automatic deletion more conservative in cases 5-6-7. --Pwendt|talk 23:13, 11 September 2016 (UTC)

Publisher names

Are there any favored stylings for publisher names in the database?

In particular, suppose the publisher name (or imprint or division name that we favor) does appear on the title page of a book, in a single line and in ASCII characters only,
is there any emerging guideline or consensus concerning alternative representations in the database such as these?

Company -- Co. -- Co
, Ltd. -- , Ltd -- Ltd. -- Ltd
& -- and
A, B, and C -- A, B and C [conjunction of three]
A, B, and Co. -- A, B and Co.
, Publishers -- Publishers [dropping the only comma]
sheer truncation of "Press", "Publishers", "Publications", "Publishing"
-- and The [dropping or including the initial article]

--Pwendt|talk 23:44, 11 September 2016 (UTC)

I don't think we have a standard at this time. Typically we just use the publisher's preferred name. However, we have to be careful because sometimes minor variations in publisher names can indicate a transition from an imprint to a separate publisher, a merger, and so on. There have been discussions re: beefing up publisher support, e.g. adding support for multiple publishers, but nothing definite has been decided. Ahasuerus 02:41, 13 September 2016 (UTC)

Reversing a canonical Author name (or work Title)

Is there any shortcut available for reversing the current choice of one Author name (or one work Title) as the canonical one?

There is a feature request to Enable mass variant title creation for Authors, but it would require a fair amount of work. There are various permutations that need to be accounted for. Ahasuerus 02:37, 13 September 2016 (UTC)
Thanks. --Pwendt|talk 16:55, 13 September 2016 (UTC)

If no shortcut, presumably available by notice to moderators, I suppose I would reverse the choice of one Author name as follows --in the simple case where there is only one pseudonym and there are no variant titles, only pseudonymous works--

  1. for the pseudonym Author, remove pseudonym ("Remove")
  2. " ", copy all biographical data and perhaps webpages from the canonical Author
  3. for each pseudonymous work Title, remove that relationship (enter "0" as parent id)
  4. " ", copy some info from the canonical Title record, if applicable
  5. for the former canonical Author, make pseudonym
  6. " ", delete all biographical data and perhaps webpages
  7. for each former canonical Title, make it a pseudonymous work if there are publications under the former canonical Author name, else delete it (should be step 8)
  8. " ", delete some or all info copied in step 4, if applicable (should be step 7)

(9. Look at linked wikipages, if any. Create new ones where applicable, with cross-references at least. I know that no shortcut covers this step.)

Steps 4 and 8 (4 and 7, as revised) alone will frequently be empty.

In steps 2 and 6, I mean move perhaps all information except

Transliterated Name, Family Name (for alphabetization only, i suppose), Working Language

--"perhaps" because some Webpages should be linked from both records.

I would paste, say, "reversing A. B. Frost as pseudonym for Arthur Burdette Frost" into every Note to Moderator.

--Pwendt|talk 20:07, 12 September 2016 (UTC)

You have the right idea, but keep in mind that title tags and series names/numbers are automatically moved to the parent title when you create a variant. There is another feature request to auto-move synopsis information. Ahasuerus 02:37, 13 September 2016 (UTC)
Amend steps 7-8: delete some former parent Titles, all but those with publications under the former canonical Author name. --Pwendt|talk 16:55, 13 September 2016 (UTC)

Publishing novels in parts

I had been looking at a few of those lately so thought I should ask. For example Hyperion. 4 of the variants are half-novels (2 of the Italian and 2 of the French ones) due to publisher splitting editions. Should the half-novels really be a variant of the whole novel? Anniemod 23:07, 12 September 2016 (UTC)

Unfortunately, there is not much else that we can do at the moment. Perhaps if we were to add a "disambiguator" field to document abridged, split, revised, expanded, etc versions of titles, it would help clarify things. Ahasuerus 02:31, 13 September 2016 (UTC)
This choice is the result of one of our rare consensus. Hauck 13:15, 13 September 2016 (UTC)
OK then :) Thanks! It still does not look logical but if that is what it is, then it is what it is. I would have thought that the serialization feature can be reused in that case but apparently not. Anniemod 21:18, 13 September 2016 (UTC)
We use SERIALs when a book is serialized in multiple CHAPBOOKs, e.g. see Donna Grant's Summary page. However, when a 900-page novel is split into two 450-page volumes it doesn't count as a serialization. Ahasuerus 22:09, 13 September 2016 (UTC)
I understand the used convention and when Serial is used. I was just thinking aloud (and using the word reused) - it does not really make a difference if you split into 12 80 pages ones (so chapbooks) and a legitimate serial or 2 480 pages ones - it is still a split novel. As I said - as that is the convention, then that is the convention. It does make the page look a bit incorrect but that is what we have. Thanks. Anniemod 22:26, 13 September 2016 (UTC)

New "recognized" sites

The following sites are now "recognized" and will be credited when we link to them:

  • Ansible
  • Aozora Bunko
  • Encyclopedia of Fantasy
  • Encyklopedia Fantastyki
  • Fancyclopedia 3
  • Library of Congress
  • NASA
  • National Library of Australia
  • National Library of France
  • SF Site
  • WorldCat

Ahasuerus 14:33, 13 September 2016 (UTC)

Evidently the online edition of Clute/Grant, The Encyclopedia of Fantasy (1897), is now served primarily as from, to which redirects. Probably it would be best to handle both. Anyway 5 leftover Author records show linkname "" rather than "Encyclopedia of Fantasy" (located via Advanced Search: Author: Webpage contains). --Pwendt|talk 18:28, 13 September 2016 (UTC)
Good catch, updated. Ahasuerus 19:27, 13 September 2016 (UTC)

Advanced Publication Search and ISBNs

The way Advanced Publication Search handles ISBNs has been adjusted. In the past only exact ISBNs and catalog IDs could be used. As of 20 minutes ago, you can also use "contains", "starts with" and "ends with".

However, keep in mind that ISBNs are tricky. Valid ISBNs are stored without hyphens and then the display logic adds hyphens as needed. When asked to perform an exact ISBN search on a valid ISBN, the software actually checks for the presence of 4 different ISBNs that match the entered search value:

  • ISBN-10 with hyphens
  • ISBN-10 without hyphens
  • ISBN-13 with hyphens
  • ISBN-13 without hyphens

This explains why the software finds the right ISBN even when the user enters hyphens.

On the other hand, when the software performs a "contains" or a "starts/ends with" search, the user typically enters a partial ISBN, which can't be used to build alternative ISBNs. For this reason partial ISBNs need to be specified exactly as they exist in the database, i.e. without the hyphens. Hopefully this make at least some kind of sense! :) Ahasuerus 22:08, 13 September 2016 (UTC)

A new cleanup report - Family Names

A new cleanup report has been deployed. It identifies authors with non-Latin characters in the "Family Name" field. Keep in mind that what we currently call "family names" are really "sorting names". They are used to build the Author Directory, so they need to use Latin-1 characters, e.g. "Sliwinska" for "Małgorzata Śliwińska", "Farmer" for "Фармер", etc. The data will become available tomorrow morning. Ahasuerus 02:43, 14 September 2016 (UTC)

A lot of Japanese family names are flagged by the report because of the macron in the transliterated family name. Should we strike the macron? If so, we should standardize on how we transliterate the family name. The macron represents a long vowel, so Satõ, could be Sato, Satoo, or Satou; or Ryūnosuke, Ryunosuke, Ryuunosuke. The most common transliteration (and less accurate) uses the vowel without the macron, so Tōkyō becomes Tokyo.--Rkihara 17:04, 14 September 2016 (UTC)
The only thing that the "Family name" value affects is the author's position in the Author Directory and how it is sorted in search results. If the first or the second character in a "Family Name" value is non-Latin, the author's name will not appear in the Author Directory. If the third (or subsequent) letter is non-Latin, it will make the author's name sort in an odd way, e.g. "Satõ" may appear before "Sata". For this reason, getting the first two letters of "Family name" values right is more important than getting subsequent letters right. Ahasuerus 17:22, 14 September 2016 (UTC)
We need to strike the macron then, since long vowels occur in the first and second letters too, like Ōhara, and Tōemon. Along those lines, should all diacritics be removed? Even if they are available in the Latin character set?--Rkihara 17:52, 14 September 2016 (UTC)
Technically, non-English "Latin-1" characters like "é" are OK to have in this field because the database recognizes and sorts them correctly. For example, "Claire Dé" is sorted as if her last name were "De" even though her "Family name" is currently defined as "Dé".
The rule of thumb is that if a character is used by a West European language, it's most likely part of Latin-1. However, it's not guaranteed, e.g. "œ" and "ij" are not present in Latin-1. For this reason it's easier to use unaccented characters in all cases -- the end result is the same and you don't have to memorize which accented characters belong to Latin-1. Ahasuerus 18:11, 14 September 2016 (UTC)
That's what I had been doing with the Central and Eastern European ones while transliterating - fix the family/sorting name at the same time to ensure that the sorting works. If the sorting can solve it itself that is fine but I am not going to think which of the 6 accented and diacritic 'a' characters are fine and which are not - so all get fixed. :) Anniemod 22:18, 14 September 2016 (UTC)
I think removing all diacritics and such from any letter in the Family Name field would be fine. Just give the appropriate Romanization without any of those and things should sort well. That's what they do on Wikipedia when creating the DEFAULTSORT field. ···日本穣 · 投稿 · Talk to Nihonjoe 18:18, 14 September 2016 (UTC)

Advanced Title/Publication Search and author data

Advanced Title Search and Advanced Publication Search have been updated. You can now restrict searches based on the author's place of birth, date of birth and date of death. For example, you can search for 2010 NOVELs by authors born in Australia. Ahasuerus 17:14, 14 September 2016 (UTC)

Advanced Title/Publication searches have been modified to allow limiting searches based on author Web pages. For example, you can search for 2010 NOVEL titles by authors one of whose Web pages contains "austlit". In addition, the names of certain selection criteria have been clarified to avoid confusion. Ahasuerus 20:00, 14 September 2016 (UTC)
That's awesome! Anniemod 22:30, 14 September 2016 (UTC)

Advanced Search for code, not display (eg, Notes and Synopsis

May this be useful to some of you. Perhaps the point can be indicated in a second line at ISFDB Advanced Search, below "Supported wildcards: * and % match any number of characters, _ matches one character".

Last week I added to the database several Notes including (as displayed) Encyclopedia of Fantasy (1897). Later I located them using Advanced Search on Title and Publication records; Notes and Synopsis fields; contains.

For this it was not adequate to search for "Encyclopedia of Fantasy (1897)" or the like; no hits. Instead search for the precise code, Encyclopedia of Fantasy</i> (1897) or the like--at least the crucial fragment, perhaps y</i> (18

--Pwendt|talk 16:55, 15 September 2016 (UTC)

A new cleanup report - IMDB links

As some of you know, the IMDB database changed its URLs a while back. Like us, they used to embed movie/series names into their URLs. Like us, they eventually realized that non-English titles like "Hólmfríður Þórhallsdóttir" caused too many issues. The difference between us and them was that we started using plain record numbers in URLs while they adopted "title codes". Their title codes always use the same format: "tt" followed by their internal title number. For example, the title code for the original Star Trek series is "tt0060028".

About 70% of our awards with IMDB links already use the new format. The remaining 30% (377 awards at last count) will need to be fixed manually. A new cleanup report has been deployed and the data will become available tomorrow morning. The software has been updated to prevent anything other than "title codes" from being entered into this field going forward.

Please check each "title code" carefully when making corrections. There have been various changes to IMDB titles over the years. The old titles that we used when entering the data 5 or 10 years ago may not longer match what's in the IMDB database today. Make sure that you are linking to the correct IMDB title and not to a similarly named one. Ahasuerus 17:56, 15 September 2016 (UTC)

So in cases like this one: (where the episode is cited), do we link to the show or to the episode? And in that specific case, as it is a two-parter episode that is nominated as one entry, what are we linking to? The first part (tt0517721),the second part (tt0517722), the show (tt0105946)? Thanks Anniemod 18:52, 16 September 2016 (UTC)
I believe that awards and nominations associated with individual episodes should link to the awarded episode. Two-part episodes are a special case which we are poorly equipped to handle, but I guess the least bad option is to link to the first part. Ahasuerus 19:58, 16 September 2016 (UTC)
And maybe adding the notes into the record for the multiparts... Thanks! Anniemod 21:01, 16 September 2016 (UTC)
Looks like there was that single two-part award (or at least I did not catch any others). Maybe the help pages should be updated to advise on the case in case more of those show up (these days every series seems to be going for multi-part episodes so we may have more of those)? Meanwhile, I think I have all of the ones that were still not fixed in the Pending Approvals list. Anniemod 01:07, 18 September 2016 (UTC)
Done. And thanks for working on this report! Ahasuerus 01:21, 18 September 2016 (UTC)

Wrong title of unverified book

Чужак and its title reference should actually be called Лабиринт. The source shows the Лабиринт title. The English editions is correct - it is just the title of the original title and publication that are messed up. It is a valid title for the work (according to ) but this is not the correct edition according to the linked source and the cover. What is the correct thing to do in this case? Fix it and add a note that Чужак is also a valid title? Make a variant? Something else? Thanks Anniemod 18:26, 15 September 2016 (UTC)

I had to work on Frei's bibliography back when The Overlook Press and Gollancz started publishing her books in English. It was a huge can of worms because her Russian publishers kept changing book titles and reshuffling collection contents. I am sure we'll sort it out when we get more editors willing and able to work on entering Russian pubs. For now I suggest we just fix the pub and add a note about title/contents changes. Ahasuerus 18:47, 15 September 2016 (UTC)
Not the only one - a lot of authors are in that state in the old Bloc countries. Some of it is copyright related. Some of it is just people not knowing what they are doing. Anniemod 20:53, 15 September 2016 (UTC)
Sounds like a plan. At least this will get the cover and title to agree... I will go see if I can fix it. :) Anniemod 20:53, 15 September 2016 (UTC)

Transliterating Authors into Hiragana and Katakana

I would like to suggest that we add transliterations of author's names in Japanese pubs in either Hiragana (Japanese names) or Katakana (foreign names). This would make it easier for native speakers of Japanese to search our database. The reason for this is that only a few kanji have unique pronunciations. Names are especially fluid, every name has multiple renderings in kanji (Satō has nearly 20), and each set of kanji comprising a name can be pronounced in several different ways. Japanese search routines will bring up all possibilities, but I don't think our searches can do this. Kana are phonetic, so a search will bring up everything we have transliterated, including authors that we don't know the kanji for.--Rkihara 23:25, 15 September 2016 (UTC)

I've done this for a few of them. We should be careful, though, that we don't introduce incorrect readings for a given name. For names, 佐藤 is pretty much the only one you'll tend to run into for Satō. There are very rare exceptions, but they are very rare. Occasionally, you'll run into 佐東. Any others are really quite rare. ···日本穣 · 投稿 · Talk to Nihonjoe 00:19, 16 September 2016 (UTC)
That's good to know. I used Satō for an example, since I think it is the most common surname. What about given names as opposed to family names?--Rkihara 16:09, 16 September 2016 (UTC)
I've never heard of Satō as a given name. If that's not what you were asking, please clarify. ···日本穣 · 投稿 · Talk to Nihonjoe 19:06, 19 September 2016 (UTC)
Assuming that we are talking about entering Hiragana and Katakana versions of author names in the "Transliterated Name" field, then we already have precedent for that. Here is what the Help text for transliterated legal names says:
  • If you need to enter multiple values in this field, click the "Add Trans. Legal Name" button and enter additional values in the new fields that will appear. This may be done in the following types of cases:
    • There are multiple competing Romanization schemes for the author's working language. For example, there are multiple Romanization conventions for the Chinese language, including Gwoyeu Romatzyh, Wade-Giles, and Pinyin.
    • The author's working language uses multiple scripts. For example, Japanese names are usually written using kanji, but in certain cases hiragana or katakana are used. In Serbian, both Latin and Cyrillic are used, although the Cyrillic alphabet is considered primary. In Azerbaijani, the Perso-Arabic script, Cyrillic and Latin have all been used at different points.
    • The author has lived in different countries which use different alphabets/scripts. For example, Alexander Lomm lived in Russia as well as Czechoslovakia and wrote SF in Russian and Czech. Since we list Russian as his working language, the Russian form of his legal name, "Кличка, Вацлав", is currently entered in the Legal Name field. The Czech version of the name, "Klička, Václav", and the fully Romanized version, "Klichka, Vatslav", appear in the Transliterated Legal Name field.
It was my understanding that the other, more recently added, "transliterated title/name" fields would be used the same way, but it wasn't spelled out in Help. Ahasuerus 00:43, 16 September 2016 (UTC)
I'll start with the authors. I think titles transliterated into hiragana would be difficult for a native speaker to read, like English written without spaces. It would have to be disambiguated in some manner.--Rkihara 17:53, 16 September 2016 (UTC)
They aren't difficult to read since you just read the characters that are there, and it's not uncommon to have them given that way as a reference in articles and so on. ···日本穣 · 投稿 · Talk to Nihonjoe 19:07, 19 September 2016 (UTC)
Someday it will be nice to have non-Latin author directory support. Uzume 10:43, 20 October 2016 (UTC)

Advanced Publication Search - verifiers

"Primary Verifier" has been added to the list of selection criteria in Advanced Publication Search. I don't think we need to add "Secondary Verifier" as a selection criterion, do we? Ahasuerus 15:53, 16 September 2016 (UTC)

Changes to "... Titles with a Latin Author Name" cleanup reports

All "... Titles with a Latin Author Name" cleanup reports have been modified to display each title's type. Ahasuerus 19:56, 16 September 2016 (UTC)

Awesome. ···日本穣 · 投稿 · Talk to Nihonjoe 20:50, 16 September 2016 (UTC)
Update: All of these cleanup reports as well as all "X Titles without Transliterated Titles" have been updated to display each title's author(s). Ahasuerus 15:37, 18 September 2016 (UTC)

Advanced Author Search - Working Language

"Working Language" has been added to the list of selection criteria supported by Advanced Author Search. Ahasuerus 15:17, 18 September 2016 (UTC)

Transliteration changes

As per FR 925, the way transliterated names and titles are displayed on record pages has been changed. They are now displayed as mouse-over bubbles instead of separate lines. The change affects author names, legal names, titles, publications, publishers and publication series. Ahasuerus 17:25, 18 September 2016 (UTC)

Opinions sought: Is "The Book of Imaginary Beings" fiction or nonfiction?

I'm sure many of you have read The Book of Imaginary Beings by Jorge Luis Borges and Margarita Guerrero; it is a compilation of information about fantastic animals and people from folklore and literature, but along with info from other people's sources Borges added oddities to it in his own style. Would you say it belongs in the fiction or nonfiction section? --Vasha77 00:45, 19 September 2016 (UTC)

Sounds like the Barlowe's Guide to Extraterrestrials. I would treat it the same way. ···日本穣 · 投稿 · Talk to Nihonjoe 04:50, 19 September 2016 (UTC)
Yes except that Barlowe thought up illustrations of other people's aliens, whereas a few of the entries in "Imaginary Beings" were entirely invented for this book... And they have sometimes been reprinted separately. Would those reprints be filed under fiction, and the work as a whole under (eccentric) nonfiction? --Vasha77 15:54, 19 September 2016 (UTC)
The illustrations, if entered individually, would be listed as INTERIORART, not fiction. If they are cover images, they would be that, not fiction. Regardless of whether he thought up some of the images himself, it sounds like the same general type of book as Barlowe's, so I would treat it as such. ···日本穣 · 投稿 · Talk to Nihonjoe 16:07, 19 September 2016 (UTC)
I read it a long time ago, so take it with a grain of salt, but it felt more like fiction than non-fiction. It was similar to Stanislaw Lem's "A Perfect Vacuum", which we list as a COLLECTION. Ahasuerus 16:58, 19 September 2016 (UTC)
I also read it a long time ago. As I recall it's a short collection of descriptions of various beings, mostly from folklore and mythological history plus one or two from science fiction and fantasy, written in the style of short entries in an encyclopedia. Disregarding their lack of fictional style – which as we know needn't be present for a work to be regarded as fiction – I reckon that categorising the book as fiction would be incorrect, even though you may be more likely to find it grouped together with Borges's fiction in a bookstore. There are probably several comparisons you could make, such as Brian Stableford's Dictionary of Science Fiction Places which has a similar style in presenting it's imaginary content, and which is also correctly categorised as non-fiction, or Star Trek: Aliens and Artifacts which again is a 'book of imaginary beings' yet categorised correctly as non-fiction. Why should we make an exception for Borges? My two cents. PeteYoung 20:32, 19 September 2016 (UTC)
OK, I will change the excerpts to ESSAY, then. --Vasha77 16:59, 20 September 2016 (UTC)

Interviewees only listed under canonical name

In this entry and this entry, I submitted a correction to change "Bob Eggleton" to "ボブ・エグルトン", since that's how he was credited in the original work. These were declined. It seems counter-intuitive to me to not list them how they are credited in the original work. Is there a policy somewhere about this? I can't find anything which states that we can only use the canonical name. ···日本穣 · 投稿 · Talk to Nihonjoe 20:20, 19 September 2016 (UTC)

See Template:TitleFields:Interviewee.The canonical name is used in order for the interview to be displayed on the interviewee's author page. -- JLaTondre (talk) 23:10, 19 September 2016 (UTC)
So I guess we need to get a feature request to display interviewees that are listed under a pseudonym. ···日本穣 · 投稿 · Talk to Nihonjoe 23:30, 19 September 2016 (UTC)

Fonts Used in Mouseover Hover Box

The fonts that appear in the box when mousing over foreign authors and transliterations are hard to read and need to be larger. At least the size used on page.--Rkihara 16:32, 27 September 2016 (UTC)

Sounds like a good idea. I'll take a look. Ahasuerus 16:35, 27 September 2016 (UTC)

Publications/Titles mismatch

Just making sure I do not mess this up. :) In cases like this one: , the title should get a variant in Hungarian and this publication should be pointed to it, correct? Thanks! Anniemod 18:45, 29 September 2016 (UTC)

Yes, the easiest way is an unmerge, followed by modification at the new title level then a variant to the english initial record. Hauck 19:20, 29 September 2016 (UTC)
Noted, thanks. :) I am surprised that the report that catches the mismatches did not flag that one... Anniemod 19:22, 29 September 2016 (UTC)
And while I was delaying getting into it, you seem to have fixed it :) Thanks Anniemod 20:08, 29 September 2016 (UTC)

Same person or two different people

Can anyone confirm that is indeed both the author and the artist with that name? I think it is two people with the same name but due to the pretty common name, I cannot find anything that proves it... Thanks! Anniemod 00:14, 30 September 2016 (UTC)

Well, it's also possible that it is an integer person. It's not uncommon to try ones talent in different areas and both of her titles aren't so far away from each other in time. Stonecreek 03:45, 30 September 2016 (UTC)
It is possible :) Something does not sound right (no idea why) so thoughts I would see if someone has an idea :) Anniemod 15:39, 30 September 2016 (UTC)

Pub-title mismatches

The Publication Title-Reference Title Mismatches cleanup report has been enhanced to include all publication types. It can be accessed by all editors, but only moderators can "ignore" records.

Please note that the report logic ignores subtitles and minor punctuation differences. The reported differences usually warrant some kind of action. Some very old publication records are translations linked to canonical titles; they will require a separate VT instead.

The data will become available in about 3.5 hours. I expect 1,400+ pubs to be flagged. Ahasuerus 01:45, 30 September 2016 (UTC)

Merge or variant; Subtitle or not

First, should a book be interpreted as having a subtitle given this title page, where the candidate follows the author credit?

I would say that is not a subtitle, but rather a description (since it is separated from the title). -- JLaTondre (talk) 00:52, 5 October 2016 (UTC)

Anyway: I added one recent publication of The Golden Book of Springfield by Vachel Lindsay, with a Note to Moderator that I would later merge with the very long title. In fact I leave them, as the two in-database Vachel Lindsay Novels, and inquire. (I depart having hastily put links to the HathiTrust digital catalog records into the early publication 578378.)

I understand that maintenance of two or more NOVEL Titles for one novel is permitted, with one the parent of all others as variants. For Through the Looking-Glass we have many, including variants so minute as comma and hyphen. Is there any agreement on when variants should be maintained, rather than merge the NOVEL Titles and leave our record of differentiation to the title fields of the NOVEL Publications?

Either way, merge or variant, there is the remaining question which version of the title under which to merge or to make parent of the other versions. In one large class, the original title is long, with one or more subtitles, and a later title is identical except to omit subtitle(s).

--Pwendt|talk 20:13, 4 October 2016 (UTC)

Title records are often entered without series or subtitle information. Publication titles can be entered with series or subtitle information. If this is the only variation in titles of different publications of the same work, then they would be merged under the same title record. In other words, if one publication title is just the title alone (ex. TITLE), another publication title contains the series (ex. SERIES: TITLE), and a third contains a subtitle (ex. TITLE: SUBTITLE), they should be merged under a since title record of the name TITLE. If the title varies for other reasons - spelling (ex. honor vs. honour), abbreviations (ex. and vs. &), punctuation, etc. - then it should technically be a variant. However, we're not always consistent on that for small deltas. -- JLaTondre (talk) 00:52, 5 October 2016 (UTC)

Two records for the same press

Phantom Press International and Phantom Press (Poland) are the same Gdansk based publisher: Phantom Press. What is the policy for these cases - do we link them via the notes or do they get merged somehow? Thanks! Anniemod 20:25, 5 October 2016 (UTC)

I've merged them under Phantom Press (Poland). Hauck 05:45, 6 October 2016 (UTC)

The Gods of Pegāna introduction muddle

What I'm about to say will be very long, because the situation is very confused and I want to be clear! Let me start by posting four images: pages v, vii, 1, and 3 of the 1905 first printing (publisher: Elkin Mathews) of The Gods of Pegāna.

page v page vii page 1 page 3

What we have there is -- p. v: the heading PREFACE in slightly larger type than the text body, separated from the body by about one line, and a paragraph beginning "There be islands..." (note that the paragraph has the first word in small-caps).
p. vii -- no heading, and another paragraph beginning with a word in small-caps, "In the mists..."
p. 1 -- the heading THE GODS OF PEGANA in type about twice the size of the body, separated from the body by about two lines, then another text beginning with small-caps: "Before there stood..."
p. 3 -- the formatting is the same as on p. v, as far as the size and separation of the header. This story, "Of Skarl the Drummer", once again begins with a word in small-caps. Every other individual story in the volume has the same formatting.

To sum up: OF SKARL THE DRUMMER is a chapter heading, and so is PREFACE. But THE GODS OF PEGANA (as indicated by its different formatting) is not a chapter heading, but rather the title of the entire collection of stories. Therefore, the section "Before there stood..." has no heading and no title. Furthermore, as indicated by the fact that it starts with a word in small-caps, "In the mists..." is not part of the Preface but another independent item, also with no heading and no title. (Note that this edition of the book does not have a table of contents.)

So there are two issues that need to be cleared up in order for these writings, and all later publications of them, to be clearly indexed in the DB.
A. How should we refer to these three introductory short texts?
B. How do later editions format, title, and refer to them, and how can what they do be made clear in the DB?

A selection of links and images should make it clear that later editions are wildly varied.

  1. 1916 Published by John W. Luce, 1916. Taking the half-title as p. i, here is how the material is arranged: v. "There be islands..." headed by "Preface" in the same format as chapter heads. vii. Contents. ix. List of Illustrations. xi. "In the mists..." 1. "Before there stood..." headed by a much larger format "The Gods of Pegāna". 3. "Of Skarl the Drummer".
    In other words, the same texts with the same titles, or lack of them, as in the Elkin Mathews edition. But how does this edition list them in the table of contents? "There be islands..." and "In the mists..." are not listed; page 1 is listed as "The Gods of Pegana", and is not apparently distinguished from chapter headings such as page 3 "Of Skarl the Drummer." Anyone who only looked at the table of contents would think that the title of "Before there stood..." was "The Gods of Pegana".
  2. also has a scan of a 1911 copy published by "The Pegana Press" in which the beginning pages are nearly identical to the ones in the 1905 edition.
  3. Gutenberg Project Gutenberg e-text. Does not state what print edition(s) it was prepared from. The table of contents lists "Preface" and "Introduction". In the body of the book, the heading "Preface" precedes "In the mists..." and "Introduction" precedes "Before there stood". The next heading is "Of Skarl the Drummer". "There be islands..." is not in the text at all.
  4. As included in Gods, Men and Ghosts edited by E. F. Bleiler (Dover, 1972). On page 219, we see image the heading "The Gods of Pegana" in the same font as is used for story titles in this collection. Below it, is PREFACE in a smaller font. This is followed by "There be islands...", "In the mists..." and "Before there stood...", each beginning with a word in small-caps and separated by a blank line. On the next page image is OF SKARL THE DRUMMER in the same font as PREFACE and starting with a word in small-caps. In other words, I would say that the three texts have been treated here as separate items, but they have together been given the collective title "Preface". The table of contents image only lists "The Gods of Pegana" (page 219), the overall title.
  5. As included in In the Land of Time, and Other Fantasy Tales edited by S. T. Joshi (Penguin, 2004). image Three sections each with a chapter-like heading: titles "Preface", "[Prologue]", and "The Gods of Pegāna".
  6. I invite anyone who's read this far to find other editions and see what they do!
Wildside has an omnibus of all of the collections. See the Look Inside for details. There, for The Gods of Pegāna, clearly titled, are Preface -- "In the mists..." -- and Introduction -- "Before there stood..." -- as two short pieces before Of Skarl the Drummer. I see no sign of anything using "There be islands...". --MartyD 17:53, 29 October 2016 (UTC)

Here is a sketch of a table showing just what standardization is up against.

I do have a proposal for improving the DB. I think that the three short introductory texts should each be given a master title entry whose title is their opening words. This will be completely unambiguous. We then can note for any publication which of them are included, and how they are titled there. I don't think that "There be islands..." should be given the master title "Preface" because there is just too much potential for confusion, being as Gutenberg uses "Preface" to refer to "In the mists..." and Bleiler uses it as the collective title for all three texts. Explanatory notes will become much simpler if we always refer to all the texts by their opening words.

What do you say? --Vasha77 21:45, 6 October 2016 (UTC)

Afterthought-- maybe these three canonical titles?
  1. Preface: "There be islands..."
  2. Prologue: "In the mists..."
  3. Introduction: "Before there stood..."
Or that for the preface, just the opening words for the others? -- Vasha Vasha77 14:48, 7 October 2016 (UTC)
I've added three additional editions that I have at hand. I don't know that we have a standard for dealing with untitled fiction (we do for poems). When I've entered untitled introductions, I've used "introduction" in lower case, with a note that the item is actually untitled. I've followed this convention with my proposals for the editions I've added. I have used preface, prologue and introduction for each of the three titles. It seems fine to use these generic titles as each item appears that way in at least one of the editions. It also may be instructive to look at how Locus1 has handled these items here and here. Locus ignores the untitled items. I'll also mention that Blieler in The Guide to supernatural Fiction describes the book as containing "31 subtitled episodes". That count only works if you assign a title to the "introduction". I'd be in favor of using "The Gods of Pegāna" when it is present on the same page, regardless of differences in font. I'm also not in favor of adding the first line as part of the title, as it doesn't match our standards. I'd prefer that the first line be quoted only in the notes, if necessary. Lastly for titles that are composes of more than one of these three titles (Dover and Gollancz), I feel that these need to be separate titles that are not merged with the main three. Unfortunately, we don't currently have a way to indicate a whole/part relationship of this nature. Well that's my $0.02. --Ron ~ RtraceTalk 02:24, 12 October 2016 (UTC)
This plan has the advantage of grouping as many instances as possible of the same text under the same title; nearly everything would fit neatly under one of "preface", "prologue", and either "introduction" or "The Gods of Pegāna" (especially if you don't use a different title for books that print "Pegana" instead of "Pegāna"). The titles would not exactly match what is on the page (in the case of untitled items, or Joshi's "[Prologue]") but your lower-case convention is intended to indicate that. The disadvantage is that this system creates a number of identical-titled items that have to be kept apart. What title would you give to the combination-of-three that the Dover edition calls the "Preface"? How would you indicate that what the Gutenberg edition calls the "Preface" is not the same text as the one everyone else calls by that name? --Vasha77 03:29, 12 October 2016 (UTC)
It's not a goal to have the same text always use the same title. Rather our policy is to reflect the title as it appears in the publication. If the same text appears under several different titles, then each should be added as they appear in each publication; one title should be selected as the canonical title, and the rest should be made variants of that parent. Unfortunately, we don't have a structure in the database for noting when several works are combined into a single work. There are several discussions in the Rules and Standards archives regarding fix-ups which is an analog for what we are discussing here. I don't have a problem with two different texts sharing the same title. They have to because that's the title each works was published as. The only thing is to ensure that the notes reflect that the two titles shouldn't be merged. The notes are compared on the merge screen, and it's clear that the two title records are unique when it's reflected in the notes. There are many examples of this in the database.--Ron ~ RtraceTalk 01:55, 25 October 2016 (UTC)
And I just noticed that you favor writing "Preface" when that word is in the book and "preface" when it is not. That adds another element of complexity that may be unwelcome. How in heck are people who want to add a new edition supposed to understand what to do?
What we don't have is a standard on what to use as a title for SHORTFICTION (as these are fictional essays), when they are published without title. We do use the first line for poems. Fiction doesn't divide into lines in the same manner. I've used "introduction" in lower case in these instances. In this case, I've suggested using "preface" and "prologue" merely because we'll need different titles within the same collection and I do think it would be OK to align the lower case names with a canonical title parent where a title actually exists.--Ron ~ RtraceTalk 01:55, 25 October 2016 (UTC)
Isn't that inconsistent with assigning names like "Introduction (The War of the Worlds)"? That isn't what's literally in the book either. By the reasoning of what you just wrote, it should be called "Introduction" and kept apart from the author's other intro's by notes. Anyway, yeah, I feel like there is a major problem going on here, so I guess taking it to Rules and Standards is the answer. --Vasha77 02:29, 25 October 2016 (UTC)
It might be nice (when using your system) to reduce things to as few variants as possible. Use only "preface", with a note on the edition as to whether the title is literally there or not; only "The Gods of Pegāna", with a note on the edition whether the macron is printed or not. -- Vasha77 03:36, 12 October 2016 (UTC)
I had in mind a convention that can be described in one sentence: exactly what's on the page, followed by the first line. The results are in pink in the table. This plan does make it easy to create titles for additional editions, and (I think) it is readable -- it tells you what's in the book without you having to look at the notes. It might create a lot of different variants, but currently only two, three, and three, respectively, for the three texts. Unfortunately it doesn't work for combinations of several texts, but at least naming them by a different convention would mean they couldn't be confused with single texts. The other major drawback is that (you say) nothing like it has ever been used in the DB before. --Vasha77 04:44, 12 October 2016 (UTC)
Additional notes above in-line. What this really comes down to, is how to title SHORTFICTION when published without a title. I've noted how I've done it for other instances, and suggested how to extrapolate that here. If you disagree strongly, you should probably start a discussion on the Rules and standards discussions board for just that issue. --Ron ~ RtraceTalk 01:55, 25 October 2016 (UTC)

Proposed solutions

Please add notes. --Vasha77 23:45, 7 October 2016 (UTC)

1905/1916 Gutenberg/
Gods, Men and Ghosts (1972) In the Land of Time (2004)
"There be islands..." book Preface absent part of Preface Preface
proposed Preface: "There be islands..." Preface[, Prologue, and Introduction] (The Gods of Pegāna)

I'm not very happy with this one. But the idea is that if some edition used the title "Preface" for just the 1st two texts, it would be indexed as "Preface [and Prologue] (The Gods of Pegāna)". -- Vasha
Preface: "There be islands..."
"In the mists..." book no title Preface part of Preface [Prologue]
proposed "In the mists..."
Prologue: "In the mists..."
Preface: "In the mists..." see Row 1 Prologue: "In the mists..."
[Prologue]: "In the mists..."
"Before there stood..." book no title Introduction part of Preface The Gods of Pegāna
proposed "Before there stood..."
Introduction: "Before there stood..."
Introduction: "Before there stood..." see Row 1 The Gods of Pegāna
The Gods of Pegāna: "Before there stood..."

Beyond the Fields We Know (1972) The Complete Pegāna (1998) Time and the Gods (2000)
"There be islands..." book Preface untitled Preface
proposed Preface (The Gods of Pegāna)
Preface: "There be islands..."
preface (The Gods of Pegāna) note that it is untitled
"There be islands..."
Preface (The Gods of Pegāna)
Preface: "There be islands..."
"In the mists..." book untitled untitled The Gods of Pegāna
proposed prologue (The Gods of Pegāna)
"In the mists..."
prologue (The Gods of Pegāna)
"In the mists..."
The Gods of Pegāna (with notes not to merge with the same title)
maybe (ugh) The Gods of Pegāna [Prologue and Introduction]
"Before there stood..." book untitled The Gods of Pegāna included as part of The Gods of Pegāna (above)
proposed introduction (The Gods of Pegāna)
"Before there stood..."
The Gods of Pegāna
The Gods of Pegāna: "Before there stood..."
not listed as it is incorporated in the earlier title

ISFDB e-mail

Please note that our e-mail service is currently down. If you click "E-mail this user", your e-mail will be accepted by the system but not sent. It looks like a replay of the problem that we had back in January when the OS was upgraded. I am currently looking into this issue. Ahasuerus 16:35, 8 October 2016 (UTC)

For now, I have changed the Wiki settings to allow registration without e-mail confirmation. Ahasuerus 18:28, 8 October 2016 (UTC)
It would appear to be a second edition of the mess that we encountered in January. I have asked Al for help. Ahasuerus 17:28, 10 October 2016 (UTC)
It might be worth asking if he's willing to help upgrade the current version of the wiki to a newer version, too. There have been a lot of fixes and updates since it was last updated. ···日本穣 · 投稿 · Talk to Nihonjoe 23:23, 12 October 2016 (UTC)
It's on Al's list of things to do, but he has been very busy with non-ISFDB stuff the last few years. Ahasuerus 00:49, 13 October 2016 (UTC)

The Collected Strange Stories, Volumes I & II options

I am not sure what is the better option here:

The double volume has awards and nominations so even if we do not have a recorded publications, I think the title should stay to keep this information. Thoughts? Anniemod 21:45, 8 October 2016 (UTC)

Standardizing capitalization in Portuguese?

Back in March, a conversation took place on Albinoflea's discussion page in which it was mentioned that there isn't just one standard capitalization practice for titles in Portuguese (and Portuguese: A Reference Guide agrees that "Capitalization is optional in titles of books"). Now, that conversation talked about just following whatever is in the printed book. That isn't what you do in English titles (which you regularize); and sometimes the title is printed in all-caps anyhow. So... Does anyone have any opinions? The main options are: A) Follow rules similar to English where small grammar words aren't capitalized; B)Capitalize only the first word, the same as in French etc.; C) Either of these rules according to circumstance; D) What's in the book, and if you don't know, pick either rule. --Vasha77 01:33, 13 October 2016 (UTC)

I'm still not sure what the answer is for Portuguese, but I would say that in general if you're not sure you can always indicate which method you've used in the Pub Notes field.
Having just had a conversation about French capitalization, I think it would be good to document somewhere what the accepted practice is for each language (to the extent that we have them), similar to what we've done for currency, since the object is to standardize and not duplicate exactly what is present in the pub for these particular fields. Albinoflea 19:26, 14 October 2016 (UTC)
Sure. In that case I would definitely vote for the non-capitalized standard for Portuguese; that way no one has to learn exactly which small words don't get capitalized. Learning that for English is bad enough. --Vasha77 04:15, 15 October 2016 (UTC)

Reminder re: using "<br>"

Please keep in mind that we no longer need to use "<br>" to separate lines in Notes. It was something that we had to do in the 2000s, but the software was upgraded to handle carriage returns correctly in the early 2010s. Ahasuerus 23:14, 13 October 2016 (UTC)

I've been specifically reprimanded for not including the breaks in notes in the last couple years. Also, the times I've not included them, the system hasn't handled them correctly. Has something changed in the last couple years? ···日本穣 · 投稿 · Talk to Nihonjoe 18:04, 14 October 2016 (UTC)
The change was made in March 2014:
  • [It is] no longer necessary to enter "<br>" to create a new line. Old "<br>"s (we have almost 85,000 of them!) should still work.
I am unaware of any problems with the way carriage returns and HTML line breaks are handled in Notes, but, of course, it doesn't mean that the affected part of the software is bug-free. Could you try to recreate the problem? Ahasuerus 19:16, 14 October 2016 (UTC)

Differences between reprint and reissue?

Hi. German distinguishes between two kinds of new editions of a work:

  1. a work is reprinted but continues the printing count (3rd printing, 4th printing etc.) from previous editions. It can nevertheless be a revised edition, for example by correcting errors (in German "Neuauflage").
  2. a work is reprinted and re-starts the printing count at 1 (1st printing). This can happen when another publisher is publishing the new edition, often accompanied by a new ISBN (in German "Neuausgabe").

Is there such a differentiation in English? Are there differences between a "reprint" and a "reissue" in English? Jens Hitspacebar 17:12, 14 October 2016 (UTC)

In the US, the main distinction is between "editions" and "printings". An "edition" can have new or changed material included, e.g.:
  • a new introduction/afterword
  • textual revisions
  • new cover art/interior art
  • a new publisher
  • a different format (hc vs. tp vs. pb vs. ebook)
  • a new ISBN
A "printing" is typically an almost exact reprint of the first printing of a given edition with some minor changes, e.g. a new price, typo corrections, occasionally a new cover. Ahasuerus 19:27, 14 October 2016 (UTC)
I see. That sounds like it's pretty similar to the distinction in German. I'm looking for appropriate translations/terms for this from German to English. To put it short my understanding of your answer is that the terms would be something like this:
  • "(revised) new printing" for any new printing which might have been revised (or "reworked"?) but is, apart from that, a reprint of previous editions (bullet point one of my initial post).
  • "(revised) new edition" for any (revised) edition that fits the criteria for "edition" you listed above (bullet point two of my initial post)
Are these correct terms in English? What about the terms "reissue" and "reprint", are they not relevant in this context? Hitspacebar 21:26, 14 October 2016 (UTC)
The word "reprint" simply means that a text has been reprinted, e.g. see Doc Savage (Bantam reprints). A reprint that is almost exactly the same as the original is called "a facsimile reprint", e.g. see Adventure House Facsimile Reprints. A "reissue" is broader than a "reprint" because you can also reissue movies, stamps and so on.
I don't think I have seen books that claimed to be "revised new printings". When a book has been revised, it usually says "a revised edition", "a new revised edition", "an expanded edition", "author's preferred version", and so on. I guess the reason for it is that the word "printing" implies that there haven't been any significant changes.
As far as restarting the printing count goes, it typically happens when a new publisher gets involved. A book first published in 1989 and reprinted by Tor in 2010 may say "First Tor printing" on the copyright page. I can think of only one example of a major publisher restarting the numbering scheme for its books. It happened at Ace at some point in the 1970s or 1980s, but I don't remember the details. Ahasuerus 22:49, 14 October 2016 (UTC)
Thanks a lot! I think that's enough info to flesh out some appropriate translations for the ISFDB:Foreign_Language_Abbreviations page. As for something like "revised new printing": I've actually seen this in German pubs (as "Überarbeitete Neuauflage"). The question is of course if the publishers always use these terms consistently and correctly... ;) Jens Hitspacebar 08:44, 15 October 2016 (UTC)

Add centimetre values to "pb" in PubFormat help

I think the Template:PublicationFields:PubFormat help for "pb" can be improved a bit regarding the centimetre values because there's a slight deviance between "tp" and "pb" sizes:

  • For tp it states: "7.25" (or 19 cm) tall, or at least 4.5" (11.5 cm)"
  • For pb it states: For books as tall as 7.25" or as wide/deep as 4.5" use "tp". This contains no information about centimetre values. I suggest to add the same centimetre information stated for "tp" to the "pb" section because otherwise one might overlook what's stated for "tp" and convert the inch values in the "pb" section to exact centimetre values (which be would smaller than the ones for "tp": 18.42 cm x 11.32 cm).

I think this addition is a no-brainer but nevertheless wanted to make sure beforehand that it's ok. Jens Hitspacebar 13:14, 15 October 2016 (UTC)

Sounds like a reasonable plan. Ahasuerus 14:30, 15 October 2016 (UTC)
Done. Jens Hitspacebar 14:59, 15 October 2016 (UTC)

Title records for stories in omnibus

Andrew Hurley translated all of Borges's story collections and published them as Collected Fictions. This is in the DB as an omnibus, though the collections (in this translation) were not previously published. I think I ought to create Title records for the individual stories, but how? There is no publication of the "collection" alone to add them to. --Vasha77 15:54, 17 October 2016 (UTC)

(Also, there‘s a lot of work to be done separating out different translations of Borges and attaching them to the correct publications. I guess I'll start on that tonight, since I've finished other things I was working on, except for The Gods of Pegāna, which I'm still waiting for feedback on.)--Vasha77 16:46, 17 October 2016 (UTC)

Any translations of individual stories that are not in the database will need to be added as new titles using Edit Publication and then varianted. Translations that already exist in the database can be imported into this omnibus using "Import Title" -- see "Import Individual Titles" on page Help:Screen:ImportContent.
Re: adding COLLECTION titles to this omnibus, I don't think we can do that if Andrew Hurley's translations had not been previously published as collections. Ahasuerus 20:38, 17 October 2016 (UTC)
OK then, I should just remove all the "collections" from the "omnibus", change it to a COLLECTION, and add the individual stories. I will get started on that. --Vasha77 20:54, 17 October 2016 (UTC)
Sounds like a plan. We may also want to add a note explaining the book's relationship to previously published collections. Ahasuerus 21:07, 17 October 2016 (UTC)
Note added: "This collection, translated by Andrew Hurley, contains the stories from the following Borges collections: Historia universal de la infamia, Ficciones, El Aleph, El hacedor, Elogio de la sombra, El informe de Brodie, El libro de arena, and Veinticinco agosto 1983 y otros cuentos. Although Hurley divided the translated collections into separate sections in Collected Fictions, he did not previously publish them separately." --Vasha77 21:25, 17 October 2016 (UTC)
Approved, thanks. Ahasuerus 21:57, 17 October 2016 (UTC)

Sorting, author directory and nobiliary particles

I have noticed we have a number of authors that seems to be sorted incorrectly in the author directory. Specifically I am referring to those with nobiliary particles, such as "van", "von", etc. For example Paul van Loon and Kees van Toorn should probably not both sort to author directory entry "Va" but rather to "Lo" and "To" respectively. I am wondering if it would be good to develop a cleanup report for such things. At the very least, this should probably be captured as a feature request. Thanks. Uzume 10:54, 20 October 2016 (UTC)

I disagree - I would look for van Toorn under "van Toorn", not under "Toorn"... So if it is under va..., it is sorting properly.Anniemod 19:02, 20 October 2016 (UTC)
Uzume is right. Nobiliary particles are - at least in some languages - not part of the sort order, even if the "extra" name part consists of more than one word. As for Dutch, "van" and "de" are not part of the sort order, and in German "von" is not (among others). Example: Rolf von der Reith has to be sorted as "Reith, Rolf von der". And for "Kees van Toorn" the correct ordering is "Toorn, Kees van". Jens Hitspacebar 19:46, 20 October 2016 (UTC)
I wonder if the ultimate solution may be making the "Family name" (soon to be changed to "Sorting name") a multiply occurring field like "transliterated titles". It may also help address the issue of multiple directories (Japanese, Cyrillic, etc.) Ahasuerus 19:57, 20 October 2016 (UTC)
Where would you expect van Gogh to sort? I would not even check under "G". Ideally these all should be foundable in both lists (v and G for van Gogh) (and Ahasuerus posted while I was typing) :) Anniemod 20:01, 20 October 2016 (UTC)
Ahasuerus's suggestion makes the most sense. I agree with Annie, all native speakers of English would search on the nobiliary particle, so van Gogh, da Vinci, Van Vogt. Maybe it would be good to look at implementing a "fuzzy search," which would also take care of multiple transliterations.--Rkihara 20:13, 20 October 2016 (UTC)
I'd definitely expect van Gogh under "G", not under "v". But as a German native speaker I'm used to this kind of sorting. The solution with multiple sort names sounds good - and can then be used excessively for authors with extreme nobiliary particles like Ute Helena Bertram Gräfin von Nimcz zu Caldaha :) Jens Hitspacebar 20:23, 20 October 2016 (UTC)
Multiply occurring would be great allowing for multiple entries in a directory as well as appearing in multiple directories (e.g., as was mentioned Japanese kana, Cyrillic, etc.). Obviously this would likely require multiple cleanup reports to go along with such (e.g., the one I brought up about nobiliary particles). I like the idea of being able to find John Norman under Norman, ノーマン, and Норман, etc. On this multilingual note, I was considering whether it was a good idea to start linking to Wikidata vs. multiple Wikipedia languages too (e.g., link to Q186273 over both 安部公房 and Kōbō Abe for A265). Uzume 19:19, 26 October 2016 (UTC)
Wikidata looks like it will be a useful addition when we add "third party identifier" support, but I suspect that it may be too intimidating for casual users. Ahasuerus 02:01, 27 October 2016 (UTC)

ISFDB downtime at 9:30pm server time

The server will be unavailable between 9:30pm and 9:35pm server time. Patch notes to follow. Ahasuerus 01:08, 25 October 2016 (UTC)

Everything is back up. The way the following Web pages:
  • My Recent Edits
  • My Pending Edits
  • My Rejected Edits
  • My Errored Out Edits
  • Recent Edits
  • Recent Approvals (moderator-only)
  • Recent Rejections (moderator-only)
  • Errored Out Submissions (moderator-only)
display their data has been streamlined to use the same column names etc.
In addition, the "Subject" column of the following submission types:
  • New Publication
  • Clone Publication
  • Add Publication
  • Add Award Type
  • Add Award Category
  • Add Award
has been changed. For submissions approved after this patch it will link to the newly added record.
If you encounter any issues, please post your findings here. Ahasuerus 01:33, 25 October 2016 (UTC)

Letters into essay series

I remember a few years ago there was a... not sure I'd call it a trend, but shall we say a prevalence among some editors towards putting letters from authors to magazines/fanzines into essay series, and not long after that a push-back from other editors that bibliographically speaking this was a Bad Idea and should not be considered an activity useful to the ISFDB. I never indulged in it myself, but I see we still have series such as Letters by Philip K. Dick and Letters: James Blish to name just two, and I'm sure there must be many others, whereas letters from other authors (Ursula K. Le Guin, Arthur C. Clarke) are not yet treated this way. Maybe some editors are still happily and methodically pursuing this practice, I really don't know. My own point of view was that if this was to be an ISFDB standard worth adopting maybe it should apply to both genre authors and essayists/letterhacks in an even-handed way. Do we expand this practice with some ground rules, such as determining who gets to be "above the 'Letters into essay series' threshold" and who doesn't, or do we formally abandon it? Thoughts please! PeteYoung 13:40, 26 October 2016 (UTC)

The first question that comes to mind is whether it would be useful to create a separate title type for LETTERs that would be distinct from ESSAYs. I am in the process of compiling a list of possible new title types (fictitious essays, plays, etc) and perhaps letters should be added to it. Ahasuerus 14:00, 26 October 2016 (UTC)
I agree with those that think that it's a "Bad Idea". IMHO the grouping into series is spiralling out of control with series like this one or that one created for items that shouldn't even be included. Let's not talk about such byzantine constructions and all these self-published novels that have their own series and subseries. In fact I'm wary of this series phenonmenon because, in my view of "classic" bibliographer, the use of series makes a mess of the (default) display. Hauck 14:07, 26 October 2016 (UTC)
I have no problem with putting letters into series or alternately a TITLE type. I think it serves a useful purpose for biographers and other researchers. I also feel we should enter all letters, and not try to sort out people who are above the threshold. Departments without identifiable contents like Brass tacks, and Calender of Upcoming Events, should not be series.--Rkihara 15:49, 26 October 2016 (UTC)
Departments can be a headache because, as it says in FR 319 Implement Magazine Departments:
  • if a title first appeared in "Analog" as a "Probability Zero" short story, it only belongs to the "Probability Zero" group of titles in the context of "Analog", but not when it is reprinted elsewhere. Currently we handle "Probability Zero" as a regular series, but it's really a "magazine department", a different type of title groupings.
Unfortunately, it would be a non-trivial enhancement, which is why it hasn't been implemented yet. Ahasuerus 17:14, 26 October 2016 (UTC)

User Feedback Forum?

It occurred to me that it might be good to open a user feedback forum, so real users can give us feedback about what they find useful, what they like or dislike about the ISFDB, and so on. A permanent forum might be a lot of work, so maybe we could let people post comments a couple of times a year, a week at a time?--Rkihara 23:28, 26 October 2016 (UTC)

Well, you have to be a registered user in order to post here. It's not hard to register, but it's still a hurdle.
I guess one way to get around this issue is to go to a place where potential users hang out and ask them "on their turf": Usenet (not that there are many people left on Usenet), Reddit, Goodreads, etc. Some places are wide open and you can post anything while certain other places may require moderator permission lest the post should be removed as self-promotion. Ahasuerus 01:58, 27 October 2016 (UTC)
Facebook might be a good place to do this, since a lot of users are probably members, and spammers are filtered out.--Rkihara 05:22, 27 October 2016 (UTC)
Facebook might be the "easiest" choice, but even though it's pretty big there are still lots of people who would never register there. Apart from that, the really bad thing about FB is that, even in the groups, there are no real threads (like the ones "real" forum software offers). Which means it can be really hard to find and resurrect older threads, or to follow certain topics. I always liked Usenet's simplicity back in the days, but alas it has become a niche, and is no real option I think. The best option might be to set up our forum software instance. This would also offer the option to migrate our wiki communication to it eventually, which could result in two main sections in the forum: the user communication and the editor communication (including several sub-sections like Rules, Help etc.). Registered (approved) user would be able to post right away, whereas anonymous posts would have await moderation to avoid spamming.
Or, a completely different idea: I've been using several StackExchange sites for my everyday IT developer life (stackoverflow, askubuntu, superuser, serverfault) and found it pretty awesome so far (though only as a reader, no idea really how it works for editors). They already have a Science Fiction & Fantasy site, and there's a site to make proposals for new sites, maybe it'd be possible to add an "" site there? Or even more easy: we use the existing Science Fiction & Fantasy site and ask everyone to use the tag ISFDB for Q&A (which of course would mean that we'd have to hang out there and check for questions tagged that way). Jens Hitspacebar 21:40, 27 October 2016 (UTC)

Help Pages update

What is the process about changing help pages such as this one? The table needs to be updated for OCLC from 2012 to 2016 (or even just "current") so that it does not confuse new editors... Anniemod 20:58, 27 October 2016 (UTC)

Fixed. As a general rule, straightforward changes like this one can be made by anyone. More substantial/complex changes to Help pages are usually made after reaching consensus on the Rules and Standards page. Ahasuerus 22:51, 27 October 2016 (UTC)
Thanks and will keep this in mind. Anniemod 23:27, 27 October 2016 (UTC)

Pseudonym or fix a typo in a publication

In Bards and Sages Quarterly, October 2009 the name of Jamie Eyberg is misspelled as Kaime Eyberg. Amazon Look inside shows the misspelling clearly. The author is claiming the story list of stories and chances are that this is a one time misspelling. How do we handle this kind of situations - pseudonym it anyway or fix the entry of the magazine and update its notes to note the misspelling? I started on the pseudonyming but decided to stop and ask first :) Thanks Anniemod 23:48, 27 October 2016 (UTC)

For authors, as per Help, "[t]he name should be entered exactly as it is actually given on the publication's title page." This includes typos. Ahasuerus 00:02, 28 October 2016 (UTC)
Which is why I went for the pseudonym but good thing I did not and stopped and came to ask - so I can actually take another look at this. Apparently the misspelling is only in the TOC - the Amazon Look Inside contains page 16 and it is Jamie there (should have looked at it earlier but did not expect that many pages to be visible). Actual story beats TOC spelling, correct? And if I am reading the rules properly, that changes the situation. All it needs now is a fixed name in the magazine is a note in the magazine notes for the discrepancy(similar to what is already done for the other misspelling). Anniemod 00:11, 28 October 2016 (UTC)
Yes, that's exactly right. The title page trumps TOC, the cover credit, etc. Ahasuerus 00:16, 28 October 2016 (UTC)
I'll go fix it then. Thanks. And Fixing the link above. Anniemod 00:19, 28 October 2016 (UTC)

Hungarian pseudonyms

We have quite a few cases where we have both the "Last First" and the "First Last" as separate entries and I had been looking at pseudonyming them properly. However - in what direction? If we look at Hungarian usage, "Last First" is the common usage. If we look at almost an other language, the common will be "First Last". So what direction should the pseudonym be done in? What would be "the most common form" for these?

If the author's working language is Hungarian (or Japanese), I would use the "Family name Given name" form as the canonical name. Ahasuerus 22:53, 28 October 2016 (UTC)
That's where I had been leaning towards as well (the same way as a non-Latin language form is canonical for the ones that use other alphabets). If this will be the rule, can we update the help page to indicate this? :) Anniemod 22:56, 28 October 2016 (UTC)

Merge or variant

If an illustration is recolored (or super-exposed or whatever you do to achieve this), do we merge or do we variant? Here is the original The Necronomicon Files and here is the second version The Necronomicon Files: The Truth Behind the Legend: Revised and Expanded Edition. The art is the same (look at the biggest available image - it is even more obvious there) - except that the colors had been shifted a bit... I went for a merge but just cancelled it and decided to ask first. Anniemod 23:57, 28 October 2016 (UTC)

Ping :) Anyone with an opinion (I suspect that one got buried under the rest so drawing attention to it) Anniemod 22:55, 1 November 2016 (UTC)
When it comes to cover art, we merge. We don't consider variations in color, mirrored images, cropping, etc. as different versions. Typically, in these cases, the artist created a single image that the publisher(s) tweaked. -- JLaTondre (talk) 23:09, 1 November 2016 (UTC)
So I was on the right path then :) I will merge them. Thanks! Anniemod 23:27, 1 November 2016 (UTC)

Server downtime - 2016-10-28 at 8:30pm

The server will be down between 8:30pm and 8:35pm. Patch notes to follow. Ahasuerus 00:20, 29 October 2016 (UTC)

The server is back up. A new report, My Recently Changed Primary Verifications, has been deployed. You will find a link in the navigation bar immediately under "My Primary Verifications". As soon as a submission changing one of your primary verified publications has been approved, you will see the word "New!" (in fashionable blue) appear next to the link. It will disappear once you have visited the report page and navigated to a different page.
The report includes Publication Edit submissions as well as imports/exports into your verified publications. The report will list all of your primary verified publications which have been changed by other editors starting with the most recent change. Each line will provide a link to the body of the submission which you can follow to see what has been changed.
Please note that the report is not retroactive, which is why it is currently empty. Also please keep in mind the limitations of the submission viewer which shows the submitted data, but can't show the data that existed in the publication prior to submission approval.
Once the dust settles and we get used to the new functionality, we will need to decide what kinds of changes still merit Talk page notifications. Ahasuerus 00:48, 29 October 2016 (UTC)

Quarto magazine page count

I was once told by a moderator that a quarto magazine would always have a page count that was a multiple of 4. I actually have run across a counter-example. In Science Fiction Chronicle August 1988, the center sheet has a fold-out page on the right-hand side. The pages are unnumbered on the sheet, but page 26 precedes the sheet and page 33 follows it. The last numbered page is 56, making 58 the last page number. Doug H 21:58, 30 October 2016 (UTC)

Sounds like a case where the weird number of pages needs to be explained in the publication notes so someone does not "fix" it in the future. :) Anniemod 22:09, 30 October 2016 (UTC)

WHSmith added

WHSmith has been added to the list of supported third party Web sites. They have a decent selection and their e-book prices are often more accurate than Amazon UK's. The latter has been known to show Amazon's discounted Kindle price but not the publisher's recommended retail price (RRP). Ahasuerus 22:19, 30 October 2016 (UTC)

Canonical name question

What should be the canonical name for cases such as this one Nguyễn Dữ? The Chinese name because he worked in Chinese? The Vietnamese one because he is from Vietnam? Anniemod 18:39, 1 November 2016 (UTC)

Typically the canonical name is the most frequently used name. Ultimately it doesn't matter all that much since alternative names/pseudonyms will take you to the canonical page. Ahasuerus 20:04, 1 November 2016 (UTC)
So do we leave this guy as is (with the Vietnamese name) or change it to the Chinese (losing the Vietnamese all along - as we have no publications under either)? Anniemod 20:25, 1 November 2016 (UTC)
It would be nice if we could enter a least some of his Vietnamese publications. That way we'd keep the Vietnamese form of his name, which would help when searching the database either directly or via Google. Ahasuerus 20:40, 1 November 2016 (UTC)
The Chinese name is already there as a Legal name. If we temporarily leave the Vietnamese as canonical, we have both on the page and search can find it... Anniemod 20:50, 1 November 2016 (UTC)
I doubt leaving it as Vietnamese for now would cause any problems. Ahasuerus 21:12, 1 November 2016 (UTC)

Author as a Publisher

This of course is nothing new but with the proliferation of self-publishing lately (we do not have enough of it to be noticeable but that will probably change), won't that end up overwhelming the Publisher's Database sooner or later and drowning the actual publishers? Maybe it is time to add a "author/publisher" or "self" flag on the publisher records so that someone can filter publishers when searching/looking at lists. Any thoughts? Anniemod 16:38, 2 November 2016 (UTC)

Yes, it's getting to be noticeable, but I don't think it will cause any significant problems. For example, there are a lot more authors than publishers, yet the Author Directory page loads almost as fast as the Publisher Directory page. In addition, the proposed flag would raise at least two design issues:
  • Some self-published authors just say "Published by John Q. Author" on the title/copyright page, but other self-published authors come up with separate names.
  • Some self-published authors eventually start publishing books by their friends and peers, at which point they become "real" publishers, e.g. Lawrence Watt-Evans's Misenchanted Press.
Ahasuerus 20:47, 3 November 2016 (UTC)
Fair enough. I will table the idea for now (until I find a place where it is a problem) :) Thanks for the answer! Anniemod 22:35, 3 November 2016 (UTC)

Verification sources for non-English publications

Are there any plans to allow secondary verification for non-English publications? The list we have at the moment is great - for English language works older than a decade. For newer English language works, no DB exists besides Worldcat (in a way ISFDB is probably the DB). But for non-English language works, there are databases around (FantLab for Russian, SFBG for Bulgarian to name a few). Is there a plan to have verifications based on the publication language (and I know that the language is on the Title level but publication always has a title so it always has a language)? If no such plans, how about putting them somewhere on the list? That will help a lot with clearing errors and making the DB a bit less English only :) Anniemod 20:12, 3 November 2016 (UTC)

There are two related Feature Requests: FR 897 Re-do primary verifications (and the FRs linked by Marc Kupper in the Discussion section) and FR 235 Add support for external identifiers. Adding support for external (or "third party") identifiers will let us link directly to other bibliographic sites' records, e.g. this one. Assuming their URLs are stable, which, as we know from years of dealing with various Locus-hosted indices, is not always the case. Ahasuerus 20:36, 3 November 2016 (UTC)
SFBG has a special set of URLs that are stable - I need to find the note about the rules for that. However - I can as well ask both if they would agree ISFDB to link back to them (when/if that become possible) and how stable the URLs are (for the same reasons)... And because of the way the DB is built, it is almost error-less and pretty complete for SF in Bulgarian (outside for periodicals for which we actually have another source). Sometimes coming from a minority language makes this kind of projects easier. And no - I am not involved in either. :) Anniemod 20:51, 3 November 2016 (UTC)
How about redoing the secondary verifications space as well - allowing either a language to drive the display (so you do not need to mark N/A for non-English books across the whole board) or by allowing a "write-in" from a list that adds one or more additional lines based on a predefined list. Automating a lot of it will be nice of course but it is also about secondary verifications that cannot exist for a work... Anniemod 20:51, 3 November 2016 (UTC)
Keep in mind that the software supports multi-lingual publications. In addition, some national sites occasionally include foreign language books. For example, Fantlab, a Russian site, has a page for this Hungarian translation of a Russian novel.
That said, there is a Feature request to:
  • set "N/A" automatically for several of the secondary sources, where the date of the pub is past the date of the verification source
Ahasuerus 22:47, 3 November 2016 (UTC)
So let's have the ability for manual addition (from a pre-defined list) - so if a Bulgarian book is found in FantLab Човекът-лъч for example), you can add FantLab as one of the secondary sources but it does not show up for all Bulgarian books to be needed to be marked as N/A everywhere. Same for multi-lingual pubs -- allow editors to curate the list of secondary sources (because adding all for all books reduces the numbers we can have for secondary verifications). The automatic N/A is a great idea as well (although a change in the date of a publication will need to be tied with making sure these are still valid)... Anniemod 22:56, 3 November 2016 (UTC)
First we'll need to implement support for external identifiers, which have the potential to change the way we handle secondary verifications. Many secondary sources, including the Reginalds and Tuck, have unique numbers. We may end up migrating their verifications to the "external identifiers" area. In addition, some secondary sources like the Clutes are not really pub-specific. We may have to create a new "title verification" page as per FR 324.
None of the listed features are particularly high priority. The only exception is "external identifiers". They will enable Fixer to support ISBN-less publication, an increasingly important area. Ahasuerus 23:18, 3 November 2016 (UTC)
Understood. And thanks again for the explanations. I was just wondering while chasing a FantLab reference.:) That gives me another idea though: I wonder if what I am thinking cannot be achieved in another way - now most of the Russian publications end up with a reference to FantLab anyway. Adding a new field in NewPub/AddPub for a web page (the way we have it for a title) will allow these to be standardized a bit, make them easier to migrate if we have external verification downstream and effectively remind editors to put them there? And that may serve as a useful and easy way to find what is linked a lot for specific languages. :) Anniemod 23:33, 3 November 2016 (UTC)

To include or not to include, this is the question.

I know we have some kind of inclusion rules for non-printed media but I can never make my mind what applies and what does not. So I thought I would ask directly. Should the Spec Fic fiction from Words Without Borders be included in ISFDB? I lean towards including it - it had been around for a long time and it does have a lot of non-originally English prose. Of course, as a non-SF publication, it will fall under the "only SF stories included" rules but is there a reason not to include it at all? Thanks for any opinions. Anniemod 22:33, 3 November 2016 (UTC)

If I understand correctly our policy: if it's downloadable, it may be in (only the fiction as you surmised). Hauck 08:00, 4 November 2016 (UTC)
And to enter it, this (a non-genre anthology) or this (a non-genre magazine) could be helpful examples, I hope. Stonecreek 08:52, 4 November 2016 (UTC)
Well, that is my problem - what is "downloadable". :) I can send the story to kindle, I can save it, I can print it. But that is valid for anything online. You cannot download an ebook of an issue as a whole (which is why I decided to ask) but once the issue is published it is there (had not changed links since at least 2010 when I started reading it and probably longer and the archive is there since 2003 (kinda like Strange Horizons which we do catalog - but then it is all Spec Fic and non-fic). It is a thin line sometimes... And that specific one is important because it is bringing stories from all over the world into English -- so skipping it makes no sense to me. Anniemod 19:23, 4 November 2016 (UTC)
Stonecreek, thanks :) I had been looking at Nature as well for an example of non-genre publication. One of those days I will just jump and enter a few (and get the moderators to help me fix the mess when I make one) :) Anniemod 19:23, 4 November 2016 (UTC)

International editors

I had been looking at the best way to start getting more Bulgarian titles in the database (especially with my books not here and not being able to verify based on them). Talking with some people back home, we kinda came up with a system - I will enter books, do all varianting and what's not and then have a second editor verify (check all the data, either make changes or more likely send me a list of changes to make) and then PV when the book matches what they have on paper. I will be working either from already copied data into text format or from scans (and with access to the person with the book during the entering of data).

The reasons for the split of actions is two-fold - no time/desire on the other editor part to learn how to enter books here (and with non-English books, the process is long and complicated) and lack of enough English. That will also allow me to ping other people with even less English under their belt to verify against real books (for other books). At the same time, it is their verification, they get the credit. :) And I would prefer not to enter from their name.

Here comes my problem - until the last fix, a change in a PV-ed publication required a notification (or advised for one anyway). So adding a note in that second editor page that someone else is editing (and/or monitoring their page) would have been enough. Now with the new handy "View changed verifications", it gets complicated -- especially if we change the policies on notifications.

I was thinking of Transient verifications (after all having access to what I need from a book is almost as good as having the book anyway) but if my reading is correct, these do not get notified on a change (or am I misreading that?). Plus I do not want to misuse the feature.

And that may not be an isolated case - with the complicated system and the mandatory English, that may be a way to get international editors to lend a hand...

Thoughts? Am I missing an easy way to get all that achieved? Thanks. Anniemod 22:35, 4 November 2016 (UTC)

I would not use a transient verification. Verification says you are working from the text. While some people use it for full scans, it's really stretching it if you are only working with partial scans. I also don't believe we have discussed changing the notification system. The auto notifications are more to catch things that fall through the cracks so to speak, in my opinion. I believe if you just make the notification to the verifier as if it was your book, but explain you are making the change for someone else, that would be fine and workable. -- JLaTondre (talk) 23:58, 4 November 2016 (UTC)
I am not worried about me notifying the other verifiers (explanations work there - and there won't be almost any anyway - these are new books that will be added). It is more about me seeing the changes if someone changes the records and make sure that the record is still correct (as the PV either won't be able to react or won't check).
I had the same thought about the Transient - thus asking and not just doing it - was just exploring options. Although a scan + access to the person that has the book is kinda the same as having a library book for a week.:) But it does not solve my issue of seeing who changes the records so it does not make a difference. :)Anniemod 00:05, 5 November 2016 (UTC)

Adding one story to a collection

A collection is published. Months or years later it gets republished and a story is added (or dropped). I think the new collection should be a separate title maybe with a note explaining the difference. But from what I had seen around the DB, we seem to be keeping that under the same title (not even as a variant). What is the policy and agreement on this exactly? Thanks! Anniemod 01:37, 5 November 2016 (UTC)

New editions with relatively minor changes are usually handled in Notes and do not result in the creation of a new title record. Major changes do. Ahasuerus 02:41, 5 November 2016 (UTC)
Define major? Adding one story to a 20 stories collection is probably minor. What about adding 1 to a 4 stories collection? :) I guess it comes down to the editor, right? :) Thanks! Anniemod 02:51, 5 November 2016 (UTC)
If I could come up with a definition of "minor" and "major" that all ISFDB editors could agree with, I would die a happy man :-)
More seriously, it occurs to me that a lot of our problems -- abridged/expanded/revised works, disambiguated "Introductions", split novels, etc -- could be addressed by adding a short "disambiguator" field to title records. Unlike Notes, it would be displayed on Summary and Publication pages, hopefully clarifying things. Ahasuerus 03:07, 5 November 2016 (UTC)
You have a better chance solving world hunger than defining minor. :) In my book, anything that adds a new fiction element (which is not an excerpt) should be a separate title. :) But I won't act on that as this is not how the DB works.
Hm... that may work (although it will require let another cleanup effort). It can also open the door for later visual changes and filters (now we have a "show only translations in these languages), similar ones can be made for other disambiguators. Anniemod 03:12, 5 November 2016 (UTC)

Publication with no title reference?

This publication Worldpool looks very weird - no title reference in sight. I think that changing the type of the collection inside of it to a chapbook (to become the title that is missing) will fix it and make it look as it is supposed to. Or am I missing something? (asking before trying so I know how to handle the next one that I see like that). Shouldn't that one have been caught in some report? Thanks! Anniemod 09:45, 5 November 2016 (UTC)

Yes, personnally I would instead change the publication type to collection to match the contents but it'll work either way. Hauck 10:16, 5 November 2016 (UTC)
Amazon says 192 pages (just checked) -- so collection it is. Anniemod 10:30, 5 November 2016 (UTC)
Just curious... Why does this record even exist-- I though the DB didn't include comics? --Vasha77 11:35, 5 November 2016 (UTC)
It seems see here that there's a kind of special dispense for this series. To be frank, it itches me to delete the lot. Hauck 11:52, 5 November 2016 (UTC)
Elfquest - I've learned to leave those alone. For now. One day they will probably go down but for now, I am just fixing them when I find them with issues. Anniemod 19:20, 5 November 2016 (UTC)

Wiki cleanup question

What should be done with the remaining ones here: Fanzine Wiki pages not linked to Fanzine records. The ones that had their fanzines under different names or disconnected for other reasons are off the list already - these are ones that have wiki pages but no issues in the DB at all (that I can find anyway). Just leave them alone? Something else? Thanks! Anniemod 01:02, 7 November 2016 (UTC)

Ideally we would create skeleton entries for the known issues, put them in a series and then migrate the Wiki page to the database side.
Let's use Fanzine:Gothic as an example. There is a Pulp Trader listing which gives the following publication dates:
  • v1#1 - June 1979
  • v1#2 - December 1979
  • v2#1 - June 1980
  • v2#2 - December 1980
The dates match the range on our Wiki page, so we can go ahead and create bare-bones publication records. The other 13 fanzines can be handled similarly; some will take more time to process than others. Ahasuerus 01:55, 7 November 2016 (UTC)
Did not think of this. Sounds like a good plan. Anniemod 02:04, 7 November 2016 (UTC)
Is this eligible for inclusion? Bobolings? Example issue is here. I am struggling a bit with what the rules of when something is eligible and when not. If I am reading the rules correctly, this one is not eligible... Thanks for any opinions. Anniemod 20:05, 7 November 2016 (UTC)
Well, it is a perzine and the author tends to ramble a bit, but it contains his reaction to Richard Lupoff's One Million Centuries, mentions SF conventions, Jerry Pournelle, Buck Rogers, etc. It wouldn't be high on the list of priorities if we didn't have it, but since we do, we might as well preserve the data. Ahasuerus 01:45, 8 November 2016 (UTC)
So just add skeleton empty issues with no content for the ones we know something about? Anniemod 02:26, 8 November 2016 (UTC)
Pretty much! Ahasuerus 02:48, 8 November 2016 (UTC)

Pub typing opinions sought

I just accepted a Fixed submission of Creepy Campfire Quarterly #4. I converted it from NOVEL (Fixer gets confused about these things) and made it MAGAZINE, but I'm having editor's remorse and am wondering if it should be ANTHOLOGY instead. It's dated to the day, and verbiage inside it refers to "issue", but the publisher's website also uses "anthology" when talking about it. So I am soliciting input and am posting here because I imagine the thinking going into what type it should be will be of general interest/use. Thanks. --MartyD 04:03, 7 November 2016 (UTC)

If there's no general content to give it a magazine stamp (letter section, review sections etc.) I guess it is a similar case as Moorcock's / Platt's / Bailey's New Worlds Quarterly, which could be seen both ways according to the rules. Stonecreek 04:24, 7 November 2016 (UTC)
Reading the introduction to the first issue (Amazon Look Inside shows it), the editors call it an anthology and explain why it is different from a magazine (not a very coherent explanation but it is there). And they explain that there will never be any columns in it. So I guess it is closer to an anthology than a magazine (although it is a thin line). Anniemod 04:46, 7 November 2016 (UTC)
I converted it to anthology. It's easy enough to change back if we ever discover a reason to do so. --MartyD 01:20, 8 November 2016 (UTC)

Format question

Is a 60 pages staple-bound publication a pamphlet or a trade paperback (due to the size of the pages, it will a tp if it is a paperback)? Anniemod 08:23, 7 November 2016 (UTC)

I'd enter it as a pamphlet like this one. Hauck 08:47, 7 November 2016 (UTC)
Trade paperback is always perfect bound. If it's saddle-stitched (stapled), it's not a trade paperback. ···日本穣 · 投稿 · Talk to Nihonjoe 19:21, 7 November 2016 (UTC)
That's my feeling as well - it's one of those self-published thingies that get on my nerves when cataloging. Decided to ask though. Thanks! Anniemod 19:35, 7 November 2016 (UTC)

Award: Prix Litteraire Mannesmann Tally

I was reading an interview for Kevin J. O'Donnell and it mentioned an award - the "Prix Litteraire Mannesman Tally". It is not referenced in his ISFDB bibliography. It's not even in the list of awards (under that name). His Wikipedia entry makes reference to it, but as "Prix Litteraire Mannesmann Tally" and the link is unresolved. It also states that it was the French translation that won, so I'm not sure if it is a translation award. Long and the short - anyone ever heard of this and is it relevant? Doug H 17:51, 11 November 2016 (UTC)

IIRC it was given at the 1987 Paris french convention that I've helped organize. The award was for the book and is not a translation award. It's even listed here (the list is not complete). Hauck 18:02, 11 November 2016 (UTC)
Was it a "one off" award? Ahasuerus 18:09, 11 November 2016 (UTC)
No, there are at least three winners. There was some outcry at the time because of the amount given which was thought by some french fans to be indecent (it was nearly twice my yearly income at the time) and because it came from "outside" the genre and thought as tainted by mercantilism. Hauck 18:19, 11 November 2016 (UTC)
Oh, I see. Should we add it to the list of supported award types then? Ahasuerus 18:39, 11 November 2016 (UTC)
To be frank, I'm not found of awards (especially when their attribution process is obscure). But it's just my opinion (such data can be added in the note field). Hauck 18:48, 11 November 2016 (UTC)
Well, the advantage of entering awards as separate award records is that it makes them appear on author-specific award pages like this one, in the Award Directory, on the 7 "Authors/Editors Ranked by Awards and Nominations" lists, etc. It also lets us record additional information about awards, creates structures lists, automatically generated hyperlinks, etc. Ahasuerus 21:03, 11 November 2016 (UTC)


There are a number of magazines that include obituary columns. Some of these record the obituaries as titles and group them as series. I've taken the route of only including these if the people are already on ISFDB. I did notice however, that there is no actual link between the obituary and person. My question is - can anyone confirm there was nothing else to do? The specific example used the person's legal name, rather than their published name. Doug H 18:26, 11 November 2016 (UTC)

There is a Feature Request to "Link from non-fiction works to subject covered" and I hope to get to it at some point. However, it will be relatively time-consuming to implement, so it's low on the list of priorities. Ahasuerus 18:38, 11 November 2016 (UTC)
Something to look forward to then. Thanks. Doug H 22:23, 11 November 2016 (UTC)

To change titles or not

All publications under the variants here have the longer names (and the longer names match so we won't have multiple new variants). The title records have the short names. So - should we equalize on the long names or is that one of the legitimate cases where we should leave the title and pub title mismatched? I vote for equalizing - there aren't so many that the page will become unreadable if the variant titles get a bit longer and it will not be left as a precedent:) Anyone with an opinion? Anniemod 22:00, 15 November 2016 (UTC)

Any opinions? Anniemod 19:41, 17 November 2016 (UTC)
One of the affected pubs has been verified, so I would suggest asking the verifier what he thinks about the proposed change. Ahasuerus 21:41, 17 November 2016 (UTC)
(Since it was sitting at my feet, I verified my third edition to go with the already verified second edition.) The title page has three blocks: (1) Reginald's Science Fiction and Fantasy Awards (2) A Comprehensive Guide to the Awards and Their Winners (3) [in italics] Third Edition, Revised and Expanded.
LCCN link says "Uniform title: Science fiction and fantasy awards; Main title: Reginald's science fiction and fantasy awards : a comprehensive guide to the awards and their winners; Edition: 3rd ed., rev. and expanded". I suggest that we should use at least the LCCN "Main Title". But these multiple editions really are new versions of the book--they differ by much more than what would justify a "variant title" for a fiction book. Chavey 02:11, 18 November 2016 (UTC)

Change proposal - series transliteration

Publication series have transliterations while regular series don't. Any plans to add the fields there as well? And if not, can I propose to add them? (I had been looking at series such as this one that will look at lot cleaner if that second part can be moved to a transliteration... Anniemod 08:59, 16 November 2016 (UTC)

Regular (title) series present unique challenges. There is no "variant series" mechanism, so we stuff all known series names into the same field using slashes as delimiters. For example:
Similarly, we enter translated and transliterated series names in the same field, e.g. とある魔術の禁書目録 / Toaru Majutsu no Index / A Certain Magical Index or Волонтеры Вечности / The Stranger's Woes.
There is a Feature Request to add support for translations. However, I think we need to consider the issue holistically and decide what we want to do with variants, translations and transliterations at the same time. Ahasuerus 16:22, 16 November 2016 (UTC)

Default Size for Kanji and Kana

The default size for Kanji, and Kana are too small to read easily, especially at my age. Would it be possible to set the default a couple of notches larger?--Rkihara 17:55, 19 November 2016 (UTC)

If I understand the question correctly, you'd like to increase the font size of Kanji/Kana characters while keeping the font size used to display Latin, Cyrillic, Greek, etc characters the same. Is that about right? If so, then it should be possible, but it wouldn't be easy. The software would have to examine each and every character that it is about to display in order to determine whether it's Kanji/Kana. Is there an easy way to change font size in your browser? For example, in the Windows versions of Firefox, IE and Chrome all I need to do is press "Control" and then either "-" or "+". Ahasuerus 00:27, 20 November 2016 (UTC)
Sorry, I sometimes think people can read my mind. Most of my reading problems with kanji and kana are in editing, so when I'm entering characters, or transliterating, the small font size makes difficult for me to read. The problem being that the difference between two characters could be a tiny stroke. Maybe a button in the edit mode that enlarges all text, not just kanji, in the editing mode for simplicity? I do use command+ (Mac) to zoom sometimes, but that's inconvenient. It's not that easy for me to read the characters on the biblio pages either, but I can live with that.--Rkihara 08:52, 20 November 2016 (UTC)
Do you use a trackpad? If so, you can use the pinch zooming on it. If you use a mouse, hold down Command or Control and use the scroll wheel. ···日本穣 · 投稿 · Talk to Nihonjoe 18:15, 21 November 2016 (UTC)
Text with mouse zoom is fuzzy, zooming with command+ gives sharper text, but I have to zoom through 2-3 steps. I learned Japanese long after I learned English, and don't use it a lot, so I don't have the instant recognition of kanji as a unit of meaning, as a native reader would. I can read small, blurry, misspelled text in English fairly easily, but I can't do that with kanji or kana.--Rkihara 20:26, 21 November 2016 (UTC)
Hmm...I've never had a problem with it being fuzzy. I wonder why yours does it that way. ···日本穣 · 投稿 · Talk to Nihonjoe 22:20, 21 November 2016 (UTC)
Most browsers let you customize font settings. Perhaps the settings used by Ron's browser result in fuzziness for certain font sizes. Ahasuerus 22:26, 21 November 2016 (UTC)
Played around with it a bit. Apparently I've been doing it wrong. Control-Scroll zooms the screen, Command-Scroll zooms and re-renders the text. A lot better, but the zoom rate is hard to control.--Rkihara 05:37, 22 November 2016 (UTC)
Command-Scroll Zoom behavior on the author edit pages is very erratic, though well behaved elsewhere.--Rkihara 05:41, 22 November 2016 (UTC)
You can adjust how they work by going into the System Preferences | Accessibility | Zoom (I'm assuming you have MacOS, based on the above). ···日本穣 · 投稿 · Talk to Nihonjoe 20:31, 22 November 2016 (UTC)
Also, there is a minor HTML bug on the Author Editor page. It's not visible to the naked eye, but it may be confusing Ron's browser. Let me see if fixing the bug may help "unconfuse" the browser. Ahasuerus 20:51, 22 November 2016 (UTC)
The Author Editor bug has been fixed. Let's see if it helps. Ahasuerus 20:58, 22 November 2016 (UTC)
Zoom with scroll still erratic on author edit page. Zooms sometimes, then doesn't respond to input.--Rkihara 23:14, 22 November 2016 (UTC)
If the browser doesn't respond to input, then it's way beyond anything we can do on the server side. Granted, there are ways to create complex Web pages which will try to exploit browser vulnerabilities and/or install malware on your computer, but our software doesn't do anything remotely like that. Ahasuerus 23:38, 22 November 2016 (UTC)
I see. I can think of 2 approaches that may help. They both involve creating a new User Preference. The first approach would add a User Preference to control the font size of all ISFDB pages, i.e. bibliographic pages, edit pages and moderator pages. The second approach would add a User Preference to control the font size of editable fields on Edit pages. Which one would work best for you? Ahasuerus 19:10, 20 November 2016 (UTC)
I'd say the latter, being able to change the font size of editable fields, though the former might be useful for other people.--Rkihara 07:03, 21 November 2016 (UTC)
OK, I'll go ahead and create a Feature Request then. Ahasuerus 20:33, 21 November 2016 (UTC)
I like this idea, though I just use the Zoom feature which is in pretty much every browser. That way I can zoom in if needed, and then zoom out afterward. ···日本穣 · 投稿 · Talk to Nihonjoe 18:14, 21 November 2016 (UTC)

Publisher page changes

The publisher page has been adjusted. If a publisher:

  • has at least one publication series associated with it, and
  • has publications that are not in a publication series

then the software will display the number of pubs not in a publication series and link to a new Web page called "Publications not in a Pub. Series for Publisher: [publisher name]". See Festa for an example.

If the number of publications not in a publication series exceeds 500, the count will be displayed, but there will be no link to the new page. Instead the software will display a message similar to "8940 publications not in a publication series (too many to display)" -- see Tor for an example. Ahasuerus 23:53, 19 November 2016 (UTC)

I like this change; thanks for implementing!
Now that I've looked at a bunch of publisher records, I'm wondering if it might make sense to either place this at the top of the list (above the "Publication Series" header) or to leave it at the bottom but without a bullet and maybe a little white space to help differentiate it? It tends to blend into the list, especially when it is a link. Albinoflea 07:49, 22 November 2016 (UTC)
That's a good point. I have moved the link to the top of the Publication Series section. How does look now? Ahasuerus 21:50, 22 November 2016 (UTC)
Now that I have re-read your comment, I see that I misread it. You wrote "place this ... above the "Publication Series" header" while I read "place this ... below the "Publication Series" header". Sorry about that! However, before I do anything else, could you please review the current layout to see how well it works? Ahasuerus 21:53, 22 November 2016 (UTC)
No worries, I think I do prefer it at the top, and maybe it doesn't need to be above the "Publication Series" header, maybe just remove it from that unordered list so it stands out as being different... so instead of:
Publication series:
  • 867 publications not in a publication series (too many to display)
  • Alpha Science Fiction
  • Archway
something more like:
Publication series:
867 publications not in a publication series (too many to display)
  • Alpha Science Fiction
  • Archway
Albinoflea 21:15, 23 November 2016 (UTC)
It so happens that I was coding the latest patch while you were typing your response :)
After sleeping on it, I decided to make the link available on all publisher pages, the idea being that it can be useful to get an overview of a publisher's activity. Prior to this change, if a publisher published 10 books in 10 different years, you had to pull up 10 pages to see everything. Now you can see all of them on the same page, nicely sorted by year.
The latest patch also implemented additional quality of life enhancements: grammar fixes, a link from the new page back to the publisher page, and better sorting on the "publisher year" page. Ahasuerus 21:48, 23 November 2016 (UTC)
Will the patch take a while to implement? I am not seeing it yet (for example, on Tor). ···日本穣 · 投稿 · Talk to Nihonjoe 22:37, 23 November 2016 (UTC)
That's odd. I see "8941 publications not in a publication series (too many to display on one page)" between the line which says "Note: Other imprints: Starscape" and "Publication series:". Are you seeing something different? Ahasuerus 23:07, 23 November 2016 (UTC)
Oops, missed that. Is there a way to view 100-200 at a time? Basically show what it would if there weren't too many publications to display on one page? ···日本穣 · 投稿 · Talk to Nihonjoe 23:16, 23 November 2016 (UTC)
A timely question indeed! I have been slowly consolidating the logic behind our publication tables to ensure that they all had the same columns, headers, etc. I have one more table to consolidate, then I should be able to review/enhance the way paging is done. Unless I ran into unexpected issues, I should be able to add paging to this publication table as well. Ahasuerus 23:50, 23 November 2016 (UTC)
Awesome! ···日本穣 · 投稿 · Talk to Nihonjoe 00:04, 24 November 2016 (UTC)

(unindent) Looking good! Thanks! I find myself missing the "show covers" link but hopefully that will be making a reappearance if you're consolidating the display logic. Albinoflea 17:06, 24 November 2016 (UTC)

Could you please provide more details? I don't think I am aware of any recent changes to "Show Covers". There are two User Preferences, "Display cover images on Title pages" and "Display cover scan indicators on Title pages", which, AFAIK, are working correctly. Although, now that I am thinking of it, the latter should be clarified to add that it also controls whether "cover scan indicators" appear on search pages. Ahasuerus 17:42, 24 November 2016 (UTC)
"Display cover scan indicators on Title pages" has been changed to "Display cover scan indicators on Title and search pages". Ahasuerus 18:03, 24 November 2016 (UTC)
Ah, sorry, what I was trying to say was the newly generated pubs_not_in_series.cgi pages don't have the "show covers" links but pubseries.cgi pages have a full compliment of links: Show last year first • Sort by series number • Show covers
Clearly the "Sort by series number" link makes no sense on pubs_not_in_series.cgi but the other two links probably would serve a purpose; the first link might be changed to "show most recent year first" if the default sort order is already set to show last year first.Albinoflea 07:15, 27 November 2016 (UTC)
Oh, I see. Sure, FR 952 has been created. Ahasuerus 22:47, 27 November 2016 (UTC)

Patch - punctuation in author names

FYI, a new patch was installed a few minutes ago. It tweaked the way punctuation (apostrophes etc) is handled in author names. The changes were not supposed to affect user-experienced software behavior in any way. If you see anything unusual, please report your findings here. Ahasuerus 23:36, 20 November 2016 (UTC)

My Pending Edits changes

"My Pending Edits" has been changed to show the name of the moderator who has your submission(s) on hold. In addition, column names have been adjusted to be more accurate. Ahasuerus 00:36, 21 November 2016 (UTC)

Title/Publication Type values in Advanced Search

Advanced Search has been modified to treat Title Type and Publication Type values in a case-insensitive manner. For example, an Advanced Title Search on "Title Type is exactly Poem", which used to result in an error message, now produces the same results as an Advanced Title Search on "Title Type is exactly POEM". Ahasuerus 21:03, 21 November 2016 (UTC)

"Suspected Duplicate Authors" now available to all editors

The cleanup report "Suspected Duplicate Authors" is now available to all editors. The ability to "ignore" suspected duplicate pairs remains restricted. Ahasuerus 22:38, 21 November 2016 (UTC)

Gilbert Morris?

Would anyone be interested in fleshing out Gilbert Morris's biblio? According to Wikipedia, he wrote or co-wrote around 200 books prior to his death earlier this year. Only a few of his series were SF, so we'll have to be careful. Ahasuerus 01:54, 22 November 2016 (UTC)

Wiki cleanup - magazines

Most of the remaining magazines in the wiki cleanup section are like this one - series exists but not all magazines that are known to exist are added as stubs (or full records) -- although more than what is marked in the wiki is added in this case, there are still a lot of missing ones (see the grid). Should I add them as stubs (so someone can add content one day if they can find it) or just leave them as they are? I lean towards adding the stubs so at least we have the records but wanted to ask for the community opinions :) Anniemod 18:04, 22 November 2016 (UTC)

I think creating stubs is the better approach. It will preserve the data that we already have and make issue grids more useful/complete. Ahasuerus 18:29, 22 November 2016 (UTC)
That's my thinking as well (and that is what I had been doing with the fanzines with no series) but decided to check if that may be an overkill. I am seeing a lot of NewPub submissions in my future then :) Anyone opposing? Just checking before I start creating them. Anniemod 18:53, 22 November 2016 (UTC)
I'm presuming that this cleanup report is a moderator report. Doug H 16:09, 23 November 2016 (UTC)
I believe it is available to all users. See [1]. -- JLaTondre (talk) 17:32, 23 November 2016 (UTC)
Wiki cleanup reports (see the "Wiki Cleanup" section) are available to all editors. However, only moderators can delete Wiki pages. Once an editor transfers all relevant data from a Wiki page to the database, the next step is to add Template:Page transferred to the top of the page and wait for a moderator to delete it. Ahasuerus 17:35, 23 November 2016 (UTC)
Considering that I am not a moderator, I can confirm that the wiki cleanup reports are visible to all editors :) Anniemod 18:04, 23 November 2016 (UTC)

Patch: Unicode characters in URLs

The software has been changed to allow Unicode characters in URLs. In the past you had to use their encoded values in order to get past the validation logic. With the latest change the only remaining requirement is to have "http" or "https" as the first characters of the URL. Ahasuerus 23:04, 22 November 2016 (UTC)

Awesome! ···日本穣 · 投稿 · Talk to Nihonjoe 19:43, 23 November 2016 (UTC)

"Other" pub format - an example

Just an interesting example of what 'other' would be used for in a publication. From a review of four titles:

"These four "classics" of the genre are available once again in ASCII format on either 5.25" or 3.5" diskettes".
Doug H 16:03, 23 November 2016 (UTC)

Eliminating "sf"/"shortfiction" during editing

FR 937 "Eliminate 'sf'/shortfiction dropdown choice", which I created a couple of months ago, reads, in part:

In the Content section of NewPub/EditPub/AddPub/ClonePub forms, the "storylen" dropdown list is typically limited to four choices: "-", "shortstory", "novella", "novelette". However, sometimes a fifth choice, "shortfiction", appears. Also, when "-" is filed for a SHORTFICTION title, the software puts "sf" in the "storylen" field. [The 'sf' code and the 'shortfiction' value are unnecessary during editing because] the software already displays the word "shortfiction" for SHORTFICTION titles without a storylen code.

I plan to change the software to eliminate all appearances of "sf" and "shortfiction" in drop-down lists. The database will be simultaneously updated to remove all occurrences of "sf" from the "story length" field. Short fiction titles without a story length code will continue to be displayed as "shortfiction", so nothing will change as far as our users are concerned.

If there is anything that I may be overlooking and that may cause issues, please post your concerns here. Ahasuerus 00:13, 24 November 2016 (UTC)

Author Directory changes

Author Directory has been enhanced in a couple of ways:

  • The main table no longer displays invalid entries for transliterated author names
  • Lower level pages use the same table layout and paging/merge logic as Advanced Author Search

In addition, all Advanced Author Searches have been enhanced to use canonical names as secondary sort values. For example, if you request that author records be sorted by "family name", Advanced Author Search will sort records by "family name" and then sort them by canonical name within each "family name". Ahasuerus 23:08, 24 November 2016 (UTC)

2016-11-27 downtime

The server will be unavailable between 6pm and 6:05pm server (US Eastern) time. Ahasuerus 22:26, 27 November 2016 (UTC)

The server is back up. 'sf'/'shortfiction' is no longer used in "story length" field. I will be checking Help shortly to make sure that there are no references to it. SHORTFICTION titles without a "story length" value will continue to be displayed as "shortfiction". Ahasuerus 23:02, 27 November 2016 (UTC)
All references to 'sf' have been removed from Template:TitleFields:Length. I also rewrote the text for readability and clarity. It would be great if more eyeballs could review the changes. Ahasuerus 23:41, 27 November 2016 (UTC)
Shortstory should really be 2 words. Anniemod 03:27, 28 November 2016 (UTC)
Well, that's the way it's been displayed on all pages since day one. We'll need to change the software before we change Help :-) Ahasuerus 04:03, 28 November 2016 (UTC)
If we are trying to match the software exactly, then all 3 should start with small letters then as they are in the dropdown :) Anniemod 17:22, 28 November 2016 (UTC)
Good point, updated. Ahasuerus 18:57, 28 November 2016 (UTC)
Two more format things: Is the double star by design in front of the short fiction categories? It is inconsistent this way. And shouldn't shortstory/novella/novelette also be with small letters to match the drop down when you add a short story? Anniemod 19:08, 28 November 2016 (UTC)
Updated, thanks again! Ahasuerus 20:15, 28 November 2016 (UTC)
Looks good. :) Anniemod 21:52, 28 November 2016 (UTC)
And the omnibus introductory sentence took me a bit to understand. Maybe move it before the "other types" so it goes: Short fiction (start on a new line for that - so the first line is just the explanation of what it is), omnibus, others. And I would say "For omnibus, the value should be built in the following way" or something along these lines. Anniemod 03:27, 28 November 2016 (UTC)
Updated, thanks! Ahasuerus 04:03, 28 November 2016 (UTC)
Looks great! Anniemod 17:22, 28 November 2016 (UTC)

Wiki cleanup - publications - opinions needed

A lot of the wiki pages for publications look like this. Some of them even have pictures of the logos and whatnot. What should be done with them:

  • Add all the information from them to the publication notes (except when there are pictures - for them only the second option seems viable)
  • Link the publication to the wiki page in the notes (because publications do not have a handy place to add a link as all other types of records do)?
  • Something else?

Opinions? Anniemod 23:31, 28 November 2016 (UTC)

I believe almost all of these pages were created by Marc Kupper back in 2008. The last time we discussed them, Marc saw value in preserving the data. It may be too elaborate and detailed for the main database record, so I would be inclined to keep these Wiki pages "as is" and link to them from the Note field. Ahasuerus 02:15, 29 November 2016 (UTC)
Links it is then. With notifications to the PVs that I am adding them (the ones that want notifications for notes anyway). Any chance to add "Link" fields into Publications at some point as we have it for everything else? (it will be really useful for linking publisher pages, online versions of magazines, pages on other sites (Fantlib, SFBG and so on) and whatnot? Anniemod 02:22, 29 November 2016 (UTC)
At one point I proposed adding the standard repeatable "Web Pages" field to publication records. At the time, there was no consensus re: its desirability, but please feel free to propose it again if you believe that the functionality would be desirable. Ahasuerus 02:43, 29 November 2016 (UTC)
I just did, didn't I? :) Especially now with removing the wiki connection so you cannot stick your links there? Or do you want me to start a new topic here? Anniemod 03:09, 29 November 2016 (UTC)
I think it would be better to start a new topic to make sure that all editors see it. Ahasuerus 04:00, 29 November 2016 (UTC)

A new cleanup report -- "Empty HREFs in Publication Notes"

A new cleanup report, Empty HREFs in Publication Notes, has been deployed. Once the nightly process runs, all editors will be able view the data. As of this morning there were 28 publications with empty HREF attributes in Notes. Ahasuerus 02:12, 29 November 2016 (UTC)

Corrected. Chavey 12:53, 29 November 2016 (UTC)
Thanks! Ahasuerus 15:57, 29 November 2016 (UTC)

Change proposal - add links fields to publications

As we are removing the wiki connections and more and more specific editions have their own web pages (at publisher's websites), not to mention magazines and fanzines, it seems to make sense to have the Web Page fields on publications the same way we have them on pretty much anything else. Any opinions? Anniemod 04:17, 29 November 2016 (UTC)

I think that would be a good idea. ···日本穣 · 投稿 · Talk to Nihonjoe 06:16, 29 November 2016 (UTC)
I don't know if this would be worthwile: publishers do fold or revamp / migrate their websites, meaning that the links will become obsolete after some time. Stonecreek 07:12, 29 November 2016 (UTC)
True - but there are stable links also - FantLab for Russian publications, SFBG for Bulgarian ones, (more for different languages), our own Wiki if there are details that are too much for the details page. And some publishers do keep stable links for years. Then there are magazines and fanzines archived all over the place - now the only option for all of these is the notes field (and the title record is not an option because it is on the year level). If we are worried about links that move/stop existing, we might as well remove all links from all kind of entries - we just went through cleanup of IMDB links for example because they changed their format. Just thinking aloud. Anniemod 07:29, 29 November 2016 (UTC)
Yes, but it will still be a problem to maintain the links: if we link one publication we can't exclude a link submitted by anyone. This would end in a pandemonium of unsupported links. Plus: I have experienced the collapse of sites that I personally would have thought stable. Stonecreek 07:56, 29 November 2016 (UTC)
How is that different from having links for Publishers and Titles? We already have those. And having them won't mean "always use them" - the ones on the titles are not used that much either. Anniemod 08:00, 29 November 2016 (UTC)
I often run into dead links, which I can then use to go to the Wayback Machine to see what used to be there. So having a specific dead URL is still a very useful piece of information. And maybe someday we'll even have an "automated system" to convert any such dead link into a Wayback link. Chavey 08:15, 29 November 2016 (UTC)
It is the sheer amount of links that would have to be maintained that's different (if the links aren't used that often why bother to add them anyway?). Maybe it'd be better to automatically add a link to a publisher's site, provided we have one and it'd be easily installed. Stonecreek 09:39, 29 November 2016 (UTC)
Why do you expect more links than we have for titles now? Most of the added links will be there because it contains information and/or is built for this specific publication. Publisher site was just an example above. I am in the process of adding a lot of links to our Wiki pages into publications (as part of the wiki cleanup - too much data to move into the notes, so it needs to stay in the wiki and to be linked) - which made me wonder why we do not have links here. Same with the fanzines and magazines I had been working on - all links need to go into the Notes field now - which is inconsistent with any other record we have here. Anniemod 09:50, 29 November 2016 (UTC)
The title links are mostly to wikipedia, which hopefully is much more stable than any other site maintained by a publisher. I'd think that one of the reasons that we don't supply links to individual publications is the instability of other sites. Stonecreek 10:57, 29 November 2016 (UTC)
I think it would be a nice feature for (perma)links to LCCN, OCLC, National Libraries etc. We now have thousands of links in the notefield. --Willem 11:17, 29 November 2016 (UTC)
This'd be ideally dealt with in the tools (found on left side), as is already for OCLC and others. Stonecreek 11:40, 29 November 2016 (UTC)
There is a Feature Request to add support for "external identifiers" which will be linked to by number (BLIC, LCCN, ASIN, OCLC, etc.) Fixer has been begging me to implement this functionality for some time now, so I will probably get to it sooner rather than later.
Having said that, there will always be unstructured links to third party sites which will require full URLs instead of numbers. I also agree with Darrah re: the Wayback Machine -- a dead link is often more useful than no link at all. We already capture third party URLs in Notes, so it's just a question of how we want to organize them; I find that a standard repeatable field is easier to use and more maintainable than HTML in Notes. Ahasuerus 13:59, 29 November 2016 (UTC)
FR 957 has been created. It mentions that "[i]t is suggested that this functionality should be implemented in conjunction with the "Support for external identifiers" FR in order to minimaize the amount of rework that will be required when moving links from Notes to the newly implemented fields." Ahasuerus 03:31, 1 December 2016 (UTC)

Publication Series with not all titles linked

Hi, I'm not in a position to do editing for ISFDB right now, but this may be something one of you would like to change:

[Series: Fiction River]

[unlinked titles]

Special Edition: Crime and Alchemy and Steam should probably be linked to the main series?

Also Pulse Pounders is linked to the main series with the Publication Series name tacked on the front; the rest of the books in the series do not have that.

Thanks, JJ 05:17, 29 November 2016 (UTC)

I've submitted updates and corrections for everything. ···日本穣 · 投稿 · Talk to Nihonjoe 06:16, 29 November 2016 (UTC)

Should these be united under a global "Fiction River" universe?

JJ 23:40, 22 December 2016 (UTC)

"Clear Form" buttons?

Does anyone use the "Clear Form" buttons that some data entry forms display at the bottom of the page? I have only clicked them a few times when I meant to click "Submit Changed Data" instead. Needless to say, the results were not what I wanted.

If my experience is representative, we may want to remove these buttons to help minimize editor frustration. Ahasuerus 18:07, 29 November 2016 (UTC)

We have "Clear Form" buttons somewhere? That should tell you how much I will miss them. Although that may explain something happening one of the evening when I though that my browser is getting crazy on me and reloading and losing content... :) Anniemod 19:19, 29 November 2016 (UTC)
Title Editor, Publication Editor, Add Variant Title, and Add Publication support this functionality. Ahasuerus 19:24, 29 November 2016 (UTC)
The first two explain my "browser" issues :) I tend to not even look at the buttons on those two - so even if I saw the clear buttons, I never used them and I had forgotten all about them. I do not think I had ever used Add Variant (I still cannot figure out a usecase where it helps reduce the number of edits) or Add Publication(I just clone and clean the differences even when there are a lot). Anniemod 20:07, 29 November 2016 (UTC)
I can't really imagine ever using "Clear Form" either. --Vasha 20:57, 29 November 2016 (UTC)
I've never used it, but I always assumed it cleared the form which I never imagined a use for. Having just tried it, if kept, it should be renamed to "Reset Form" which is a clearer description of what it does. I have had a use for that, but canceling the edit and starting over (which I've done in the past) is just as easy for as infrequently as it is needed. I'd be fine if it's removed. -- JLaTondre (talk) 23:27, 29 November 2016 (UTC)
Thanks, FR 956 has been created. Ahasuerus 02:57, 1 December 2016 (UTC)
Done. Ahasuerus 03:04, 1 December 2016 (UTC)

Changing "Non-Genre" and "Graphic Format" to checkboxes?

The Title Editor page contains four fields which let editors select a value from a drop-down list. Two of them, "Language" and "Title Type", are OK as is. The other two, "Non-Genre" and "Graphic Format", have drop-down lists which are limited to "No" and "Yes". I propose changing them to check-boxes. They would be similar to what we use on the User Preferences page and hopefully make the page cleaner, especially if we add more "Yes/No" choices in the future. For example, the way the software works right now, you can't use "jvn" or "nvz" in the "Storylen" field if the title type is SHORTFICTION. There is a feature request to create additional fields for these values, which will increase the number of "Yes/No" drop-down boxes. Ahasuerus 18:23, 29 November 2016 (UTC)

Yes please. :) I would also say that "jvn" and "novelization" probably should also be checkboxes that may populate length if we really want them to but are clearer on their own. Anniemod 18:49, 29 November 2016 (UTC)
I like this idea. ···日本穣 · 投稿 · Talk to Nihonjoe 19:15, 29 November 2016 (UTC)
I agree on all counts. --Vasha 21:17, 29 November 2016 (UTC)
Agreed. -- JLaTondre (talk) 23:52, 29 November 2016 (UTC)
Done. Unfortunately, the way most browsers treat disabled HTML check-boxes is somewhat iffy. You can see the odd coloration if you try to edit an INTERIORART or COVERART title. The "Graphic Format" check-box will be disabled, but the difference may not be apparent. Ahasuerus 01:15, 1 December 2016 (UTC)
It turns out that I forgot to update the "New Publication" form, which caused all NewPub submissions to set these flags. The bug has been fixed and the data has been corrected. Sorry about that! Ahasuerus 13:47, 1 December 2016 (UTC)
Speaking of drop-down lists, how come the options for story lengths are in the order novella-shortstory-novelette? shortstory is the commonest and should be at the top. --Vasha 06:11, 1 December 2016 (UTC)
I am not sure whether the current order, which was implemented in 2004 when I was on hiatus, was an attempt at alphabetizing the "story length" values or whether it was random. Once the code has been rewritten and streamlined, it should be easy to change the order. Ahasuerus 13:46, 1 December 2016 (UTC)
That'll be good, thanks. --Vasha 16:00, 1 December 2016 (UTC)

(unindent) The help pages for those two need changing -- they still reference dropdown boxes. Anniemod 05:59, 3 December 2016 (UTC)

Updated, thanks. Ahasuerus 23:58, 3 December 2016 (UTC)

Splitting edit forms?

Some of our data entry forms let you do multiple things. It seemed like a good idea at the time, but it's been my experience that it can confuse new editors. Here are the main offenders:

  • "Make/Remove a Pseudonym" lets you:
    • create a new pseudonym, OR
    • delete an existing one
  • "Make This Title a Variant Title or Pseudonymous Work" lets you:
    • turn the current title into a variant of an existing titles, OR
    • turn the current title into a variant of a new title, OR
    • break the current variant relationship
  • "Import Content" lets you:
    • import all titles from another publication, OR
    • import one or more explicitly specified titles

The advantage of the current system is that it limits the number of options in the navigation bar on the left. However, its disadvantages may outweigh its advantages. I am too used to the current system to be a good judge of how intuitive it is, so I'd like to ask other editors, especially relatively recent editors, whether the current system was confusing when they started editing here. Ahasuerus 18:39, 29 November 2016 (UTC)

I am not sure I agree with you about Import Content. It kinda makes sense to stay together - the page is small and clear enough and the two options are pretty clearly separated. If anything, that is one of the most intuitive actions around here -- once you remember it exists :) Anniemod 18:53, 29 November 2016 (UTC)
What prompted me to ask this question was a fairly experienced editor recently mentioning how tedious it was to import individual titles into a publication. It turned out that he didn't realize what the bottom half of the page was for. Ahasuerus 19:02, 29 November 2016 (UTC)
Ah, I do not remember if someone pointed it to me or if I figured it out but I still think it is one of the most intuitive pages for me. I never get confused on that one (some of the others still require me to think a bit on what I am trying to do really). Your mileage can vary of course. Anniemod 19:17, 29 November 2016 (UTC)
The Make/Remove ones, I agree, can get a bit confusing - not to mention that the design of the Make pseudonym that allows ID or a name is confusing as well (ask me how often I was pasting the id in the wrong field or how long it took me to realize that if I have the ID, I do not need the name as well? Finding an ID is easy and fast - why do we have the ability to use a name? :) . Anniemod 18:53, 29 November 2016 (UTC)
I believe you are the first editor to say "Finding an ID is easy and fast - why do we have the ability to use a name?" Everyone else who has commented on this data entry form over the years said something like "I don't think I have ever used author IDs. Why do we support them here?" :-) Ahasuerus 18:59, 29 November 2016 (UTC)
Well... with the DB becoming international, it should be self explanatory now why we need them. :) I am surprised though - the system is complicated enough without the possibility of a misspelling to attempt to create a wrong pseudonym. But then it may be my IT background kicking in and demanding IDs. Or the years of editing non-latin DBs and sites :) Anniemod 19:17, 29 November 2016 (UTC)
As a recent editor, I didn't know until this moment that "Make/Remove a Pseudonym" existed. One of my biggest sources of learning has been simply finding all the things I could and should do. For that reason, I think having a long, long list of similarly-named tasks would be a hindrance; I'd rather have them grouped together, similar tasks combined in one page.
I agree that "Import Content" is good and would only be weakened by putting the two functions on separate pages. As for varianting, there's a case to be made for separating the making/breaking of relationships between existing titles, on the one hand, and creating a new variant title, on the other hand. I find that the filled in form at the bottom of that page attracts my eye and draws me to click on the button underneath it even when I entered an ID in the top section. I've done that a zillion times, though not recently (well, only once recently...) I actually think that page is confusing. Maybe a two-stage process? Enter the ID you want to link to; if there is none, "click here" to bring up the form for a new one. --Vasha 19:32, 29 November 2016 (UTC)
Instead of new pages, perhaps look at re-laying out the existing ones. Import content, in particular, would be a good case for this. Separate the sections with a bit more space and add a title to each section to make it clearer what each one is for. -- JLaTondre (talk) 23:59, 29 November 2016 (UTC)
Thanks, FR 955 has been created. Ahasuerus 01:27, 1 December 2016 (UTC)

Link to pending edits should be on all pages?

While you're discussing changing the interface, can I request that the links to Pending and Recent Edits be on all pages, just as My Messages is? Sometimes I submit a change and I immediately want to go cancel it, or check that it isn't a duplicate of one I previously submitted; so the link to Pending ought to be on the "submitted" page, and for that matter, everywhere. --Vasha 02:08, 30 November 2016 (UTC)

I agree. It would be nice to always have the "Logged In As" section visible (with all the user-specific "My xxx" links in it). After submitting anything, a person has to go to the Home Page to get that section to reappear. ···日本穣 · 投稿 · Talk to Nihonjoe 18:25, 30 November 2016 (UTC)
Agree. And the "Cleanup Reports" link. It is very annoying to try to get back to it after a submission when I forget to open a new tab or when I am working from my phone :) Anniemod 18:58, 30 November 2016 (UTC)
Thanks for the feedback! There are some limitations due to the way the software is structured, but what you are describing should be doable. FR 953 has been created. Ahasuerus 19:28, 30 November 2016 (UTC)
The requested changes have been made. The "My..." links are now displayed on all pages. "Cleanup Reports" has been moved to the bottom of the "Other Pages" section and should also appear on all pages. Ahasuerus 23:39, 30 November 2016 (UTC)
Speedy delivery! Thanks! ···日本穣 · 投稿 · Talk to Nihonjoe 23:44, 30 November 2016 (UTC)

L. M. Montgomery's novels?

We list two novels written by the famous children's author Lucy Maud Montgomery who is usually credited as "L. M. Montgomery". As near as I can tell, one of the novels, Jane of Lantern Hill, contains no speculative elements. The other one, Magic for Marigold, apparently includes references to "childhood magic", basically daydreaming.

Would anyone with first hand knowledge of these novels be able to clarify whether they have any speculative elements? If not, then we may want to delete them since she is "below the threshold". We may also want to change her canonical name to "L. M. Montgomery". Ahasuerus 23:31, 30 November 2016 (UTC)

Magic for Marigold has an imaginary friend and a lot of daydreams but there is no way to stretch it to be able to call it a fairy tale, let alone speculative fiction. Jane of Lantern Hill doesn't even have that - no speculative elements there (I don't even remember daydreaming on any scale). And I agree that the canonical name should be the L. M. variant - it is actually considered her pen name :) Anniemod 23:43, 30 November 2016 (UTC)
Looking through the links on L. M. Montgomery's page led me to the anthology Saved by the Belle, a book of stories about women that has exactly one speculative entry ("The Shell of Sense" by Olivia Howard Dunbar); although a couple others are otherwise in the DB due to appearing in anthologies of adventure and suspense. Should I remove all the stories except "The Shell of Sense" from the contents? --Vasha 01:03, 1 December 2016 (UTC)
Checking submission history, I see that this collection was fleshed out by Darrah Chavey. We'll need to check with him to see what he knows about it. Ahasuerus 01:34, 1 December 2016 (UTC)

Name for the "Contents notes for omnibus titles" field?

I am ready to start working on FR 163, "Create new Title fields to move storylen values to". I am not going to add support for "suggested age range" for juvenile titles, at least at this time, but otherwise I plan to do exactly what the FR requested. We will have three new fields for:

  • novelizations (a check-box)
  • juvenile titles (a check-box)
  • contents notes for omnibus titles

The two check-boxes will have self-explanatory names, but I am not sure what we want to call the "contents notes for omnibus titles" field. Once a field name has been established, it's somewhat time-consuming to change it, so it would be nice to get it right the first time. Not only should the name be clear, but ideally it should also be flexible in case we decide to expand its use later, e.g. to cover collections. Any ideas? Ahasuerus 03:26, 1 December 2016 (UTC)

Why not just call it Content(s)? Or Content Expression if we want it more descriptive? Anniemod 05:48, 1 December 2016 (UTC)
OK, "Contents" may work. It's the same as the name of the "Contents" section in New/Edit/Clone/Add Publication forms, but I doubt it will cause problems. Ahasuerus 13:42, 1 December 2016 (UTC)

Novel contents: is there a way to...

I have been entering the volumes of Carlos Fuentes's Obras reunidas (Collected Works), most of which are omnibuses. One of them, Volume VI, though, is a single novel, accompanied by copious editorial materials, and here's the problem: The title of the volume is Obras reunidas VI: Nuevo mundo narrativo. Terra nostra. The title of the novel, which begins on page 29, is Terra nostra. Given that the record has the title type NOVEL, is there any way I can get the Contents to display correctly? Currently the NOVEL listed in the contents has the title of the volume. It probably wouldn't be a good idea to import a record for the novel alone with the correct title... --Vasha 16:54, 1 December 2016 (UTC)

The way you had done it is how it is usually done - see Complete Writings of Oscar Wilde, vol. X: Dorian Gray for example. The Variant connection shows the actual name so it looks fine to me. Adding notes in the publication is probably your best bet :) Anniemod 17:31, 1 December 2016 (UTC)
Yeah, you're right, the way it is now is 100% comprehensible, so that's fine... it just offends my sense of order and aesthetics! :-) --Vasha 17:47, 1 December 2016 (UTC)
Yeah, I know. :) Now - technically you can import the properly named novel and remove the one with the ugly name. But that will make the publication and title titles different and as the way it looks now is clear enough, I am not sure that it is worth it. Plus it will show on the report for the mismatch and more likely than not I will come to your page to ask you which one you like more because the difference is not worth the exception. :) By the way - you may want to edit your publication to put the page number for the novel into the entry - you cannot do that when you are adding but once it is accepted, you can (so it goes in the proper place in the table of content).:) Anniemod 18:03, 1 December 2016 (UTC)
If I merged this with the title Terra nostra instead of varianting, I would get the result I wanted at the cost of the title and publication records having different titles. As you said, that is one of those things that is "just not done" so I think better not to. Is there some technical thing that this would mess up, too? I have seen some titles that do have publications with somewhat different titles, usually versions with and without a subtitle. --Vasha 18:17, 1 December 2016 (UTC)
Yes, result will be the same (and with less edits) :) No, it won't break anything, it will just show on one of the cleanup reports(that I had been working very hard to clear - thus my half joke above about me coming to ask you which one you want to keep:) ). But it can be ignored by a moderator in case there is a legitimate reason for the difference - at which case it won't be a problem at all. Now - subtitles are a special case - there is a suppression of the flags when subtitles are involved (that is why there is a standard for the subtitle position and so on - so they can be recognized.) If you have two different subtitles on the pub and title, it gets flagged but when one of them has none. Not consistently implemented - in some cases we have the title without allowing all pubs to stay under it cleanly; in some we have variants for US and UK separate ones. It comes down to conventions, preferences and timing I guess. Anniemod 18:32, 1 December 2016 (UTC)
By the way - if you really think that it will be better to have the novel name properly set inside (and considering that you have the content inside, it does make sense), just submit the merge. And when tomorrow it shows on the report, I will remember to ask a moderator to ignore it :) Anniemod 18:36, 1 December 2016 (UTC)
I will wait until tomorrow to see if anyone else objects. If they don't, I'll submit the merge. --Vasha 18:45, 1 December 2016 (UTC)
On second thought, no I won't. Better to have a separate title record for that volume so (for example) it can be found on a search of titles. --Vasha 19:01, 1 December 2016 (UTC)
Every action has a side effect :) Anniemod 21:17, 1 December 2016 (UTC)

Bug with REVIEWs/INTERVIEWs and non-genre/graphic format check-boxes

There is a bug in the software that handles REVIEWs/INTERVIEWs and non-genre/graphic format check-boxes. I expect to fix it shortly. I will also fix the affected REVIEW/INTERVIEW titles. Ahasuerus 20:26, 1 December 2016 (UTC)

The software and the data have been fixed. Ahasuerus 20:45, 1 December 2016 (UTC)

Fixer's Future

(moved from the Help Desk)

Between the increased number of SF books being published (how dare they!) and Fixer's data acquisition logic becoming better, my workload has jumped significantly in the last few years.

Fixer used to find just a few thousand new ISBNs per month and it took me a week+ to process them. Now he finds anywhere between 6,000 and 8,000 ISBNs depending on the month. It takes me more than 2 weeks to process them, sometimes closer to 3 weeks. The workload is becoming unmanageable and I am not getting any younger either (8-10 hours a day is pretty much my limit now.)

We'll need to do something about this issue if we want development to continue at a reasonable rate. I have been toying with different procedural and software-based solutions, but they all have different flaws. I am kind of hoping that your work on publishers may serve as a template for a more permanent solution. Ahasuerus 01:51, 1 December 2016 (UTC)

Single threaded processes are bad processes. That relies on people being available monthly - and despite good intentions, life happens. So how about delegating part of the work in a slightly different way? I do understand that Fixer itself cannot be delegated but why not delegate some of the manual checks and then have Fixer (or another robot) do the Add/ClonePub records. My understanding is that every record that Fixer adds is manually inspected to verify that it is in scope and to fix the things he is not good at (determining type for example). So enlist the rest of the editors to help with the manual checks: once Fixer has its catch, get it to parse and populate tables/sections (will explain the plural in a second) with all needed elements for Clone/Add and post them so people can help fill the missing details and get the "is it a real book and is it ours" check. Add a last column "ready for adding" and get the robot to add based on that column
This way you eliminate the longest work by far in working through a publisher/list of ISBNs from the editors' tasks and you have data that can be added automatically. Editors will mark the book as chapbook, novel, collection or whatever it is properly, mark the cover artist if known and so on - as long as the table shows the empty spaces where Fixer does not know and have the data populated for the ones it can so they can be verified, that should be easy enough. Different types will need somewhat different follow-up work - novels and non-fiction can go directly (no content to add after that). Same is valid for chapbooks that have only one story and the name of the chapbook and story are the same. Collections/omnibus/anthologies and annoying chapbooks where the story has a different name are all special cases but they can be handled the old way or something new can start getting developed. Even if just the novels and the easy chapbooks and the non-fiction gets handled this way, it will eat a chunk of the work.
The multiple tables/sections are to allow different editors to claim a section so they can work on it without stopping on each other's toes (and we have the nice way to sign in wiki). If you have time and want to process the old way, claim the group for you and do it - or do not post all of them to start with.
Am I missing something that will make that non-workable? Anniemod 02:41, 1 December 2016 (UTC)
I have considered using Wiki pages to facilitate Fixer processing, but before we get any further, let me post a link to this explanation of how Fixer works and another link to the associated moderator Help section. Once we are on the same page, we can discuss which steps could be adjusted/farmed out to other editors.
I should also point out that the original plan was that I would process Fixer's queues and supervise submission creation, leaving submission approval to other moderators. Unfortunately, experience shows that Fixer's submissions require a significant amount of manual TLC, which moderators are not always in a position to provide. You have to research misspelled author names, omitted/variant publishers/imprints, series, publication series, bindings (is it a pb or a tp?) and so on. Moderators are busy people and in many cases they don't have the time needed to investigate these issues thoroughly. Ahasuerus 00:29, 2 December 2016 (UTC)
After we started talking about Fixer, I went and read all I could find (including those two). I just reread them again though and I think we are on the same page (except for non-documented stuff I do not know :) ).
That's exactly my point though - make the robot (Fixer or Fixer2.0 or whatever) submissions easy to moderate. If you farm the verification, data cleanup and preparation out, then moderators can approve the submissions from the robot with no major verifications -- the problematic ones that cannot be found won't show up and if a book need something special to be found, the notes for it can be in the wiki as well. Include the cleaning editor name if you want so if there is a bad batch of data, they are foundable after the first few bad ones and help the editor understand what they are doing wrong. It will take awhile at the start I suspect but it will get streamlined sooner or later. And allowing a data cleaner to add a flag saying "manual/more details needed" can exclude anything that needs to be added but is just too... complicated for the automation or that they cannot figure out how to clean properly. I have no idea how many of the 6000/8000 are "easy" short of basic cleanup but if it is more than half, I would be surprised. All those checks already happen - except that you are the only one doing them...
Let me give you an example - when I am adding a book, I need to find its Amazon cover link (click on it, clean the field, run it in a browser window to be sure I did not misspell it and then paste in the field. For the Publisher - copy the string if I do not know it already, search in the DB to see if we have a record, get the proper name. And so on. If the list posted for cleanup uses automation to pull the picture URL so you need just to verify, pull the publisher (and flag the ones that do not exist in ISFDB?) and so on - basically fill as much as it can in the tables, it becomes a verification work with less copy/paste unless if it is needed... And if the parser does not work properly and dumps all in one field? Well - it can be moved around from whoever is clearing for addition.
I am not even trying to find a way to automate everything -- it is about getting enough titles out from the queues and finding more automated ways to do things so the queues are manageable. And that same process will be reusable for lower priority queues as well... Anniemod 01:51, 2 December 2016 (UTC)
Why not run the Fixer processing under cleanup reports? Put a series of okay/reject checkboxes next to each manual process. If not rejected, or fully edited, the title reappears in the cleanup report tagged, e.g., 1/6, 3/6, etc., in ascending order. When 6/6 is reached it either goes to the submission queue or is edited and approved. Editors starting with new Fixer submissions could do the tasks they find the most interesting first, then return later to complete the rest.--Rkihara 16:28, 2 December 2016 (UTC)
I am not 100% sure that I understand the proposal correctly, but it looks like you are referring to (some of) the first 9 steps of the Fixer processing. I am afraid they can't be delegated to other editors. I have updated Fixer's User page to explain that "The software known as "Fixer" resides on a development server which is different from the main ISFDB server. It can't be moved to the main server without a complete rewrite due to software compatibility issues." Ahasuerus 16:45, 2 December 2016 (UTC)

(unindent) I have further clarified Fixer's logic on his User page. The important things to note are that steps 1-10:

  • have to be run by me because they occur on the development server
  • have been highly optimized over the years

Our current problem is with step 11:

Moderators review and approve submissions created using Fixer's account. The robot maintainer reviews and approves submissions created using his account. See this Help section for a discussion of the challenges associated with processing robotic submissions.

Alas, processing Fixer-generated submissions is the most labor intensive and time-consuming step.

The way I see it, Fixer performs the following high level tasks:

  1. Find target ISBNs
  2. Find detailed bibliographic information for each target ISBN
  3. Build ISFDB-compliant submissions and add them to the ISFDB submission queue using the Web API

Each task is valuable since it saves editors a lot of time. I would like to preserve this functionality if at all possible.

The biggest problem with adding Fixer-built records to Wiki pages -- instead of building submissions and adding them to the ISFDB submission queue -- is that it forces editors to create submissions manually, which is time-consuming. It also makes it much harder to keep track of what has been rejected vs. submitted, which Fixer needs to know in order to avoid creating duplicate submissions.

After reviewing Annie's and Ron's responses, I am thinking that perhaps the best compromise solution would be something like the following:

  1. Fixer adds a submission to the main ISFDB submission queue
  2. A moderator clicks the "Approve" button
  3. The software adds the new record to the database as per the current process
  4. The software recognizes that the approved submission was generated by a robot (in this case Fixer) and adds the ID of the newly created publication to a special cleanup report
  5. All editors have access to the new cleanup report. An editor can put one or more publication IDs "on hold" the way moderators can put submissions on hold.
  6. Once the editor has massaged the "held" publication record to his or her satisfaction, he or she removes the publication from the cleanup report.

If we adopt this approach, we will need to think of a way to ensure that new editors do not accidentally stumble on this functionality and mess things up due to inexperience. Perhaps we could limit access to this cleanup report to editors with more than a certain number (1,000?) edits? Ahasuerus 17:11, 2 December 2016 (UTC)

Maybe total edits to New Pub+Pub Update>1000?--Rkihara 17:28, 2 December 2016 (UTC)
I don't have that many from those two categories I think :) Maybe a waiver if you have a certain number overall? There are legitimate cases where you know what you are doing but you had not added or updated too many new publications. Now I am back to reading the whole thing more carefully - comments to follow:) Anniemod 18:22, 2 December 2016 (UTC)
I used the proposed number, but narrowed down to the two categories, at 200-250 edits you pretty much know what you're doing.--Rkihara 20:27, 2 December 2016 (UTC)
BTW, at this time the list of top contributors is a raw HTML page generated by the nightly process. The only way to determine how many edits of different types an editor is responsible for is to parse the HTML. We'll need to rewrite the logic to put this data into the database. Luckily, it shouldn't take long. Ahasuerus 20:41, 2 December 2016 (UTC)
Unless I'm missing something, this is parsed out on the "Stats" page.--Rkihara 21:51, 2 December 2016 (UTC)
Our "stats" pages and the cleanup reports are regenerated nightly. However, there is a difference between the way the former and the latter are handled. All cleanup data is put in a proper database table, which is used by our cleanup reports during the day. On the other hand, the "stats" data is formatted as HTML tables and then dumped to text files. When you access one of the "stats" pages, the software simply reads in the data from the appropriate text file and displays it as a Web page.
What this means is that if we want to know how many approved edits an editor has to his or her credit, we need to open the "stats" file and parse the data out of the HTML table. It's not hard, just ugly. It can also cause errors later if the layout of the HTML tables ever changes. Alternatively, we can just count the number of approved submissions in the submission table, which is more resource-intensive. Ideally, we'll simply rewrite nightly processing and make it store the "stats" data in the database. Have no fear, one way or another we'll get it done :-) Ahasuerus 22:29, 2 December 2016 (UTC)
So basically you want to have the Fixer output always approved and THEN fixed instead of cleaning the data before addition? I suspect that this will be easier to implement in the short term but I still would like the other idea more (clean beforehand, then have a robot create the submissions from it - by basically removing the 3rd action from Fixer and giving it to another robot. I do not want to shift the creation of the newSub to the editors - just the cleanup of the data:) And maybe it will be a two step process and a post-addition cleanup is the way to start. Anniemod 18:29, 2 December 2016 (UTC)
What my idea was (and still is) is to inject the review between these two steps:
*Find detailed bibliographic information for each target ISBN
*Build ISFDB-compliant submissions and add them to the ISFDB submission queue using the Web API
Fixer finds all it can but instead of posting the submission, it creates a list somewhere (new UI or wiki) where someone can do the verifications before it hits the queue. Once that is done, get Fixer (or another robot) to use the cleaned data to post the submission. Wiki is the technology we have here. But it may as well be an "updateable record" where Fixer creates the submission but does not submit it and instead puts it a list that editors can get from, massage and send out when all is verified... In all cases that will reduce the work needed during the moderation, shifting the work ahead of time and allowing more people to help... I know that means more development but that will make it a lot more manageable in the long run. Anniemod 18:42, 2 December 2016 (UTC)
I see. I think there may be a way to do what you are trying to accomplish without creating new tools.
At the moment all submissions are put in the ISFDB submission table. Each submission can be in one of the following states:
  • New (and optionally "On Hold")
  • In Progress (aka "Errored Out" if it remains in this state indefinitely)
  • Approved
  • Rejected
We could add another submission state. Let's call it "Automated" for now. "Automated" submissions will not appear on the "New Submissions" page which moderators have access to. Instead they will appear on a separate Web page, which will be open to a subset of experienced editors. The page will display a list of all "Automated" submissions with an "Edit" button next to each.
When an editor clicks on the "Edit" button, two things will happen. First, the submission will be put on hold, which will prevent other editors from viewing or editing it. Second, the body of the submission will be loaded in the "New Publication" edit form. Once the editor has massaged the data and is satisfied that everything looks OK, he or she will click "Submit". The submission process will do the following:
  • Create a regular new submission independent of the original Fixer-generated one
  • Optionally add a "Note to Moderator" to the new submission. It may say something like "The original data was compiled by [robot name] and was modified by editor [user name] prior to submission".
  • Change the state of the original Fixer-generated submission from "Automated" to "Processed" or some such.
How does it sound? Ahasuerus 19:24, 2 December 2016 (UTC)
P.S. In addition, editors will need to be able to remove their holds and to mark "Automated" submissions as "deletion requested". Ahasuerus 19:33, 2 December 2016 (UTC)
I like that plan a lot -- it clears the submission queue from barely identified issues, it clears your time from the need to inspect each and every request and it allows more people to try to help.
And if Fixer messed up and started a NewPub instead of AddPub for something, the editor can add a moderator note "to be varianted under <ID>" or "to be merged with <ID>" so the handling moderator can do that immediately after approval. The special case where it is AddPub under the wrong title will need to either go the "deletion and manual add" or allow it throw and ask for an unmerge and followup merge/variant". Or alternatively this new thing may have a button to allow conversion from AddPub (requires parent ID) to NewPub and vice versa? Or replace the parent title when Fixer get a bit confused where a title belongs. Anniemod 19:54, 2 December 2016 (UTC)
If we decide to go with this approach, we'll probably implement it for NewPubs first since they take much longer to process. Once we are confident that things are working smoothly, we can add support for AddPubs. Ahasuerus 20:00, 2 December 2016 (UTC)
I was thinking aloud on what else may need changing while cleaning up a request. :) Small steps are better than no steps at all. Anniemod 20:03, 2 December 2016 (UTC)
As a general rule incremental changes tend to be safer. There are times when you can't avoid drastic changes, but the risk tends be high. Luckily, all we have is a single server, so we can deploy hundreds of minor patches without having to worry about patch management. Ahasuerus 20:36, 2 December 2016 (UTC)
True that - I wish it was possible in my day job as well. :) Anniemod 20:44, 2 December 2016 (UTC)

(unindent) I have come up with a design document for the proposed change. It involves a bit more than what we have discussed so far, but it looks doable. I will set other things aside and start working on it tomorrow. Thanks for the feedback! Ahasuerus 00:00, 4 December 2016 (UTC)

Back to the eligible editors - can you set up another security flag on ID's? Then it's a one-time load based on whatever criteria you want and infrequent updates thereafter? Doug H 20:05, 4 December 2016 (UTC)
Let me make sure that I understand the proposal correctly. Is the idea to explicitly give individual editors permission to access automated submissions? If so, then yes, it's possible, but it would involve considerably more work than the process described above. I think it should be reasonably safe to use the number of approved edits this way since all submissions will be subject moderator review. Ahasuerus 23:54, 4 December 2016 (UTC)
The discussion indicated to me that eligibility criteria were a best guess. Incorporating the criteria in the code does not allow for exceptions - either good editors or bad ones. And it ties a security attribute to a set of measures that may change, e.g. I saw a count of approved edits, but not a count of disapproved edits or any time factor. I was just wondering if a degree of separation could be achieved at little or no cost. Doug H 13:40, 5 December 2016 (UTC)
At the moment we use the standard MediaWiki software, which runs our Wiki, to handle almost all user management tasks. This includes account creation, logging on/off and moderator privileges. In addition, there are a couple of places in the software where we basically say "Restrict this functionality to users number 12345 and 67890". It is not very elegant, but it gets the job done as long as the number of affected users is low.
Upon reflection, you raise a valid question. There is a difference between being able to enter the book in your hands using Help as your guide and being able to work with secondary sources which present additional challenges. Perhaps we could start by limiting access to automated submissions to a subset of editors and then slowly expand it. Ahasuerus 18:26, 5 December 2016 (UTC)

Chinese book help

I posted over on Linguist's talk page, but it'd good to get a few eyes on this. Anyone else know Chinese? ···日本穣 · 投稿 · Talk to Nihonjoe 02:55, 3 December 2016 (UTC)

Proposal to create a rule changelog wiki page


Create a changelog of every change made to the ISFDB rules and standards. This changelog has nothing to do with the way a rule change is determined and the discussion which leads to a rule change. It only comes into play after that. I created an example how it could look like which contains one of the latest rule changes: User:Hitspacebar/Rules_and_standards_changelog

Why do we need this page?

Counter-question: if you go on holidays or a longer hiatus and come back here - how do you find out if, which and how the rules and standards have changed? Do you (want to) read all discussions on the rules and standards wiki page? That can be a lot to read. Also consider that older discussion are deleted from the wiki eventually.

Another reason: I had at least two situations where two moderators had a different "current practice" regarding submissions I had made. Therefore we definitely need a means which helps us to ensure that editors, moderators, documentation and current practice are as much as possible based on same current rules and standards.


It's possible that the changelog will be incomplete because nobody thinks about adding an entry after a rule change has been decided. See the following thread which could help to ensure a complete changelog.

See also

This thread is based on my initial proposal here and a subsequent discussion about it on Ahasuerus's talk page.

Jens Hitspacebar 20:37, 4 December 2016 (UTC)

A changelog will provide a quick overview of rule changes, which will hopefully help returning editors get up to speed quickly. Ahasuerus 23:57, 4 December 2016 (UTC)
Since there's no additional feedback I think we should start using this rule changelog. If there's no objection I'd move the page to its correct place at Rules_and_standards_changelog and add links to it here:
As for the separate template which I proposed in the thread below and which should help ensure that the rule changelog is always complete: I drop that proposal. Jens Hitspacebar 18:57, 7 December 2016 (UTC)

Create a standardized process for rule changes

I originally wanted to propose a new wiki template which should be added to each thread started on the Rules_and_standards_discussions page and which should help to keep track of the status of each discussion. It's goal is that with this template it's a lot easier to see if something has to be added to the rules changelog (see previous discussion thread) after a discussion is finished. Here is my proposal for this template: User:Hitspacebar/Template:RuleDiscussionStatus. This template can be seen as something similar to the status fields used in many ticket and bug tracking systems. They usually have a "status" ("open", "rejected", "resolved", "closed" etc.) and some also have a separate "solution" field ("done", "not relevant any more" etc.).

However, Hervé intervened and suggested that before we start using such a template we should discuss the way rule changes are decided - which is currently not a standardized process but more like an ad hoc process based on whoever coincidentally participates in a discussion.

In contrast to Hervés point I think that the template I'm proposing can be used regardless the way we decide about rules changes because this template is not a part of the discussion or the decision-making process. Whatever this process is or will be, the template is sitting on top of (or next to) it in order to keep track if there has been an outcome at all and help ensure that the rule changelog is updated (as long as we don't move these discussions to a completely different software, of course).

Maybe Hervé can chime in with his own words :)

So there are actually two things here:

  • If we want to implement the rule changelog I proposed in the previous thread: how do we ensure that each rule change is logged there? Is the template I am proposing for that sufficient, and if so, could we use this template without agreeing an a standardized rule change discussion result determination process before?
  • What should be the standardized process to determine the outcome of a discussion about rules and standards?

See also the former discussion about this on Ahasuerus's talk page.

Jens Hitspacebar 20:38, 4 December 2016 (UTC)

Thanks for the invitation, but I decline. Hauck 07:47, 5 December 2016 (UTC)
That comes as a surprise. You had a valid point regarding a missing standardized process for rule changes and thought you might have some ideas for an implementation. I hope there's nothing wrong (or even offensive) with my wording above which led you to your decision (if that's the case: that wasn't my intention). Hitspacebar 08:51, 5 December 2016 (UTC)
No, no, don't worry, there's nothing wrong with your intervention. I just don't want to enter a Nth debate without an outcome or worse, with an outcome based on an "ad hoc process" that is a complete mystery. Hauck 09:02, 5 December 2016 (UTC)

Entering Title notes at New Pub entry time?

Ahasuerus requested that I solicit the community's opinions about this. So please opine! :-)

Do people want/need the ability to supply title-level notes in the New Pub screen?

I recently noticed a number new pub submissions that placed what should have been Notes in the Synopsis field. The latter is made available on the New Pub screen, while the former is not. Some other recent edits moving information from Synopsis to Notes suggests a fair number of editors have done the same thing, and many times neither editor nor moderator has remembered/noticed to go move the information after the title record was created. I realize the sheer number of notes-like fields on New Pub is already daunting, but I'm wondering if empirical evidence isn't suggesting we ought to have provision for title notes entry there.

Ahasuerus pointed out FR 152 Add a new field for Title level Notes in the New Publication forms, but only display it when it's checked in User Preferences.

Considering the way that is worded, I suggest commenting on two aspects: (1) Should one be able to supply Title Notes in the New Pub screen and (2) Whether the resulting "notes" clutter matters and any thoughts about mitigating that clutter.

This isn't an R&S discussion, so we don't need to agree! All comments are grist for Ahasuerus' implementation mill. --MartyD 11:33, 9 December 2016 (UTC)

Yes, please! At least it would save editing the title entry just for adding the translator who belongs in there if there's one to credit. Stonecreek 12:11, 9 December 2016 (UTC)
I have been thinking about this issue for the last few days. I think the advantages of adding this field to the New Publication form, especially when dealing with translations, would outweigh the disadvantages, i.e. the extra clutter.
As far as the issue of making it a User Preference goes, I suspect that it would only add an unneeded layer of complexity. We can always add it later if requested. Ahasuerus 13:08, 9 December 2016 (UTC)
Heck yes, I say it's badly needed.
As for "some other recent edits moving information from Synopsis to Notes", that was me; I searched for mentions of editors or translators in Synopsis. No doubt there are other things needing moving that I didn't find that way. --Vasha 13:10, 9 December 2016 (UTC)
Just dropping in to say "Yes please". PeteYoung 14:25, 9 December 2016 (UTC)
Another "yes, please". I would also ask for the storylen to show up on the newPub (because of omnibi) but one battle at a time. Anniemod 17:15, 9 December 2016 (UTC)
"Yes!" from me. This would be very nice to have. I also like the idea of being able to add the storylen on new pub omnibuses (omnibi?). ···日本穣 · 投稿 · Talk to Nihonjoe 20:51, 9 December 2016 (UTC)
[OT: it is "omnibuses" because it is derived from the dative case form of the latin "omnis" (nominative case singular). The plural ending -i is only correct if the singular nominative case ending is -us.] --Vasha 21:48, 9 December 2016 (UTC)
Now, now, why inject logic in such a nice conversation :) I kinda know it is not omnibi but I like the word a lot so I am pretending that it is my bad English that makes me say it :) Anniemod 22:28, 9 December 2016 (UTC)
"Logic" has nothing to do with it when it comes to Latin. That's why I'm glad that English always allows us to form plurals by adding -es; we only do it the Latin way when we are trying to be fancy ;-). --Vasha 23:18, 9 December 2016 (UTC)
Unless it's "datum" or "bacterium" :-) Ahasuerus 23:37, 9 December 2016 (UTC)
Just to demonstrate that learning the Latin forms can't be done in five minutes, here are two quiz questions.
1. What is the Latin plural of "consensus"? Answer: "consensus". That's because it does not have the -us ending of words that form their plurals with -i; instead the ending is -ūs (with a long vowel), plural -ūs, but Latin spelling doesn't actually mark vowel length.
2. What is the Latin plural of "octopus"? Answer: "octopodes" (well... that isn't Latin, read on). This is actually a Greek word (ὀκτώπους) which, transcribed into the Latin alphabet using Latin spelling conventions, becomes "octopus". The Greek plural of that word is ὀκτώποδες (octopodes).
So... even though I know this stuff, you won't catch me showing off by using those forms... it's plain ol' "consensuses" and "octopuses" for me. --Vasha 23:18, 9 December 2016 (UTC)
I know enough Latin to get in trouble - had been planning to pick it up again and never seem to find the time or the will. Bulgarian has a lot of ways to form plural - there are basic rules based on gender but then there are so many exceptions that... And some words have double plurals depending on how you use it. When I started learning English, I was waiting for the other shoe to drop on the plurals and genders for a long time. Then I realized it just is that easy. Anniemod 23:24, 9 December 2016 (UTC)
Yes! It would save me a lot of additional edits. --Willem 21:09, 9 December 2016 (UTC)

(unindent) The motion passes by acclamation! :-)

The software has been modified and "Title Note" fields should be available in all New Publication forms. Re-checking things, I see that the field is missing the standard mouse-over Help, but that should be easy to fix. If you find any other problems, please report them here. The "story length" field will be added once the recently approved changes to its values are done. Ahasuerus 22:32, 9 December 2016 (UTC)

Can we please fix the help page here to mention the new field? Anniemod 22:45, 9 December 2016 (UTC)
Certainly. The "Title Note" template has been cleaned up and linked from Help:Screen:NewPub. The software has been updated to display a mouse-over bubble and link to the template. Ahasuerus 23:12, 9 December 2016 (UTC)
Awesome! ···日本穣 · 投稿 · Talk to Nihonjoe 23:44, 9 December 2016 (UTC)
Great, thanks! --Vasha 00:02, 10 December 2016 (UTC)
Having now tried it out... could you make the box maybe one line taller?
ETA: I just noticed the draggable corner to expand the box. That's handy! So never mind about the initial size. --Vasha 01:05, 10 December 2016 (UTC)
However, I think I use Notes a lot more often than Synopsis, so if that's true of everyone, the initial size of the two boxes should be reversed (Notes larger). --Vasha 01:09, 10 December 2016 (UTC)
Well, the original idea, which we discussed way back, was that synopsis data tends to be long-ish, hence the default size. However, it was mostly speculation based on our (then) limited experience. If the editors who actively use this field in 2016 need more space, it will be easy to accommodate their needs. Ahasuerus 01:22, 10 December 2016 (UTC)
My idea was that if people are using Notes 10 times more often than Synopsis, it ought to be the more visually prominent one so that no one accidentally puts their notes in the wrong box. As for needing more length, it's true that synopsises are long, but the draggability takes care of that. --Vasha 01:26, 10 December 2016 (UTC)
I'll be more than happy to change the default size if there is popular demand for it. Speak up, folks! Ahasuerus 22:24, 10 December 2016 (UTC)

(unindent) I post a note about a feature. I go to work. I come home from work. The feature's implemented and active. Magic! I could get used to this.... :-) --MartyD 01:58, 10 December 2016 (UTC)

Actually, it was merely sufficiently advanced technology. There would have been an extra charge for magic :-) Ahasuerus 22:24, 10 December 2016 (UTC)

NewPub/ClonePub/AddPub and transliterations - new user preference proposal

I would like to propose a user preference that allows an editor to see/add transliterations to the transliteratable elements while adding a publication. At the moment, if I am adding a novel in English or any other Latin-1 language, I need to make one submission. If I am adding one in Bulgarian, I need about 8 (the initial one, transliterations for the title, the publication, the publisher, the publication series, the cover art, the author, the artist for the cover). If I have introduction or any other essay inside or internal art, I need 1 or 2 per element(1 if we already have the author). For collections, we have 1 or 2 additional one per story... And that does not include varianting when needed, pseudonyming or any of those additional elements that you may have in English as well.

That's the same for any languages that use alphabets that have no Latin-1 characters (some of the Central European may be as bad but they have a chance to have some strings not needing transliterations :) I understand that most editors won't need to see these fields (thus the user preference) but anyone working in any of the non-Latin1 languages, that will be very helpful... I suspect that it may not be easy to implement but any improvements in adding those titles will be a good change. Thanks! Anniemod 02:25, 10 December 2016 (UTC)

Oh yeah, I guess it would be really nice to have the ability to do the transliterations at the same time as everything else when working in Turkish. At least for titles; names seem like less of an issue, since new names are added less often (and you then have to edit the profile in any case, to make sure that the working language, last name, etc. are correct). I think having the option (in preferences) to have a transliteration field below all titles wouldn't clutter things up too much. Oh, but Anniemod reminds me that I haven't been adding transliterations to publishers, pub series, and cover art titles! Yes, it would be nice not to have to go back and do all that. --Vasha 03:03, 10 December 2016 (UTC)
I've taken care of most of the ones you added lately :) They show up on a report, Turkish is easy enough for transliteration. You may want to check the cover art though - if it is not marked as Turkish, it does not show up on the report and I may have missed some. :) Anniemod 03:06, 10 December 2016 (UTC)
Thanks very much -- I noticed that some of them had been done, and didn't know it was you. --Vasha 03:33, 10 December 2016 (UTC)
I agree that it would simplify things considerably. However, as Annie correctly guessed, it would be a fairly significant change. More importantly, there are some design issues to consider. For example, at this time there is no easy way of telling if the publisher/publication series/author that you are about to enter already has a transliterated name on file.
At one point we discussed creating a second Web page which would be optionally displayed after you click "Submit". It would only appear if some of the entered values needed transliterations or if there were identical Contents titles on file which needed to be merged.
The approach seems to be desirable, but we'll have to consider all possible permutations. For example, what happens if one of the "to-be-merged" Contents titles is no longer in the database when the moderator reviews the submission? Will it make the submission "unapprovable"? That could cause a lot of painstakingly entered data to be lost, something that can't happen with NewPub submissions under the current system. Ahasuerus 03:34, 10 December 2016 (UTC)
An idea -- allow adding of transliterations where it makes sense on the forms. If there is already transliteration, allow the moderators to chose if they want to keep the old (the default) or replace with the new. Or just discard the new one and keep the old one in case it is there? A second page puts a lot on the order of submissions - what if I have mine waiting and someone else works on whatever I am merging into? This way, we won't lose work that is already done -- and if there is transliteration, it will stay in place.
So if the publisher is already in and I had checked that, I can leave the transliteration field -- and the one from the DB will be there. But if it is not there or I see it does not have a transliteration, I can as well take care of it with a new publication. I do understand that it is not trivial - but I am looking at a few thousand Bulgarian titles that a friend had agreed to work with me on (me entering, him verifying) and that will be a lot of waiting time and editing :)Anniemod 03:57, 10 December 2016 (UTC)

(unindent) Let's consider the use case that Annie mentioned earlier:

If I am adding one in Bulgarian, I need about 8 (the initial one, transliterations for the title, the publication, the publisher, the publication series, the cover art, the author, the artist for the cover)

The way New Publication pages currently work, "transliterations for the title, the publication" can be entered as part of the original submission, which leaves us with the following 5 fields:

  • Publisher
  • Publication series
  • Cover art
  • Author
  • Cover artist

Since we use the publication title as the cover art title, I think we should change the software to use the submitted transliterated value for the cover art title as well.

As far as the other 4 fields go, chances are that there will be more than one submission for each publisher, publication series, author and cover artist. All of them will be caught by the nightly cleanup reports and will require one submission per record to fix. If we were to add 4 new "transliterated value" fields to the NewPub form, editors would be prompted to enter the same information over and over again, which seems counterproductive.

That said, adding a "transliterated title" field to the Contents section may save time since Contents level titles are usually unique. It may be possible to add a button which, when clicked, would add "Transliterated Title" fields to the page. Ahasuerus 22:43, 10 December 2016 (UTC)

Last time I tried, a transliteration added to the title did not carry to the publication (I have a test in the queue if someone can approve it - in case my memory deceives me). Anniemod 22:59, 10 December 2016 (UTC)
I have approved the submission -- the results are here. (BTW, experience suggests that we need to be extra careful with test records since they have a nasty tendency to stick around much longer than expected.) Ahasuerus 23:08, 10 December 2016 (UTC)
It carried through so I was wrong on that reducing my old 8 to 6:) Still a lot of edits :) For the test record - I am aware so I have a note to delete it -- the pub deletion already submitted; title to follow, If I was planning to be away for awhile, I would have put my name in the title to make sure I can find it when I am back :) Anniemod 23:15, 10 December 2016 (UTC)
I have approved the submission and deleted the title record, so we should be in good shape now. Ahasuerus 23:23, 10 December 2016 (UTC)
Changing the software to pick up the transliterated cover art together with the work makes sense. And adding a button to add transliteration for content will make it manageable. Anniemod 22:59, 10 December 2016 (UTC)
OK, FR 958 has been created to take care of Cover Art titles. Ahasuerus 23:14, 10 December 2016 (UTC)
FR 958 has been implemented. Ahasuerus 17:24, 18 December 2016 (UTC)
I am reasonably sure that "Add Transliterated Title" should be doable, but we'll need to clean up the filing code first. At the moment it's so messy that adding more complexity may break it. Ahasuerus 23:14, 10 December 2016 (UTC)
FR 959 has been created. It's deliberately vague to allow multiple implementation paths. Ahasuerus 19:10, 11 December 2016 (UTC)

Transliterations for works with no language

I am not sure if one of the current reports need to be tweaked or we need a new one but I keep stumbling upon titles that need transliterations and which are nowhere to be seen on our reports - they are usually coverart, occasionally interior art records (I am not sure if there had ever been other types) and always without a set language. Can we get these on a report? I think that anything that contains a non Latin-1 character and has no transliteration should be on a report so someone can go, set the language(in the case when it is not set) and add a transliteration. This way we will know that all the old records are now cleared at least for that :) Anniemod 00:48, 11 December 2016 (UTC)

That's a good point. The cleanup reports that handle transliterated values are organized by language code. That way all Japanese titles are together, all Russian titles are together, etc. Then there is a "catch-all" report which handles less common languages. However, it ignores titles without a language code, an oversight. Bug 640 has been created. Thanks! Ahasuerus 01:02, 11 December 2016 (UTC)
The affected cleanup report has been updated to include titles without a language code. The data will become available tomorrow morning. We are looking at roughly 80 additional titles. Ahasuerus 17:32, 16 December 2016 (UTC)
Once everything in the queue is approved, these are all done. :) Anniemod 21:50, 17 December 2016 (UTC)

Add transliteration to "Make variant"

Can we have the transliteration field in the "Make variant" page (the lower part where you create a new title)? We already have the language and date here - if we can add the transliteration as well, we may not need an additional edit all the times when the original is in Russian/Bulgarian/Japanese and so on :) Thanks! Anniemod 22:26, 12 December 2016 (UTC)

Another good point -- FR 960 has been created. Ahasuerus 23:21, 12 December 2016 (UTC)
Done. Ahasuerus 01:08, 18 December 2016 (UTC)

New Cleanup reports proposal

I would like to propose 2 new cleanup reports:

  • Language mismatch in the same compound title - we do have billingual books and they can be ignored but we also have a lot of cases where the novel is set to French but the coverart either have no language or is set to English for example. This should pinpoint a huge ammount of the remaining non-English titles (short fiction, essays, arts records) - as soon as the main title language is set, these will be highlighted. That may even be an automated job - if no language is set, set the same as the container; if a language is set but is different, put on a report.
  • Container titles with no language (collection, omnibus, anthology, magazine, fanzine). A lot of those will be English but we will find a lot of non-English as well.

Thoughts? I am trying to find as many non-English titles as we can without needing to go through thousands of English titles. Anniemod 18:38, 16 December 2016 (UTC)

Let me start by re-posting something that I wrote on my Talk page a few days ago:
... the plan has always been to add languages to titles in two phases. During the first phase we were going to assign language codes to the "low hanging fruit" manually. During the second phase we were going to assign 'English' to the rest of the titles automatically.
The first phase has been a success in that only 46.3% of our titles do not have a language code assigned.
A few months ago I decided to check if we were ready for the second phase. I ran a few reports and found hundreds of authors like Philippe Druillet whose language-less titles are a mix of English and French records. Clearly we weren't quite ready for "phase 2".
I think the next step is to use this cleanup report, which finds author/language mismatches, as a template and build on it. What we need is a cleanup report to find language-less titles by non-English authors. There are 18,290 of them, which may seem like a lot, but it's just 3% of the total number of language-less titles. Once they have been upgraded, we can move on to "phase 2". And there will be much rejoicing! :-) Ahasuerus 04:26, 15 December 2016 (UTC)
FR 961 has been created. Ahasuerus 04:41, 15 December 2016 (UTC)
A further review revealed that we already had a cleanup report which showed "language-less titles by non-English authors". However, it was limited to non-art titles. I didn't want to add art titles to the same report because it would have made it difficult to find non-art titles. Instead I created a new cleanup report and deployed it earlier today. The data will become available tomorrow morning.
I also like Annie's idea re: automating some parts of the language assignment process. If a language-less art (COVERART/INTERIORART) title is associated with a publication whose "reference title" has a language code, I think it should be safe to assign that code to the art title automatically. I'll need to run a few checks to see if I am missing any potential issues. Ahasuerus 22:33, 16 December 2016 (UTC)
How about short fiction and essays in that second case (and even interviews, reviews, poems and other internal elements)? Any reason they cannot automatically get the language code as well? The only problem will be bilingual editions and these won't have a language set on their reference title (or if they do, most likely their content will also be set). That will eliminate a huge amount of now language-less titles allowing only true unknowns to remain behind... Anniemod 23:11, 16 December 2016 (UTC)
If the guinea pigs do well, the rest of the menagerie will be next :-) Ahasuerus 23:19, 16 December 2016 (UTC)
One more thing - magazines and fanzines. I was looking at Amazing Stories and none of the editors from before 1993 that I checked have the language set. Fanzines and magazines do not change their language during their runs. Any way to be able to mark a fanzine/magazine language so that some kind of automation can set the editors (and then the inside titles if the plan work out)? Anniemod
I assume the algorithm would look something like:
  • Examine the EDITOR titles for each series
  • If none of them have a language code assigned or if there are 2+ language codes, skip the series
  • Otherwise retrieve the language code and assign it to all language-less EDITOR titles in the series
I can't think of a scenario where the algorithm would fail. Ahasuerus 22:57, 17 December 2016 (UTC)
Sounds good. Then it will come down to assigning language to 1 EDITOR only (if the series has none - which may be true for some of the older series that had not been worked on lately) and that is manageable. Maybe throw on a report the ones for which language cannot be determined(so someone can go and set 1 record properly) and the ones where there are different languages assigned (so someone can see if that means two magazines that need to be split or a wrong language assigned or something like that? The only thing I would be worried if a magazine with no language has one editor wrongly set to English and that gets populated everywhere but... short of checking all magazines, I do not see how we can prevent it - and we do need some automation. Although if the editor has their language set correctly, that would show on the language mismatch report (it includes EDITOR, correct?) so it should be caught before all of them are set. Annie 23:07, 17 December 2016 (UTC)

(unindent) So, should I work on those 500 art pieces that need language now in the new report or should I give you a few days to see if we will be able to set it automatically (most of them are attached to publications that already have a language)? Anniemod 20:59, 17 December 2016 (UTC)

Well, the new report was written before I saw you proposal re: automatic language assignment. My current thinking is that there is no harm in cleaning up a few dozen titles manually. Perhaps we'll see additional patterns that may help us come up with a better auto-assignment algorithm. Ahasuerus 22:31, 17 December 2016 (UTC)
Or a few hundred - if I start on them :) They are perfect for what I need right now - not too much thinking, no external verifications. I will count how many would have not be caught by an automated "assign the reference title language" script to see how it looks like. By the way, all of the no-language transliterations would have been assigned languages properly this way (just as a reference point) - they would have still needed transliterations but languages would have been there. Anniemod 22:43, 17 December 2016 (UTC)
Now that the majority of language-less titles have been handled automatically, the number of outstanding art titles has dropped to 551. Consequently, the latest and greatest version of this cleanup report is no longer limited to 500 titles per day. Ahasuerus 23:12, 20 December 2016 (UTC)

Import Content and Make Variant changes

The layout of the data entry forms "Import Content" and "Make Variant" has been changed as per FR 955. The goal was to make it clear that each form supports two different options/paths. There was no other change in functionality. Ahasuerus 00:17, 18 December 2016 (UTC)

I like it. --Vasha 23:27, 18 December 2016 (UTC)

Availability update - 2016-12-17

There are some pesky "real life" issues which may affect my availability in the near future. In the unlikely event that I suddenly become unresponsive and stop updating ISFDB Downloads, please contact Al von Ruff who has full access to the server. I have created ISFDB:How To Create a Public Backup File to help ensure backup continuity. Ahasuerus 01:59, 18 December 2016 (UTC)

Best wishes to you as you try to deal with this. My thoughts and prayers are with you. Chavey 06:51, 18 December 2016 (UTC)
Thanks! Ahasuerus 13:50, 18 December 2016 (UTC)

Multilingual publications

A new cleanup report, "Multilingual Publications", has been deployed. The data will become available tomorrow morning.

The report looks for publications which contain titles in multiple languages, e.g. French and Spanish. A number of them are legitimately multilingual (and moderators have the ability to ignore) but the majority are not and will need to be fixed. For example, a few dozen Italian magazines are still using the old, pre-2011 data entry standard for translations. Ahasuerus 23:10, 18 December 2016 (UTC)

Awesome. A lot of those have their coverart set to the wrong language (default most likely for the entering).:) Is that catching a "no-language, English" pairing or is it only for proper language codes? Annie 18:12, 19 December 2016 (UTC)
It's the latter. Language-less titles are next on the chopping block... er, list. Ahasuerus 18:50, 19 December 2016 (UTC)
Makes sense. Let me go and take a look at one of the ones I was working on then - it looked as if there was one missing language only but because it was the anthology itself, none of the stories were showing languages - I suspect that once approved, I will see which one complains without the need to open each story to verify. Annie 19:21, 19 December 2016 (UTC)
How is cover art going to have a language? If a book gets published somewhere around the globe and a picture of e.g. Dutchman Hieronymus Bosch (died 1516) is used, does the picture suddenly have a language other than that of the painter when he made his piece of art?--Dirk P Broer 01:03, 20 December 2016 (UTC)
It shows in what languages the art had been used - the art does not change but the usage is different (and its title may be different -- otherwise you will end up with a Latin letters name on a Bulgarian book when that same one is used on a Bulgarian book. Or a Japanese lettered one on an English one and so on... Annie 01:17, 20 December 2016 (UTC)
So the language of the art can be defaulted into the language of the publication record it is used in?--Dirk P Broer 02:08, 20 December 2016 (UTC)
Yes, that's how the software works now. When you create a NewPub submission, you specify a language. That language is assigned to every title in the new publication, including its COVERART title(s). Ahasuerus 02:17, 20 December 2016 (UTC)
We once had the idea to disassociate most of the art from languages (apart from maps, comics and cartoons). I do think that it's still worthwile. In principle, Dirk is right: art in general has no language, only it's title has (whereas fiction has a defining language). We could spend our time with other things than determining the language of a piece of art (which has none). Stonecreek 05:21, 20 December 2016 (UTC)
So there is Interior Art called "Some Book [2]" in a book. I get the book and I identify it as the painting "Момиче със слънчогледи" of some Bulgarian artist. Do you propose that we use the original art name and set the language to Bulgarian (or leave it language-less if language is removed) but using the original name in the original language instead of varianting into the original record instead? Cover Art is different - as it follows the title for the most part (although why would we treat the two differently?)Annie 19:21, 20 December 2016 (UTC)
I find that having a COVERART title's language displayed can be handy. For example, let's consider this COVERART title by Josh Kirby. It has 4 COVERART VTs and 2 INTERIORART VTs. Of the 6 VTs, 2 are associated with English books, 2 with German books, 1 with a French book and 1 with a Dutch book. The ability to see each VT's language on Title and Bibliography pages is very nice and I wouldn't want to lose it. Ahasuerus 20:14, 20 December 2016 (UTC)
I am trying to make the same argument for INTERIORART as well - being able to see what languages the art is used in is something that I will hate to lose. Annie 20:21, 20 December 2016 (UTC)
Personally, I'd think it'd be enough to display the variant titles (after all, we are at work to catalogue fiction, art somehow was an addition). I think most of us are able to have an idea about the language of the book the piece was published in. So, are there more opinions on that topic? Stonecreek 04:54, 21 December 2016 (UTC)
Also, it's really meaningless to assign a language to a painting by, say, Karel Thole or Salavador Dali (if they spoke one language in their works it's the one of surrealism). Thole's work was reproduced widely in many publications from different countries. For him and most other artists we'd have such a vast amount of work of differentiating to do. If a cover for an issue of Urania is reproduced in a German publication (and there are quite a lot of, and they are titled as such), we'd have to set the language for that publication to Italian, others to Dutch, some also in English, still others would remain German). And all that while the art itself has no language, only the editor or the publisher decided to give a title to it. Stonecreek 05:07, 21 December 2016 (UTC)

ISFDB Downtime - 2016-12-20

The ISFDB server will be unavailable between 5pm and 5:20pm server (North American Eastern Standard) time. There is a chance that the downtime may last a bit longer than estimated. Ahasuerus 21:41, 20 December 2016 (UTC)

The server is back up. The language auto-assignment process has been run and we are down to 112,171 (8.96%) language-less titles. Ahasuerus 22:16, 20 December 2016 (UTC)
Plus the 238 undefined :) Annie 22:20, 20 December 2016 (UTC)
Indeed. I am not sure how they reproduce, but I suspect that it has to do with title merges and/or unmerges. I hope that they will disappear once we complete the language assignment process. Ahasuerus 22:31, 20 December 2016 (UTC)
And here is one small side-effect: Russian Publications with Latin characters - probably because of some of the art pieces being the only language-assigned items - I will clean these now but outside of that the auto-assignment seems to have run pretty well. Annie 16:25, 21 December 2016 (UTC)
Clarification because the example won't be there for very long. The language in The Baum Bugle - 1980 got set to Russian because of the Russian illustrations in one of the magazines under it being the only title with language. Which caused it to change it on all 3 publications and its titles. Nothing else bled into the non-Latin languages - it may be possible that some Latin cross-assigning happened where there was a single wrong title with a languages. Annie 16:57, 21 December 2016 (UTC)
Yes, that's exactly right. If we end up with 100 invalid language auto-assignments out of 433,000, we'll be in pretty good shape :) Ahasuerus 17:35, 21 December 2016 (UTC)
I do not disagree - I just saw it and thought I should mention it :) Annie 17:47, 21 December 2016 (UTC)

EDITOR with no language report proposal

Based on what we had been discussing about EDITOR records not having languages (and a lot of language-less titles being parts of them), can we get a report of all EDITOR titles with no language so the non-English ones can be picked out, leaving only English ones (and then language can be set automatically for them enabling then cascading the language)? Same for Collections and Anthologies I would think if the numbers are not too great... Annie 23:22, 20 December 2016 (UTC)

Here is where we stand at the moment:
| title_ttype  | count(*) |
| INTERVIEW    |      288 |
| SERIAL       |      393 |
| OMNIBUS      |      448 |
| EDITOR       |      788 |
| ANTHOLOGY    |     1067 |
| POEM         |     1467 |
| CHAPBOOK     |     1490 |
| COLLECTION   |     1845 |
| REVIEW       |     3296 |
| NONFICTION   |     4449 |
| ESSAY        |     9872 |
| INTERIORART  |    10453 |
| SHORTFICTION |    17046 |
| COVERART     |    22374 |
| NOVEL        |    37300 |
The plan is to create a cleanup report which will identify language-less "container" titles of the following types: OMNIBUS, EDITOR, ANTHOLOGY, COLLECTION. We will probably do it one title type at a time to make the report more manageable. Once these container titles have been cleaned up, we will re-run the "automatic language assignment" script and re-evaluate where we are. Ahasuerus 00:39, 21 December 2016 (UTC)
How about adding CHAPBOOK as well so we can clear those as well? Maybe after Collection and EDITOR and after the automation is run based on that? Annie 00:58, 21 December 2016 (UTC)
Chapbooks are typically reprints, so I am not sure how much we will gain. Once the 4 big ones have been taken care of -- BTW, it occurs to me that we should run the auto-assignment script after each one -- we can review what's left and determine the best course of action. Ahasuerus 01:19, 21 December 2016 (UTC)
Version 1 (OMNIBUSes) of the report has been coded and deployed. The data will become available tomorrow morning. Ahasuerus 01:24, 21 December 2016 (UTC)
Yes, that's why I said after the first 4 - a lot of their stuff will be caught before that but there may be some remaining (especially short European novels that are not novels really:) ). And running the script after each is probably a good idea - this should eliminate more titles eventually... Annie 01:28, 21 December 2016 (UTC)
With any luck, the auto-assignment script will make everything (but NOVELs) much more manageable. NOVELs... well, there may be ways to identify non-English novels and then assign "English" to the rest. Ahasuerus 01:34, 21 December 2016 (UTC)
And Non-fiction... The cleanest way I can see for those 2 categories may be to start reviewing short-ish lists of them (alphabetically or whatever), assign non-English, skip the English and clearing them for assignment of English to the whole group... Annie 01:55, 21 December 2016 (UTC)
Well, the brute force method is our last resort, but there may be other things we could do first. For example, we may be able to auto-assign English to titles associated with pubs published by established US/UK publishers. Ahasuerus 17:38, 21 December 2016 (UTC)
That sounds like a better idea :) Or even identify the publishers that have more than X non-language NOVEL/NON_FICTION and figure out the language for them . In case the assignment based on the known names leaves too many that is... Annie 17:49, 21 December 2016 (UTC)

Writers of the Future and Illustrators of the Future as awards

Okay, trying this again since the last discussion never really decided anything. (^_^)

For the Writers of the Future, there should be the following awards:

  • "Golden Pen" (annual, picked from the quarterly winners, should have a note indicating it's been called the "Grand Prize" at various times, too)
  • "Quarterly First Place"
  • "Quarterly Second Place"
  • "Quarterly Third Place"
  • "Quarterly Finalist"

For the Illustrators of the Future, there should be the following awards:

  • Golden Brush" (annual, should have a note that it's been called the "Grand Prize" as various times)
  • Finalist

If there are some that are slightly inconsistent, we can make a note on them. ···日本穣 · 投稿 · Talk to Nihonjoe 19:09, 21 December 2016 (UTC)

Availability update - 2016-12-21

My computer time will be very limited for the next week or so. I should be able to handle the weekly backups, but otherwise development and Fixer will be on hold. Ahasuerus 20:38, 21 December 2016 (UTC)

Hopefully not for a bad reason. Have a good holiday week. Annie 22:50, 21 December 2016 (UTC)
Have a good break --Vasha 09:36, 22 December 2016 (UTC)

Dating of Granta

Here'a a slightly complicated one. Granta (which sometimes publishes genre stories) has two versions with completely separate tables of contents, one print/ebook and one online. Both of them have the same issue number. The print version (example preview) is dated according to season (Autumn 2015). The online one also has its TOC listed on a page with that date. However, the stories are posted online gradually over time, with each one having a posting date on its own page. Example: "Click" states that it's published in "Granta 133: What Have We Done | The Online Edition / Fiction / 14th January 2016". So, which date to use? Is "Click" a 2015 story (according to the issue date) or a 2016 one (according to the date on its page)? --Vasha 09:55, 22 December 2016 (UTC)

Personally, I'd use the earlier date (2015) and would consider the posting as a reprint with its own later date. Hauck 10:05, 22 December 2016 (UTC)
Reprint? But it wasn't in the print edition, the web is the first appearance. --Vasha 10:11, 22 December 2016 (UTC)
In this case, the two editions (print and online) should be separated (like we do for Clarkesworld, IIRC) and the story be present (with a 2016 date) only in the "electronic" issue. Hauck 10:34, 22 December 2016 (UTC)
Even though the table of contents of the online edition also had the date Autumn 2015? Are there other examples where stories have a date attached to them which is not the date of the issue they are in? --Vasha 10:42, 22 December 2016 (UTC)
Well, than the example story is to be dated as a 2016 publication with an accompanying note why the magazine's dating is misleading. Stonecreek 13:55, 22 December 2016 (UTC)

(unindent) I think I have made up my mind. It's the usual practice to give magazines the cover date instead of the date they actually were placed on the newsstands, right? I am going to say that the "date" on the story pages is the latter, they are in the Autumn 2015 issue and "went on sale" another date -- so then they should be in the DB as Autumn 2015. Does anyone disagree with that? --Vasha 19:04, 22 December 2016 (UTC)

I would look at awards eligibility for proper dating I think. A story in Analog January 2016 for example is eligible for 2016 even if it is published in November 2015. This online story is dated on its own and as such is eligible for 2016 according to how I read the rules for eligibility - so it should be a 2016 story here I think... - with a note in the publication on why some stories are not as old as the issue (as Stonecreek proposed above). Annie 21:14, 22 December 2016 (UTC)
If the web-only version doesn't have a downloadable version (PDF, ebook, etc.), then it shouldn't generally be included here per ISFDB:Policy#Excluded. ···日本穣 · 投稿 · Talk to Nihonjoe 22:18, 22 December 2016 (UTC)
The story will make it into the DB sooner or later as it is from a print collection anyway. :) But I think I will start a separate discussion on what "downloadable" means these days - "Strange Horizons" do not have a downloadable form either for example and it is definitely in scope here. Plus there are plugins that create ebooks from a story (and send it to your Kindle automatically these days). So we do need to precise that language a bit... Annie 22:34, 22 December 2016 (UTC)
Good point on awards, Annie-- but couldn't some eligibility rules disagree with each other? To choose only the award most relevant to this particular story, the Stoker Awards say this: "In the case of works first appearing in magazines or other periodicals, the date of publication which appears on the magazine, commonly referred to as the 'cover date' -- rather than the date of distribution -- will determine eligibility. If there is no cover date the date of first release will apply." I argue that the online edition of Granta 133 DOES have a cover date-- it's on the page with the Table of Contents. By Stoker rules, that takes precedence over all other dates.
@Joe-- the story was published in a collection later in 2016, so I should determine the date of first appearance to put in the title record, even if the online publication isn't catalogued. --Vasha 22:39, 22 December 2016 (UTC)
I am not sure that it does actually - the rule is written with printed magazines in mind and I have a suspicion that this story will get its eligibility from the 2016 date on the story itself (as it is not part of the printed magazine that came out earlier). I guess sooner or later someone somewhere will need to make a determination on these cases (because if you look at the Riles forum, I have a similar question for another journal and an online-only story there (thankfully with a clear year) :) The new world of publishing is a headache-inducing mess... Annie 22:47, 22 December 2016 (UTC)

Cleanup for Imprints of Publisher Grimbold

I am not sure what the best way to handle this would be. Would an experienced editor be willing to do this cleanup as appropriate?

These two should be merged:

website: Kristell Ink | The Sci Fi and Fantasy imprint of Grimbold Books

This one should be corrected:

website: Tenebris Books | Dark Fiction imprint of Grimbold Books

Thanks much. JJ 23:28, 22 December 2016 (UTC)

Story Lengths for The Starlit Wood

ISFDB: The Starlit Wood: New Fairy Tales

In response to a query from me, I received the following message from Saga Press: "No novellas, but the following titles are novelettes (7,500 words - 17,500 words):"

  • Pearl by Aliette de Bodard
  • The Briar and the Rose by Marjorie Liu
  • Spinning Silver by Naomi Novik
  • The Other Thea by Theodora Goss

The rest are short stories. If an experienced editor is willing to make that change, I'd much appreciate it, and then I'll mark the whole entry "Verified". Thanks much! JJ 23:35, 22 December 2016 (UTC)

I've submitted the change. As soon as a moderator approves it (which may take a bit of time), all of the stories in the book will have their proper length specified. Annie 23:49, 22 December 2016 (UTC)
And now approved. Annie 03:15, 23 December 2016 (UTC)

Bulfinch's Mythology

I have been looking at the last two volumes of Bulfinch's Mythology and it seems to me that they are Nonfiction rather than Collection -- that is, the recounting of legends is done as chapters, not all of which stand on their own. I plan to change them if no one disagrees. Also, does anyone have a copy of the first volume, The Age of Fables a.k.a. Stories of Gods and Heroes, and can say if it's organized the same way? --Vasha 17:08, 24 December 2016 (UTC)

I have a copy of the 1978 Avenel Books omnibus. It has these major sections and organizations:
  • First sections, having chapters:
    • "Stories of Gods and Heroes" -- Chapter I (Introduction) - Chapter XLI (The Druids -- Iona)
    • "King Arthur and His Knights" -- Chapter I (Introduction) - Chapter XXIII (Morte d'Arthur)
    • "The Mabinogeon" -- Introductory Note + Chapter I (The Britons) - Chapter XIII (Taliesin)
  • Subsequent sections, not having chapters:
    • "Hero Myths of the British Race -- Beowulf - Robin Hood
    • "Legends of Charlemagne" -- Introduction - Ogier, the Dane (continued)
Despite the "Chapter" designations, each piece is a story in itself. And it looks like the final volume did not use chapters. --MartyD 12:15, 25 December 2016 (UTC)
Fair enough! Would you be willing to enter the contents for these collections, then? --Vasha 20:06, 25 December 2016 (UTC)

Working language question

Our help page description of "working language" says, in part, "For bilingual authors enter the more frequently used language." Even though Eduardo Barreto's first language is Spanish, every one of 6 credits is listed as being in English. Those are all illustrations, so "language" may be more-or-less irrelevant, but still it seems that his "working language", by our definition, is English. And that we should have a Cleanup Report that finds authors like this. Chavey 03:36, 26 December 2016 (UTC)

But we do not have all his credits in the DB - the DB is still mostly English. So if we use only what we have here, a lot of authors will end up English when they never really worked in English. And we do not always have the translation credit especially in older content. Illustrators are obviously more complicated but I would say that if they lived in a non-English country and were born there, the language follows them. Annie 06:16, 26 December 2016 (UTC)
I agree with Annie. If we can determine what their actual language is (especially if their work is only known through reprint or translation), we should set their language to that. Perhaps this needs to be moved over to Rules and standards discussions so it can be properly discussed in the right venue? ···日本穣 · 投稿 · Talk to Nihonjoe 18:04, 26 December 2016 (UTC)

(unindent) It sounds like there are at least three separate questions here:

  1. Should we have a new cleanup report to identify authors whose "working language" is not the same as the language of the majority/plurality of their canonical titles? (My take on it is that it would be useful.)
  2. If implemented, should this cleanup report allow moderators to "ignore" authors? (My take on it is that the report should support the "ignore" capability. The fact that our coverage is of necessity partial is a sufficient reason.)
  3. What should we do about artists whose art has appeared in multiple languages? (My tentative take on it is that the "most frequently used" rule still applies, but we have to be careful because it's harder to identify the original language of artwork.)

Ahasuerus 16:43, 27 December 2016 (UTC)

As stated before: While it may be useful to have a working language for artists, I don't think it's useful to have individual artworks assigned a language. Multi language artists like Karel Thole (born in the Netherlands, having lived most of his life in Italy, original artwork published with Dutch, Italian, English and German publications, multiple reproductions at least in Italy & Germany, with varying language titles, but it's always the same piece of art). We really could reduce the amount of time for handling 'languages' of art pieces, where they only get a title attached. Stonecreek 16:53, 27 December 2016 (UTC)
Well, the number of "language-less" titles is now down to 8.9% of the total number of titles. At the rate we are going, we'll hit 0% some time in early 2017. Once that happens, there will be no way to create new "language-less" titles, so the problem will go away. Ahasuerus 18:08, 27 December 2016 (UTC)
With a lot of art, the language is "irrelevant", but there are occasional pieces of art that have text within them, where the language is relevant. So I don't think I would want to remove language from art, but it may be useful to have a "language" option that is "irrelevant" (or some synonym of that). Chavey 22:56, 27 December 2016 (UTC)
Even if it is irrelevant, it makes it possible to find easy enough how many Thule illustrations had been used as covers in Bulgarian books for example. If we mark those as irrelevant, that will require inspecting all his works to see if a Bulgarian book is somewhere in the list. I know we are not Art Central but if we have the information, we may as well make it searchable easily enough. And while a German title on a painting is fine in another Latin language, it will be confusing on a non-Latin one or when the cover or interior art of an English book ends up with something with Russian letters or Japanese letters. Annie 23:47, 27 December 2016 (UTC)
The pieces where language is relevant were reduced to cartoons and maps in a discussion ca. two or years ago. But even for those the difference would be reflected in the titles. We seldomly would record more than the title for them (but if somebody wishes to add the details, such as countries or places for maps: as before, it'd be possible to do so in the notes).
What it does lead to is highly irrelevant and futile questions: 1) What language would a reproduction of Kepler's Somnium, seu Opus posthumum de astronomia lunari, divulgatum a M. Ludovico Kepplero, .. have assigned in a Bulgarian publication? 2) Which language should be assigned to Barclay Shaw's cover for Neuromancer, as it was published on English and German publications?
I'd also think that the question of usage for an artist like Thole could only be solved meaningful computationally, and it shouldn't be a concern for a database devoted to fiction. It's also still quite simple not correct to assign a language to a piece of art. Stonecreek 15:56, 29 December 2016 (UTC)
I don't want to enter a philosophical debate about art and language (just as an aside we're not talking here about pieces of art but about book covers or interior illustrations). For the record, I'm very attached to the concept of language for COVERART or INTERIORART (for the same reasons cited by Annie) and would be very unhappy to see all this work and all this painstakingly entered data written off. Such pages are extremely valuable to study the visual marketing of SF and the diffusion of SF stock images. Hauck 17:39, 29 December 2016 (UTC)
Sorry, Hervé, but philosophical questions about the language of a piece of art is exactly what one will get when we stick to the idea that language has to be added to cover art and interior art. It's absolutely not philosophical to see that a piece of fiction is made out of language, and a piece of art is not! And as we have artists like Pablo Picasso, Salvador Dali and Edvard Munch (to name a few prominent ones) in the database we are not only talking about cover & interior art but about the original pieces of fine art as well. (Aside: the cover art of Tim White, as an example, has been published in Germany on covers as well as interior art; the latter using the original English titles as well as made-up German ones; Thole's art was published similarly).
Has Munch's The Scream really the language English? What language has the original? Should it be Norwegian (as Munch was of this nationality), French (as it seems to be painted in & the first exhibition seems to have been in France), or German (as Der Schrei der Natur was the initial title chosen by Munch)?
Also, one should think that a short glimpse on the list of variant titles suffices to get an idea on the languages of publications the piece was published in (which could be in turn in a variety of languages: original, adapted, untitled etc.). As the scope of the database is growing in pieces as well as in languages we really will run into multiple heated debates on the language of art pieces that we could easily avoid (because it's futile and none of our business). Stonecreek 05:48, 30 December 2016 (UTC)
(outdent) I think we have to agree (and update documentation appropriately) on what the language field means. If we agree that for artwork, the language field merely represents the language of the publication the artwork appears in, then this concept is manageable. If it is an attempt to assign language to the art itself separate from the publication, then this becomes the unmanageable mess Stonecreek describes. I always envisioned the first definition and so have been okay with the field. If the latter, then I would object as well. It's clear we don't have agreement on that yet. Not just Stonecreek's comments, but also Chavey's "there are occasional pieces of art that have text within them, where the language is relevant" comment above. If a cartoon in an English language magazine happens to use a foreign phrase, it would still be English by the first definition. As for Ahasuerus's question three above, for artists I would recommend the working language be the artist's "native" language and not the language of the publications they most often appear in within ISFDB. Caspar David Friedrich was 19th-century German Romantic landscape painter, but publications in the database using his work are, except for one French, English. -- JLaTondre (talk) 14:22, 30 December 2016 (UTC)
I'm afraid that the first version won't work: See pp. 34-36 of this publication (Staatenschiffe der Tarnii KOLTOROC, artwork & shortfiction): if this would be reprinted in an English publication (for example in a nonfiction work on the Perry Rhodan series) the artwork would be English but the accompanying shortfiction German (both pieces bearing the same title proper). We would deal with artwork completely contrary than with texts.
If we stick to that idea we really are down the road of endless discussions on the language of specific artworks AND WE REALLY HAVE TO DO MORE MEANINGFUL THINGS WITH THE FICTION PART OF THE DATABASE.
The idea of sticking a language to an artwork may look nice but it does make things very complicated. Stonecreek 06:31, 31 December 2016 (UTC)
Yes, it does work. The vast majority of publications are single language. A English language publication is most likely going to use a translation of the German shortfiction; not the original German so there is no disconnect. For the few (relatively speaking) multi-language publications, a simple corollary to the original definition solves any issues: if the artwork record is for the publication as a whole (cover art, a single record entered for all artwork, fep, etc.), use the publication language. If the artwork is for a specific content in the publication, use the language of the that content. While I'm sure a more convoluted hypothetical can be imagined, I'd wager it would actually be an issue with our ability to only assign a single language to a text, not specific to artwork in general. -- JLaTondre (talk) 13:43, 31 December 2016 (UTC)
I suspect that your proposal with work in 99.9% of all cases. The only time a question may arise is when dealing with art books and other publications which reprint covers/interior art. For example, The Baum Bugle has reprinted Russian and Japanese illustrations from Oz and Oz-related books. They even kept the original captions in some cases. Should a Russian/Japanese/etc cover reprinted in an English/Polish/etc book keep the language of the original book or should it be assigned the language of the publication? Ahasuerus 20:47, 31 December 2016 (UTC)
I concur, "If we agree that for artwork, the language field merely represents the language of the publication the artwork appears in" is exactly (and no more) what I'm defending. I'm only interested to know that an illustration is used by such and such publication and in what language are these publications, the whole data being avaliable at a glance (that is without having to go to each publication title -that's two clicks away at best-). In fact, I'm just simply not interested to know in what language "are" Foss' paintings, just the language of the books they appear on. Hauck 14:59, 31 December 2016 (UTC)
Still, it would suffice in 99.9 % of the cases to look on the titles displayed to grasp what language it is. There are oh so few cases ("Solaris" or "Neuromancer") where you have to dig a little deeper.
There are ever more artists who aren't bound to one national market, and tend to title their works in English (like Volkan Baga). Their works would be assigned French in a French publication, German in a German publication & English in an English language publication and varianted. Quite often it'd come out that not the English publication (say, in the Spectrum series) was the first, but one in another language: you'd have to revise all the varianting, instead of just merging the identical titles which would represent identical artworks. That'd be quite absurd!
To say it again: texts are made out of language, art is not! We'll run into trouble assigning language to art - we already have (absurd for we aren't an art database).Stonecreek 15:47, 5 January 2017 (UTC)

Availability update - 2016-12-27

Sorry about the somewhat cryptic "availability updates" earlier, but I didn't want to be too specific until I knew the prognosis.

Basically I ran into some unexpected health issues a couple of weeks ago. It looks like things should be back to normal in the near future although my productivity may be on the low side for a while. I don't expect a permanent change in cognitive abilities at this point. Well, aside from the fact that, as John W. Campbell, Jr. once pointed out, no one is getting any younger :) Ahasuerus 23:33, 27 December 2016 (UTC)

Take care of your health - anything else can wait. Wish you a speedy recovery (or improvement if full recovery is not in the cards). Annie 23:43, 27 December 2016 (UTC)
I hear you. Take your time and get better soon. (that almost sounds contradictory) ···日本穣 · 投稿 · Talk to Nihonjoe 23:55, 27 December 2016 (UTC)
That's too bad -- best wishes. --Vasha 01:21, 28 December 2016 (UTC)
Thanks for the kind words! There has been no discernible permanent damage, so full recovery is likely. Looking forward to more patches in 2017! :-) Ahasuerus 17:32, 28 December 2016 (UTC)
While aching with a minor illness I hadn't covered the wiki (& the db) for the last days. I wish you the very best for 2017 (and the rest of 2016!). Christian Stonecreek 17:37, 29 December 2016 (UTC)
Glad to hear that things appear to be developing in the right direction (uphill). Keep on that road! (I say, writing from the emergency room where they seem to be getting my wife on that road as well.) Chavey 17:06, 30 December 2016 (UTC)
Thanks! Hopefully things will work out for the best for all of us in 2017! Ahasuerus 17:55, 30 December 2016 (UTC)

Malzberg's Gather in the Hall of the Planets

We have this as a NOVEL, but by all calculation that I made it really can only be a NOVELLA by our standards: even with 300 words per page (which I'd doubt) in its first publication this wouldn't add up to 40,000 words that'd be necessary to characterize it as a NOVEL. It also doesn't stand up to that size with the German publication from 1985. Are there any differing (higher) calculations (else I'd make the necessary transformation)? Stonecreek 14:24, 30 December 2016 (UTC)

Personal tools