ISFDB:Moderator noticeboard/Archive 16
New cleanup report - Omnibuses without Contents Titles
A new cleanup report, "Omnibuses without Contents Titles", has been added to the Cleanup menu. It looks for Omnibus pubs without at least one Novel, Collection or Anthology title. Ahasuerus 02:40, 1 July 2014 (UTC)
SFE3 reconciliation update
I have sent Dave Langford a cross-reference list of ISFDB author pages linking to SFE3. Hopefully he will be able to update the relevant SFE3 pages in the foreseeable future.
P.S. There are 500+ unreconciled URLs on the SFE3 cleanup page, but I figure it will take a while to process them since most of them are for authors that ISFDB doesn't know about. I will send Dave another list once the SFE3 cleanup page is clean. Ahasuerus 03:19, 1 July 2014 (UTC)
- Because the remaining authors (out of an original list of 5000!) are those that don't have publication records in the ISFDB, I don't see how they can be reconciled. In fact, the title of the script, "Authors with Missing SFE3 URLs", is no longer valid. It would work well as a project for some enterprising soul. The majority fall within one of these categories: Pre-20th century authors of lost race and utopian works, authors of soft-core/erotic works with sfnal elements, and non-English language authors (mostly Asian authors). There were also some authors of manga and graphic novels, and persons from the fields of film and television that fall outside the ISFDB's policy of inclusion. I tried to "reconcile" those from the list when I was working it, but there may be some that slipped through. Mhhutchins 03:41, 1 July 2014 (UTC)
- True, it's more of an "add some books and authors that we are missing" project than a "reconciliation" project at this point. Perhaps we could copy the outstanding SFE3 URLs to a Wiki page where all editors could work on them. Ahasuerus 04:30, 1 July 2014 (UTC)
- I would be interested in working on the pre-20th century authors. I've spent a reasonable amount of time on pre-19th century stuff, and am willing to work forward as well. Chavey 00:05, 2 July 2014 (UTC)
- If you're going to tackle the job of adding those authors to the database, I recommend the first volume of Reginald's reference work. I started a project of adding missing titles from Reginald a few years ago, but it stalled after about 50 pages. Other things got in the way, but I hope to eventually get back to it. Mhhutchins 02:27, 2 July 2014 (UTC)
Another reminder about CHAPTERBOOK submission by Fixer
Fixer doesn't automatically create a content record for the SHORTFICTION in a CHAPTERBOOK record. So moderators handling Fixer submissions must keep that in mind when accepting a CHAPTERBOOK typed submission and add a content record after acceptance. (This is true for all submissions, since most new editors aren't aware of it either.) There are currently about a dozen Fixer-submitted pub records which don't have a content record. (These can be found in this clean-up script.) Thanks. Mhhutchins 15:21, 5 July 2014 (UTC)
I have cleaned up Help:Screen:EditAward and Help:Screen:AddAward to the best of my ability. I have also created Help:Screen:LinkAward, which I believe covers everything that editors will need to know when entering/editing awards. Could someone please review the Help pages and see if anything stands out? I have been moving things around for a few hours and it's all a blur at this point, so I may well be missing something obvious.
Oh, and before I forget, as far as the "IMDB title" field is concerned, we may want to discuss it further. At one point we linked directly to IMDB's title pages, but then IMDB changed their naming conventions (twice, I think.) Having been bitten once or twice, we adopted the current approach, which is less user-friendly, but also less likely to cause broken links in the future. Perhaps now that IMDB's URLs have been stable for a number of years, we can start entering URLs in this field again. Ahasuerus 00:16, 6 July 2014 (UTC)
Fixer and publishers
When you see that Fixer is using an uncommon form of a publisher's name, e.g. "Jabberwocky Literary Agency, Inc." instead of "Jabberwocky Literary Agency", could you please let me know so that I could regularize Fixer's logic? TIA! Ahasuerus 03:13, 17 July 2014 (UTC)
An editor and/or moderator has created/accepted submissions which credit Alisa Krasnostein as the author of five collections. I'm assuming this was done because she was nominated for awards as the editor of the five titles. This is not the proper way to handle editors and award nominations. I don't know how to link an editor to an award nomination for a collection, or even if it is possible with the current software, but we can not bypass the ISFDB standards by creating false parent records to acknowledge an editor for a collection. If someone can not come up with a reasonable solution on how to handle this, I will be deleting the titles from the database, and relinking the award records to the correct title record. Thanks. Mhhutchins 14:41, 19 July 2014 (UTC)
- Same thing has been done to other editors, e.g. this record. This should have been discussed before undertaken. They all appear to be for Australian awards. Mhhutchins 14:41, 19 July 2014 (UTC)
- I completely agree. I would recommend that the editor be indicated in the notes on the title record for the collection, which is how this should have been handled. In some cases, it also appears that not only do we have two title records (with and without the editor), but also two awards with one pointing to each title. In those cases, I'd recommend that we delete the award pointed to the erroneous title record and then merge the two title records into one without the editor and with no parent title. If there is only one award, I'd have to see if merging the collection titles ends up with the award pointed correctly. If it doesn't we may have to re-point the awards to the correct title after the titles have been cleaned up. --Ron ~ RtraceTalk 20:46, 19 July 2014 (UTC)
- I will give the original moderator a few more days to present a case for keeping the records as they currently are. If there's no response, I'll clean-up the records based on ISFDB standards. Mhhutchins 16:37, 23 July 2014 (UTC)
- Discussion noted. Please give me a day or 2 more to respond. --clarkmci / j_clark 01:59, 29 July 2014 (UTC)
- [Later] From my experience with the Aurealis Awards and the Ditmar Awards, I've noticed 3 situations where an award nomination record could/should link to more than one other ISFDB record:
- (i) More than one title is cited, as a group, for 1 award. (usually in the same series, but not always the full series) (multiple title records)
- (ii) Author and illustrator of a work are cited for 1 award (also multiple title records but slightly different relationship)
- (iii) The editor of a collection is also cited for the award (title record of the collection, plus an "author" without a specific title).
- Adding a dummy title record that includes the editor (with reason in the Notes field) was an attempt at a temporary work-around for the 3rd situation. Is the preferred temporary solution to enter 2 award records to capture this award information for the title - one linking to the normal title record, and the other an "untitled" award with the editor? (Note that the Ditmars add the editor to their collection citations, Aurealis doesn't, so get the situation where I linked the Ditmar award to the dummy title & the Aurealis to the normal title record - maybe I should have put the dummy as a vt, not the other way around?.)
- Thanks for the guidance (and for the recent work on the Awards module - Aurealis & Sir Julius Vogel categories now have a display order set up, Ditmar category ordering is "in progress".) --clarkmci / j_clark 22:30, 30 July 2014 (UTC)
- Given the number of possible permutations, it sounds like we may need to add a "Note" field to Award records. Would that help? Ahasuerus 22:33, 31 July 2014 (UTC)
- I would still advise against creating any dummy title records as variants just to link these awards. A Note field to Award records would be very much appreciated, as we can add more data about the specific award.
- There is a similar situation with awards given to series and art books which credit both the author and the artist. See this clean-up list. I've gone as far as I can to clean-up this list. We can only link awards to titles and not series, and if we link the art book awards to the book's title record, we would be losing data about the artist nomination/win. In the second case, is there a possibility of creating a second link from an award record? This would also solve the situation above with editors being recognized with award nominations. Mhhutchins 23:06, 31 July 2014 (UTC)
- OK, I have created a feature request to add a Note field to Award records.
- As far as linking an award record to 2+ titles goes, that would be a major change. I'll have to think about it. Ahasuerus 02:19, 1 August 2014 (UTC)
- There will always be oddities in the awards, which are best handled by untitled awards. But having a notes field to explain what's odd about them would be very useful. Chavey 02:05, 17 January 2015 (UTC)
Post-submission pages and moderator review pages
As of last week, all of our post-submission pages display the same data as their "moderator review" counterparts. Upon reflection, I wonder if it may be better to let moderators skip the post-submission page and go straight to the moderator review page when they create a submission. Are there any reasons not to do this? Ahasuerus 03:31, 22 July 2014 (UTC)
- I'll take it as "no objection" :) Ahasuerus 04:51, 30 July 2014 (UTC)
- Missed this when it was posted for some unknown reason. But I have no objection to removing the review page for moderators. Much better if it goes straight to the moderator page. For the first few dozen submissions I thought I was doing something wrong. Mhhutchins 08:06, 30 July 2014 (UTC)
Fixer and 2012 e-AddPubs
Now that 2013 and 2014 have been cleaned up, I am starting 2012, which is comparable to 2013 at 5,900 ISBNs. FYI, here is how many e-AddPubs Fixer has in his queues for the following years:
- 2012 - 5,900
- 2011 - 5,300
- 2010 - 4,000
- 2009 - 3,000
- 2008 - 1,650
- 2007 - 1,290
- 2006 - 590
- 2005 - 520
- 2004 - 440
- 2003 - 390
At the rate we are going, I gather that we should have all of the queued up e-AddPubs done by the end of the year. Ahasuerus 19:36, 31 July 2014 (UTC)
- January has been submitted. Ahasuerus 16:47, 1 August 2014 (UTC)
- February has been submitted. Ahasuerus 22:32, 3 August 2014 (UTC)
- March has been submitted. Sorry about the volume -- Fixer had run into certain technical issues and I needed to flush some of his queues to get everything back to normal. Ahasuerus 01:49, 5 August 2014 (UTC)
- April has been submitted. Ahasuerus 03:47, 7 August 2014 (UTC)
- Fixer's logic has been further adjusted to reduce the number of superfluous parenthetical "suffixes". Ahasuerus 06:25, 8 August 2014 (UTC)
- Wonderful. I noticed that my follow-up edits have been greatly reduced. Thank you. Mhhutchins 19:41, 8 August 2014 (UTC)
- Fixer has to walk a fine line between chopping off extraneous data and deleting important things like embedded publication series designations. Hopefully he hasn't gone too far... Ahasuerus 22:02, 9 August 2014 (UTC)
- May has been submitted. Ahasuerus 17:38, 8 August 2014 (UTC)
- June has been submitted. Ahasuerus 22:02, 9 August 2014 (UTC)
- July has been submitted. Ahasuerus 01:12, 11 August 2014 (UTC)
- August has been submitted. Ahasuerus 21:20, 12 August 2014 (UTC)
- September has been submitted. Ahasuerus 18:57, 15 August 2014 (UTC)
[unindent] Are we finished with 2012? I see the latest set of Fixer submissions is more recent publications. Mhhutchins 15:52, 17 August 2014 (UTC)
- Not quite -- there are still 100+ October 2012 ISBNs in one of Fixer's queues plus November and December. However, the other day Fixer completed a monthly run against the usual suspects, so he is now aware of additional e-AddPubs. I have already submitted 2013 and now I am in the process of submitting 2014. It's not a lot, only 200+ ISBNs. Ahasuerus 17:15, 17 August 2014 (UTC)
- October has been submitted. Ahasuerus 21:37, 17 August 2014 (UTC)
- Newly discovered 2014 e-AddPubs have been submitted. Ahasuerus 00:27, 18 August 2014 (UTC)
- November has been submitted. Ahasuerus 01:19, 19 August 2014 (UTC)
- December has been submitted. Nicely done! Ahasuerus 05:45, 20 August 2014 (UTC)
Consider Phlebas, PV'ed by BLongley
There are no other verifiers for this pub, so I am writing my question here, as per instructions on BLongley's talk page. I noticed that the pub record has 1997 as the publication year, but the reprint history cited in the pub notes indicates that the year should be 1996. Can I change this? Thanks, Patrick -- Herzbube Talk 11:27, 13 August 2014 (UTC)
- Yes, please do. Mhhutchins 13:31, 13 August 2014 (UTC)
Post-approval page for unmerge submissions
The post-approval page for unmerge submissions has been changed to display links to newly created title record(s). Ahasuerus 04:47, 17 August 2014 (UTC)
Bad HTML in Notes
The script formerly known as "Bad HTML in Publication Notes" has been renamed, enhanced to find mismatched angle brackets in all record types (Publications, Titles, Series, Pub Series, Publishers, Awards, Award Types, Award Categories) and moved to a "Notes" section at the bottom of the page. In addition, its performance has been improved. Ahasuerus 04:41, 19 August 2014 (UTC)
Mismatched double quotes in Notes
A new cleanup script, "Mismatched Double Quotes", has been added to the Notes section. Some of them are relatively harmless, but a few can cause display problems if they interfere with embedded HTML. Ahasuerus 15:58, 19 August 2014 (UTC)
New cleanup script - Titles without pubs
A new cleanup script, "Titles without Publications", has been added to the menagerie. At the moment it's limited to pub-less COVERART and INTERIOART records. These two title types are our low-hanging fruit since in most cases they can be safely deleted. I plan to add the other title types one type at a time since they will require additional research to identify and enter missing pubs. Here is a snapshot of how many different pub-less title records we currently have in the database:
+--------------+------+ | ANTHOLOGY | 37 | | COLLECTION | 35 | | COVERART | 60 | | INTERIORART | 230 | | EDITOR | 168 | | ESSAY | 513 | | NOVEL | 804 | | NONFICTION | 244 | | NONGENRE | 93 | | OMNIBUS | 15 | | POEM | 156 | | SERIAL | 35 | | SHORTFICTION | 1306 | | CHAPTERBOOK | 10 | +--------------+------+
In addition, "Omnibuses without Contents Titles" has been moved to the "Publication" section of the cleanup menu. Ahasuerus 02:47, 20 August 2014 (UTC)
- I've been working on this off and on for the past few days and got the art titles down to less than 100. Before you add the other types, I was wondering if there's a way records can be marked to keep them from reappearing once a moderator has determined they are good records which should remain in the database. We can always add a note telling other moderators not to delete them, but how can we can rid them from the clean-up script. Thanks. Mhhutchins 23:41, 22 August 2014 (UTC)
- It's certainly possible and it seems desirable, although the implementation will be a little trickier than it may appear. Let me add it to the list... Ahasuerus 00:44, 23 August 2014 (UTC)
- The art records have been cleaned. Please add a few more types when you get a chance, preferably the EDITOR, CHAPTERBOOK, SERIAL, OMNIBUS, ANTHOLOGY, and COLLECTION types. (In that order if you're unable to do all of them at once.) Thanks. Mhhutchins 04:59, 28 August 2014 (UTC)
- Will do! I still have 120 ISBNs to process to wrap up October (the last 100-200 are always time-consuming due to the way they are queued up), but then I should have a solid 2 weeks to work on development. Ahasuerus 05:50, 28 August 2014 (UTC)
- Feel free to allow Fixer to submit some of those October titles, and I'll gladly work them. Your call. These Fixer ebook submissions are just too damn easy. :) (Walks away whistling.) Mhhutchins 14:50, 28 August 2014 (UTC)
- Done (see below), although in this case "October" actually means "August" :-) When Fixer's data acquisition process is run, it typically finds a fair number of recently published books which were either (a) not listed earlier or (b) not classified as SF. The latter are particularly problematic because many juvenile books are only listed as "juvenile" or "juvenile adventure" even though the "adventure" may have take the protagonist to Alpha Centauri. Similarly, paranormal mysteries are often listed as "mysteries". Ahasuerus 18:11, 28 August 2014 (UTC)
January has been submitted. February is in progress. Ahasuerus 05:24, 21 August 2014 (UTC)
- February has been submitted. March is in progress. Ahasuerus 22:21, 22 August 2014 (UTC)
- March has been submitted. Ahasuerus 03:13, 25 August 2014 (UTC)
- April has been submitted. Ahasuerus 05:48, 28 August 2014 (UTC)
- May has been submitted. Ahasuerus 07:30, 30 August 2014 (UTC)
- June has been submitted. Ahasuerus 04:17, 2 September 2014 (UTC)
- July has been submitted. Ahasuerus 03:10, 3 September 2014 (UTC)
- August has been submitted. Moving right along! Ahasuerus 04:50, 4 September 2014 (UTC)
- September has been submitted. Ahasuerus 06:10, 6 September 2014 (UTC)
Moderator review logic for NewPubs has been modified to display yellow warnings about new/pseudonymous/disambiguated Contents authors. ClonePub is next. Ahasuerus 21:33, 22 August 2014 (UTC)
- Great. I've thought about that, but always thought it would be too complicated to implement. Sorry for underestimating your software savvy. Thanks. Mhhutchins 21:58, 22 August 2014 (UTC)
- Well, it's mostly a question of man-hours or, in some cases, man-years. For example, it took Microsoft less than 5,000 man-years to develop Windows 7 (25 teams times 40 people per team times a few years) and I am reasonably sure that we could get the ISFDB software in really good shape if we could invest 5,000 man-years :) Ahasuerus 22:26, 22 August 2014 (UTC)
(unindent) Question: Do we want to display moderator warnings for pseudonymous/disambiguated Contents authors if the title record in question is being auto-merged? It seems kind of unnecessary since the title is already present in another publication. Ahasuerus 02:22, 23 August 2014 (UTC)
- ClonePub is done. Edit Pub is next. Ahasuerus 02:54, 23 August 2014 (UTC)
Fixer's dead tree submissions
As per Michael's suggestion above, I have asked Fixer to create a dozen NewPub submissions for dead tree editions from August 2014. They are a mixed bag: one book is by a previously self-published author signed up by 47North, another collects stories previously published in minor magazines, a third needs to be added to a series, etc. The bottom line is that they need varying degrees of TLC and may present different types of challenges. If the approving moderator(s) have questions for Fixer, please post them here and I will try to explain how certain decisions were made.
Oh, and something else that I noticed a few days ago. When entering/approving books scheduled for release in the future, please double check the data currently on file. UK/US editors love to add/drop articles, e.g. "End of Days" vs. "The End of Days", which means that we need to create VTs. Ahasuerus 18:04, 28 August 2014 (UTC)
- If the title or author is unfamiliar, I would suggest checking the author's summary page to look for similar titles, series, etc. Mhhutchins
- Only 20 2014 dead tree NewPubs left in Fixer's queue! Ahasuerus 04:01, 29 August 2014 (UTC)
(unindent) Now that 2014 has been processed -- thanks, Michael! -- we need to decide what we want to do with Fixer's AddPubs and NewPubs going forward. AddPubs are easy since I can simply continue submitting e-AddPubs for 2011, 2010, etc until we process them all. As far as NewPubs go, we have various choices. Here is what Fixer's main internal queues look like as of this evening:
Mon 2014 2013 2012 n-p n-e 1-p 1-e| n-p n-e 1-p 1-e| n-p n-e 1-p 1-e| ------------------------------------------------------------------------ NON 42 0 0 0| 0 0 2 0| 360 0 0 0| Jan 0 1920 0 115| 0 1293 2 19| 611 891 23 9| Feb 0 1269 0 41| 0 1163 2 14| 4 902 104 14| Mar 0 1173 0 98| 0 1264 0 8| 1159 990 4 21| Apr 0 1470 0 10| 0 1374 3 12| 1134 995 2 8| May 0 1358 0 26| 0 1193 7 16| 1190 1067 4 12| Jun 0 1640 0 17| 0 1177 4 16| 1086 977 1 17| Jul 0 1405 0 39| 0 1225 1 15| 935 1186 2 13| Aug 0 657 0 32| 0 1103 6 17| 1218 1129 6 13| Sep 0 233 0 51| 0 1152 40 21| 1375 1270 7 15| Oct 0 307 0 4| 0 1582 158 24| 12 1424 209 15| Nov 513 160 60 0| 0 2243 62 13| 11 1118 180 10| Dec 357 132 22 0| 0 2596 103 23| 15 1235 133 27|
The first column to the right of the "month" column is "n-p" or "new paper books", i.e. unprioritized paper-based ISBNs. The second one is "n-e" or "new e-books", i.e. unprioritized e-books. The third one is "1-p" or "priority 1 - paper-based" and the last column is "1-e" or "priority 1 - e-books".
As you can see, we have a lot of unprioritized "new e-books" and some "priority 1" e-books published in 2014 and 2013. The reason most of the "priority 1" e-books haven't been submitted yet is that they lack price information. If I submit them, the approving moderator may have to spend some time looking for price information. Alternatively, I could spend some time on prioritizing "new ebooks", moving the cream of the crop to the "priority 1" queue and then submitting it. Or I could submit some of the 400-ish "priority 1" paper-based books published in 2013 that have been waiting their turn. (I have to warn you, though, that older "priority 1" books tend to be lower profile because Fixer doesn't identify them until some months after publication.)
So I guess I have two questions. First, what should be the AddPub/NewPub ratio? Second, what kinds of AddPubs would you like to see next? Ahasuerus 02:06, 30 August 2014 (UTC)
- Ideas? Suggestions? :-) Ahasuerus 16:59, 1 September 2014 (UTC)
- Sorry, missed this update to the message. My suggested order of release to the queue (thanks for asking):
- 2014: 1-p, 1-e
- 2013: 1-p, 1-e
- 2012: 1-p, 1-e
- 2014 - 2012: n-p
- 2012 - 2014: n-e
- From this order, you can see that I believe that the non-prioritized ebooks should be the last submitted. (I can't imagine when we'll be able to get to them anyway, unless a few more moderators take a small portion of their ISFDB time to assist.) Mhhutchins 05:29, 2 September 2014 (UTC)
- It looks like my explanation of the "unprioritized" (wasn't it kind of a joke word just a few years ago? and look at it now!) queue was too cryptic.
- What happens with Fixer is that every month he ventures out and finds all kinds of ISBNs that may be of interest to us. He adds them to the unprioritized queue and then I have to review them and assign priority. If an ISBN is of no interest to us, e.g. a Star Trek calendar or a D&D module, I "reject" the ISBN. "Rejection" means that the ISBN is moved to a special "rejected" queue which Fixer checks at data acquisition time to ensure that it doesn't get re-added to the unprioritized queue.
- Another choice that I have is to move the ISBN to the "suspended" queue, which I primarily use for picture books for small children and for foreign language books. It's not that I have anything against the latter, but our Anglophone sources are usually inadequate to create robust records for them. Well, except for some Spanish language books published in the US.
- Finally, I can assign an ISBN to queues 0, 1, 2 or 3. As described earlier: "Queue 1 is the one that gets submitted to ISFDB. Queue 2 is "we'll get to these guys eventually" and includes Ellora's Cave, amateur reprints of the classics, new books by self-published authors some of whose works are already in ISFDB, etc. Queue 3 is the nether regions of the self-publishing world. Queue 0 is where incomplete records go; it gets revisited once every few months to see if better data may have become available."
- Here is another Fixer report showing how many ISBNs we have in queues 0, 1, 2 and 3 (no breakdown by e-book vs. paper book in this case):
Mon 2014 2013 2012 new 0 1 2 3| new 0 1 2 3| new 0 1 2 3| ------------------------------------------------------------------------ NON 42 21 0 1 0| 0 126 2 2 21| 360 3 0 2 0| Jan 1920 115 115 406 761| 1293 28 21 358 483| 1502 7 32 81 55| Feb 1268 48 41 315 523| 1163 20 16 404 464| 906 4 118 208 288| Mar 1173 50 97 398 618| 1264 23 8 426 471| 2149 2 25 130 15| Apr 1470 84 10 367 628| 1374 36 15 507 519| 2129 0 10 6 4| May 1357 116 26 421 579| 1193 19 23 363 576| 2256 1 16 11 8| Jun 1640 86 17 428 508| 1177 32 20 3401116| 2062 0 18 8 5| Jul 1405 138 39 329 485| 1225 32 16 356 860| 2121 3 15 4 5| Aug 657 143 32 87 129| 1103 28 23 307 508| 2347 1 19 0 4| Sep 233 136 51 28 7| 1152 30 61 336 542| 2645 3 22 10 8| Oct 307 125 4 43 1| 1581 68 182 368 727| 1436 37 224 344 488| Nov 673 7 60 3 0| 2241 46 75 323 639| 1129 32 190 287 504| Dec 489 0 22 1 0| 2596 97 126 314 575| 1250 29 160 350 486|
- Now, the problem that I have with prioritization is that it can take a surprisingly long time. If Fixer finds a new SF book by "Adrian Tchaikovsky", it's almost certain that it's by "our" Adrian Tchaikovsky and it only takes a second to add it to Queue 1. However, if Fixer finds a book by "Bill Smith" (to use a random example), it can take some time to determine whether it's by "our" Bill Smith or by some self-published author and needs to be moved to Queue 3. And that's just one type of problem that I have to deal with.
- Since Fixer finds up to 5,000 new ISBNs every month and since up to 2,000 of them have to be processed manually, it can take a while to prioritize everything, usually 1-2 days per month. That's why we have so many unprioritized ISBNs prior to 2013 -- it wasn't until after my retirement in late August 2012 that I had the time to make Fixer's handling of these queues more robust and that I had the time to prioritize all of the new ISBNs as they come in. I still need to review 1990-2012 and prioritize the 300,000+ unprioritized ISBNs that we have on file, but that's a daunting task. I try to do it in small chunks, one month at a time, in between software improvements and processing new ISBNs. And I have barely scratched the surface with e-books, a relatively recent addition.
- So, to bring it back to the issue at hand, "unprioritized" doesn't mean low priority. It just means that a priority hasn't been assigned yet. Ahasuerus 16:44, 2 September 2014 (UTC)
- Thanks for explaining the difference. The sheer number of Fixer's unsubmitted records is overwhelming, mind-numbing in fact, and a task whose completion appears to be unachievable. All I can suggest is that you submit them based on the priority you've already established, regardless of whether they're print or electronic, starting with the most recently published.
- Imagine if we had ten moderators who could commit to handling a batch of 50 submissions, which would take less than an hour. Three batches a week by each moderator would handle 15,000 Fixer submissions in ten weeks. That would clear the first two of your prioritized queues and make headway into the others. Mhhutchins 17:39, 2 September 2014 (UTC)
- The good news is that we are now able to keep up with new ISBNs. Prior to early 2013, we had been slowly falling behind because Fixer had no prioritized queues.
- Also, I have been improving Fixer's logic to make the prioritization process less time-consuming. Recently I have added the ability to separate AddPubs from NewPubs and also the ability to separate ISBNs associated with "known authors" and "established publishers" from everything else. When I start working on prioritizing a new month, I usually sort out the AddPubs first, which doesn't take long. Then I do the NewPubs for "known/established authors/publishers", which takes longer because some "known authors" are really self-published authors added to ISFDB when Fixer was younger and less experienced. Finally, I prioritize the remaining ISBNs, which are easy because in 95+% of all cases they can be safely relegated to Queue 3. With these improvements in place, I hope to be able to clean up a significant part of the backlog in the foreseeable future. Just need more hours in the day :-)
- Anyway, I guess I will continue submitting e-AddPubs for 2011 and then 2010, 2009, etc. I will also start submitting small quantities of p-NewPubs for 2013. Perhaps 10 NewPubs for every 50 AddPubs? Ahasuerus 20:14, 2 September 2014 (UTC)
Wild Cards series
While checking on book six in the series Ace In The Hole I came at the series via the editor George R. R. Martin. There seem to be some inconsistencies in his being credited as the editor. Volume 6 has a note crediting him, but no reference. I haven't a copy of volumes 10, 12 and 20. Doug 21:05, 28 August 2014 (UTC)
- Look at the note field of the title record. It explains that Ace in the Hole is a collaborative novel by various authors, edited by Martin and Snodgrass. Novels are credited to the authors and not the editor. Volumes 10 and 12 are solo novels, and 20 is another mosaic novel by various authors. If you click on the series link, you'll see all the titles, regardless of their credited author or editor. Not all works in a series will be displayed on all the authors' summary pages. Only those which they wrote, or in the case of anthologies, edited. Hope this helps. Mhhutchins 22:05, 28 August 2014 (UTC)
- I only have the two copies - I and VI so can only speak to the confusion between the two of them. Both are identified as Mosaic Novels, both are edited by George R.R. Martin and both are copyrighted to Martin. He also contributed to the first as an author and is rightly listed as such. But why would I give him an editor credit but only a note for the VI? Doug 15:06, 29 August 2014 (UTC)
- Because the first volume is an anthology. It has distinct stories, each with their own title page, and each crediting the individual author. This is the ISFDB definition of an ANTHOLOGY. Ace in the Hole has none of that. It is a seamless novel, written by five different authors. There is no division of the credit. Editors are not credited in the author credit field of a record typed as NOVEL. Editors are credited in the author credit field of a record typed as ANTHOLOGY. There is a clear distinction between the two publications and each are handled according to ISFDB rules. Since Martin is not credited as one of the authors of Ace in the Hole, he is not credited in the author field of the ISFDB title record. Mhhutchins 16:56, 29 August 2014 (UTC)
Destructive Wiki Editing and Responsiveness to Primary Verification Inquiries.
Another editor took an irrational, to my mind, dislike to me nearly a year ago. I shan't rehash it again, but the original discussions are here and here. At the end of the second thread, Michael appealed to us to be more civil in our discussions, and I have kept that in mind in my subsequent posts. Since the other editor, Bill, has the second highest total of primary verifications, it is inevitable that I am going to want to make changes or have inquiries about those he has verified. Over the past year, he has not responded to a single one of my inquiries or requests. This has forced me to rely on other verifiers to reply, or to leave publications in an incomplete or incorrect state. Earlier tonight, I started a discussion about a book where he is the first primary verifier with myself and two others also listed. Because he was first, I started the discussion on his talk page leaving notes on the others' pages to join the discussion. Alas, he no longer seems content to simply ignore my questions. This time he deleted my inquiry. This, of course, also broke the links to the discussion that I had informed him I was leaving on the other verifiers' pages. I am at a loss as to how to deal with this sort of destructive editing. I believe there is an implicit contract when verifying a publication that the verifier agrees to field inquiries about the publication. If an editor refuses to do so, should he remove his verification? If an editor does not respond to inquiries about suggested changes, should that be treated as tacit assent? Clearly the best solution would be resolve whatever the issue is that has caused him to behave this way. However, since he has refused to engage in the discussion since my last post responding to his accusations, I don't see how that can be accomplished. I know that we don't exactly have a hierarchy here, but I would ask that the moderators urge Bill not to make destructive edits of other editors' posts going forward. Also, since he won't engage with me, if any other editor wants to try to mediate in order to resolve his underlying issue with me. I would welcome that effort. For the post that he deleted, I will re-host the discussion on my own talk page (as the second primary verifier). He is welcome to participate in that discussion, but if he does not, I will assume that we have his assent and that by deleting the original discussion, he chose not to participate and left the decision to the other verifiers. Thanks. --Ron ~ RtraceTalk 04:11, 30 August 2014 (UTC)
- Well, the good news is that we keep the last 50+ versions of each Wiki page, so anything deleted can be accessed or recreated easily. However, this also raises a larger issue. Moderators need to be able to communicate effectively as described in Moderator Qualifications. At the very least moderators should be able to communicate with other moderators or else our editing model may become unworkable. I'll leave a message on Bill's Talk page to see if we can resolve this issue. Ahasuerus 05:50, 30 August 2014 (UTC)
Changes to Note-specific cleanup scripts
The two Note-specific cleanup scripts have been changed as follows:
- The body of each report will be recreated nightly from now on
- It should take less than a second to access each page
- If you resolve a problem, the fixed record will immediately disappear from the list when you refresh the Web page
- The list of records with "mismatched double quotes" has been limited to records that contain "http" or "HTTP" in the Note field. We have about 600 affected title/pub records, although in many cases the missing/additional quote character is not related to "http"
Ahasuerus 05:42, 30 August 2014 (UTC)
- Thanks for the change in the logic of the script. While I was the one who requested that mismatched double quotes be the search criteria, it never occurred to me that the use of quotes other than in HTML links was so much more prevalent. In the hundreds of "mismatched double quotes" I corrected, there may have been half-a-dozen that involved bad HTML, which was the purpose of the script. The others were simply used to offset titles in the note field. I'm glad you're able to better define the search, even though I suspect of those 600 records, less than 25 will have the bad HTML we're looking for. In any case, we'll be able to clean those along with the other mismatched quotes. Thanks. Mhhutchins 06:18, 30 August 2014 (UTC)
- I too was at first surprised by the number of mismatched double quotes. I then tried a few approaches to make the logic smarter, but most of them didn't work because there were too many possible permutations resulting in invalid HTML. Eventually I settled on the presence of "http" in the text as the only additional check, which eliminated about 50% of false positives. Oh well, one step at a time :) Ahasuerus 06:56, 30 August 2014 (UTC)
Publications with Suspect Images
I am in the process of re-checking the existence of the cover scans that we link to. Non-Amazon images appear to be very stable -- not a single failure in the first 6,000 images checked. Amazon-hosted images may be a different story, but it will take a few weeks to check all of them since I don't want to flood Amazon with queries (and they would just block my IP anyway.) Ahasuerus 17:39, 30 August 2014 (UTC)
- All 14,000+ non-Amazon images have checked out. I am currently in the process of verifying Amazon-hosted images. As expected, things are not quite as rosy, but not too bad overall -- 14 missing images out of the first 10,400 images checked. Ahasuerus 16:53, 1 September 2014 (UTC)
Forty Signs of Rain artwork
I'm about to perform some edits to the artwork credits of Forty Signs of Rain; both of the UK versions are credited to Dominic Harman in ISFDB, but a pub note makes it seem that this is based on conjecture.
The US version is credited jointly to Dominic Forbes and Tom Hallman. I've written to both Dominic Forbes and Tom Hallman and believe I can assert proper credits now: Dominic Forbes did the original design but Tom Hallman added the Washington DC skyline to the US version, so Forbes should be credited with the UK artwork.
I'm not sure who among the various editors and verifiers made the initial attribution to Dominic Harman, but I'm leaving notes on all of the talk pages and the Moderator Noticeboard (due to an inactive participant). Please feel free to weigh in. Thanks, Albinoflea 05:23, 31 August 2014 (UTC)
- Neither of the UK edition publication records should have credited Harman based on the slim evidence of "matching" the art of previous works in the series. That conjecture should have been left in the Note field and the Cover Artist field should have been blanked. That having been said, the primary verifiers should discuss the matter and have the final say in how to credit the record. Mhhutchins 07:12, 31 August 2014 (UTC)
- Sounds like proof enough to credit Forbes for the UK editions. I agree with Michael, the credit should have been blanked. --Willem H. 09:31, 31 August 2014 (UTC)
- I'll ask Dominic Harman just what his input was, exactly. PeteYoung 12:36, 31 August 2014 (UTC)
- Thanks Pete for checking in with Dominic Harman. I think that completes the circle of possibilities.
- It has been a week, so I suspect I've heard from everyone I'm likely to hear from at this point and I will go ahead and start making changes. Thanks to all for your input. Albinoflea 03:11, 8 September 2014 (UTC)
Duplicate of Arthur C. Clarke's "Die letzte Generation"
This primary verified record (verifier not active here anymore) is a duplicate of this unverified record, both being the first printing of that edition. The unverified one can be deleted. Moreover, the verified record uses English titles for the forewords instead of the German translations, which should be corrected as well (I just submitted a new record for the third edition which contains the correct German titles). Year of the verified record should be set to "2003-12-01".
How's the procedure for records of non-active primary verifiers? Do only the moderators do the changes after getting a notice here, or is it ok if I go ahead and change the records after your ok here?
Addition: same problem with these two records of Joe Haldeman's "Der ewige Krieg": http://www.isfdb.org/cgi-bin/pl.cgi?292249 and http://www.isfdb.org/cgi-bin/pl.cgi?367172 Hitspacebar 19:47, 31 August 2014 (UTC)
- Make the necessary changes to the verified records, and delete the unverified. Note this discussion in the Note to Moderator of the submissions. Thanks. Mhhutchins 21:47, 31 August 2014 (UTC)
Changes to cleanup scripts
The list of cleanup scripts (which I think we should call "cleanup reports") and "Variant Titles-Pseudonym Mismatches" have been improved as follows:
- The 5 reports that are regenerated nightly show the number of outstanding records next to their names. These numbers are as of the time the report was last run and may be higher than the current number. As I move more cleanup reports to the nightly job, the number of reports showing how many records are outstanding will grow. Eventually, a moderator will be able to look at the list and quickly determine which reports need attention.
- As per Michael's suggestion, I have added the ability to permanently mark questionable records as "Resolved". When you click on a "Resolve this record" link, the system sets a database flag indicating that the record is "OK as is". The nightly job which looks for potential problems will ignore "resolved" records. At this time this functionality is limited to the "Variant Titles-Pseudonym Mismatches" report, but I plan to add it to a number of other reports like "Authors with Invalid Names", "Authors with Invalid Spacing" and "Titles without Pubs".
The changes were not earthshaking, but I had to tweak a number of things to get everything to work, so if you see anything unusual, please report it here.
P.S. The "Mismatched Angle Brackets" reports is apparently missing 4 problem records. I suspect that the missing records are titles with problems in the Synopsis field and will look into it next. Ahasuerus 03:13, 1 September 2014 (UTC)
- The "Cleanup Scripts" menu has been renamed. Ahasuerus 03:23, 1 September 2014 (UTC)
- Added a Synopsis section to the "Mismatched Angle Brackets/Double Quotes" reports. The next run of the nightly job should be clean. Ahasuerus 23:12, 1 September 2014 (UTC)
Series - Legends of Ethshar - sequence error?
- No longer. Feel free to make corrections when you see errors in non-primary verified records. Thanks. Mhhutchins 16:12, 2 September 2014 (UTC)
Additional submission queues/states?
At this time a submission can be in one of three states: New, Integrated (i.e. approved) or Rejected. In addition, new submissions can be put on hold. There are three moderator-only Web pages which let moderators view new submissions, integrated submissions and rejected submissions respectively.
We have a couple of feature requests to tweak this behavior, namely:
- FR 298: "Defer Fixer submission", which would add another submission state, "Deferred", a new Web page page to go with it. The functionality seems moderately useful although we don't have that many deferred submissions at this time.
- FR 55: "FLAG (Instead of HOLD) for Attention Of other MOD". Perhaps instead of having a special "FLAG" state, we could simply expand the functionality of the "HOLD" state, but that's implementation details. Again, it seems moderately useful.
Also, if we are going to add the ability to defer submissions, do we also want to create a separate submission state (and a separate Web page) for robotic submissions by Fixer and friends? That way I could ask Fixer to submit, e.g., 200-300 AddPubs without overwhelming the main submission page.
In addition, would it be advantageous to change the term "Integrated" to "Approved"? The former seems to be hardly ever used by editors. Ahasuerus 21:53, 2 September 2014 (UTC)
- 1. DEFERRED - This would work if the moderator deferring the submission will go back later and handle it. Or at least occasionally choose to work the deferred queue. It's hard enough getting moderators to work Fixer's submissions in the main queue. It wouldn't work for me if the moderator is only trying to get a submission out of the main queue. If this is implemented, the deferring moderator would have to provide a reason for doing so.
- 2. FLAG - I'm not certain that this is necessary. If a moderator opens a submission and determines they're unable to handle it, simply leave it in the queue. Another moderator will eventually handle it. That's the current procedure. If I open one that requires the attention of a moderated primary verifier, I'll place it on HOLD and leave a message on their talk page. Unless a FLAG can do that and allows the moderator to specify which other moderator should handle the submission, I don't see the point of it. Mhhutchins 22:13, 2 September 2014 (UTC)
- Yes, that's pretty much what the creator of this FR suggested:
- Dropdown list of MODS to select the person who needs to review. ... Benefits: Allows a Mod to see items requiring their attention in the new submission screen. Allows an Editor to see that Mod X has flagged the submission for review by Mod Y. Allows other MODs to see that a submission is awaiting someone to log in for review (A submission doesn't get reviewed and then ignored by 5 mods waiting for mod 6 to log in - Only Mod 1 reviews and flags and time is saved for Mods 2-5)
- It's not terribly difficult to implement, perhaps a day or two of development time. Ahasuerus 22:35, 2 September 2014 (UTC)
- 3. Yes, change "Integrated" to "Approved". Thanks. Mhhutchins 22:13, 2 September 2014 (UTC)
- Sounds good. Also, what do you think about a separate Web page for "Robotic Submissions" as mentioned earlier? Ahasuerus 22:35, 2 September 2014 (UTC)
- I'm not sure it's needed from a moderator point of view. The rates aren't flooding the queue. However, if it's easier on you to submit larger batches at once, than I think that's a reasonable reason. The only downside is the potential for it to be more easily ignored. -- JLaTondre (talk) 23:06, 2 September 2014 (UTC)
- Sounds good. Also, what do you think about a separate Web page for "Robotic Submissions" as mentioned earlier? Ahasuerus 22:35, 2 September 2014 (UTC)
- [after edit conflict that agrees with my sentiment] Related to what I said earlier, if its only purpose is to separate Fixer submissions so that other moderators can avoid looking at them, then I would advise against it. It's hard enough to get them out of the queue as it is. Moving them to their own queue would only increase the "out-of-sight / out-of-mind" mentality. Mhhutchins 23:08, 2 September 2014 (UTC)
(unindent) "Integrations" have been changed to "Approvals" and "Rejects" to "Rejections".
As far as the issue of a separate page for "robotic" submissions goes, well, sometimes we have a number of moderators working on them at the same time. When that happens, a batch of 30-50 AddPubs can be processed in less than an hour, so I have to check the queue on a regular basis to maximize moderator productivity. And, of course, when I am asleep, there is a gap of up to 8+ hours. It would be more efficient to submit 200-300 AddPubs and not have to worry about them for at least a few hours, but then the queue would get unwieldy since it's limited to 200 submissions per page.
Re: the "out-of-sight / out-of-mind mentality", one thing that comes to mind is that we could make the navigation bar on the left show the number of currently outstanding robotic submissions. Ahasuerus 23:43, 2 September 2014 (UTC)
- If it's easier for you, then add 200-300 Fixer submissions to the regular queue. They're going to be worked by the same moderators anyway, so it doesn't matter how many are there. The other moderators are working their own submissions without even laying an eye on the queue, so it's not going to be burdensome to them, regardless of how many submissions are queued. Mhhutchins 00:11, 3 September 2014 (UTC)
- Thanks for the feedback! I have submitted roughly 100 e-AddPubs -- let's see what we learn from this experiment and then go from there. Ahasuerus 01:42, 3 September 2014 (UTC)
- Wow, that was fast! OK, the next batch is on its way and I will submit another one as soon as I finish assigning country codes. Ahasuerus 02:49, 3 September 2014 (UTC)
Authors with Invalid Last Names
"Authors with Invalid Last Names" has been moved to nightly processing. If a last name identified by this cleanup report is actually valid, you can click on the "Ignore this last name" link and the author will be permanently ignored by this cleanup report.
As a reminder, similar functionality was added to the "Variant Title-Pseudonym Mismatches" report a few days ago. For consistency's sake, I have changed the name of the link to "Ignore this title". Ahasuerus 01:31, 3 September 2014 (UTC)
- Great! Love the ability to permanently remove a record once it's been determined to be correct. Thanks for adding this feature. Mhhutchins 01:41, 3 September 2014 (UTC)
- The current plan is to move all of the cleanup reports to nightly processing and add this functionality where feasible. This will kill two birds with one stone: make these pages come up almost instantly and allow moderators to mark records as "OK as is". Once this project has been completed, I will add a summary table at the top of the cleanup menu listing the reports that have one or more problem records. Ahasuerus 01:56, 3 September 2014 (UTC)
The image shows up fine on this pub and clicking on the image gets larger view. However clicking on the ISFDB source link gives the message
Error creating thumbnail: /var/www/html/wiki//bin/ulimit4.sh: line 4: /usr/local/bin/convert: No such file or directory
Doug 15:35, 3 September 2014 (UTC)
- You need to reset your wiki preferences. Click on the "My Preferences" link. Click on the "Files" tab. Change each of the two options to the highest setting. Those images you're referring to have a height of 640 pixels, which is beyond the requested limit. (Some uploaders don't feel they have to use the same standards as the rest of us.) By changing your settings, the wiki will be able to "create a thumbnail" and allow you to see it. If this doesn't help, let me know. Thanks. Mhhutchins 18:03, 3 September 2014 (UTC)
- Worked fine. Someone's been busy the last couple of years. Doug 22:43, 3 September 2014 (UTC)
- BTW, when posting a new message to a wiki page, don't use the "Edit" tab. Click on the "+" tab, which allows you to post a new topic without having to edit the whole page. Doing the "Edit" method may conflict with any other user posting simultaneously. Also using the "+" method displays the topic of your post in the list of Recent Changes which most moderators monitor and which will let them know when a new topic has been posted. Editing the entire page doesn't do that. Thanks. Mhhutchins 18:09, 3 September 2014 (UTC)
- I don't think I did use the Edit. Certainly not my intention. Tried clicking it just now and the full edit doesn't look familiar. I'll try watching where I'm going though. Doug 22:43, 3 September 2014 (UTC).
- It could be different based on what "skin" you're using. The default skin uses the "+" tab to start a new topic. The skin I use (Cologne Blue) has a link labeled "Post a Comment" to start a new topic. Mhhutchins 03:40, 4 September 2014 (UTC)
Authors with Invalid Spaces
The cleanup report "Authors with Invalid Spaces" has been moved to nightly processing and the ability to ignore author names has been added. I also changed the algorithm to look for author names with double quotes, but forgot to change the name of the report to reflect the additional functionality. It will be changed in the next patch. Ahasuerus 22:53, 3 September 2014 (UTC)
- There were two authors with double quotes, and when I changed them to single quotes, the names were removed from the list. But the report page still shows there's one on the list. Is there one that's not visible? Mhhutchins 01:04, 4 September 2014 (UTC)
- Well, as the message at the top of the cleanup menu says, "The numbers are as of the time the report was last run and may be higher than the current number." When the cleanup menu is displayed, the logic tries to account for any records that have been marked "Ignore" since the last run, but it only works for some reports depending on how they are compiled. In any event, I have re-run the nightly process and the count has been changed to 0. Ahasuerus 02:35, 4 September 2014 (UTC)
- The report page showed "6" on the "Authors with Invalid Spaces" button (maybe "7", I'm not completely sure). I either fixed or "ignored" all of them, and returning to the report page it showed "1". If it had remained "6" (or "7") I would have understood that it would not be cleared until the nightly re-run. It was the "1" that threw me. Thanks. Mhhutchins 03:28, 4 September 2014 (UTC)
- The logic that calculates the displayed number takes "Ignored" records into account, but it doesn't know about "fixed" records. If you told the system to ignore 6 records and fixed 1, then it would display "1". However, as I recall, there were two author names with double quotes, so something appears to be off by one. Oh well, it's all good as long as the number is down to "0" after the nightly process runs :-) Ahasuerus 04:21, 4 September 2014 (UTC)
- Only one of the double quoted authors had to be fixed. The other already existed in the db with single quotes, so I changed the title record's credited author from double to single. That automatically deleted the double author without having to make a submission to change him. (I couldn't do it if I tried, since the single author already existed.) Now it's all explained. Thanks. Mhhutchins 04:39, 4 September 2014 (UTC)
- And now we know... the rest of the story :-) Ahasuerus 04:49, 4 September 2014 (UTC)
- As an aside, double quotes in author names should be automatically converted to single quotes at data entry time, but apparently there is a flaw in one of the data entry forms. Ahasuerus 01:29, 4 September 2014 (UTC)
- The name of the report has been changed. Also, "Reviews That Create Stray Authors" and "Variant Titles in Series" have been moved to nightly processing. Ahasuerus 02:35, 4 September 2014 (UTC)
(unindent) "Pseudonyms with Canonical Titles" has been moved to nightly processing. Ahasuerus 22:20, 4 September 2014 (UTC)
Fixer's NewPub submissions
As per discussion above, I have asked Fixer to submit the contents of Queue One for the first half of 2013. The resulting submissions are not terribly exciting, but it's a start. Ahasuerus 21:45, 4 September 2014 (UTC)
- Paper-based NewPubs for January-September 2013 have been submitted. I have updated the publicly available snapshot of Fixer's queues. Ahasuerus 21:32, 5 September 2014 (UTC)
Prolific Authors without Defined Language
"Prolific Authors without Defined Language" has been changed to ignore the following authors: unknown, Anonymous, various, uncredited, The Readers, The Editors. The cleanup menu has been changed to display the current number of authors with missing SFE3 URLs. Ahasuerus 00:34, 5 September 2014 (UTC)
- "Prolific Authors without a Defined Language" has been moved to nightly processing. Ahasuerus 22:44, 5 September 2014 (UTC)
- Unless you intend on running another check to match SFE3 authors and ISFDB authors, the title of "Authors with missing SFE3 links" should be changed to "Authors on SFE3 who don't have title records in the ISFDB", or something similar. They can't have links to SFE3 if they're not in our database. Mhhutchins 05:58, 6 September 2014 (UTC)
- That's a good point, although I expect that the report title will need to cover different types of mismatches since both ISFDB and SFE3 are constantly adding new authors. Perhaps something as simple as "ISFDB-SFE3 author mismatches"? Ahasuerus 06:17, 6 September 2014 (UTC)
- If you do intend on running another check, keep in mind that HUNDREDS were removed from the original list for several legitimate reasons, including the fact that we don't link pseudonyms to their SFE3 pages. Also, many SFE3 pages are stubs with no substantial data, so there's no worth in linking to them. The idea of doing a manual check between the two databases for those same authors is something I would not look forward to doing again. Mhhutchins 05:58, 6 September 2014 (UTC)
- Have no fear, all "resolved" SFE3 records are stored in a special table and will be ignored by the next reconciliation run :) Ahasuerus 06:17, 6 September 2014 (UTC)
- An audible sigh of relief! And "ISFDB-SFE3 author mismatches" covers both scenarios. Mhhutchins 15:30, 6 September 2014 (UTC)
- OK, the name of the report has been changed.
- BTW, this is a good example of why certain seemingly simple tasks can take longer than one might expect. Not only did this report need to have a way of tracking "resolved/ignored" SFE3 authors to facilitate subsequent runs, but it also needed to support other sites in addition to SFE3. The way the software is written, we can easily create similar reports for other sites that we want to synchronize with. Ahasuerus 17:30, 6 September 2014 (UTC)
- Way above my pay grade, but I acknowledge and respect your ability to accomplish such tasks. Thanks. Mhhutchins 18:42, 6 September 2014 (UTC)
- I just noticed you've removed all authors from the "Prolific Authors without Defined Language" report. Was 47 titles enough to be considered not prolific? Mhhutchins 04:08, 7 September 2014 (UTC)
- Hm, apparently the nightly job that I forced to run earlier decided to error out. Which is curious since it had run successfully on the development server. Let me take a look... Ahasuerus 04:23, 7 September 2014 (UTC)
- This was the first time that the new and improved "Pseudonyms with Canonical Titles" report could find no matching author records. As it turned out, the "improved" logic couldn't handle this condition and the nightly process aborted, thus preventing the rest of the reports from being regenerated. Oops! All fixed now. Ahasuerus 05:46, 7 September 2014 (UTC)
Info from MIT Science Fiction Society
The MIT Science Fiction Society librarian has e-mailed me the following information:
Christopher Andrews (http://www.isfdb.org/cgi-bin/ea.cgi?155121) has
more books listed at:
Chris Beckett (http://www.isfdb.org/cgi-bin/ea.cgi?11286) lists the
2008 version of MARCHER, but there appears to also be a later "Author's
Preferred Text" publication, on paper at:
(I *think* this is the same as the ebook "Preferred" edition; both
are from the same publisher, both in August 2014.)
Andrez Bergen (http://www.isfdb.org/cgi-bin/ea.cgi?179003) has more
books and at least one of them, Tobacco-Stained Mountain Goat, looks sfnal:
Jim Bernheimer (http://www.isfdb.org/cgi-bin/ea.cgi?137588) has more
books, some on paper:
L. J. Cohen (http://www.isfdb.org/cgi-bin/ea.cgi?203069) has three
novels on paper, see:
Joe Ducie (http://www.isfdb.org/cgi-bin/ea.cgi?197939) has more
books, some on paper:
Martin T. Ingham (http://www.isfdb.org/cgi-bin/ea.cgi?97617) has a whole bunch more on-paper books: http://www.amazon.com/Martin-T.-Ingham/e/B003TZYRHU/ref=dp_byline_cont_book_1 With two books in it, "West of the Warlock" is a series. From the descriptions, PRISONER OF TIME looks like a sequel to VIRTUAL WILES.
Daniel Price (http://www.isfdb.org/cgi-bin/ea.cgi?200913) also has a non-genre book:
Jeremy Robinson (http://www.isfdb.org/cgi-bin/ea.cgi?134882) has more books: http://www.amazon.com/Project-Nemesis-Thriller-Jeremy-Robinson/dp/0984042393 http://www.amazon.com/Project-Maigo-Thriller-Jeremy-Robinson/dp/0988672561 http://www.amazon.com/Beneath-Origins-Jeremy-Robinson/dp/0983601704 http://www.amazon.com/Last-Hunter-Collected-Jeremy-Robinson/dp/0988672553 http://www.amazon.com/Insomnia-Seven-More-Short-Stories/dp/0983601755 http://www.amazon.com/Kronos-Jeremy-Robinson-ebook/dp/B004AHKCO8
Randolph Beck (unlisted) has one book:
Jonathan P. Brazee (unlisted) has a bunch of books on paper: http://www.amazon.com/Jonathan-P.-Brazee/e/B007E4W0GC/ref=dp_byline_cont_book_1 At least four of them (the United Federation Marine Corps" books and the werewolf and wererat ones) are clearly sfnal; some others like the "The Return of the Marines" series are set in "the near future" but don't seem to have sfnal elements.
Marion G. Harmon (unlisted) has a bunch of books, four on paper: http://www.amazon.com/Marion-G.-Harmon/e/B007EG8KVY/ref=dp_byline_cont_book_1
Jack Hessey (unlisted) has a bunch of books, two on paper:
Russell F. Moran (unlisted) has at least two sfnal books:
The subtitle and description of THANKSGIVING GANG suggests that there
are more by him out there somewhere but not on Amazon.
G. S. Jennsen (unlisted) has one book:
Chris Kohout (unlisted) has one book:
Blake M. Petit (unlisted) has one book:
Kevin Gerald Rau (unlisted) has six books: http://www.amazon.com/gp/search/ref=sr_adv_b/?search-alias=stripbooks&field-keywords=&field-author=kevin+gerald+rau
Thomas Settimi (unlisted) has three books: http://www.amazon.com/Thomas-Settimi/e/B007JX7Q08 Of these, CONVERGENCE is clearly sfnal. ROSWELL 1947 includes a character from CONVERGENCE but I can't say for certain from the description whether *it's* snfal too. And I can't tell from the BEAK OF THE TURTLE description whether it's sfnal fiction or just whackadoodle "ancient aliens" stuff that the author believes is true.
Will Wight (unlisted) has six books, three of which are on paper: http://www.amazon.com/Will-Wight/e/B00D9S1IMO/ref=dp_byline_cont_book_1
Other: Someone named Peter Grant has five books, four of them sfnal, on paper: http://www.amazon.com/Peter-Grant/e/B00CS8MHJE/ref=dp_byline_cont_book_1 I don't know whether he's the same person as the artist Peter Grant who's listed at http://www.isfdb.org/cgi-bin/ea.cgi?166262. (I suppose someone could ask him; he participates in the "Customer Discussions" at his Amazon page.)
- Cohen's books added. I'll try working on the others as well, but it will be intermittent. -- JLaTondre (talk) 23:55, 5 September 2014 (UTC)
- Thanks! I have put <strike> tags around Cohen's books to help keep track of what's been completed and what is outstanding. Ahasuerus 00:06, 6 September 2014 (UTC)
- A few more done. -- JLaTondre (talk) 12:53, 6 September 2014 (UTC)
- A few more done. -- JLaTondre (talk) 14:43, 1 November 2014 (UTC)
- Another one. -- JLaTondre (talk) 22:44, 24 November 2014 (UTC)
Unexpected change in font size and face
The font size and face for entering this message has been reduced. The same thing is happening on the Note field of publication entry and update forms (but no other fields.) Is it just my browser, or has there been a change on the server end? Mhhutchins 19:17, 5 September 2014 (UTC)
- No change that I know of. Have you tried "Ctrl/+" to change the size of the font used by your browser? Ahasuerus 19:30, 5 September 2014 (UTC)
- Yep, that was the first thing I tried. But I know that wouldn't have changed the font face. And why would it only affect the Note field and not any other entry field? Mhhutchins 19:35, 5 September 2014 (UTC)
- Synopsis, Note and Moderator Note fields are defined as "text area" fields. Browsers handle "text area" fields somewhat differently, so it's possible that this change, whatever it was, affected only these three types of fields. However, I am not seeing any changes on my end and I haven't installed any patches since yesterday. Ahasuerus 19:54, 5 September 2014 (UTC)
- Also, the search box is now wider than the table that contains it, so that part of it is outside of the left-side menu column. Strange. Mhhutchins 19:35, 5 September 2014 (UTC)
- Are you using Google Chrome? Depending on the selected font size, the search box may not fit perfectly within the left-side "navigation bar" under Google Chrome. Ahasuerus 19:54, 5 September 2014 (UTC)
- I've been using Google Chrome for quite awhile, and the search box always fit until today. There may have been an update to Chrome that affects display. Mhhutchins 02:46, 6 September 2014 (UTC)
- Chrome 37 made some changes to font rendering, one of which is a change from Courier to Consolas as the default monospace font, which is what gets used in large text areas by default on most web sites, so you're not going crazy.
- If it really bugs you, you should be able to change it back in Settings --> Advanced Settings --> Languages --> Web Content --> Customize Fonts Albinoflea 03:06, 8 September 2014 (UTC)
- Thanks for confirming that it's not just me. I'll return the settings to the original font. Mhhutchins 16:06, 8 September 2014 (UTC)
"Magazines" cleanup reports
All cleanup reports in the "Magazines" section have been moved to nightly processing. Ahasuerus 02:32, 6 September 2014 (UTC)
- The two "Series" cleanup reports have been moved to nightly processing. Ahasuerus 02:42, 7 September 2014 (UTC)
- Nice. Another report that I can tell before hand whether to check it. This new procedure is really saving me time (and probably much server time as well.) Thanks. Mhhutchins 03:12, 7 September 2014 (UTC)
- Four of the five "Titles" reports have been moved to nightly processing. The only exception is "Titles with Bad Ellipses", which seems to be on hold for now. One step at a time :) Ahasuerus 17:13, 7 September 2014 (UTC)
- I've not worked on the "Titles with Bad Ellipses" report, since some editors believe that there is a fundamental difference between ". . ." and "...", a difference I'm unable to fathom. Mhhutchins 20:20, 7 September 2014 (UTC)
- We already had this discussion. For some editors (I suppose that I'm one of those targeted) a title like "De purs esprits . . ." is as annoying and as incorrect as "The moon is a harsh mistress" is to an anglophone. Hauck 20:57, 7 September 2014 (UTC)
- If you feel that strongly, and believe that the choice should be based on the extent to which a standard annoys the sensibilities of the editor, then the standard should be eliminated. There are other standards which I find extremely annoying but haven't chosen to ignore or disregard. Flexible standards based on an annoyance factor? Sounds great. Mhhutchins 22:20, 7 September 2014 (UTC)
- One thing that we could do would be to limit this cleanup report to English titles if that works for everyone? Ahasuerus 21:18, 7 September 2014 (UTC)
- So only English titles should be fixed, regardless of how annoyed an English-speaking editor may find it? Why don't we just remove the standard, kill the clean-up report entirely, and allow each editor to determine which one to enter? Mhhutchins 22:20, 7 September 2014 (UTC)
- I think it's not unreasonable to have language-specific punctuation standards. For example, Spanish uses inverted question/exclamation marks (¿/¡), e.g. ¿Dónde estabas el pasado Pluterday? but you wouldn't want to use them in Arabic, which has what is known as the "reverse question mark" (؟). Similarly, English allows both "..." and ". . .", but in French only "..." is valid.
- Different squids for different kids :-) Ahasuerus 22:40, 7 September 2014 (UTC)
- But the form of the ellipsis as given in the standards is not a language-specific punctuation. There is no English standard that states whether there should be spaces between the periods of an ellipsis. That's something the users of the ISFDB agreed to well before most of us even started working here. It has nothing to do with the English language. The standard was established, I assume, to prevent editors from creating variants based on how another editor chose to enter an ellipsis. Again, the ISFDB standard for the ellipsis is not language-specific. Mhhutchins 16:14, 8 September 2014 (UTC)
- When this issue originally came up in 2006 or 2007, it was because some editors were using ". . ." and some "...". Clearly it made searches more difficult and error-prone, so we needed to pick one.
- Now, it so happens that both ". . ." and "..." are acceptable in English -- the former is supported by The Chicago Manual of Style while the latter is supported by the AP Stylebook. Somewhat arbitrarily, we chose to go with the Chicago Manual of Style, which is very much an English-language standard. If 90% of our titles and users had been French 7-8 years ago, the only possible choice would have been "...", as we discussed earlier this year. And if 90% of our titles and 90% of our user base had been Japanese, I guess our choices (apparently) would have been "…" and "……".
- Let me take a step back and make a comment about The Big Picture. In most cases databases are models of (some part of) the world around us. Whether they drive Amazon or IMDB, they need to reflect the field than they are modeling. And if that field is going through changes, the database needs to change in order to continue to reflect things accurately. In our case it may mean adding things like various DVD and electronic bindings as they emerge.
- Similarly, if the organization serviced by a database decides to expand its coverage -- e.g. the way Amazon has gone from a "book-selling company" to a "lots-of-things-selling company" over the last 20 years -- the underlying database has to change in order to support the new areas. In our case it may mean accommodating different languages, e.g. the way we had to change the software to support "right-to-left" languages, or migrating to language-agnostic data entry standards in place of the original standards which were based on various Anglophone standards. And if a language-agnostic standard is not possible, as is the case with ellipses, I believe we need to allow language-specific standards, like we already do with capitalization. Ahasuerus 23:14, 8 September 2014 (UTC)
[unindent] I have to reiterate my original point, which it seems is being misunderstood. As you said there was a reason why one form of the ellipsis was chosen (for search purposes). It had NOTHING to do with language. Bringing language into the discussion only muddies the water. Mhhutchins 02:04, 9 September 2014 (UTC)
- As I recall, we didn't think in terms of language when we picked the Chicago Manual of Style (". . .") standard as the ISFDB standard when entering ellipses. All we were thinking of was how to make searches less error-prone and how to make the data look more consistent. However, even though we weren't thinking of it, we were still choosing between two competing English standards because that's what our editors were familiar with. We never considered that other languages may have different standards and that ". . ." would be invalid in French, German, Italian, Romanian, Spanish, Finnish, Russian, Kazakh, Hungarian, Polish, Danish, Azerbaijani, Norwegian, Swedish, Slovak, Slovenian, Serbian, Croatian, Turkish, Thai, Arabic, Hebrew, Bulgarian, Esperanto, Lithuanian, Latvian, Dutch, or Ukrainian (as my earlier review suggested.)
- By the way, a similar thing has happened with capitalization. Early in the game, we had to decide whether to capitalize all words in our titles/pubs. In English, you can do it either way, so we picked one way almost randomly. We didn't think that we were being Anglocentric, but the choices that we had were dictated by the practices that we were familiar with, which were Anglophone. It wasn't until much later that French, German, and Polish editors pointed out that other languages' capitalization rules were more restrictive and we all agreed that our records should use the capitalization rules of the language that the book is written in. Thankfully, our searches are case-insensitive, so there was no impact on searches.
- I should also note that this is not an ISFDB-specific problem. In a prior life I was on a couple of national standards committees. Even though the standards we were responsible for were national in scope, they also affected people in other countries because they used "our" technology. Occasionally we would get requests from countries like Japan asking us to allow certain language-specific things that we wouldn't have even considered otherwise. We usually said something like "Oh, you really need that? OK, sure, let us add it to the next version of the standard". Ahasuerus 20:48, 9 September 2014 (UTC)
Regardless of the language, a form was chosen 7 or 8 years ago. And I followed that standard for the years I've been here, again, regardless of the language. (Don't forget, you're the one who wrote two clean-up scripts to find titles and pubs with "bad ellipses"!) Mhhutchins 02:04, 9 September 2014 (UTC)
- That's right, the addition of these two cleanup reports in early February is what started the original discussion. The fact that I hadn't thought of it myself is kind of embarrassing since I studied French back in the day, but apparently I never learned the difference. Ahasuerus 20:48, 9 September 2014 (UTC)
If you want to change the standard to three periods without spaces, then do it. Make a universal change to all records and the discussion would be over. Mhhutchins 02:04, 9 September 2014 (UTC)
- If we can determine that all relevant (i.e. represented in ISFDB) languages are OK with "..." instead of ". . .", then I agree that it would be the logical thing to do. After thinking about it, I am increasingly inclined to go with this solution. However, are we sure that some languages do not use, e.g., " ..." (note the extra space) instead of "...", which would affect searches? Ahasuerus 20:48, 9 September 2014 (UTC)
The same reason for having one standard 7 years ago is the same reason to have one today. It's not the form of the ellipsis I'm arguing about. It's the decision to choose one, make it a standard, and stick to it, regardless of the language. Mhhutchins 02:04, 9 September 2014 (UTC)
- Well, the main reason to have a standard was to make searches less error-prone. However, this is only an issue because English speakers may enter either form of the ellipsis in the Search box. A French/German/Polish/etc user would presumably never do that when searching for French/German/Polish titles, which suggests that separate language-specific data entry standards may be viable, although, admittedly, also more difficult to document and enforce. Oh well, it will all be moot if we change the standard to "...". Ahasuerus 20:48, 9 September 2014 (UTC)
Nothing in my argument suggests that all ISFDB standards should follow the English language, without regard to the language of the publication. To imply that is my point of view, only makes me out to be a language bigot. Mhhutchins 02:04, 9 September 2014 (UTC)
- I don't think any of us are "language bigots", but our perception of the "realm of the possible" is often shaped by the languages that we are familiar with. For example, I had a hard time wrapping my head around certain software problems that were raised last year when we added support for "right-to-left" languages like Arabic. Apparently my brain just doesn't work that way :-( Ahasuerus 20:48, 9 September 2014 (UTC)
A "language-agnostic" standard is possible: choose the one that is an acceptable form in both English and French, which is the one without spaces, according to at least one French-speaking editor. Or better yet, drop the standard entirely, and give the editors the choice about which form to use. I can't imagine anyone giving ellipses in a search form, if that was the only reason for having a standard in the first place. Mhhutchins 02:04, 9 September 2014 (UTC)
- My current thinking is that changing the standard to "..." and mass-changing the existing data would be the easiest thing to do. I will probably sleep on it and post a suggestion on the Standards page tomorrow. Ahasuerus 20:48, 9 September 2014 (UTC)
- Thank you for the reasoned response. Mhhutchins 00:07, 10 September 2014 (UTC)
A new section for Fixer
Due to the increased frequency of Fixer-related updates (which is great!), I have created a new section, Fixer's Status. Hopefully it will make it easier to keep track of where Fixer is. Ahasuerus 17:44, 6 September 2014 (UTC)
Cleanup reports - "Variant Titles"
The "Variant Titles" section has been moved to nightly processing.
Also, one of the reports, "Series of Variant Titles", has been decommissioned: its sibling report, "Variant Titles in Series", is sufficient for our purposes. Ahasuerus 02:04, 8 September 2014 (UTC)
All Awards-related cleanup reports have been moved to nightly processing. In addition, the following changes have been made to these reports:
- It turns out that the patch that made the "Suspect Untitled Awards" report ignore certain harmless untitled awards also introduced a bug in the logic. The bug was responsible for masking recently added untitled award, which would have been otherwise identified by this report. Now that the bug has been fixed, we can see additional "suspect" awards. (There is only one at this time and it's harmless.)
- "Empty Award Types and Categories" has been split into "Empty Award Types" and "Empty Award Categories".
I was going to add the ability to "ignore" untitled awards to "Suspect Untitled Awards", but I ran out of time. I will work on it tomorrow. Ahasuerus 05:28, 9 September 2014 (UTC)
- The way the "Suspect Untitled Awards" report works has been changed. The logic no longer includes special "if-then-else" conditions for "suspect" untitled awards (like some 1964 Hugo awards) which are known to be legitimate. Instead moderators can choose to "Ignore This Award" and the award ID will be added to the list of awards to ignore. Due to this change, the "Suspect Untitled Awards" report currently shows 43 suspect award records, but I expect that all of them will be marked "ignore" in the next day or two. Ahasuerus 04:23, 11 September 2014 (UTC)
- Next day or two? How about 5 minutes? They were all for either 1) series, for which we can't link awards, or 2) illustrated works which recognized both the author and the artist, and the ISFDB doesn't credit artists in the author field of a title record for an illustrated work, so it's not linkable without losing half of the award credit. (Most of those were for specific publications of a work, and not for the title alone, e.g. the 50th Anniversary edition of Animal Farm illustrated by Ralph Steadman placed in the Best Art Book category of the 1997 Locus poll.) So it was easy to clear the list. Mhhutchins 04:59, 11 September 2014 (UTC)
- To quote the last great spaghetti Western, "Nobody was faster on the draw" :-) Ahasuerus 05:29, 11 September 2014 (UTC)
merging two publications of same ebook
- Publication records can't be merged. If you're certain that only one ebook edition of this work exists, then delete one of them, and update the other one to create a more complete record. Keep in mind that Smashwords is not a publisher. They facilitate authors to self-publish their work, similar to Createspace, an Amazon company. Does your copy have an explicit publisher credit? Thanks. Mhhutchins 03:25, 10 September 2014 (UTC)
New version of "Publications with Suspect Images"
Titles without Pubs -- Part 2
Added a section for CHAPTERBOOKs and the the ability to "ignore" titles. Ahasuerus 05:35, 11 September 2014 (UTC)
"Series with CHAPTERBOOKs"
"Series with CHAPTERBOOKs" has been moved to nightly processing. Ahasuerus 17:16, 11 September 2014 (UTC)
- P.S. The other three CHAPTERBOOKs-related cleanup reports show "(0)" next to their names, but they haven't been moved yet. I expect to do it later today. Ahasuerus 17:17, 11 September 2014 (UTC)
- All CHAPTERBOOKs-related cleanup reports have been moved to nightly processing. Ahasuerus 18:27, 11 September 2014 (UTC)
A couple of minor changes have been made:
- Moderators are now able to see a link to the main Cleanup Reports menu when viewing regular ISFDB pages.
- Clarified the message on the Key Maintenance page, which may have been confusing to new editors.
Ahasuerus 22:44, 11 September 2014 (UTC)
"Pre-2005 pubs with ISBN-13s and post-2007 pubs with ISBN-10s"
I am working on moving publication-specific cleanup reports to nightly processing and I have a question about "Pre-2005 pubs with ISBN-13s and post-2007 pubs with ISBN-10s".
Do we need to add the ability to "ignore" some of these ISBNs once they have been confirmed as accurately entered? As I recall, some publishers didn't switch to ISBN-13 until 2008, so it would seem like the ability to "ignore" post-2007 ISBN-10 pubs would be beneficial. Ahasuerus 00:01, 12 September 2014 (UTC)
- Yes. Eventually, we'll want to add 2007 ISBN-10s to the list, so we'll need to have the ability to "ignore" any publications that don't follow the standard. It would also be nice to have the records numbered like other reports. I've been working on this one forever, correcting several hundred records, and it doesn't seem to be getting any smaller. If they're numbered at least I'll be able to see the progress. Thanks. Mhhutchins 00:12, 12 September 2014 (UTC)
- Done and done. We are down to 349 2008 ISBN-10s and just one 2009 ISBN-10, so things are looking up. Less encouragingly, I see that there are 4,000+ 2007 ISBN-10s, but we'll cross that bridge when we get to it. Ahasuerus 02:33, 12 September 2014 (UTC)
- "Duplicate Publication Tags" has been moved to nightly processing. Ahasuerus 03:35, 12 September 2014 (UTC)
- Ditto "Publication Authors that are not the Title Author". Ahasuerus 04:07, 12 September 2014 (UTC)
- All cleanup reports have been moved to nightly processing. The only exceptions are the 2 "ellipsis" reports which are in limbo pending a Standards discussion (see above) and the 2 reports that are regenerated once every few months: "ISFDB-SFE3 Mismatches" and "Suspect Images".
- The next step will be to modify the main cleanup menu not to display reports without problems. Ahasuerus 00:29, 13 September 2014 (UTC)
The cleanup menu has been modified not to display problem-free reports. In addition, the "ISFDB-SFE3 Mismatches" and "Suspect Images" reports as well as the two ellipsis reports have been moved to a new section, "Reports That Are Not Regenerated Nightly". The report previously known as "Uncommon Publication Bindings" is now "Invalid Publication Bindings". Finally, since almost all of the reports are now regenerated nightly, the note explaining this fact has been moved from individual reports to the main cleanup menu. Ahasuerus 04:49, 13 September 2014 (UTC)
- I have made a significant number of changes to the way the cleanup reports work internally. There should be no user-experienced differences, but I thought I'd mention it here in case anyone should encounter bugs or other unusual behavior. Ahasuerus 05:21, 14 September 2014 (UTC)
Changes to "Omnibuses without Contents Titles"
The "Omnibuses without Contents Titles" cleanup report has been changed as per the Help change which allowed publications with NONFICTION titles to be classified as OMNIBUSes. Ahasuerus 20:49, 14 September 2014 (UTC)
The trimphant return of "Publications with Duplicate Titles"
"Publications with Duplicate Titles" has been re-added to the cleanup menagerie. It was decommissioned in early 2014 due to the excessive load it puts on the server every time it runs, but now that it runs as part of the nightly process, it shouldn't cause problems. Ahasuerus 04:06, 15 September 2014 (UTC)
"Pseudonyms with Canonical Titles"
"Pseudonyms with Canonical Titles" has been tweaked to display the standard "No records found" message once all problems have been resolved. Ahasuerus 16:23, 15 September 2014 (UTC)
"Interviews and Reviews not in Pubs" folded into "Titles without Pubs"
"Interviews and Reviews not in Pubs" has been folded into "Titles without Pubs". In addition, the latter has been modified to cover SERIAL titles without pubs. Ahasuerus 18:54, 15 September 2014 (UTC)
Ellipsis-specific cleanup reports changed
The two ellipsis-specific cleanup reports have been changed to look for ". . ." rather than "...". They have also been moved to nightly processing and will no longer appear on the main cleanup menu if there are no "problem" records on file. Ahasuerus 22:55, 19 September 2014 (UTC)
More title types added to "Titles without Pubs"
The cleanup report "Titles without Pubs" has been modified to include OMNIBUS titles. Ahasuerus 17:44, 20 September 2014 (UTC)
- A COLLECTIONs section has been added. A cursory review suggests that a number of pubs need to be moved to these pub-less titles to get everything in sync. Also, there is a pub-less Hungarian translation which I will work on later tonight. Ahasuerus 23:34, 20 September 2014 (UTC)
- The report has been further modified not to display title records that have had publications added to them since the last nightly run. Ahasuerus 00:16, 21 September 2014 (UTC)
- Added support for ANTHOLOGY records. Ahasuerus 01:24, 22 September 2014 (UTC)
- Added EDITORs. Ahasuerus 02:58, 23 September 2014 (UTC)
- Added NONGENRE. Ahasuerus 16:48, 23 September 2014 (UTC)
- I suppose there should be a discussion about the titles in the latest category. They're all publess NONGENRE titles, and many of them are from NONGENRE authors. Should we leave them in the database or delete them? I'd go with the latter, since NONGENRE works by NONGENRE authors (authors "below the threshold") are not eligible for the database anyway. Why keep a list of titles of their other works, and only encourage editors to add publications to them? For example, the nongenre work of Wilkie Collins, whose supernatural stories were a small percentage of his entire output. Mhhutchins 17:12, 23 September 2014 (UTC)
- As it happened, the list was no problem to clear, since most turned out to be NONGENRE novels by marginally genre writers, so I chose to ignore them. The problem with Collins' novels still persists. It looks like someone just went through Project Gutenberg and added all of his novels to the ISFDB, creating publication records for the PG editions, regardless of their spec-fic content (or lack thereof.) A few of them have any other pubs attached to the title. That's why most of them didn't show up on this cleanup report. Mhhutchins 17:54, 23 September 2014 (UTC)
(unindent) NONFICTION records have been added. Ahasuerus 19:57, 23 September 2014 (UTC)
Fixer 2012 revisited
Now that all paper-based ISBNs for October, November and December 2012 have been processed, we can start moving back in time on that front as well.
I have spent much of today sorting out the 1,375 paper ISBNs for 2012-09. After consigning more than 80% to various low priority queues or rejecting them outright, I ended up with 123 AddPubs and 132 NewPubs in Queue 1. That's not bad, especially considering that Fixer had more ISBNs captured for 2012-09 than for any other other month in 2010-2012. (Some months in 2007-2009 have more outstanding ISBNs because Fixer didn't become fully operational until some time in 2009.)
I plan to continue mixing and matching e-AddPubs for 2010 with paper-based AddPubs and NewPubs for 2012. At the rate we are going, we should have e-AddPubs finished in the next few months and then we can start e-NewPubs, which are a whole 'nother story. Ahasuerus 01:31, 23 September 2014 (UTC)
- And one that I'm not likely to read. The e-AddPubs are taking their toll, and I may be sitting out for the e-NewPubs. So don't expect them to move as quickly as these. Mhhutchins 03:49, 23 September 2014 (UTC)
- Yes, burnout is very much a concern with the kind of volume that Fixer has been responsible for lately. There is no need to rush it -- Fixer tells me that he doesn't mind if his submissions stay in the queue for a few days :-) Ahasuerus 20:16, 23 September 2014 (UTC)
Two New Cleanup Reports
A new section, "Reviews", has been added to the Cleanup Reports menu. It contains two new reports, "Reviews without Reviewed Authors" and "Reviews not Linked to Titles". The former is fairly straightforward since REVIEW records without reviewed authors (there are 7 of them as of this afternoon) are invalid and need to be fixed. The latter is a mixed bag including:
- Reviews of books that we already have on file, so we just need to create a link
- Reviews of books that we do not have on file, but clearly want to add
- Reviews of books that we do not have on file and probably do not want to add. These can be "ignored" using the "Ignore" link.
- Incorrectly entered interviews. There is no easy way of converting a REVIEW record to an INTERVIEW, so we just need to delete the existing record and create a new one.
There are a few Hungarian and Polish reviews which I plan to work on later today. Ahasuerus 20:10, 23 September 2014 (UTC)
- Rather than ignore the "Reviews of books that we do not have on file and probably do not want to add", I would suggest that the REVIEW records be replaced by an ESSAY record. It's going to take a little work, but at least there's a list which we work from to do it. By using the "ignore" button, I'm afraid these titles will be lost until there's another script written to find them. Why not just make that determination now instead of removing them from the list? Mhhutchins 20:41, 23 September 2014 (UTC)
- I guess the first question to answer is whether we want to have non-linked REVIEWs at all or whether they should all be converted to ESSAYs. The "Reviews" section of the NewPub Help page says:
- "Note also that only books, magazines, and short fiction are entered; if the column reviews fanzines, you don't need to enter the review records for these, only the ESSAY record. Non-sf works should be entered but if an onerous number of non-sf-related works are reviewed in a column you are entering, discuss the situation on the Bibliographic Rules page to decide what can be eliminated."
- The second sentence appears to suggest that editors should create REVIEW records for non-genre works reviewed in SF publications (magazines etc) even when the associated title record is not eligible for inclusion based on the "non-genre threshold" criterion. For example, take this review record. It's a Hungarian review of "A Fekete Kalóz" (1898, original Italian title "Il Corsaro Nero", English title "The Black Corsair"). The author, Emilio Salgari, did write some SF, e.g. "I figli dell'aria" ("The Children of the Air", 1904 -- see "Encyclopedia of Italian Literary Studies: A-J", p. 1718), but "Il Corsaro Nero" was an adventure story about pirates and contained no SF elements. Assuming that Salgari is "below the threshold" for non-genre purposes, should we enter the review as an ESSAY or should we create title records for Il Corsari Nero and its Hungarian translation? Ahasuerus 22:07, 23 September 2014 (UTC)
- That section of the standards that allows the entry of reviews for nongenre publications is OK ... until it comes to nongenre publications by nongenre authors. This allowance created hundreds of titleless authors in the database. It took me more than a year to clear the cleanup script that found such titles. In almost all cases I had to convert the REVIEW record to an ESSAY. Otherwise those hundreds of titleless authors would still be in the database and showing up on the cleanup report. You can see that there is a basic conflict here in the standards: we either allow such REVIEW records to be created and don't care if they are not linked to a publication record, or we don't have two different cleanup reports that find 1) titleless reviewed authors and 2) unlinked REVIEW records. By and large, most editors do not go back to publication records to see if the REVIEW records they created was linked to a current title in the database. Thus, we have the cleanup report you just created. 90% of the titles on this report are nongenre publications by authors who barely squeaked into the database by having another title in the database, regardless of its genre. (If they had no titles, they would have appeared on the other cleanup script.) If we're not going to change the standards (as you've cited above), then we need to think about encouraging editors to enter reviews of nongenre books as ESSAYS, just as we require that they do so for films, recordings, graphic novels, etc. The fact that this exception crept into the rules without deeper discussion of the consequences is a failure of those who pushed for the change. Mhhutchins 23:15, 23 September 2014 (UTC)
- Now to answer your question about the Salgari work: the REVIEW record should be converted to an ESSAY, i.e. delete the review and replace it with an essay. Mhhutchins 23:15, 23 September 2014 (UTC)
- You'll also notice that the list includes many issues of magazines and fanzines. According to the rules, those should never have been entered as REVIEWs, but as ESSAYs. Obviously some moderators were allowing such reviews and editors weren't making attempts to "link" these reviews. (Because the merging of EDITOR records into annual records doesn't allow such links.) Mhhutchins 23:20, 23 September 2014 (UTC)
- I haven't been following the evolution of this standards area very closely, but the Help page linked above appears to instruct editors to enter reviews of NONGENRE/NONFICTION books and magazines (sic) as REVIEWS rather than as ESSAYS. If we ask the verifiers to link their verified REVIEWS before we have a stable standard, they will all do it differently and then it will be even harder to get everything in sync. So I guess we need to get a consensus and clarify Help before we tackle this subset of unlinked reviews.
- I can think of at least three possible ways to handle reviews of "non-genre books by primarily non-genre authors":
- Create a REVIEW record (as per the current version of Help) and leave it as an "unlinked review".
- Create a REVIEW record, then enter the associated NONGENRE/NONFICTION title and link the two.
- Enter the review as an ESSAY; do not create a REVIEW record.
- Based on your comments above, you believe that the last approach would work best. Overall I agree; I don't have a great deal of experience in this area, but I don't like the idea of entering title records for, e.g., Encyclopedia Americana, and unlinked reviews are also unappealing.
- Do you want to post a new version of the previously quoted Help paragraph on the Standards page? Something along the lines of:
- For reviews of books and short fiction that are eligible for inclusion in the database based on the Rules of Acquisition enter the review information in the Review area. For reviews of media products (films, TV shows, games, plays, etc), magazines, fanzines as well as of books that are not eligible for inclusion in the database enter a separate ESSAY record in the "Content" section of this data entry form, but do not enter anything in the Review section.
- ? Ahasuerus 00:37, 24 September 2014 (UTC)
- Sounds like a reasonable solution to the problem(s). I'll post it on the Rules & Standards Discussions page tomorrow. Thanks. Mhhutchins 05:44, 24 September 2014 (UTC)
- Re: the concern that "these titles will be lost until there's another script written to find them", we are actually in fairly good shape there. There will be no need to create another cleanup report to list "ignored" reviews since the current report can be easily run in a "show ignored reviews" mode. I was going to wait until we finished the first pass before enabling this mode, but it would be trivial to enable it now. Ahasuerus 22:07, 23 September 2014 (UTC)
- That's good to know. Mhhutchins 23:15, 23 September 2014 (UTC)
- Also, is it possible to give a link to the verifying editors of the publication records containing these non-linked reviews? It would be best for them to repair, link, or "ignore" them. I'll gladly do it for records by missing verifiers, but if they're still active, the primary verifiers of the records should take on the task. Mhhutchins 21:12, 23 September 2014 (UTC)
- I also would encourage the approach of "If it doesn't belong in the database, then a review of it should be an essay only, not a REVIEW". Chavey 12:30, 25 September 2014 (UTC)
New cleanup report - Reviews of uncommon title types
A new cleanup report, "Reviews of Uncommon Title Types", has been added. It goes hand in hand with the previous change, which disabled auto-linking reviews to CHAPTERBOOKs. There may be scenarios where a CHAPTERBOOK review is legitimate, but my guess is that the vast majority of our 1,025 CHAPTERBOOK reviews will need to be relinked.
P.S. I realize that adding a lot of "stuff" to the plate in a short period of time can feel overwhelming, but I am trying to wrap up the Cleanup Reports menu so that I could move on to other areas. Ahasuerus 22:46, 25 September 2014 (UTC)
- There are legitimate cases where a review is of a CHAPTERBOOK in toto and not just its fiction content, for example here. In this case (and others similar), I'll "ignore" the record, but will relink those obvious others to their SHORTFICTION content. The same situation is present in the art reviews written by Karen Haber for Locus. In many cases, she would review the new artwork for reprints of novels, for example here. In this case, it wouldn't make sense to link her review to J. M. Barrie's NOVEL title record, but to the title record for Raquel Jaramillo's INTERIORART. As in the above I'll "ignore" it. If in either of the circumstances there arises a question of what is actually being reviewed, I'll message the ISFDB editor if the work has been primary verified. I ask that any other moderator working on this cleanup report to be careful regarding similar cases. Mhhutchins 03:07, 26 September 2014 (UTC)
- That's exactly why I changed this report's title from "Reviews of Invalid Title Types" to "Reviews of Uncommon Title Types" :-) Ahasuerus 06:30, 26 September 2014 (UTC)
- If we don't have the actual review present, it seems it would be very difficult to determine when a review is of the Chapterbook itself, and when it is really a review of the story within it. My suspicion is that we should have a "default" that we go to if there is insufficient data, i.e. link it as a review of the story, and when someone believes it's a review of the Chapterbook itself, they should be expected to put a note of that in the review title rec. Chavey 18:33, 28 September 2014 (UTC)
- That's pretty much how the auto-link logic works at the moment since CHAPTERBOOKs are ignored for auto-linking purposes. If an editor determines that a REVIEW record should be linked to the CHAPTERBOOK rather than to the SHORTFICTION record, he has to create a Link Review submission. Perhaps we should update Help to explain that REVIEWs of CHAPTERBOOKs should have a note attached to the REVIEW record. Ahasuerus 23:40, 28 September 2014 (UTC)
- How about this, new text in italics:
- "Note also that only books, magazines, and short fiction are entered; if the column reviews fanzines, you don't need to enter the review records for these, only the ESSAY record. Most reviews of chapterbooks should be entered as a review of the short fiction item within the chapterbook, unless the review specifically addresses aspects of the chapterbook binding itself, such as the cover, on making a hard-to-find title available, etc., in which case the reason should be included in the Review "Notes" field. Non-sf works should be entered but ..."
- Are there other common examples of an actual chapterbook review that should be included? Chavey 13:23, 29 September 2014 (UTC)
- How about this, new text in italics:
- Reviews that pay particular attention to the interior art should also be given as an exception where the review should be linked to the CHAPTERBOOK record and not the SHORTFICTION record. Artwork may not have been used in previous editions of a work of fiction, or may not be used in subsequent editons. Mhhutchins 16:39, 29 September 2014 (UTC)
Fixer - 2014-10 cleanup
Earlier today I finished integrating Fixer's catch for 2014-10 and 2014-11. Fixer knows of another 131 ISBNs for 2014-08 and 2014-09 that are not in the database and I plan submit them on Fixer's behalf later in the week. Ahasuerus 20:42, 5 October 2014 (UTC)
- All done thanks to the cleanup crew! Ahasuerus 04:31, 9 October 2014 (UTC)
Edit Title tweaks
The review page for Edit Title has been changed to display a warning if the name entered in the Series field is new. A number of other review pages have been modified to display a "Warnings" column, which is currently blank in most cases. Ahasuerus 05:29, 7 October 2014 (UTC)
Publisher-specific cleanup reports
Two new cleanup reports have been added: Publishers with Identical Names and Publishers with Similar Names. The latter is currently limited to publishers with " LLC"s and ", LLC"s in their names. Once the low hanging fruit has been taken care of, we can enhance the logic to look for other patterns. Ahasuerus 01:56, 11 October 2014 (UTC)
Edit Publication - review page changes
The review pages for Edit/Clone/etc Publication have been changed to display warninga when an invalid ISBN or a non-existent publisher/publication series are entered. Ahasuerus 04:25, 11 October 2014 (UTC)
"Titles without Publications" updated
"Titles without Publications" has been updated to include NOVEL titles. Ahasuerus 05:38, 14 October 2014 (UTC)
- P.S. I will take care of the Polish, Russian and Hungarian titles once I am feeling better. Ahasuerus 06:06, 14 October 2014 (UTC)
- Wracaj do zdrowia / выздоравливай скорее / Jobbulást! (courtesy of Google Translate.) PeteYoung 10:19, 14 October 2014 (UTC)
- Thanks, I'll do my best! :) Ahasuerus 22:59, 14 October 2014 (UTC)
New moderator warning
A new warning has been added to all moderator review pages. It is displayed if the submitted image is hosted by a site that we do not have permission to link to. Ahasuerus 04:24, 15 October 2014 (UTC)
Edit Title enhancements
The moderator review page for Edit Title has been enhanced to display a list of affected publications and their verification status.
I am a bit concerned about titles with hundreds of associated pubs, e.g. The Hobbit. If you find the list too long, I can reposition the "Approve/Reject" buttons to appear immediately after the main Title table. Ahasuerus 19:49, 15 October 2014 (UTC)
- P.S. It occurs to me that the verification information is available on the Title page. Perhaps all we need on the moderator review page is a warning about primary verifications existing for the about-to-be-modified title record? Ahasuerus 20:01, 15 October 2014 (UTC)
New moderator warnings
All author/reviewer/interviewer/artist fields have been upgraded to display standard moderator warnings about new/pseudonymous/disambiguated authors. Hopefully I didn't miss anything. Ahasuerus 00:55, 17 October 2014 (UTC)
- The display logic has been tweaked not to display warnings for existing authors/artists/etc. In addition, authors are now properly hyper-linked when linking a title to an existing parent title. Ahasuerus 01:37, 19 October 2014 (UTC)
New cleanup report - mismatched VT title types
A new cleanup report has been added to the VT section of the menagerie. It finds VTs whose title type doesn't match the parent's title type. SERIAL/NOVEL, SERIAL/SHORTFICTION, SERIAL/COLELCTION, SERIAL/NONGENRE, INTERIORART/COVERART and COVERART/INTERIORART are recognized exceptions.
In addition, the "Titles without Pubs" report has been modified to ignore 8888-00-00 titles. Ahasuerus 01:12, 21 October 2014 (UTC)
- Thanks for the requested cleanup report. Much appreciated. I've already starting asking verifiers to confirm mismatched titles, and I'm sure there's going to be many more requests. Mhhutchins 22:05, 21 October 2014 (UTC)
- Sure thing! I guess one could argue that we ought to change the data entry forms to disallow mismatched variants. However, there are so many possible permutations when performing multi-step operations that it may require additional and fairly sweeping changes like enforcing publication-title consistency. Ahasuerus 22:28, 21 October 2014 (UTC)
- Would a simple moderator warning be too difficult to implement? Even if any of the mismatches are allowed under the current standards? It would give the moderator the option to dismiss the warning (because the mismatches are allowed) or discuss the mismatch with the submitting editor. Mhhutchins 23:24, 21 October 2014 (UTC)
- That's a good point and I have changed the software to display appropriate moderator warnings except for SERIAL/NOVEL, SERIAL/SHORTFICTION, COVERART/INTERIOART and INTERIORART/COVERART mismatches. However, I suspect that most of the existing mismatches were created when an editor modified a title record's type via Title Edit without changing associated variants. Oh well, that's what we have the cleanup reports for :-) Ahasuerus 01:49, 22 October 2014 (UTC)
- Thanks! I just made an "uncommon" variant, and the warning popped up. Wonderful! Mhhutchins 04:33, 24 October 2014 (UTC)
- Yup, we are getting there :-) Ahasuerus 06:35, 24 October 2014 (UTC)
- Once I've cleared the 300+ records on the current report, it's going to be easier to find them after they're in the database. The notifications being given during the cleaning process will also make editors and moderators more conscious of types when varianting. I have found some exceptions that would require the option in other reports to "ignore this title". (For example, some reviews are later reprinted as essays in some nonfiction collections, often with new titles that would get lost if they retained the REVIEW type.) I hope that's possible for this report as well. Mhhutchins 23:24, 21 October 2014 (UTC)
- It would be easy to add the "Ignore" functionality, but could you please post a few examples first? I wonder if it may be possible to discern a pattern and tweak the report logic to ignore certain types (or perhaps sub-types) of mismatches. Ahasuerus 01:49, 22 October 2014 (UTC)
- The ESSAY / REVIEW variants in the first section are good examples. (I admit that 11 of the 12 are of my own devising, when I edited a collection of nonfiction by Michael Bishop, giving titles to some of his previously published reviews.) If you go further down the list, the SERIAL / POEM combination may have to be made into an exception. (A single, long poem was serialized, but in this particular case, it might be better to change the SERIAL to POEM, since they could be considered excerpts.) I've had a few discussions with editors who feel exceptions should be made: in one case a long poem by Perrault which was translated into a story version, and another discussion about a prose poem by R. E. Howard that was translated into a story. Mhhutchins 04:33, 24 October 2014 (UTC)
- I see, thanks. I guess we'll have a better idea of what the "realm of the possible" looks like once all the "regular" inconsistencies have been eliminated. Ahasuerus 06:35, 24 October 2014 (UTC)
Fixer: September/November 2014
The last batch of Fixer-acquired NewPubs is almost done, but some lower priority ISBNs (<180) are still outstanding. I will be submitting them on Fixer's behalf over the next few days. Ahasuerus 21:19, 26 October 2014 (UTC)
- The last ISBN was submitted a few minutes ago. Thanks to the moderators who have been diligently working on Fixer's submissions! Ahasuerus 05:26, 30 October 2014 (UTC)
New cleanup report -- EDITOR Records not in MAGAZINE/FANZINE Publications
A new cleanup report, "EDITOR Records not in MAGAZINE/FANZINE Publications", has been added. An ad hoc run of the nightly process has found only 3 eligible EDITOR records, so we are in reasonably good shape there. Ahasuerus 23:25, 26 October 2014 (UTC)
The Cleanup Reports menu has been tweaked. You can know choose to view a full list of cleanup reports. Ahasuerus 00:26, 27 October 2014 (UTC)
New Cleanup Report - Title Dates after Publication Dates
A new cleanup report, "Title Dates after Publication Dates", has been deployed. For a title to qualify for this report, its year needs to be at least two years later than the year of its earliest publication. Once the current crop (1,059 records) has been cleaned up, we can adjust the compilation logic to be more inclusive. Ahasuerus 02:00, 28 October 2014 (UTC)
Typo in edit request for House of Chains
I've mis-spelled an artist name in notes and reference - Neil Gower as Neil Glower in this edit. I'll submit a fix rather than retract and re-type if you don't mind. Thanks. Doug 18:57, 28 October 2014 (UTC)
New cleanup report - "Series with Numbering Gaps"
A new cleanup report, "Series with Numbering Gaps", has been deployed. As of this afternoon we have 715 series with various types of gaps, e.g. J. T. Edson's Bunduki. Please note that some series cannot be fixed because one or more of their titles also belong to another series, e.g. Tarzan at the Earth's Core is book 4 in the Pellucidar series as well as book 13 in the Tarzan series. If a series can't be fixed, please explain the circumstances in the Series record's Note field before clicking on the "Ignore" link. Ahasuerus 20:07, 28 October 2014 (UTC)
- There are also series where not all books in the series are genre (and the author is not above the threshold to include non-genre books), so they should not be filled out. An example is the Avery Sisters Trilogy, only 1 book of which appears to be genre. (I've marked that one as "Ignore", and already had the note for it.) Chavey 03:46, 30 October 2014 (UTC)
Publisher and pricing.
I am verifying my set of Steven Erikson Malazan books. They are all Bantam Books, but generally have a price in £, so I've been verifying those publications. They list the publisher as Bantam Books (UK). I've been adding the Canadian price to the notes. Then I found some with only Canadian prices. One with only the £. All printed in the UK (except the one printed in Germany). They do all have one of two different UK URL's in the UPC area where the price is printed.
Do I list them all as Bantam Books (UK), and put the UK price where I have it, and a Canadian price (C$) where I don't? Doug 21:13, 28 October 2014 (UTC)
- Do the ones with only Canadian pricing indicate the address of the publisher? If it's a UK address, keep the publisher as "Bantam Books (UK)", and note that it is Canadian priced only, so was probably intended solely for the Canadian market. There may be printings with UK pricing, but we'll leave that until a primary verifier comes along. If the books give a Canadian address only, create a new publisher "Bantam Books (Canada)". Mhhutchins 01:40, 29 October 2014 (UTC)
New cleanup report - Publications with Invalid Catalog IDs
A new cleanup report, "Publications with Invalid Catalog IDs", has been added. As of this evening, there are 805 Catalog IDs that do not start with a "#". Some of them are ISSNs, some are malformed or truncated ISBNs, etc. There will be another cleanup report to find ISBNs with invalid checkums. Ahasuerus 00:35, 29 October 2014 (UTC)
- I have been moving ISSNs to the Note field when I come across them. I will continue to do so, using this list, but will not be notifying the primary verifiers. (This standard was discussed many moons ago, and should have already become the standard practice.) Mhhutchins 01:42, 29 October 2014 (UTC)
- The format has been tweaked a bit and a "Count" column has been added. Ahasuerus 00:48, 30 October 2014 (UTC)
New cleanup report - Publications with Invalid ISBNs
A new cleanup report, "Publications with Invalid ISBNs", has been deployed. At the moment "invalid" is defined as failing checksum validation. Of the 87 identified ISBNs, a dozen are actually ISSNs that look like ISBNs to the report logic, while the rest are ISBNs with checksum issues. This report supports the ability to ignore confirmed "bad" ISBNs. Ahasuerus 00:47, 30 October 2014 (UTC)
Suspect Amazon date
While handling Fixer's submission for this publication, I stumbled over an open contradiction. While Amazon (and Fixer) had the date as 2014-08-30, the blurb clearly stated Fans of the bestselling Belador urban fantasy series are anxiously awaiting book 5 - DEMON STORM - on October 20, 2014 and Demon Storm (book 5) was released Oct 2014. So I changed the dates of publication & title according to this statement. It leaves the question: Which dates supplied by Amazon are reliable in general? Christian Stonecreek 15:35, 30 October 2014 (UTC)
- Projected publication dates are generally supplied to Amazon by publishers. If a book is subsequently delayed and the publisher doesn't inform Amazon about it, Amazon's records can get out of sync.
- Also, it's been my experience that if a book is "indefinitely delayed", Amazon's reaction can be unpredictable. Sometimes the ISBN is deleted from Amazon.com's database, but remains in Amazon UK's database. Other times the publication date is changed to 2020, 2025 or 2030 (sic!).
- After processing close to 100,000 Amazon records over the last 6 years, I have developed fairly good heuristics that identify 90%+ of suspicious ISBNs. Nonetheless, a few slip through the cracks now and then, as you have discovered. Ahasuerus 16:15, 30 October 2014 (UTC)
- Thank you very much for the answers, Ahasuerus! It's good to know I don't have to give up on Fixer! Stonecreek 16:25, 30 October 2014 (UTC)
Terry Pratchett's WIntersmith
A proposed change to this pub verified by BLongley. I have a copy (unlikely to keep it).
There is an Author's Note on a final unnumbered page (last page of novel is also unnumbered), so I wanted to add + to page count and content of Author's Note (Wintersmith). This same note is in a paperback edition I'm adding as well. Note - my copy has no dust jacket, so I can't verify the artist / pricing. Doug 16:02, 30 October 2014 (UTC)
- Proceed to add the content and adjust the page count field. Since there is a primary verification which includes the artist credit and price, it's not necessary to note that the artist or price can't be verified by a second PV. If you had been the first verifier, you could have noted that, and then waited for a second verifier to do the verification of those fields and remove your note. Thanks. Mhhutchins 20:06, 30 October 2014 (UTC)
New cleanup report - Publications with Identical ISBNs and Different Titles
A new cleanup report, "Publications with Identical ISBNs and Different Titles", has been deployed. A few things to keep in mind:
- Only "major differences" are reported for now. If one title is a subset of another, e.g. "Title: Series" vs. "Title", neither pub will appear on the report.
- The publication count displayed on the main cleanup menu is 1132, but the number of suspect ISBNs is much lower, around 370.
- Some "ISBNs" are not real ISBNs, e.g. "0005-6677" used by "The Baum Bugle" is an ISSN. These pubs will disappear from this report once the ISSNs have been moved to the Note field, so there is no need to "ignore" them.
- In some cases publishers reused ISBNs on unrelated books or changed the title on subsequent printings. There isn't much we can do about that aside from "ignoring" the affected pubs.
- Many discrepancies are due to data entry errors (typos, using the cover title vs. the title page title, etc) or inconsistencies (hyphens vs. em-dashes, etc)
Happy fixing! :-) Ahasuerus 19:33, 30 October 2014 (UTC)
- Great list. Is it possible to move the "Ignore this pub" link into its own column? I'm afraid I'll inadvertently click on it while trying to get to the pub record. (I say that because among my first attempts, I almost clicked it!) Thanks. Mhhutchins 20:10, 30 October 2014 (UTC)
- Sure. Does it look better now? Ahasuerus 20:39, 30 October 2014 (UTC)
- Yes, better. Thanks. Mhhutchins 22:14, 30 October 2014 (UTC)
- One further request: is it possible to add a column giving the publication date of the record? This will make it easier to quickly determine if there is a duplicate record for the same edition/printing. Mhhutchins 00:46, 31 October 2014 (UTC)
- Done! Ahasuerus 00:57, 31 October 2014 (UTC)
- You deserve a raise. Who do I speak to? :) Mhhutchins 01:37, 31 October 2014 (UTC)
New cleanup report -- Publication-Title Type Mismatches
A new cleanup report, "Publication-Title Type Mismatches", has been deployed. At the moment it only checks that each publication has one (and only one) title of the matching type. Once the current crop of 144 "problem" publications has been cleanup up, we can tweak the logic to find other publications with invalid title types, e.g. NONFICTION titles in CHAPTERBOOKs. Ahasuerus 02:45, 3 November 2014 (UTC)
New cleanup report - Authors with Duplicate Pseudonyms
A new cleanup report, "Authors with Duplicate Pseudonyms", has been deployed. There was only one duplicate pseudonym-author relationship and I have already fixed it, so no additional action is required for now. Ahasuerus 23:24, 3 November 2014 (UTC)
New cleanup report - Container Titles in Publications with no Contents
A new cleanup report, "Container Titles in Publications with no Contents", has been deployed. At this time it's limited to COLLECTION and ANTHOLOGY titles which have at least one publication with fiction contents and at least one publication without fiction contents. Ahasuerus 23:35, 4 November 2014 (UTC)
2 New Cleanup Reports - HTML in titles
Two new cleanup reports, "Title records with HTML in Titles" and "Publications with HTML in Titles", have been deployed.
As far as I can tell, this substantially completes the process of converting the ad hoc reports/queries which were originally posted on the Bibliographic Projects in Progress, ISFDB:Data Consistency and Talk:Database Schema pages to nightly processing. Of the 55 nightly reports, a couple will need to be tweaked once the currently identified "problem" records have been cleaned up, but I am not aware of any other requested reports. Ahasuerus 00:48, 5 November 2014 (UTC)
User DESiegel60 has not responded to a query on his page since July 2012. I wonder if we should add the "This editor is no longer actively participating" banner to his Talk page? Chavey 18:16, 8 November 2014 (UTC)
- Sounds like a good idea. Ahasuerus 18:27, 8 November 2014 (UTC)
Coming submission approval changes
As per FR 549, I am working on changing the software to "Let moderators skip the post-submission page and go straight to the approval page". I have already changed and consolidated much of the underlying code, so I expect that the patch will be deployed tomorrow or on Friday. Since it will change the way moderators process their own submissions, I thought I'd warn everyone ahead of time to help minimize confusion. Ahasuerus 03:29, 13 November 2014 (UTC)
- Will the post submission option still be available?Kraang 04:12, 13 November 2014 (UTC)
- Do you mean a user preference that would allow a moderator to land on the post-submission review page after each submission, i.e. to keep the current behavior? I had no plans to do it since I didn't realize there was demand for it, but it can certainly be added as part of the patch. Ahasuerus 04:24, 13 November 2014 (UTC)
- It would be helpful when making major changes I like to review them and have the option of rejections. For simple changes the straight to approval would be very helpful.Kraang 03:46, 14 November 2014 (UTC)
- Post-change, you will still be able to review submissions and reject them. Currently, you see the "submission review" page twice, once after you create a submission and then again when you click on the "Moderate submission" link. Post-change, you will only see this page once, but you will still be able to review and, if necessary, reject your submissions.
- Oh well, it's too late anyway since I have already coded a User Preference to let moderators keep the current behavior. It will become visible on moderators' My Preferences pages when the rest of the changes are activated, probably in a couple of days. (I am in the process of changing all 30+ submission types and it's taking longer than expected.)
- However, please keep in mind that it's an "all or nothing" User Preference. If you turn it on, you will see the post-submission review page for all submission types. If you turn it off, you will go straight to the approval page. Ahasuerus 04:05, 14 November 2014 (UTC)
(unindent) Done! A lot of old junky code was removed/replaced in the process and additional checks for uncommon error conditions have been added. If you run into any issues, please post them here. Ahasuerus 02:26, 17 November 2014 (UTC)
New cleanup reports -- Suspected [...] Authors without a Language Code
The following cleanup reports have been added to the pool:
- Suspected Dutch Authors without a Language Code
- Suspected French Authors without a Language Code
- Suspected German Authors without a Language Code
- Suspected Other Non-English Authors without a Language Code
Hopefully they will make the process of assigning language codes to authors more efficient. Ahasuerus 05:31, 17 November 2014 (UTC)
- Hopefully you'll get a better response at cleaning these than I did in asking for help with this report. Mhhutchins 05:55, 17 November 2014 (UTC)
- Well, the records are not going anywhere, so one way or another we'll get to them. Eventually :-) Ahasuerus 07:33, 17 November 2014 (UTC)
- To Michael => Why didn't you just ask? [User:Hauck|Hauck]] 15:12, 17 November 2014 (UTC)14:51, 17 November 2014 (UTC)
- Re-read the message. I did ask for help, otherwise I wouldn't have said that I'd asked for help. Mhhutchins 17:47, 17 November 2014 (UTC)
- A few days ago Michael posted a request on the Community Portal, but it presumably got lost in the shuffle. It's not an uncommon problem when a project gets so busy that it becomes hard to keep track of everything that is going on. I am not sure how we could improve the process... Ahasuerus 16:32, 17 November 2014 (UTC)
- (some of these errors appeared when cloning backwards in time a publication first entered in a later printing, my mistake, others came from books not PVed nor created by me). Done now. Hauck 15:12, 17 November 2014 (UTC)14:51, 17 November 2014 (UTC)
- Thanks! I have changed the threshold for French authors from 6 titles to 4 and forced the compilation logic to ignore "The Readers". Ahasuerus 17:07, 17 November 2014 (UTC)
- P. S. There was a minor bug in the compilation logic which inflated the title count for some authors. It basically counted reviewed titles twice, so some authors with 1-3 titles appeared on the list. I have fixed the logic and the data should be corrected when all reports are recompiled in 12 hours. Ahasuerus 17:39, 17 November 2014 (UTC)
- Most of the suspected Dutch authors are (surprise) Dutch. Some are Belgian and there are a few I don't recognise. I'll be working on this list the coming days. What happens if you set the limit to one Dutch title per author? --Willem 20:02, 17 November 2014 (UTC)
- There are 453 suspected Dutch authors. 296 of them have only one Dutch title and the other 157 have 2+ Dutch titles. I figured that it would be easier to clean up authors with multiple titles first, hence the limit. Similarly, there are 1415 suspected French authors and 1465 suspected German authors, so I chose (hopefully) appropriate limits/thresholds to make these reports more manageable. As authors get cleaned up, language-specific limits will be lowered until all authors have language codes assigned.
- Also, please note that I have found a couple of English language authors who didn't have a language code assigned and who popped up on these reports. Some of their translated titles had language codes assigned and were not set up as VTs. Ahasuerus 20:47, 17 November 2014 (UTC)
- There are only a few left on the Dutch list that need some more research. Can you set the limit to 1? --Willem 19:30, 20 November 2014 (UTC)
- Done. The other thresholds have been adjusted as well. Ahasuerus 20:37, 20 November 2014 (UTC)
- The thresholds for suspected French and German authors have been further adjusted. Ahasuerus 00:38, 1 December 2014 (UTC)
Fixer: November 2014
Fixer has completed his monthly run and is now ready to start creating submissions. There are more ISBNs than last month (it fluctuates from month to month), so I expect that it will take a number of days to process everything. I won't know the exact number for a few days since I need to do a lot of pre-processing.
As we discussed earlier, I have been trying to increase the share of ISBNs submitted on Fixer's behalf, although I expect that I will continue to handle certain categories for now. Ahasuerus 01:30, 18 November 2014 (UTC)
- New 2013 ISBNs have been sorted out and the eligible subset has been submitted. Ahasuerus 03:15, 18 November 2014 (UTC)
- All 2014 NewPubs (except for October and November, which add up to 109 ISBNs) have been submitted. Ahasuerus 05:37, 4 December 2014 (UTC)
- I've noticed that Fixer has submitted ebook AddPubs as far back as 2008-01 (and very likely to go back to the date of the format's invention), but that submissions for paper AddPubs only go back to 2012-09. Is there a reason for giving preference to ebooks over paper? Or have paper AddPubs before 2012 already been handled? Mhhutchins
- Well, we have already processed the vast majority of e-AddPubs, so I'd like to get the remaining 3,200 out of the way. Unfortunately, my ability to multitask has been somewhat stretched lately, what with all the different types of Fixer submissions, software changes, etc. Removing at least one variable from the equation will be nice :-) Of course, we'll still want to process 2014+ e-AddPubs going forward. Ahasuerus 16:20, 4 December 2014 (UTC)
- If it's a matter of prioritizing, why not set aside all Fixer submissions for the time being, and concentrate on software changes? Unless you believe the multitasking keeps you from away from Boredom City. :) Or if you think taking a break will cause a logjam in clearing Fixer's queued submissions. Believe me, those of us who work on Fixer submissions will find plenty of stuff to do until you start him up again. I may even get a chance to get back to my long-neglected Tuck and Reginald projects. Thanks. Mhhutchins 16:45, 4 December 2014 (UTC)
- There are various considerations here. One problem with Fixer's submissions is the way the software is written. Fixer has permanent prioritized queues plus a special temporary queue for "about-to-be-submitted" ISBNs. This has many implications. For example, suppose that I am in the middle of submitting new ISBNs for January 2015. I am about to create a submission for book 3 in a trilogy when I notice that we are missing book 2. When something like this happens (which it does fairly often), I have to delete the "about-to-be-submitted" queue, run a special "ad hoc" query for the missing ISBN(s), submit them to ISFDB, then rebuild the temporary queue. It's not a big deal, but every iteration takes time and brain cells to go through the motions.
- Because of this and other software limitations, it's easier to create certain types of submissions (like "e-AddPubs for 2007-12") than other types of submissions. At the end of my day, i.e. some time around 1am or 2am, when I am no longer able to work on the software, I usually check how many "New" submissions we have in the queue and do a simple Fixer run.
- Then there is the issue of keeping the database up to date. If we were to stop adding "forthcoming" and recently released ISBNs, even for a few months, it would make the database less useful overall. To prevent this from happening, I have to run Fixer's data acquisition phase starting on the 10th of each month. The process is mostly automated, so I can do other things in parallel, but it still requires a fair amount of babysitting. Then, starting on the 15-16th of each month, I have to prioritize the 5,000+ new ISBNs that Fixer finds. Typically, about 3,000 can be prioritized automatically, but the other 2,000+ take a few days to sort out. We usually end up with fewer than 600 "priority 1" ISBNs, although it varies from month to month -- August-September are particularly intensive. Of that number, about 150-200 are AddPubs, which do not take long to submit and approve. The rest are NewPubs, which I used to submit on my behalf, but lately I have been trying to shift them to Fixer. And so it goes :) Ahasuerus 21:58, 4 December 2014 (UTC)
(unindent) NewPubs for 2014-11 have been submitted. There are 63 2014-10 ISBNs left in Fixer's NewPubs queue. Ahasuerus 02:05, 8 December 2014 (UTC)
- NewPubs for 2014-12 have been submitted. Ahasuerus 00:31, 9 December 2014 (UTC)
The Editor's Notebook (Fantastic Adventures)
I am putting a bunch of Ray Palmer's editorials into their proper series including this piece as the primary verifier is long gone, the note on his page states that I should list the disambiguating of this title on this page. MLB 05:45, 18 November 2014 (UTC)
- Submissions have been accepted. The placement of title records into series is not a change in a publication, but in the title. So it isn't mandatory to inform PV editors of the publications when you're doing this. But it wouldn't hurt to ask if you feel there may be a question about such changes. Thanks. Mhhutchins 06:29, 18 November 2014 (UTC)
- The disambiguation of generic titles doesn't have to be discussed as well, unless you find that an editor is consistently forgetting to do it. Mhhutchins 06:31, 18 November 2014 (UTC)
Cleanup report requested
We currently have hundreds of titles that are listed in the format "Title : Sub-title", violating the standard format of "Title: Sub-Title". There are a few cases where "<space>:" is correct in a title, e.g. when used as a ratio, such as in A : b : O, but most of the time a space followed by a colon is a standards violation. A cleanup report looking for those could be helpful. We might combine this with a search for wrong formats of ellipses in titles, "curly quotes" or "curly apostrophes". Chavey 22:52, 20 November 2014 (UTC)
- This is true as far as English language titles go, but in some other languages (French and ?) "Title : Sub-title" is the standard. I guess we can make the report language-aware. Ahasuerus 23:19, 20 November 2014 (UTC)
- The space before and after is also a standard used by the librarians at Worldcat. It's possible that this format was used for entering data from OCLC without changing it to match ISFDB standards. As for ellipses in titles, that's been discussed ad infinitum and I think we'd best let that issue rest for awhile. Mhhutchins 02:30, 21 November 2014 (UTC)
- I thought we had finally reached a consensus on the ellipsis format, i.e. that we were switching to 3 dots, no spaces in between. That at least was my reading of the result of this conversation. Obviously, if I've read that wrong, then we wouldn't include that as part of a cleanup that was trying to check title format standards. Chavey 06:11, 21 November 2014 (UTC)
- As far as french titles are concerned, I've taken two courses of action. When the title was in the "title subtitle" format (without ":" anywhere) I used the ISFDB standard (no space before ":" like here). When the ":" was present in the title, I followed the rules for french typography (a hard space before and after ":"). Note that some time ago someone took the liberty to discreetly modify this policy (by suppressing spaces that were clearly in the title like here which was transformed into _Enjeu: Le monde_) without notification. I found this attitude particularly annoying and lacking elementary politness. You'll understand easily that I'm against such automatic change. Hauck 10:26, 21 November 2014 (UTC)
- I thought we had finally reached a consensus on the ellipsis format, i.e. that we were switching to 3 dots, no spaces in between. That at least was my reading of the result of this conversation. Obviously, if I've read that wrong, then we wouldn't include that as part of a cleanup that was trying to check title format standards. Chavey 06:11, 21 November 2014 (UTC)
New Cleanup report - Titles with Invalid Story Length Values
A new cleanup report, "Titles with Invalid Story Length Values", has been deployed. In addition, the thresholds of "Suspected French Authors without a Language Code" and "Suspected Other Non-English Authors without a Language Code" have been adjusted. Ahasuerus 03:00, 27 November 2014 (UTC)
Facsimile Reprints I
There seem to be some gaps in older magazines when it comes to content. Is it alright to import content from a facsimile reprint as long as a note is included that this is where the content has come from and that it cannot, until later, be authenticated. MLB 01:51, 27 November 2014 (UTC)
- I'm asking this question here on this page as a moderator would have to okay any importing, and a moderator is the person to ask. MLB 01:55, 27 November 2014 (UTC)
- I am a bit nervous about this. If memory serves, a few reprint magazines may have included some (non-fiction?) material from other issues. Ideally we would have another independent secondary source in these kinds of cases. Ahasuerus 03:13, 27 November 2014 (UTC)
- Fine. Just asking. I have noted some discrepancies in some facsimiles, including missing ads, columns, artwork, letters, etc. Leading me to assume that most facsimiles are not always a reliable source. This is why I always list them now as anthologies. MLB 00:58, 1 December 2014 (UTC)
Facsimile Reprints II
Just an idle thought to those in charge. As there are, at least, four companies out there doing facsimile reprints of old magazines, perhaps the ability to clone magazines should be considered? MLB 01:53, 27 November 2014 (UTC)
- Unfortunately, this limitation is embedded in the software, which assumes that there can be only one edition of a magazine issue. I have created a Feature Request to change this behavior, but at this point I am not sure how extensive the required changes will be. Ahasuerus 03:09, 27 November 2014 (UTC)
The post-approval pages for New/Add Pub, Edit Pub and Clone Pub submissions have been changed. A new link, "Check for Duplicate Titles", has been added. Ahasuerus 00:13, 1 December 2014 (UTC)
- This is great! --MartyD 14:29, 2 December 2014 (UTC)
- I agree! :) Ahasuerus 23:43, 2 December 2014 (UTC)
- Good idea! Chavey 14:56, 3 December 2014 (UTC)
Multi-format e-book w/varied ISBNs
I accepted this Fixer submission after it checked out against the Look Inside. But then I noticed this one and Harry's notes in it about the various ISBNs for each format. Do we want one consolidated entry? --MartyD 14:28, 2 December 2014 (UTC)
- My overall approach is that anything that reduces the number of searchable/displayed ISBNs is likely suboptimal. In this case we also have a different ASIN to deal with, so I would favor keeping the new pub as a separate record and perhaps clarifying the Note field of the original record. Ahasuerus 23:49, 2 December 2014 (UTC)
- Unfortunately, Harry's notes indicate that there were 5 different ISBNs and 13 different formats, so it will be difficult to determine format-ISBN links :( Ahasuerus 00:26, 3 December 2014 (UTC)
- Whoah! Mhhutchins 00:45, 3 December 2014 (UTC)
- The notes have several minor issues. First to note, is that some ISBNs are listed multiple times in ISBN-10 and ISBN-13 format (so the list is bigger than five but there are in fact only five unique entries). Based on the notes it seems like the ISBN list comes from the "February 2008" "Publisher/Vendor records" (web site maybe). Finally, this verified pub entry seems to be for a multiformat pub (looking at this entry on Google Books lists it as "Google eBook"). Ebooks often have different ISBNs for different formats much like printed books but one complication is the so called "multiformat". This is a common format for ebooks where you buy access to the book and can download/access it in one of a number of formats instead of locked to one/few (notice the notes state verified against the PDF version of this edition). The fixer entry is likely a Kindle only format. Uzume 17:05, 23 January 2015 (UTC)
Fantastic Adventures, September 1939
I would like to add some information dealing with Fantastic Adventures, September 1939 based on the facsimile that I own, including the illustrator for "Romance of the Elements: Antimony", an article and a map for the story "Golden Girl of Kalendar" by F. Orlin Tremaine (the article is a full page in length giving the history of Kalendar), the illustrator's name (one of the illustrations is signed 'RF' which I believe is Robert Fugua, and I can replace the cover image with one that shows the back cover illustration by Paul along with the front cover artwork. I could add a note saying where this information came from. If this action is acceptable I would add the facsimile as a new publication and then import the contents from the magazine. MLB 02:12, 7 December 2014 (UTC)
- All fine, except do not replace the cover with an image containing both the front and back covers. It would confuse users to see that the "cover" isn't credited to both artists. Since we only credit the front cover of a publication, and upload cover scans of front and back if it's a single wraparound work by the one artist. Mhhutchins 06:27, 7 December 2014 (UTC)
- Done, and awaiting acceptance. MLB 07:51, 7 December 2014 (UTC)
Fixer - paper-based AddPubs for 2012
By popular demand, Fixer has shifted to paper-based AddPubs for now. It's going surprisingly quickly: there aren't as many of them as I had expected and many are CreateSpace reprints of public domain books, which get relegated to Queue 2. At this time, there are only 437 2012 potential AddPub ISBNs left and I expect that there will be fewer than 400 actual submissions. Ahasuerus 17:18, 9 December 2014 (UTC)
- As of a few minutes ago, there are 192 2012 AddPub ISBNs left in Fixer's queue. Ahasuerus 05:18, 10 December 2014 (UTC)
- All paper-based AddPubs for 2012 have been submitted! Please note that a number of supposed UK editions appear to be US editions in disguise. Ahasuerus 04:50, 11 December 2014 (UTC)
The thresholds for "suspected" German and "other non-English" authors have been adjusted. Ahasuerus 06:19, 12 December 2014 (UTC)
- A request : It may be interesting to have a value like "unknown" (or else) for some cases where choosing the correct language is only a matter of pure luck. Hauck 15:46, 12 December 2014 (UTC)
- Would you happen to have an example? Admittedly, we have a number of bilingual authors like Sam Lundwall and sometimes it can be hard to tell what the author's primary working language is/was. On the other hand, the only thing that the "Working Language" field affects is whether a canonical title has the name of its language displayed in square brackets on the Summary page.
- Let's use Andrej Dmitruk as an example since we have only 2 of his stories listed. The first story was originally written in Ukrainian and the second one in Russian. Because Dmitruk's working language is currently set to Ukrainian, the Russian title has "[Russian]" displayed next to it while the Ukrainian title does not. If we were to change Dmitruk's working language to Russian, then the Ukrainian title would have its language displayed while the Russian title wouldn't.
- If we were to add "unknown" to the list of supported working languages and change Dmitruk's working language to "unknown", then both stories would have their languages displayed on the Summary page. Is this what you had in mind? Or is there another scenario that may benefit from adding "unknown"? Ahasuerus 20:12, 12 December 2014 (UTC)
- It's mostly problematic with artists, like this one who is likely Russian but only known by her work in a French title. I've also attributed a language (German) to an author with only two publications in two different languages (french and German) where I couldn't find any bibliographic data (there was no indication of an original title in the french publication).Hauck 11:10, 13 December 2014 (UTC)
- I like the idea of having some version of author language that amounts to saying "label everything for this author by language". I'm not sure that "unknown" is the right name for that "language" -- we "know" what's going on (evenly divided). WorldCat's "Undetermined" might be slightly more accurate. Chavey 15:11, 13 December 2014 (UTC)
- Just for the record : Elena Gondik is Ukrainian; I've just updated her record (I thought I had already !). As far as the “unknown” problem is concerned, there are obviously two issues here : 1, as it has been said, the need of a special option (e.g. “diverse”) when a writer uses different languages; and 2, another one, which could be “unknown” or “uncertain”, when such is the case. As for me, I have only met the former, but the latter could crop up anytime (the Gondik case was just an omission on my part). Linguist 16:28, 13 December 2014 (UTC).
- If I recall correctly, two of the reasons for starting the project to assign languages to authors was 1) to automatically (non-manually) assign a language to all non-assigned TITLE records and 2) to remove language from all COVER ART and non-maps INTERIORART records. I don't see the point in trying to assign a language to authors who have less than a dozen title records. If there is one, please make it clear why some of us are going through the effort to assign language in the first place. (BTW, I have assigned a language to literally thousands of authors since the main clean-up report was created, working on a list of authors who had a 1000 or more titles down to the current list with authors who have 27 or less titles.) Mhhutchins 04:30, 14 December 2014 (UTC)
- It sounds like there may be two separate questions here. The first question is whether we should add a new language choice (I think I like "various" better than "diverse") to the list of supported languages in the Language drop-down list in Edit Author.
- (A little detour just to make sure that we are all on the same page: The original reason for adding the "Working language" field to author records was to enable the display of the language of translated books with no analog in the original language. For example, consider the Collection and Omnibus sections of Philip K. Dick's Summary page, e.g. "Ubik [German] (2003) [O]", or Hervé's Summary page. If we were to add "various" to the list of Author languages, it would force the software to display languages for all canonical titles.)
- The second question has to do with the "Prolific Authors without a Defined Language" cleanup report which was added in April and its language-specific cousins added last month. The original reason was, as Michael said, to automate the process of assigning languages to titles. As far as COVERART/INTERIORART goes, as I recall, it was a separate feature request.
- While we are at it, here is where we stand as of this morning:
- Titles with a language code: 402701
- Titles without a language code: 608637
- Authors with a language code: 20988
- Authors without a language code: 104250
- While we are at it, here is where we stand as of this morning:
- I am thinking that it may be desirable to change the software to assign languages to new authors based on the language of the title that they are associated with. Ahasuerus 05:57, 14 December 2014 (UTC)
[unindent] 104K authors without an assigned language? (And all having 27 or less titles in the db.) My head is still spinning trying to comprehend that figure! Yes, let's start automatically assigning authors a language based on the majority of their titles which have an assigned language. And let the chips fall where they may, with corrections being made as each erroneously assigned author is discovered. Otherwise, at the rate the current project is going, it will be a decade before completion. I can't imagine an error rate of more than 1% by going the automatic assignment route. Mhhutchins 06:48, 14 December 2014 (UTC)
- Well, at some point -- hopefully soon -- we will have to bite the bullet and assign "English" to all unassigned authors. However, I don't think we are quite done with the low hanging fruit yet. The "Suspected [non-English] Authors" reports have proved useful and, at the rate we are going, should be cleaned up in a few weeks. Once that's been done, I can create a new cleanup report to find pubs which simultaneously contain:
- non-English canonical titles or canonical titles without a language by authors whose working language is not English, AND
- canonical titles without a language by authors without a language
- Basically this report will look for pubs which contain non-English (confirmed or suspected) canonical titles. It will then check if any other authors in these pubs do not have a language assigned and report them. Since it seems likely that in most cases all titles in these pubs will be non-English, the hit rate will be hopefully high. Ahasuerus 16:25, 14 December 2014 (UTC)
- P.S. A question for Hervé: I believe I understand the intended functionality of the proposed "diverse"/"various" language code, but what kind of effect would "undetermined" have? Ahasuerus 16:27, 14 December 2014 (UTC)
- It may prove useful as a temporary measure, like in the case of the Ukrainian author cited above, before we acquire more solid evidence. Hauck 16:59, 14 December 2014 (UTC)
- I take it that it will have the same effect on the way titles are displayed as "diverse"/"various", right? Ahasuerus 04:24, 15 December 2014 (UTC)
- Oh, better yet: "multiple" instead of "diverse"/"various"! Ahasuerus 07:32, 15 December 2014 (UTC)
Smart Pop books
While discussing Lawrence Watt-Evans' bibliography with the author, I discovered that our coverage of "Smart Pop" books was rather spotty and inconsistent. We have 3 publisher records for "Smart Pop": Smart Pop, Smart Pop / BenBella and Smart Pop / BenBella Books. We also have a Smart Pop publication series under BenBella Books. Some Pop Smart pubs are not identified as such, e.g. Revisiting Narnia: Fantasy, Myth and Religion in C. S. Lewis' Chronicles. Many are missing contents data, sometimes by major authors like Silverberg, Resnick, Connie Willis, etc., e.g. see this Pop Smart page where they list the contents of War of the Worlds: Fresh Perspectives on the H. G. Wells Classic.
I am currently busy trying to rewrite the Summary Bibliography page (which is not only time-consuming, but also rather painful), but if anyone wants to clean up this area and reconcile it with the publisher's Web site, I will provide moral support :) Ahasuerus 19:25, 12 December 2014 (UTC)
- Looking at the OCLC records and the Amazon "Look Inside", it appears that Smart Pop was a publication series for the first 6 or 7 years published by BenBella. The first publication to show it as "an imprint of BenBella Books" was in 2010. I think we should keep those titles already in the database as a publication series as they appeared and not to rewrite history. I'll merge the publisher "Smart Pop" into "Smart Pop / BenBella Books", and look at adding titles when I get a chance. Mhhutchins 04:44, 14 December 2014 (UTC)
Dark Discoveries #23 , Spring 2013
I just entered the contents of this magazine to this site, and as usual I did a ham-fisted job of it. I misspelled two authors names. If this posting is accepted I will correct the spellings. Although it's been a long night, there is no excuse. MLB 08:55, 16 December 2014 (UTC)
- Hi. I accepted it. I didn't try to correct anything or do anything else with it. --MartyD 12:36, 16 December 2014 (UTC)
Are we allowed to link to author photos from Amazon? I hadn't thought of this before, but we currently have a submission to do so, and I just don't know the answer. Chavey 21:15, 16 December 2014 (UTC)
- I would assume that's implied, but have never seen explicit permission. I asked the same question here a while back about photos on Goodreads (an Amazon property), but never got a response. Mhhutchins 00:58, 17 December 2014 (UTC)
- Unfortunately, I don't know the answer. We could always ask Al, but I am not sure he remembers/knows the details either. Ahasuerus 01:01, 17 December 2014 (UTC)
A stupid question
I've just noticed that when an author entry is wiewed in chronological order, all the books in a fiction series seems to disappear. Compare here and there and that only some titles (part of series but in other languages) appear on the alphabetical list there. Is it nomal ? Hauck 19:08, 21 December 2014 (UTC)
- Strange. I've never noticed that before. Perhaps it's the results of the recent update affecting title display on a summary page. I'm pretty sure it hasn't always been like that. Thanks for finding the bug. Mhhutchins 19:14, 21 December 2014 (UTC)
- Yes, this bug was introduced in the last two patches which modified the way series are displayed. I have already fixed it in the next ("non-genre") patch, which I hope to deploy Real Soon Now (tm). The coding is done, but I am still running tests since the changes affect 35 existing scripts. Ahasuerus 19:21, 21 December 2014 (UTC)
Unless they have always been that way it looks as though someone has changed the title years of the Tarzan titles to the date of the first serial part rather than the date of the first book publication. The same also seems to be true of the Mars novels. Since the rules are so easy to access while editing pubs I hope that is not happening too often now.--swfritter 22:33, 22 December 2014 (UTC)
Fixer's 2014-12 run
Newly identified AddPubs for 2014 have been submitted. Ahasuerus 05:23, 24 December 2014 (UTC)
- Over the next few days Fixer will be submitting batches of newly identified 2014 NewPubs. There are a bit more than 100 2014 ISBNs in the queue, so nothing massive. Ahasuerus 21:40, 24 December 2014 (UTC)
- All outstanding ISBNs have been submitted. Ahasuerus 01:58, 31 December 2014 (UTC)
Cleanup report requested - Mixed type series
If a title series contains both books and editor records, then in most cases it's probably an accidental combination of what should be two (or more) different title series. I just disambiguated three different series named "Aurora", but I left untouched the example of Chaos, which is easily seen to be two different series. This suggested cleanup report would catch this one (and would have caught "Aurora"). Chavey 23:22, 24 December 2014 (UTC)
- An interesting question. Checking the database, I see that we have 16 series which include EDITOR and non-EDITOR titles. They seem to fall into 3 different categories.
- The first category, which includes Chaos, Lords of the Abyss, The Dark and The Sixth Sense, covers series which clearly need to be disambiguated.
- The second category includes series like Planet Stories: Adventure House Reprints, Thrilling Mystery: Adventure House Reprints' and Weird Tales Facsimiles. Their titles are a mix of EDITOR and ANTHOLOGY records, but the title type assignment appears to arbitrary and may need to be made uniform.
- The third category includes the likes of Galaktika, Tales of the Unanticipated and Fantasy Magazine. Their titles are also a mix of EDITOR and ANTHOLOGY records, but the distinction appears to be legitimate and presumably needs to be preserved.
- The first and the second categories can be easily handled. The third category is more problematic. Should we keep the current title assignment by adding an "ignore this record" option to the cleanup report? Or should we make a policy decision that "mixed type" series, when they occur, should be broken up into separate type-specific sub-series (within the same super-series)? Ahasuerus 00:39, 25 December 2014 (UTC)
- I see no reason why a series can't contain both book and magazine publications. There are legitimate reasons why they should be allowed and not arbitrarily assumed to be an error. Take Postscripts for example. If you must create such a clean-up script, please allow the "ignore this record" option, and moderators must be certain before changing the records just to remove them from the clean-up report. (The issue about facsimile reprints is a different matter and has never been discussed fully, at least to my satisfaction.) Thanks. Mhhutchins 04:57, 26 December 2014 (UTC)
- I assumed there would be cases where the mix of EDITOR and non-EDITOR records would be appropriate, hence my wording that "in most cases it's probably an accident", and I also suggest that we add the "ignore this record" button. As with Michael, I think we need to take a somewhat cautious approach to this script, but I do think it would help us find a number of incorrectly mixed series, and prevent the introduction of others. The "Aurora" case, for example, I know was a single series a year ago when I last worked with it, but this year it had become a mix of three series. Chavey 23:49, 26 December 2014 (UTC)
- Your wish is my command! :) The new cleanup report has been coded and deployed. (Sorry, something has come up over the last couple of days and I have been somewhat less active than I expected to be.) Ahasuerus 06:57, 27 December 2014 (UTC)
Another stupid question
Hello, when using the Alphabetical display, there's perhaps a problem with pseudonyms : in this case, the titles existing by Andrevon and its pseudo Brutsche appears twice without explanations. Is it normal? Hauck 10:36, 26 December 2014 (UTC)
- It looks like it's related to Bug 491:
- Alphabetical Biblio page doesn't always show collaborators. This happens because it has been fixed to show VTs, but the logic that retrieves collaborators doesn't look at VTs' collaborators.
- Now that the core series logic has been revamped, I can start working on other Author Biblio issues like this one. Ahasuerus 17:08, 26 December 2014 (UTC)
- Thanks. Hauck 17:58, 26 December 2014 (UTC)
The last two thresholds have been eliminated, so now it's "what you see is what you get" for all languages. Ahasuerus 07:11, 27 December 2014 (UTC)
Missing Title Reference
Hello. There seems to be no Title Reference on the Publication Record of Una temporada en el infierno, the Spanish translation of Rimbaud’s Une saison en enfer. There might be a simple explanation to it (simple, but beyond my grasp so far), but just in case this might be the beginning of an alien bug invasion, I thought I'd post this notice here. Thanks. Linguist 11:06, 27 December 2014 (UTC).
- It's because the publication type and the title type do not match. The publication is NONFICTION, but the matching title is POEM. This can be seen be editing the pub and looking at the contents. Do you know if both are NONFICTION or is this actually a CHAPBOOK with a POEM? Thanks. -- JLaTondre (talk) 11:28, 27 December 2014 (UTC)
- It is indeed a poem in prose, which could be taken for something else by an unsuspecting eye. I'll go and correct the type. Thanks a lot. Linguist 11:40, 27 December 2014 (UTC).
- I accepted the change to CHAPTERBOOK. However, that alone wouldn't fix the issue as there wasn't a matching CHAPTERBOOK content record. I've gone ahead and added that content type so all is good. There always needs to be a content type (under the content section when in publication editor) that matches the publication type or the title reference will be missing. Thanks. -- JLaTondre (talk) 12:41, 27 December 2014 (UTC)
- Thanks ! Linguist 13:20, 27 December 2014 (UTC).
- I accepted the change to CHAPTERBOOK. However, that alone wouldn't fix the issue as there wasn't a matching CHAPTERBOOK content record. I've gone ahead and added that content type so all is good. There always needs to be a content type (under the content section when in publication editor) that matches the publication type or the title reference will be missing. Thanks. -- JLaTondre (talk) 12:41, 27 December 2014 (UTC)
- It is indeed a poem in prose, which could be taken for something else by an unsuspecting eye. I'll go and correct the type. Thanks a lot. Linguist 11:40, 27 December 2014 (UTC).
Library of Doom chapterbooks
Fixer has submitted the missing "Library of Doom" chapterbooks. (And in this case "chapterbooks" actually means "chapterbooks for elementary school children".) Please note that although Amazon lists some of them under "comics and graphic novels", reviews indicate that the stories are extensively illustrated, but the text can stand alone. Ahasuerus 16:24, 31 December 2014 (UTC)
- Since many of the pubs in this series were already entered as CHAPTERBOOKS (per the ISFDB definition) with SHORTFICTION contents, I fixed these records to conform with the others. Mhhutchins 23:05, 3 January 2015 (UTC)
- Thanks! Ahasuerus 23:10, 3 January 2015 (UTC)