ISFDB:Community Portal


(Redirected from Community Portal)
Jump to: navigation, search

ISFDB Noticeboards
Before posting to this page, consider whether one of the other noticeboards might suit your needs better.
Help desk
Questions about doing a specific task, or how to correct information when the solution is not immediately obvious.
• New post • Archives
Verification requests
Help with bibliographic, image credit, and other questions which require a physical check of the work in question.
• New post • Archives
Rules and standards
Discussions about the rules and standards, as well as questions about interpretation and application of those rules.
• New post • Rules changelog • Archives
Community Portal
General discussion about anything not covered by the more specialized noticeboards to the left.
• New post • Archives
Moderator noticeboard
Get the attention of moderators regarding submission questions.
• New post • Archives • Cancel submission

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

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


Roadmap 2017

As of late, I have been tweaking the software based on user feedback and in support of various cleanup projects. It was a good match while I was recovering from the most recent unpleasantness over the last couple of months.

Going forward I hope to get back into more heavy-duty development work. With that in mind I have compiled a list of software projects that are currently close to the top of my list of priorities for 2017. I'd like to post them and solicit feedback re: their desirability and relative priority. If a description seems too vague or incomplete, please don't hesitate to ask for a clarification. Detailed suggestions/feedback re: individual FRs may need to be spun off as separate Community Portal sections. Additional requests welcome!

Continuation and further development of currently active projects

  • Language cleanup:
    • Assign languages to all titles
    • Assign languages to all authors. Should we use the manual approach that we used with titles? Or can we automate some parts of the process based on the language of each author’s titles?
      I think that for the ones remaining (with 3 titles or less), we can just assign automatic English if all the titles they have are English (after all titles have languages). We will miss a few that are not English but we will also miss a lot of them if we go manually - they will require a lot of research at this point. If a non-English title ever appears, we can always change it. That won't be much different to what will happen now if someone adds a magazine in English and do not change the author language manually after the addition... Annie 20:38, 19 February 2017 (UTC)
      Or if someone is too worried about doing it, going the manual way will always work of course. Annie 05:38, 21 February 2017 (UTC)
      I guess the "low-hanging fruit" would be identifying language-challenged pseudonyms of language-enabled authors and auto-assigning the parent author's language to the pseudonym. Ahasuerus 16:07, 21 February 2017 (UTC)
    • Yes, I think we even talked somewhere about that. If those can be done, this will eliminate a pretty good chunk of what is remaining... Annie 16:25, 21 February 2017 (UTC)
  • Wiki cleanup
    • Move series/magazine-, publisher- and publication-specific etc Wiki pages to the database
    • Move Bio and Bibliographic data to the database
Globally OK.Hauck 20:00, 19 February 2017 (UTC)
Along with this (though it might not qualify as part of your currently active projects), I would like to add, it would be very good to try and untangle the main DB from the wiki DB (particularly with respect to login credentials and sessions, etc.; I have a few ideas on this). This would facilitate upgrading our MediaWiki software which we are sorely behind on (including many security/bug fixes as well as features). Uzume 20:41, 19 February 2017 (UTC)
Unfortunately, my knowledge of the MediaWiki software is very limited. Last we talked about development, Al was going to look into this issue, but he has been unavailable for a long time. I could try and get myself up to speed, but, unfortunately, learning new things gets harder as you get older. Ahasuerus 16:18, 20 February 2017 (UTC)
It is already untangled in terms of sessions - I get asked to login in wiki while my session is fine on the site and vice versa. So it is just the users that are the same. Annie 05:38, 21 February 2017 (UTC)

New projects

  • Create a separate queue for automated submissions, which will enable non-moderators to work on them (and let me concentrate on development). Rather time-consuming to implement. Discission moved here.
  • Internationalization:
    • Support multiple prices per publication.
    • Support multiple publishers per publication. May require more thought re: imprints.
      I wonder if that cannot be solved in a different way - allow books published from a joint venture to show on 3 publisher pages - a combined one and the two separate publishers ones. Or if we go for multiple publisher, we should have a way to see which books are from both publishers at the same time without the need to compare lists manually Annie 20:52, 19 February 2017 (UTC)
      An interesting point. I have never considered it, but the proposed behavior seems desirable. We'll have to think of the best way to implement it. Ahasuerus 02:18, 20 February 2017 (UTC)
      And here comes a live example that I and KarenHunt had been investigating: Братья по оружию. It is a joint venture between the Moscow based Астрель and the Minsk based Харвест with separate ISBNs for both (9785271439186 for Астрель and 9789851815919 for Харвест) and usually all copies of the books carry both publishers and both ISBNs (I've had a few similar ones). If you create a fictitious "Астрель / Харвест" (incorrect it is not an imprint) or "Астрель & Харвест", it will remove the book from the two individual publishers' pages. Adding two separate copies (one per ISBN) is not exactly correct either as it is the same book (ha - this is one of the other projects below - multiple ISBNs). The cleanest seem to be to go for one of them and note the other in the notes. None of those scenarios are very good though. Annie 01:09, 25 February 2017 (UTC)
      And just to illustrate why it is a bad idea to use the double name, this wonderful example of Russian thought has 4 valid ISBNs on the same publication - it is again a joint venture between Харвест and this time another of the publishing group АСТ's imprints/daughter companies - this time Хранитель instead of Астрель. But just to make it funnier, it also got ISBN for the parent as well and an ISBN for the pure Moscow "АСТ Москва" which was consolidating their list. So we will need yet another joint version for this one. As you can imagine, there are more of those. :) Annie 01:25, 25 February 2017 (UTC)
      We had a discussion of Russian publishers a while back, which is what pushed this FR up the list of priorities. There are other publishers who use multiple ISBNs, but Russian language publishers account for the majority of our "problem cases". Ahasuerus 01:34, 25 February 2017 (UTC)
      Well - I saw an example, decided to post it so it is in the thread. Sorry if it was repeating already known data - I do not remember stumbling on this discussion when I was reading archives. Annie 01:52, 25 February 2017 (UTC)
      Not a problem at all! I was just sharing additional background information. Part of my job description :-) Ahasuerus 02:03, 25 February 2017 (UTC)
    • ISBNs (mostly to support internationalization, but will also help in other cases):
      • Support multiple ISBNs per publication
      • Move catalog IDs to a separate field, which will help with pubs that have both an ISBN and a catalog ID, e.g. book club publications.
      • Add two checkboxes next to each ISBN: “derived” and “corrected”. The latter will let us capture two versions of invalid ISBNs, the stated one and the corrected one. Discussion moved here.
    • Add support for translators (as well as other roles like single-author collection editors?) Discussion moved here.
    • Move “Add Authors/Transliterated Titles/etc” buttons to the right to free up screen real estate. This will become more important with the addition of new multiply occurring fields (see immediately above.)
    • Change “Family Name” to “Directory Entry”. If we allow multiple values to support different alphabets/scripts, how will we sort by this field? Discussion moved here.
    • Separate translations from VTs on the Title page. Allow users to limit the list of publications to the ones associated with the canonical title.
      Won't that also allow us to group same language translations? If not, then can we have that as well? I would love to be able to see that there are 30 different language translations without seeing that there are 10 French ones until I really want this information :) Annie 20:52, 19 February 2017 (UTC)
      I don't recall this functionality mentioned in the past. I would suggest creating a separate section to explain what you would like to see and to gauge other editors' reaction. Ahasuerus 16:23, 20 February 2017 (UTC)
      I'd like to see this one. -- JLaTondre (talk) 01:22, 20 February 2017 (UTC)
  • Re-do verifications. This will let us have more than 5 primary verifications as well as multiple transient verifications per publication.
    I'd also like to see this one. -- JLaTondre (talk) 01:22, 20 February 2017 (UTC)
  • Create a history of changes to primary-verified publications by storing a snapshot of the way each verified pub looked like right before it was changed.
  • Support for third party IDs (OCLC, BLIC, LCCN, BNF, ASIN, Goodreads, etc). This will enable Fixer to submit ebooks without ISBNs, an increasingly common scenario. Discussion moved here.
  • Make the Quick Tags list user-definable.
  • Add support for additional title types like PLAY and FICTITIOUS ESSAY (requires discussion to come up with additional values.) Discussion moved here.
  • Cleanup reports:
    • Make additional cleanup reports available to non-moderators
    • Review old/incomplete FRs for cleanup reports, add missing title types
  • Add a “non-genre” field to author records; create a cleanup report to find non-genre titles by non-genre authors. Discussion moved here.
  • Add a disambiguator field to title records (requires discussion)
    Don't understand. Hauck 20:08, 19 February 2017 (UTC)
    Is that for the ability to add the reason for the variant (change in name, change in author, translation, abridgement)? If so, I'd love this one. If not - What are talking about? Annie 20:52, 19 February 2017 (UTC)
    The problem that we are trying to address here is that some authors have identically named titles. For example, H. P. Lovecraft's Summary page shows three "The Lurking Fear and Other Stories" collections published in 1947, 1964 and 1971. They are all different, but you can't tell until you drill down to the title level and check each title's Note field. Similarly, there are 2 version of the story "The Rats in the Walls" (1924 and 1956), 4 "[To Albert A. Sandusky]" poems and 9 (sic!) poems whose title is currently entered as "[To ?]". Analogously, Clark Ashton Smith wrote 4 different "Ennui" poems, 2 different "A Sunset" poems, 3 "sonnet" poems, etc. A "disambiguator" field would be used to capture a brief summary of what's unique about each title, e.g. "1964 version", or perhaps the first line if it's a poem. It would be displayed on the author's Summary page. In addition, we will need to decide whether to use this field to capture other information like "abridged", "revised" or "greatly expanded". There were different suggestions the last time we talked about it. Ahasuerus 16:53, 20 February 2017 (UTC)

Ahasuerus 19:50, 19 February 2017 (UTC)

Additional Proposed Projects

  • Add a "printing rank" field to order multi-reprinted titles without pub dates.Hauck 19:57, 19 February 2017 (UTC)
    Yes, please! And not only for dateless - even if we know the dates, a field for edition and printing will be awesome (so if I have 2nd edition, 3rd printing, I can look for 2.3 or something like that and find it without looking through 200 publications). Annie 20:52, 19 February 2017 (UTC)
    There is a Feature request to "Add a 'printing' field to publication records". It would be relatively easy to implement, but we need to decide what kind of data we want to capture. Unfortunately, "editions" and even "printings" do not always follow the standard numbering scheme, so "2.3" won't work. For example, this 1989 edition of Heinlein's "Have Space Suit—Will Travel" says "First Ballantine Books Edition: December 1977 Thirteenth Printing: October 1989". On the other hand, this 2005 edition simply says "First Pocket Books trade paperback edition [no printing number]". This is very common, so you can't expect "editions" to be sequentially numbered. We could limit this field to numeric printing numbers, but there can be any number of "first printings" from different publishers. It doesn't mean that we won't want to implement this field, but we need to understand its limitations. Ahasuerus 04:07, 20 February 2017 (UTC)
    The 2.3 was mostly an oversimplification for the multiple printing scenarios - not necessary requesting it to be numbers only. But we do need some guidelines for formatting in that field - or we will end up with unsearchable field. It will still be better than the current way to find where we are. Even if it is just printing in the field - the edition is kinda clear from the ISBN unless if it gets reused. Annie 04:13, 20 February 2017 (UTC)
    What!? I cannot put Chinese characters in the printing field? But I know of some publishers that mark their printing this way (OK, I am facetiously joking but it is true). Uzume 04:58, 20 February 2017 (UTC)
  • Add a new title type for excerpts, and one for plays/scripts. Neither of those fit nicely in SHORTFICTION. --Vasha 21:02, 19 February 2017 (UTC)
  • Another useful feature, mainly for links from external site to the ISFDB: stable IDs. Currently, if a title is merged into another its ID ceases to exist in the database. As a result, links from external sites to such a title record will become broken links. The software should instead keep track of merged IDs and redirect to the new ID (there was a discussion about this some time ago but I just couldn't find it). Jens Hitspacebar 21:10, 19 February 2017 (UTC)
  • [Discussion of third party IDs moved here.]
    1. The next item I would like to see is revamping verifications. I am not sure how valuable it is but it has plagued us for a while now. Uzume 21:36, 19 February 2017 (UTC)
    2. Finally, author and publisher directory updates seems like a good item. This could bed sooner but probably will be easier after more internationalization/localization is completed so doing it a bit later might prove less problematic. Uzume 21:36, 19 February 2017 (UTC)
  • There had been some discussion about reviews of magazines here. I think the discussion petered out rather than concluded, but did get into alternate ways of dealing with magazine series. I'm not pushing for it, just reminding. Doug H 21:59, 19 February 2017 (UTC)

Editor review of submissions: "automated" submissions sidebar

Create a separate queue for automated submissions, which will enable non-moderators to work on them (and let me concentrate on development). Rather time-consuming to implement. Ahasuerus 00:10, 20 February 2017 (UTC)

Yes, please :) I know that it is a time consuming change and so on but it should help in the long run... Annie 20:52, 19 February 2017 (UTC)
I am for editor review of submissions (by Fixer, etc.), however I question the need for a separate queue. Uzume 23:02, 19 February 2017 (UTC)

(unindent)I have some reservations about "automated submissions" (but that maybe based on my lack of understanding) referred to above in #Roadmap 2017. Why do we need another submission type like this? It seems to me this FR is just to allow editor reviews of submissions (like a voting system to help moderators close pre-reviewed submissions). Is there some reason some submissions should not allow editor review? Uzume 21:45, 19 February 2017 (UTC)

Because at the moment Ahasuerus is doing all of them. This is the process by which all newly published books are added to the system - so we do not wait for people that have them to add them. The idea is to allow the queues that the automated robot finds to be worked on by other people as well - which includes verifying that it is indeed eligible, cleaning the record based on our standard and so on. It is not about approval of entries from people, it is about adding the entries in the DB so the moderators can approve them. Annie 21:50, 19 February 2017 (UTC)
I do not see a difference with this. It is true bot generated submissions maybe of lower quality but then so are submissions from new users. I do not think we need a new field specifying the origin of the submission. I like the editor review of submissions concept (perhaps with comments and queued results to be approved by moderators). In the general sense, consistent high volume and high quality editor reviews of submissions could be a very good way of selecting prospective new moderators. Uzume 21:55, 19 February 2017 (UTC)
We do not have real bot submissions now - Fixer is manually operated by Ahasuerus, title by title, every month. That idea is to allow these queues to be processed by other people as well. Annie 22:03, 19 February 2017 (UTC)
I understand he is doing significant manual work on his side of bot before making submissions and to lighten his load, we want to be able to have his bot make lower quality submissions with better review from our community (instead of just him) and I agree with and would not argue on that point. However, it seems to me a generalized editor submission review process is a better solution than a one targeted just at this problem (which adds unneeded extra levels and fields for such). Uzume 22:10, 19 February 2017 (UTC)
And how is that different from making more people moderators if there are people that can do reviews like that? :) Plus - the process of accepting edits is already slow, do you really want to add more layers? A new editor will get discouraged fast. Plus - that won't solve the issue with Fixer and the sumissions - no matter who reviews, they need to get into the system one way or another. If you want all that it finds to be just added and someone reject it if it is not genre and/or not eligible, that would add a lot more work on the moderators... A Fixer, non-massaged entry is missing series information, have (or can have) a weird spelling on author and publisher, title (Amazon adds a lot of junk there) and it may need to be merged somewhere. No matter how many layers of whatever we have, if those just get dumped into the regular queues, the real updates will get delayed with days until the pure deluge of those is dealt with. Annie 22:18, 19 February 2017 (UTC)
I never suggested making editor review mandatory—just generalized. I do not understand why the normal queues would cause delay. Submissions are not required to be moderated in any particular order (I understand there are order issue with regard to moderation however). Uzume 22:36, 19 February 2017 (UTC)
No but delaying a new contributor edit while we are dealing with internal ones is a bad idea. And if you have 12000 Fixer submissions in the regular queue, how will a moderator find a new one. And even if they can, working on those ratty ones will be a longer and more involved process than on user-created ones. Thus the separate queues and the ability to do something before they hit the moderators. :) Annie 23:13, 19 February 2017 (UTC)
That sounds to me more like better management of the current queue is needed not another queue. Make the submission queue searchable and a moderator can for example just say "I do not currently want to see submissions by Fixer" (or say any user with a bot flag; this list is not very long Special:Listusers/bot). Uzume 23:40, 19 February 2017 (UTC)
The main difference is allowing editing before submitting - allowing the entry to be cleaned up before it goes into the proper queue for approval. This is not something moderators can do now - they can only accept and then edit after that. If you are saying that moderators should be able to directly edit ANY submission before acceptance, then yes, better management is all that is needed. But I am not sure we really need that. Annie 23:45, 19 February 2017 (UTC)
OK, so you are saying you want a separate queue and allow reviewing editors to edit/transmute the submission as part of the review process? That would be very powerful but also extremely difficult to code (since submissions are actually only semi-structured XML documents). Uzume 00:39, 20 February 2017 (UTC)
Thus the comment above that it will be very complicated :) If you read the FR, this is what it describes. Annie 00:40, 20 February 2017 (UTC)
Maybe Annie means a tiered submission process where editors can review automated submissions converting them to normal submissions or some such. But then in my mind "automated" just means "I want editor review first". Uzume 22:02, 19 February 2017 (UTC)
Not really. The current automated process is; Fixer finds the titles (wonderful), Ahasuerus submits them to the DB (from his name or from Fixer's name). Anything with low priority just stays in the queues until someone gets to them. Or forever. It is not about magazines, it is really about what the robot finds and needs to be processed. And you do not want to add to the moderator load all the "cleanup" that is needed - there is a LOT of work to be done on those submissions. And just dumping them into the DB without cleanup, even with a flag will dilute and make the DB useless fast.Annie 22:09, 19 February 2017 (UTC)
I am not arguing against editor review for Fixer submissions. I am arguing this is more elegantly solved with a generalized process of editor review not specific to Fixer submissions. Uzume 22:13, 19 February 2017 (UTC)
Baby steps. Plus Fixer (and any other robots that can find more entries). It is not Fixer specific really - it is the ability to handle those kinds of submissions in a slightly different way. And I still argue that if a person does the submission, you do not need it to be delayed further. :) Annie 22:20, 19 February 2017 (UTC)
So you are saying the "automated" flag adds another tier delaying such submissions. I do not think that is necessary (and if you read the FR a moderator can directly moderated such without editor review anyway). The point is we do not need the flag can just implement generalized editor review of submissions. Uzume 22:35, 19 February 2017 (UTC)
If someone new does not understand the page, they may click it, sending the submission in a long(er) queue. What we need is a double queue - one that only moderators can deal with (the ones that had been seen by a human and just need the second set of eyes) and one for any "ratty" submission where more people can help with. Which is what this FR is for - allow a new process for submissions. Annie 23:09, 19 February 2017 (UTC)
I am all for editor review of submissions to help with such "ratty" submissions. I do not think we need another queue (chosen at submission time), just better management of the current queue should suffice.

OK, after some discussion...if we are talking about a totally different process where someone or something provides suggested changes (I am intentionally avoiding the word "submission") which are reviewed and potentially generate submissions (a sort of reviewed and potentially modified transmutation from the original suggestion), that is a totally different beast and I am not at all sure it makes sense to leverage the current submission system for this sort of thing. I believe Fixer and other bot like "animals", normally do not make arbitrary submissions but only very specific types (mostly NewPub; someone tell me what other types it can make). As such, it might make much more sense to have separate NewPub (suggestion? submission?) queue. A reviewer could then load such NewPub items into the NewPub editor, potentially make changes, and make the submission (or mark the suggestion rejected). This would greatly simply such an implementation (which would still not be trivial). It might also make sense for bots to mark items in this NewPub queue as completed (or actually delete them) when it noticed the items appear in the main database that match the NewPub suggestion it previously made. For this purpose, there should probably be a way for bots to query this queue (with search requests like "give me all outstanding NewPub items I provided but I have not marked as completed/deleted"). Frankly, I see this as more of a remote bot NewPub queue than an automated submission queue. Uzume 01:24, 20 February 2017 (UTC)

AddPub as well when it manages to recognize where a new publication belong. Fixer founds a few thousand new ISBNs every month. They need to be added to our DB - but they do need a lot of cleanup. What is your proposal - how those say 6 0000 publications get added without the new process and without burning out all the moderators? If Fixer can just dump them in a new queue where someone can go, edit and then submit to the moderators, this is exactly what the FR is proposing - call it automated submission queue, call it "list of possible NewPub/AddPub", call it whatever you want :) So do you just disagree with the name? What you describe here is exactly what the FR is proposing... It is not about changing the current process, it is about adding a new process BEFORE the current one that will allow these ratty ones to be streamlined for the process. Annie 01:39, 20 February 2017 (UTC)
That is not how the FR is written "Change the submission table to support a new submission state, “A”utomated" implies modification of the current submission queue. I do not think that is appropriate if what is really needed is a remote bot queue with limited input types. Uzume 02:11, 20 February 2017 (UTC)
Look at point 10 in the FR - that is what ties it back into the standard process. Annie 02:13, 20 February 2017 (UTC)
I see that now, however, I still think that unduly complicates the submission system with little value. I believe we would be better off with a new different specialized queue than trying to shoehorn this functionality into the current submission queue. Uzume 02:20, 20 February 2017 (UTC)
The linked FR would effectively create a new specialized queue. Editors would then use "automated"/robotic entries in that queue to create regular submissions. Implementing various supporting mechanisms -- like putting the automated submission on hold as described in the FR -- would be somewhat challenging, but I think it should be doable. Ahasuerus 04:58, 20 February 2017 (UTC)
I imagine it would be challenging both on the API server side of things as well on the bot client side of that (what does/should Fixer do with such a remote queue item marked on hold?). Uzume 05:23, 20 February 2017 (UTC)
When Fixer sends a submission via the Web API, he marks the ISBN as "submitted" in his internal database. From that point on he ignores the ISBN.
When Fixer gets the next weekly backup, he updates his internal database with the data from the backup. At that point the status of moderator-approved ISBNs changes from "submitted" to "already in the ISFDB". Moderator-rejected ISBNs remain "submitted" forever. It doesn't really make a difference because Fixer ignores "already on file" ISBNs just like he ignores "submitted" ISBNs. Adding a new queue on the server side shouldn't change this logic on the client side. Ahasuerus 15:16, 20 February 2017 (UTC)

(unindent) If there's a separate queue, that queue will probably back up (esp. once the novelty wears off). If it does back up, we'll end up in a situation where more of its entries turn out to be duplicated by live-editor submissions that were created without consulting that queue. It might be better to get more moderators and let Fixer submissions flow through without any sort of pre-qualification/filtering that's being done now. We are not particularly aggressive about recruiting moderators. --MartyD 12:03, 20 February 2017 (UTC)

Unfortunately, the main problem is not that we don't have enough moderators. The main problem is that processing robotic submissions requires a significant amount of research and TLC. Some moderators have the time and the energy to do the necessary research. Others don't and simply click the "Approve" button if the submission looks OK. This is the main reason why I only submit on Fixer's behalf if:
  • I am 99% sure that the submission (usually an AddPub) is truly OK "as is", or
  • it's a relatively low priority ISBN and it's better to have it on file in a potentially suboptimal state than not to have it at all
Over the years, I have tried a number of strategies to improve the situation, all of them unsuccessful.
The "automated submissions queue" is the latest stratagem in a long list. The idea is that having human editors do the requisite research up front would result in higher quality submissions to be reviewed by moderators. Admittedly, if a moderator wants to pull up the new queue and simply approve everything in it, it won't stop it.
If there is a better solution to this problem, I'd love to hear it. The current situation is barely sustainable as it is and will continue to deteriorate as my productivity declines over time. (Not to mention the increased number of SF books being published.) Ahasuerus 19:15, 20 February 2017 (UTC)
That makes more sense and I prefer that idea. This is why I am sort of against the idea of another queue (particularly if it is shoehorned in the the existing one as the first part of the current FR describes) despite its possible value in being able to modify the entries at review time (it seems like much work where this is the only key value). That said, it might make sense to add generalized editor review for the main submission queue (it could be limited in some way if there are security concerns). Such reviews could help moderators more quickly decide when closing submissions. As Marty points out such things may turn into a novelty (but then same could be said of moderators; look at how many inactive or barely active ones we have at the moment), but it might ebb and flow with new or return editors, etc. Aside from the possible value of possible valuable reviews to speed in moderation, it seems like a better tool to use when hunting for moderator candidates (so though this too might be complex and end up perhaps with smaller value this part seems to me the most valuable part). Uzume 18:24, 20 February 2017 (UTC)
If we go this way, we really should have a prioritization of some type available so that live editor updates do not get backed up when Fixer dumps its submissions. It is very annoying to wait for a moderator to approve something so you can send the next update (especially when you are new) :) Annie 05:32, 21 February 2017 (UTC)

Adding support for translators and possibly other roles

Add support for translators (as well as other roles like single-author collection editors?) Ahasuerus 23:45, 19 February 2017 (UTC)

Ok for the former, no to the latter, why not editor of novels? Hauck 20:08, 19 February 2017 (UTC)
A collection editor usually chose the stories, a novel editor does not chose content in the same way. I think we should allow such editors be somehow recognized. Annie 20:52, 19 February 2017 (UTC)
I am with Hervé here. I would like to added more contribution types supported when we have no way to capture such now but I would be wary of taking what we already have and dividing it down (into different types of authors/editors or different types of artists, etc.). I definitely want to see translators (since our current rules have no fields to capture this short of ad hoc notes). I am curious why we cannot just add translators as authors of translation title records and just reach through the VT to find the original authors instead of adding a new contributor type (we do not need one for artists they just tend to have different types of works; a different language VT work is a different type of work). If we implement such, we could have a cleanup report for translations where the canonical authors (jumping through pseudonyms) match authors of the original work allowing us to convert these over time (translations without translators report). Uzume 23:20, 19 February 2017 (UTC)
So you are proposing the Bulgarian variant of Dune to have as an author the translator? How would you separate on the translator page the ones he did write from the ones he translated? And if a user does not understand out variant system, they can misunderstand who the writer is. Plus this way we are losing the ability to have the actual author name in the language of the variant... Artists are different because they are different works - the translation by definition has two authors - the original one and the translator. Annie 23:24, 19 February 2017 (UTC)
The software can display these differently by reaching through the translation variant and author psuedonyms to properly credit authors vs. translators. Uzume 23:33, 19 February 2017 (UTC)
How about an example. Dune by Frank Herbert is translated as Дюн by Франк Хърбърт, translated by Виолета Чушкова. Which name goes where in your idea? Annie 23:37, 19 February 2017 (UTC)
I am assuming we are referring to T1442269. I am suggesting we just add Виолета Чушкова to this record as an author and the software just credits the translators differently than the authors by chasing the authors of this record from A237335 to A30 (since A237335 is a pseudonym) and whatever number Виолета Чушкова ends up (since that is not currently in the DB) and comparing that to the authors from T2036 (which is the variant translation parent of T1442269) which is A30. The software can clearly detect that A30 is the original author and whatever number Виолета Чушкова ends up is not (and thus is a translator). In a similar vein, cleanup reports could be generated for translations without translators. It should be noted that all of the above data is already required to display and properly link these entries anyway. This proposal just suggests handling that data differently to allow for and credit translators (there is already an FR900 for displaying translations differently; I just want to add translator credit support to that and propagating it to more than just title pages). Uzume 00:10, 20 February 2017 (UTC)
None of your links work by the way - you may want to look at the templates again:) The example was from my library. I get where you are getting with that. But... what about the case where it is not just a translation but also an abridgement. This is when the second author gets added now. And that will make the translator author page a nightmare for the software to work out - in which cases the author is a translator, in which cases he/she is a co-author, in which case an author. Not to mention that the searches for translators will become overly complicated. Annie 00:14, 20 February 2017 (UTC)
Currently our rules rule out using the VT mechanism for any derived (where the contents is significantly different) works except translations so that should not be an issue. In partially translated works and the like there should be no VT parent and whether an author is an translator or not becomes muddled to begin with. I believe it would be best to list all contributors as authors and specify who did what in the notes (much as is done with certain types of authorship now). Uzume 00:26, 20 February 2017 (UTC)
Not true. Abridgements are variants - in a way a lot of translations are abridgements, especially the older ones. As are two parts of a split novel. Revised novels also get varianted... Annie 00:42, 20 February 2017 (UTC)
We did not used to allow VT title records except to change the title or author and then later allowed translations (i.e., changes in the language). If indeed we now allow other derived type VT title records, I would prefer a translation flag marking a VT title record as a translation over adding new contributor role flags (add a field to title records instead of to author records). The problem with using the VT mechanism for generalized derivatives is: what constitutes a derivative? Is something a derivative if it is barely inspired by something else? As you can see this gets rather sticky quickly and the resulting connections are more then the VT system can support (since that would actually require a generalized graph construct). Uzume 01:38, 20 February 2017 (UTC)
We are already using them for that (as much as I dislike it in some cases - split novels for example) - the question is how to make it a bit more manageable. And as for the flags - if a variant is an abridgement AND a translation from different people, how do we determine who is who? We can use additional authors but we will need a flag on the author as well... Annie 01:52, 20 February 2017 (UTC)
You cannot use VT to specify abridgement AND a translation by different people anyway. That would require more than a two title records to express (one would have to link the translation to the abridgement and the abridgement to the original; our VT mechanism specifically does not allow generalized graph structures). Uzume 02:08, 20 February 2017 (UTC)
You cannot have a variant of a variant - so if my Bulgarian book is an abridgement and a translation, it goes under the main work directly. You need one title to express - you just need to add notes. Annie 03:21, 20 February 2017 (UTC)
Exactly, and in the same notes that qualify such you can qualify the contributors as well. If you cannot have the titles linked perfectly (outside of notes) it makes little sense to try and credit people perfectly. Just as you have to pick the parent (even if it is not right) you can pick if it is a derived work or a translation (even if neither of those is entirely right either). If we ever support such a linking then the credits will be able to be right too. I see little reason to hack around credits to solve the real problem. Uzume 04:12, 20 February 2017 (UTC)
And here we ended up on a circle. I cannot see how your idea is easier than a new role (considering all the additional work that will be needed to populate summary pages. And if it is not separate, something will need to be done on the publication and title pages to identify them as translators/whatever they are anyway because new contributors will be confused. Annie 04:20, 20 February 2017 (UTC)
Easier solutions do not always make for better solutions (in fact the easy way often ends up being hard long term). I am not sure how new contributors would be any more confused than by which is the proper variant parent (which cannot be set properly too). Uzume 04:54, 20 February 2017 (UTC)

(unindent) The issue of translator support has been open for close to 5 years now, so I don't expect us to come up with a perfect solution in a few hours. There are various aspects to consider, e.g. the fact that there are awards given to translations rather than to titles.

The reason that I mentioned "roles" is that whatever design we come up with will ideally also support other "roles" like single-author collection editors. The exact roles that we will eventually choose to support are less important than the decision to make the design extensible -- or not, if we find that translators are sufficiently different from other "roles". Ahasuerus 05:11, 20 February 2017 (UTC)

Yes, I was not suggesting the translators issue would be solved anytime soon. I am against dividing our author records into roles however, and just wanted to state that there are potential solutions for the translators issue without depending on the introduction of divided contributor role tags. If we do end up supporting such role tags, I would recommend corollaries for publisher roles (or any other authority control item), etc. Uzume 05:16, 20 February 2017 (UTC)
I don't know how useful it is to researchers, but I like the idea of being able to capture/catalogue various roles. Part of the challenge right now is that the nature of the relationship is embedded in one of the two ends (mostly in the publication/title end, but sometimes in the author end). Thinking about the relationship in terms of a generic "contributor", I could see typed Contributor relationships instead -- the nature of the relationship stored on the relationship. There could be many of these, and a pre-defined set of types. One type would be "Author", another "Editor", another "Translator", and so on. The bibliographic summaries for people could be organized by roles, then title types. To keep things simple and avoid big mistakes, the software could provide some specialized handling for required roles (e.g., it could still present Author (for single works, collection, and omnibus) or Editor (for anthology and magazine)) and then "other contributors" or something like that for everything else (just like we get the Title record but then other Contents). --MartyD 17:42, 20 February 2017 (UTC)
I have been thinking along similar lines for a while now. However, redoing the author logic (and the variant title logic and a lot of other things) would be a daunting proposition. It may be less painful to declare authors and anthology/magazine editors a "special privileged case" and implement a generalized "contributor/role" mechanism for translators, other types of editors, etc, separately. Once the new mechanism is in place and everything is working smoothly, we can look into migrating "special case" authors/editors to the new system.
Granted, a two-step approach is not without its dangers -- I once worked on a multi-billion dollar project which took this route and never completed the second step -- but we may be better off even if we never complete the second step. Ahasuerus 23:59, 20 February 2017 (UTC)
Well, we need to solve it if we really want to claim that ISFDB is international. I do not want to add 3000 works just to have to edit them in a year to get the translators sorted... Just saying :) And I like Marty's idea. Annie 00:04, 21 February 2017 (UTC)

Non-Genre flag for authors

Add a “non-genre” field to author records; create a cleanup report to find non-genre titles by non-genre authors. Ahasuerus 00:04, 20 February 2017 (UTC)

This will open the gates to well-intentioned persons that will want to enter such authors (as with the "graphic" or "non-genre") as it will look possible. Bonus problem: how to define a non-genre author.Hauck 20:08, 19 February 2017 (UTC)
[Discussion of author birthdays and dates of death displayed on the front page moved here.]
I believe adding a genre field to authors should be fairly easy and provide good value (like moderator warnings on submissions when adding non-genre or graphic content to non-genre authors, etc.). @Hauck: all authors are non-genre by default and those "over the fold" get promoted to "genre" and we allow (nearly) all content for them. Uzume 21:36, 19 February 2017 (UTC)
Sure, and promoted by who? We're back to the remarkably precise definition of "over-the-threshold" (or "over-the-fold"). This may sound funny but trying to explain for the upteenth time to another new contributor all these perfectly nebular notions may prove an interesting expreience for those who do not practice it. Hauck 21:55, 19 February 2017 (UTC)
Promotion (and demotion) is via the moderated submission process (changing the author genre/non-genre checkbox). It could be restricted to moderators too (like author canonical renaming, etc.) though I am not sure that is necessary. Education is always useful and we already have the rule. The only difference is having a field to declare such. Uzume 22:58, 19 February 2017 (UTC)
I really do not like the idea. I do not think that we should have non-genre works even by the over the threshold authors but that ship had sailed. And how we will determine who is genre author? If someone has a story/essay here, they are an author with genre works. Anyone that has at least one non-genre work is again an author with genre works. Trying to draw the line in the middle is going to cause a lot of unnecessary edits.
I can think of a few Bulgarian authors that had written both SF and Crime works and are considered masters of both. I would not add their crime works here or call them non-genre because they just happen to also had written outside the genre. Even the ones that under the usual "rules" will be over the threshold. Annie 00:10, 20 February 2017 (UTC)
This is already specified in our Rules of Acquisition. I agree some cases are not clear cut. If you want to challenge the rule (or champion a specific author for promotion or demotion) that is a different discussion than implementing it in software. Uzume 00:20, 20 February 2017 (UTC)
It does not determine what a non-genre author is. Again - where would you like to draw the line? If you want it to be based on what an editor wants to nominate, it becomes meaningless... Annie 02:14, 20 February 2017 (UTC)
What you are saying is the current Rules of Acquisition are meaningless in this regard (and I agree that could be argued) but that is a different (though valuable) discussion. Uzume 02:38, 20 February 2017 (UTC)
No, what I am saying is that the rules determine which works are added but do not qualify authors as genre or non-genre ones (above the threshold is not a determination of that).Annie 02:52, 20 February 2017 (UTC)
This seems a pointless designation. If an author has written a genre story, they are by definition a genre author. If it is attempt to designate authors who write primarily genre stories, what is the point? If it is an attempt to designate authors who are 'above the threshold', that is not the same as authors who primarily write genre stories so the name is deceiving. And as Hauk has pointed out 'above the threshold' is a nebulous thing that is not well suited to a database field. -- JLaTondre (talk) 01:15, 20 February 2017 (UTC)
I am curious why you think this is not suited to a database field? I agree an author could be promoted or demoted to threshold levels over time (an author starts out in one genre and gravitates to another). Should instead every editor and moderator know and potentially reevaluate each author upon every submission? That would make for very complex submission criteria and potentially break the rules of acquisition as different people came to different conclusions at each point in time. Uzume 01:58, 20 February 2017 (UTC)
Everyone has a different opinion of 'above the threshold'. People already come to different conclusions. A database field is not going to change that nor is it going to change that different editors and moderators apply it differently. It would just be a field they can now edit war over. -- JLaTondre (talk) 02:17, 20 February 2017 (UTC)
Edit warring over that single field is considerably more noticeable than generalized edit warring over all the pubs and titles of an author which what we are already at. If there is an edit war over that field it will be noticed and brought to community discussion more easily and thereby a consensus can be reached. I agree we could also likely generate some sort of automated heuristic to flag if an author should be above the threshold (by counting title and/or pub records) but it would not take into account exceptions by consensus. But the proposed flag is meant to hold the current value of community consensus (which obviously can change). Uzume 02:32, 20 February 2017 (UTC)

Derived and corrected ISBNs

Add two checkboxes next to each ISBN: “derived” and “corrected”. The latter will let us capture two versions of invalid ISBNs, the stated one and the corrected one. Ahasuerus 00:12, 20 February 2017 (UTC)

Not frequent enough to clutter the screen, better use the note field instead.Hauck 20:08, 19 February 2017 (UTC)
I agree with Herve on that one. Annie 20:52, 19 February 2017 (UTC)
The proposed check boxes would only appear on Edit/Add/Clone Publication pages. Regular Publication pages would continue to display "(Bad Checksum)" for invalid ISBNs like they do now.
The advantage of having invalid ISBNs captured in the ISBN field vs. in the Notes field is that it enables searches. For example, if you search for the ISBN stated in Alpha 5 (0-345-24140-X), you won't find it because it's only available in Notes at this time. When dealing with invalid ISBNs, we have always had to choose whether to enter the stated version or the corrected version in the ISBN field. The other version was relegated to Notes where it became unsearchable. With the addition of support for multiple ISBNs, we will no longer have to choose -- we will be able to eat our cake and have it too! :-)
As far as derived ISBNs go, there are hundreds, if not thousands, of them in the database, mostly from the mid-1970s when the industry was transitioning from catalog IDs to SBNs and then to ISBNs. It's always made me uneasy that we claim, e.g., that Alpha 4's ISBN is 0-345-23564-9 even though it says "345-23564-9-125" in the book. If we add a check box to the edit page, the main Publication page will be able to indicate that the ISBN value was derived. Ahasuerus 03:15, 20 February 2017 (UTC)
I too agree and believe implementing ISBNs as identifiers (a different FR) will likely solve much of this problem anyway. The key is to differentiate identification (e.g., cover price, etc. which can be erroneous) from identifiers (which probably should not be erroneous). Uzume 00:23, 20 February 2017 (UTC)
I still believe most usages of this can be covered by treating valid ISBNs as identifiers (in the as yet unimplemented identifiers FR) and invalid ISBNs as catalog numbers. Here is some potentially pertinent information on this subject: MARC ISBN. Uzume 03:47, 20 February 2017 (UTC)
Back when I was working on the design of the "third party identifiers" FR, I considered making catalog IDs and ISBNs just another type of identifier. Eventually I decided against it for a number of reasons:
  • Unlike third party identifiers, catalog IDs and ISBNs are stated in publications
  • Also unlike third party identifiers, catalog IDs and ISBNs are displayed on various "Publication Listing" pages
  • ISBN searches are "privileged" in the sense that they are available via the main search box
  • ISBN searches use a special ISBN-10/ISBN-13 conversion algorithm so that a search on an ISBN-10 finds a matching ISBN-13 and vice versa
  • ISBNs can be "derived" or "corrected" as described above
In the end I decided that catalog IDs and especially ISBNs were sufficiently different animals to merit separate places in the database. Ahasuerus 04:17, 20 February 2017 (UTC)
Those are good points, however, an ISBN as an identifier can be converted to a 13 digit number in all cases (and left in original ISBN-10, SBN or other brokenness as stated in the publication in other cases like the catalog number field). When searching ISBN identifiers, ISBN-10s to search for can be converted to 13 digit numbers before comparisons (so there is no issue). It would not be so hard to have a combined search for these and the catalog field(s) (matching original input against catalog fields and any convertible ISBN-13 against the identifier fields). Though very common, all (and sometime any) ISBNs for an edition are not always stated in its printed publications (e.g., I have seen some that only have an GTIN-14 which contains the ISBN-13 and another digit). I am not sure it is worthwhile to implement ISBN as both catalog numbers and identifiers and to accomplish such (but it is something to think about). Another issue is broken punctuation in ISBN. I know sometimes an ISBN is stated in a publication with improperly placed hyphens (I have no idea how we handle this if at all; I am guessing we do not maintain this original statement). Uzume 04:49, 20 February 2017 (UTC)
There is a great deal (2427 lines to be precise) of ISBN formatting logic in this module. It's yet another way in which ISBNs are different from third party identifiers. Ahasuerus 15:20, 20 February 2017 (UTC)

Changing "Family Name" to "Directory Entry"

Change “Family Name” to “Directory Entry”. If we allow multiple values to support different alphabets/scripts, how will we sort by this field? Ahasuerus 02:39, 20 February 2017 (UTC)

Switching to unicode and then sorting by it will be cleanest. Until then I kinda like it as it is (as long at non-Latin entries actually have their transliterations properly and not getting defaulted to the name of the author in English. Annie 20:52, 19 February 2017 (UTC)
I have been looking forward to the DB moving to some Unicode encoding like UTF-8 for a long time (currently it is only pseudo-Unicode by way of using HTML entities). Multiple directory entries solves more than just multiple script/language issues (see ISFDB:Community Portal/Archive/Archive41#Sorting, author directory and nobiliary particles). Multiple scripts can be handled by multiple directories. Uzume 00:51, 20 February 2017 (UTC)
Multiple scripts/multiple directories becomes a bit of an issue with author names with letters from two separate scripts - Macedonian or Ukrainian names for example which combine Cyrillic (somewhat creative on some letters) and Latin letters. Which directory will have these forms? Or do we start a separate directory for each of them? Annie 02:44, 20 February 2017 (UTC)
This is not necessarily their real name anyway. I would suggest making entries for both (all Cyrillic and all Latin even if their real name is a mixture of the two). Uzume 03:03, 20 February 2017 (UTC)
For Ukrainian and Macedonian authors, it IS their real name. And we all hope we will get more international entries, right? Saying that they do not deserve their real names in a directory while we have everyone else's name there will be a bit... uninclusive. Annie 03:19, 20 February 2017 (UTC)
I do not see how that is an issue. The above mentioned "nobiliary particles" discussion proposed having multiple entries based on different cultural conventions in different places. Uzume 03:29, 20 February 2017 (UTC)
So separate directory for each language we support (I am just trying to make sure I understand the idea) Annie 03:37, 20 February 2017 (UTC)
No, not for every language. One per script (all Latin in one, all Cyrillic in another, etc.) implemented as time and demand allows. This means for example English language authors would be found in the same directory as German ones. Another example of directory entries that do not match real names would be Japanese (which would have kana entries) where I do not think we want to consider a kanji Chinese character directory (I have no clue how we would do such for actual Chinese names). Uzume 03:53, 20 February 2017 (UTC)

(unindent) Let's take a step back and examine the uses (current and proposed) of the "Family Name" (soon to become "Directory Name") field. At this time it is used to do 2 things:

Making this field a "multiply occurring field" will make it possible for an author record to appear on multiple Author Directory pages. For example, if we specify "de la Roche" as well as "Roche" as Mazo de la Roche's Directory Entries, his name will appear on two different Author Directory pages. Once we implement non-Latin Author Directories, it will also enable us to make the same author appear in multiple directories.

So far so good. However, once an author has multiple "Directory Entry" values, we will have no way of using them for sorting purposes. Unless we make one of them our "preferred sorting value", I guess. Ahasuerus 03:49, 20 February 2017 (UTC)

Agreed. And as such instead of having a "preferred" tag, why not treat this like transliterations. You have a Family name with "transliterated" Directory names on top. This would allow a mixed (potentially collating) name as a family name (we could have Cyrillic and Latin or kanji and kana in this field) and single script directory names on top. Uzume 03:57, 20 February 2017 (UTC)
I am not sure I understand what "on top", "mixed name" and "potentially collating name" mean in this context. It sounds like you may be suggesting that we keep the current "Family name" field and add a new multiply occurring field for "Additional Directory Entries". The latter would be used for names like "Roche" in the "Mazo de la Roche" example above as well as transliterations. The former would continue to be used for sorting purposes. Is that about right? Ahasuerus 15:45, 20 February 2017 (UTC)

Add support for additional title types

Add support for additional title types like PLAY and FICTITIOUS ESSAY (requires discussion to come up with additional values) Ahasuerus 02:41, 20 February 2017 (UTC)

Not enough candidates to warrant new types.Hauck 20:08, 19 February 2017 (UTC)
Maybe not for fictitious essays; --Vasha 00:52, 20 February 2017 (UTC)
"Fictitious essays" or "fictional essays" are basically "in-universe" essay. They are written as if the events of the work that they are associated with were true. They have prompted many discussions since it's not clear whether they are "essays" or "short fiction". Help currently says:
  • Some books contain fictional essays, purporting to written by a character in the book, as introductions or afterwords. There is no "FICTIONAL ESSAY" title type, so you have to choose whether the title is better described as SHORTFICTION or ESSAY.
There have been repeated request to implement a separate title type for them. Ahasuerus 02:48, 20 February 2017 (UTC)
Well then, if there have been repeated requests, clearly there are enough cases to justify adding the new type. And I will say it definitely sounds useful to me. --Vasha 03:37, 20 February 2017 (UTC)
but for plays, even if there aren't that many of them, the difference between them and SHORTFICTION is so obvious that they need their own title type. --Vasha 00:52, 20 February 2017 (UTC)
I would also like to see a separate title type for plays, but we will need to decide how to handle related cases first. For example, are movie scripts, TV scripts, etc separate types or are they all the same type? If they are the same, should we come up with a more generic term than "play"? Ahasuerus 02:51, 20 February 2017 (UTC)
"script" covers plays, audioplays, tv scripts, and even librettos I think. --Vasha 03:37, 20 February 2017 (UTC)
(And for the other one I would like to see a new type for, excerpts, there really are a lot.) --Vasha 00:52, 20 February 2017 (UTC)
The vast majority of excerpts are works of short fiction. Should we treat them as a separate title type or as a separate "length" value like novellas, short stories and novelettes? Ahasuerus 02:52, 20 February 2017 (UTC)
I'd like a separate length value for these - this will separate them from the real short fiction. Annie 02:54, 20 February 2017 (UTC)
Keep in mind that if we add "excerpt" as a new "length" value, it won't separate excerpts from other works of short fiction on Summary pages. Ahasuerus 03:18, 20 February 2017 (UTC)
I believe you mean it won't currently separate such (the software could be made to do such without huge issues though I am not convinced that is a good idea). I vote for an "excerpt" "length" value. Uzume 03:21, 20 February 2017 (UTC)
I do not like seeing novel excerpts mixed in with the actual short fiction on the summary pages. --Vasha 03:37, 20 February 2017 (UTC)
It does not do it now anyway (See Asimov - the third and the forth are excerpts... If we go for separation one day, we may start talking about separating all lengths from each other anyway. What I meant upstream was that it will allow the excerpts to be seen easier. And allow searching them - and searching without stumbling at them. Annie 03:40, 20 February 2017 (UTC)
Agreed. Though I am not convinced about separating such on author pages. One can create custom pages via search that show just what you want. Uzume 04:17, 20 February 2017 (UTC)
I would be strongly against separating all lengths on author pages. It's not of interest under most circumstances, plus every author has lots of unspecified lengths. --Vasha 04:21, 20 February 2017 (UTC)
I am not exactly vying for it but it might be interesting to consider user preferences allowing one to select which lengths one wants to see (much as language selection is possible this way). Uzume 04:25, 20 February 2017 (UTC)
I like that, it would speed up searching for titles on prolific author pages. Maybe checkboxes for all title types on the author page rather then setting it in user preferences.--Rkihara 18:23, 20 February 2017 (UTC)
The question wasn't choosing which title types to display, but rather displaying ss/novelette/novella separately. It's this that I think would not be good on summary pages. As a preference option, OK maybe. --Vasha 20:55, 20 February 2017 (UTC)
If the intent is to separate excerpts from other SHORTFICTION titles on Summary pages, then I believe they need to be turned into a separate title type. Using "length" values to create separate Summary page sections would go against the basic "one title type = one Summary page section" principle. Ahasuerus 23:43, 20 February 2017 (UTC)
Excerpts are short fiction -- it does not really make sense to have them as a separate type any more than novellas are separate type. At least I cannot see a difference... If we split them in a separate type, then why would not things like half a novel published in a volume go into that same new type? Where do we draw the line? At least as a part of the short fiction, that line is clear. Annie 17:05, 21 February 2017 (UTC)
Yes, sometimes excerpts are carefully selected so that they form a complete short story; more often they actually don't seem complete in themselves. --Vasha 17:39, 21 February 2017 (UTC)

(copied from a side discussion in another section) If excerpt got added to the mix [of length options] it would be hard to "sort" that with the others. Uzume 01:28, 25 February 2017 (UTC)

I think you may be the only one who likes the idea of "excerpt" being a length... everyone else was talking about it being a title type --Vasha 01:32, 25 February 2017 (UTC)
That was not how I read things (but perhaps I misunderstood). As title type seems that seems problematic. As a length, I could usefully apply it to essays and nonfiction (like an excerpt of a bibliographic index, etc.) as well. Uzume 01:43, 25 February 2017 (UTC)
I read it as a support for Length as well :) However to your latest point here: length may be showing when you are working on publications for all titles but if you read the documentation and look at editTitle for example, it is valid only for Short Fiction. Making it usable elsewhere will probably require a lot of other changes (even if it becomes a length). Annie 01:46, 25 February 2017 (UTC)
OK, I can sort of see excerpt as a length, in that I hesitate to assign a length to excerpts -- they don't strike me as "really" being a short story or novelette. But that is the very reason that they seem to me to be a separate kind of thing, not short fiction. Besides, I like the idea of having excerpts be a separate section on summary pages, for which they'd have to be a title type. But maybe I am the only one who wants that. --Vasha 02:15, 25 February 2017 (UTC)

Third party identifiers

Support for third party IDs (OCLC, BLIC, LCCN, BNF, ASIN, Goodreads, etc). This will enable Fixer to submit ebooks without ISBNs, an increasingly common scenario.

Good to have some way of identifying ebooks but most of those you listed are ephemeral. Particularly ASIN. (I don't put ASIN in the notes field of an ebook because there are probably numerous ones most of which will disappear soon.) Goodreads has the advantage that they keep records for out-of-print ebooks (it's their policy not to delete) but they're very unsystematic, their records can get merged with each other, etc. --Vasha 17:49, 20 February 2017 (UTC)
Library IDs (OCLC/WorldCat, BLIC, LCCN, BNF, etc) are generally stable, at least as stable as our IDs. We currently link to them in Notes, which requires crafting URLs manually, an error-prone and time-consuming process. In addition, manually built links can become defunct if/when the third party changes its URLs conventions, as we have seen over the last few years. ASINs may not be quite not as stable as, says, LCCNs, but in most cases they work even years after the data was captured. They are frequently the only way to link to the source of our data. Even more importantly, without ASIN support Fixer is unable to submit ebooks without ISBNs, which are increasingly common. Even major authors publish e-stories without ISBNs these days and Fixer can't do anything about them. Ahasuerus 18:10, 20 February 2017 (UTC)
I would certainly like to be able to use library IDs as identifiers for non-ISBN print books. Having support for that makes sense. But Worldcat doesn't list ebooks -- what about other libraries? --Vasha 18:30, 20 February 2017 (UTC)
Some do, but the intent of this FR is to support third party IDs regardless of whether they are for paper books or ebooks. Ahasuerus 18:46, 20 February 2017 (UTC)
As for automatically importing ebooks, that opens a can of worms. We had an inconclusive discussion a while back about what constitutes an "edition" of an ebook. Personally I would like to see them condensed not too many; the way Goodreads has so many is bewildering and ugly. So if you automatically import them there's have to be manual combining. --Vasha 18:30, 20 February 2017 (UTC)
Fixer has been generating ISBN-based ebook submissions for many years. Submitting ASIN-based ISBN-less ebooks should be no different than submitting ISBN-based ebooks once we add support for third party IDs. Where it will get challenging is if we start using another source of ISBN-less ebooks. Suppose Fixer creates a database record for an ISBN-less ebook whose ASIN is "ABC". 6 months later an editor (or another robot) wants to add a record for a Barnes & Noble ebook whose B&N ID is "XYZ". How can we tell whether it's same version of the ebook, especially if the publisher is not specified and the publication dates are close? OTOH, as long as we stick to ASINs, we should be in reasonably good shape since different ASINs are generally associated with different editions. Ahasuerus 18:56, 20 February 2017 (UTC)
Amazon actually doesn't have every single ebook. Graydon Saunders is a fairly notable example of an author who choose not to offer his books via Amazon because he objected to their terms. --Vasha 20:21, 20 February 2017 (UTC)
Oh, definitely. However, as long as we add these books manually, we are less likely to run into problems. It's the robotic submissions that worry me. I have seen many cases where an ebook was originally added by a human editor without an ISBN -- because Amazon doesn't display ISBNs for ebooks even if they exist -- and then Fixer came along and tried to add the same book using its ISBN. (This particular problem will be alleviated by adding support for third party identifiers.) If we were to start using other online sources to grab ISBN-less ebooks, the problem would become more severe. Ahasuerus 20:35, 20 February 2017 (UTC)

(unindent) I believe the next most interesting item is adding support for identifiers. This is a bit sticky but useful concept. The sticky part comes in where we apply these. I believe you were mostly considering at the pub record level, however, the identifiers in this area are themselves usually defined for the concept of an "edition" (which we have no real records for since our pub records most closely correspond to a printing, save for SFBC). This probably is not a major issue but something to consider. Uzume 21:36, 19 February 2017 (UTC)

True, different catalogs and Web sites create records at different levels. For example, an OCLC record may cover multiple ISBNs. Alternatively, multiple OCLC records may cover the same ISBN. I don't think there isn't much we can do about it. Adding support for third party IDs is the best I can think of. Ahasuerus 19:50, 20 February 2017 (UTC)

Next, there is the issue of untangling ISBNs. Currently, our pub record ISBN field represents an identifying mark of the publication (along with cover price, etc.) and we just sort of play games with it to allow is to link to other sites by using it as an identifier. Uzume 21:36, 19 February 2017 (UTC)

Well, ISBNs were originally supposed to be unique and they are still used as "mostly unique" by the industry, e.g. when ordering books. Of course, as we know, publishers can tweak things like cover art and price while keeping the same ISBN. And sometimes they mess up and reuse an ISBN which they didn't intend to reuse. Still, ISBNs are the only IDs that can be used cross-catalog and cross-site with a reasonably high chance of success. Our use of ISBNs to link to other sites is the way they were intended to be used. It's not perfect -- see the OCLC discussion above -- which is one of the reasons why adding support for third party identifiers is useful.
For example, suppose an editor wants to record that some of our data originally came from OCLC records. At this time the only choices are either to:
  • enter something like "Corrected page count from OCLC 1234567, artist from OCLC 987653", or
  • manually craft URLs pointing to the OCLC records
The second options provides more value but is error-prone and time-consuming, not to mention that URL structures can change over time. Once we add support for third party identifiers, editors will be able to enter OCLC record numbers directly. Ahasuerus 20:05, 20 February 2017 (UTC)

An easy way to untangle this is to run some code on our ISBN fields and move them to ISBN identifiers and keep the original field for verbatim publication catalog markings (including wrong/broken ISBNs, etc.). This way we change the current field to a verbatim publication identification field and move identifiers and external linking to use the newly introduced identifiers (including derived/corrected ISBNs etc.). Then we just have to update documentation to keep the fields separate (I recommend renaming the current field as catalog number and documenting it to apply to cover or copyright page catalog numbers, etc.; we could expand it to allow multiple values but frankly I would say that unneeded and can be covered in the note field). Finally, identifiers are also useful at other levels such as author and/or publisher authority control (e.g., VIAF, LCNAF, etc.), and for tags with regard to subject indexing terms (e.g., LCSH, FAST, etc.). Some identifiers also apply to the work (our title record) level (e.g., Open Library works, Wikidata, etc.). Uzume 21:36, 19 February 2017 (UTC)

Yes, author-based and title-based identifiers are something to consider. Currently we enter VIAF links as "Web sites". If and when their URL structure changes, we will have to change the URLs manually or write a conversion script. If we had author-based third party identifiers, the problem could be resolved in 10 minutes. I agree that they would be nice to have, but I see it as a much lower priority than publication-based identifiers. Ahasuerus 19:32, 20 February 2017 (UTC)
Maybe, but we could for instance move Wikipedia links to go through Wikidata identifiers with links like enwiki/Q982133, frwiki/Q982133 and jawiki/Q982133 for John Norman allowing the links to be translatable (without storing and maintaining all the various links) not to mention other possibilities like enwikiquote/Q982133 (provided the sitelink for such articles exist). Uzume 00:47, 25 February 2017 (UTC)

(unindent) OK, I have been thinking about this -- how about this for an addition to the edit-publication page. Below the ISBN field, the ability to add fields for an indefinite number of other identifiers (a button that says "Add Identifier" or whatever), and next to each of these fields a dropdown list to specify what the identifier is: ASIN, OCLC number, publisher's catalog code, however many options seems useful. (Naturally, the submission would be flagged if you try to submit a number without specifying its type -- and maybe, at least for some of them, could be flagged if it's not a valid format for the type specified.) Would that work? --Vasha 01:52, 2 March 2017 (UTC)

That's actually extremely close to the implementation that I sketched out last year :-) Ahasuerus 03:11, 2 March 2017 (UTC)

OCLC access

As of a few minutes ago, OCLC seems to have launched a new version of itself. So far it's impenetrable .... they may simply have ignored older operating systems, but repeated attempts to log on have been futile. Can anyone get on there and tell them their new system just doesn't work????? --~ Bill, Bluesman 00:19, 3 June 2017 (UTC)

Which OCLC page are you trying to access? When I go to their home page, it looks unchanged. The record that I happen to have displayed at the moment also looks OK. Ahasuerus 00:59, 3 June 2017 (UTC)
It may have been a brief test. Looks normal today, though the sign-in page is different. Yesterday it took forever [several minutes] for a single page to partially load, and then it would just quit. The entire look of the 'new' version was utterly different. And I'm sure I wasn't hallucinating ... been off those meds for awhile ........ ;-) --~ Bill, Bluesman 16:08, 3 June 2017 (UTC)
Oh, I see. A partially loaded Web page can be very different from its fully loaded version. Basically, your browser receives partial instructions re: how the page should be displayed, so the end result can look odd.
It's less of a problem with simple Web pages, which simply look incomplete. On the other hand, a complex page can be totally messed up. It's like trying to build a car with half the instructions :-) Ahasuerus 16:20, 3 June 2017 (UTC)

Public backups: hosting changes

The May bill for making our public backups available online was 6 times the average for the previous year. In a way it's good news because it means that there are people out there who are interested in our data. And, of course, the more people have the latest copy of our backups, the easier it will be to recreate the site if something bad happens to the administrators.

On the other hand, there is no way of telling how much more expensive it will get going forward. In addition, the current hosting solution depends on me paying the bill every month. If I become unavailable and the bill is not paid, the account will be suspended.

For these reasons I am going to move the backup files to Google Drive, which is free as long as the total amount of data is under 17Gb. I expect the migration process to be completed some time tomorrow. Older backup files will be unavailable until we can find a good home for them. I'll update this post once everything has been migrated. Ahasuerus 16:55, 3 June 2017 (UTC)

Do you need someone to help you foot the bill? Hauck 17:24, 3 June 2017 (UTC)
Thanks for asking, but I am fine for now. The problem is that there is no way of telling what may happen in the future. In April it was $N, in May it was $NN, but what if two months from now it's $N,NNN? For example, a couple of days ago I received an e-mail request to help this blog extract certain data from the database. Apparently it's a very popular blog and its readers are interested in data and databases. Once the owner massages our data and publishes her results, we may get thousands of new visitors, some of whom may decide to download our backups. It's best to be prepared :) Ahasuerus 17:49, 3 June 2017 (UTC)
How big are the backup files, and how often are they created? I might be able to work out something to help that would be more flexible than Google Drive. ···日本穣 · 投稿 · Talk to Nihonjoe 22:52, 3 June 2017 (UTC)
There are 2 types of publicly available backup files: "database backups" and "image backups".
"Database backups" contain non-graphical ISFDB data: titles, pubs, authors, etc. They do not contain submissions or detailed user data. Each file is available in 2 different formats, one per supported MySQL version. Each file is 145MB compressed. A new version is created every Saturday. At the moment only the latest version (2017-06-03) is publicly available, but we have an archive going back to 2006.
"Image backups" contain all of the images hosted by the ISFDB. The total size is almost 10GB. Only one version is available at any point in time, although I have earlier versions on DVDs. New versions are upload every 4-6 weeks.
The files are linked from the ISFDB Downloads page. Ahasuerus 01:17, 4 June 2017 (UTC)

(unindent) The image files have been moved to Google Drive and are once again downloadable. Ahasuerus 14:36, 4 June 2017 (UTC)

Audie Awards

Can we add the Audie Awards? They are not always given to specfic titles, but those often win awards. The lists of winners are here:

Just to give a couple sources. ···日本穣 · 投稿 · Talk to Nihonjoe 15:53, 4 June 2017 (UTC)

I am not familiar with these awards, but many of them seem to be given to the casts of dramatizations. For example:
  • The Graveyard Book: Full-Cast Production
  • by Neil Gaiman
  • Narrated by a full cast including Neil Gaiman, Derek Jacobi, Robert Madge, Clare Corbett, Miriam Margolyes, Andrew Scott, and Julian Rhind-Tutt (HarperAudio)
The first link above claims that they are "an equivalent to the Oscar for the spoken-word industry". If so, then wouldn't they be "out of scope" for a fiction database just like the Oscar awards are ineligible? Ahasuerus 15:46, 5 June 2017 (UTC)
Regular, narrated audiobooks are given the awards, too. It might be limited to just a few categories, though, as there are (as you said) a number of awards which would be outside our scope (the dramatizations). ···日本穣 · 投稿 · Talk to Nihonjoe 15:52, 5 June 2017 (UTC)
I see. Are the awards given to narrated audiobooks given to their authors or to the narrators? Ahasuerus 16:37, 5 June 2017 (UTC)
They are given to the work. ···日本穣 · 投稿 · Talk to Nihonjoe 03:43, 6 June 2017 (UTC)
Do they have a Web page describing the selection/nomination process, who gives the award, etc? Ahasuerus 23:24, 6 June 2017 (UTC)
Not that I can see. ···日本穣 · 投稿 · Talk to Nihonjoe 23:35, 6 June 2017 (UTC)
Hm... It's difficult to decide what we want to do about this award with so little information. I wonder if there may be a place where we could ask. Perhaps some Wikipedia editors specializing in media research may know? Ahasuerus
It is an APA award. The site has only a "Publishers and rights holders enter titles in various categories for recognition of achievement. Finalists are selected, and from that group of finalists one winner is awarded." note. This PDF is the call for entries for this year containing also the eligibility and contest rules. However - I disagree that this is an award for a title - it is an award for a publication, not for the title itself. :) Annie 18:04, 7 June 2017 (UTC)
Which is why I used the word "work". :) ···日本穣 · 投稿 · Talk to Nihonjoe 23:23, 7 June 2017 (UTC)
In my mind "work" means the title here - not the publications. :) Annie 23:27, 7 June 2017 (UTC)
Poe-tay-toh, poe-tah-toe, and all that. :) ···日本穣 · 投稿 · Talk to Nihonjoe 16:19, 8 June 2017 (UTC)

(unindent) The Wikipedia page for the award says that the award is given to "outstanding audiobooks and spoken-word entertainment", and seems to emphasize the quality of the reader. Nevertheless, the awards themselves seem to be given to specific books and authors, and the reader is not generally named in the award. That makes it look to me like the emphasis is on the book itself, hence the award would fall within our scope. Chavey 00:31, 8 June 2017 (UTC)

Exactly, other than those awards that specifically mention something like "best dramatization" or similar, where it's obvious the award is for a cast of characters or narrators. ···日本穣 · 投稿 · Talk to Nihonjoe 16:19, 8 June 2017 (UTC)
Checking their category descriptions, I see that the genre-specific ones say that they are "for excellence in narration, production, and content of a [SF/mystery/romance/etc] audiobook". "Audiobook of the Year" says:
  • This award recognizes a title with high-quality content and production values that stands as a benchmark of excellence for the industry. Additional items weighed by the jury include success of marketing and publicity, visibility, impact, commercial success, and recruitment of new listeners to the format. The Audiobook of the Year should ser ve as a worthy ambassador to new and current listeners, and be a paragon of the audiobook art."
Although the category-specific ones mention "content" as one of the selection criteria, it would appear that for the most part these awards are given for the narration/production component and therefore probably ineligible. Ahasuerus 04:24, 9 June 2017 (UTC)
Well, they are awards for an entirely audio industry, so that make sense. :)
Still, I think if the work in question is simply narrated/read rather than dramatized, it should be eligible here. That seems to be the line I've seen in the past. Basically, if there is one (or two at the most) narrator(s), it should be fine. When it starts getting a cast, that's where we should draw the line. ···日本穣 · 投稿 · Talk to Nihonjoe 05:37, 9 June 2017 (UTC)
Narrated stories/novels are eligible for inclusion as publication records. However, I don't think that an award which considers narration/production, not to mention "success of marketing and publicity, visibility, impact, commercial success, and recruitment of new listeners to the format", can be said to be given to titles. Ahasuerus 15:35, 9 June 2017 (UTC)
Which was my point above - it is given to publications, not titles. So unless if we want to start having new variants for each narrator (as we do for translators) or even for each separate production, I am not sure that we really can include the award here... And I really do not want to see the DB go that path - translators change the text after all; narrators just read it... Annie 15:56, 9 June 2017 (UTC)
Okay, whatever. ···日本穣 · 投稿 · Talk to Nihonjoe 16:08, 9 June 2017 (UTC)
Maybe start adding notes in the publications that win? This way we will have the record and if one day there are pub-level awards here, it can be converted. We may need a new generic awards template (something like {{Award|Name_of_award|Category/Year}}... This will be very useful for other cases as well anyway (international awards that are not eligible to be added as awards here but are still important in the context of the local fandom and speculative fiction. Annie 18:03, 9 June 2017 (UTC)
That could be useful. Can we get that created to be used for awards not currently in the system? ···日本穣 · 投稿 · Talk to Nihonjoe 21:08, 9 June 2017 (UTC)
I am afraid that ISFDB templates -- as opposed to Wiki templates -- can take only one parameter at this time :-( Ahasuerus 01:58, 13 June 2017 (UTC)

Changes to "View Submission" pages

All "View Submission" pages have been modified to display a link to the newly created record. Assuming that one was created, of course :) Ahasuerus 01:56, 13 June 2017 (UTC)

New cleanup reports for external identifiers

The following "external identifiers in notes" cleanup reports have been added:

  • SFBG (26 pubs)
  • Deutsche Nationalbibliothek (5,000+ pubs, limited to the first 500 pubs for performance reasons)
  • FantLab (2 pubs)
  • Direct Amazon links (700+ pubs)
  • BNF (1,100+ pubs)
  • Library of Congress (11,000+ pubs, limited to the first 500 pubs for performance reasons)
  • OCLC/WorldCat (74,000+ pubs, limited to the first 500 pubs for performance reasons)

The data will become available tomorrow morning. Ahasuerus 20:57, 13 June 2017 (UTC)

Any way to automate the last two at least partially (and DNB as well)? Otherwise that will be one very long project... Annie 21:34, 13 June 2017 (UTC)
Well... The reason why I was able to create a script to auto-move 90%+ of our ASINs to the new field was that they had been originally entered by Fixer. Now, Fixer is a well-behaved robot, even if I say so myself. He has always used the same Note format, which made it easy to extract ASINs from Notes and move them to the new field.
On the other hand, our OCLC/LoC/DNB/BnF/BL/etc links were originally entered by humans. Admittedly, humans have their uses, but consistency isn't one of their strengths. A few random examples:
I have run a bunch of tests and it looks like it may be possible to auto-move certain LCCNs and OCLC numbers safely, but I'll need to spend more time on the logic to make sure that I don't accidentally zap something. In the meantime I have decided to make the data available for cleanup in case anyone wants to tackle it. Ahasuerus 22:50, 13 June 2017 (UTC)
How about semi-automatic (as we did with the title languages)? Identify a syntax, get them on a report for clearing (so all that do not match whatever we are looking for exactly but matches because of humans being humans, get manually fixed); the remainder get direct replacements. That should reduce the numbers (leaving the stragglers and non-patterned messes to deal with last) and reduce the moderators work on the ones that standard (when an editor and moderator are not the same person)... Annie 22:56, 13 June 2017 (UTC)
Hm, it may just work! I'll take a closer look once I finish the cleanup patch that I am currently working on. Ahasuerus 23:13, 13 June 2017 (UTC)
Having just moved several hundred of the OCLC links, I think it's safe to say that most of them fit under a very specific format, with a single OCLC link on a single line, with no additional text -- slight variation in whether it's "OCLC" or "OCLC:", whether the line ends in a period or not, and the format of the URL. Those should be able to be moved semi-automatically. Other links include additional text on the line ("OCLC xxx, which gives a page count of 289") that necessarily have to be moved by a human so as to keep the additional information. Some links don't match the "usual" format, such as the editor who likes to have the linking text be the word "WorldCat" instead of the actual OCLC number. But if you could pull out those standard formatted OCLC (and LCCN) numbers, it would put the remainder within much easier reach. Chavey 07:13, 15 June 2017 (UTC)
(Added) I would focus on notes in which there is a single line of text containing the OCLC reference, and the OCLC reference fits one of the most standard patterns. I think you could even write a regular expression that matches such lines. Chavey 07:42, 15 June 2017 (UTC)
Yup, those are the patterns that I have been looking into :) Ahasuerus 14:47, 15 June 2017 (UTC)
I did about 600 of the OCLC records yesterday, using a semi-automated script that required me to look at everything before submission. Partly the idea was to see if there were other issues that would interrupt the usefulness of the script you're working on. I identified fairly few additional issues, but I'll mention them. It was very uncommon to see any of the "easily" converted OCLC listings on anything other than a single line in the Notes. (We have one editor who likes to enter notes as a continuous flow of text with no manual line-breaks, but that's quite rare.) The beginning of an appropriate line can start in various ways, almost exclusively (deleting angle brackets) li, br, br•, or one of these with an extra space. Appropriate OCLC lines end with a manual return, a period, /li, or br (again, ignoring brackets and spaces). Almost any other ending probably requires a human to look at it. If there is more than one potential OCLC # on a line, the record should be left for humans. These oddities are so infrequent that, IMHO, it isn't worth the effort to try to find/correct them in the script; the lines meeting the criteria above will do 95% of the OCLC external IDs. Chavey 16:52, 16 June 2017 (UTC)
Many thanks! Ahasuerus 18:13, 16 June 2017 (UTC)
One other little point: In a substantial number of cases, removing the line with the OCLC number (or that and the LCCN #) leaves a single line of notes, but still inside the ul-li-/ul paradigm. While that's not an actual problem, the majority of our editors don't use the ul-li format or the br• format when there's only one note, leaving that single note to be on the same line as the Notes: header. Doing things manually, I converted all such Notes to the single line format, which would not be natural when doing it via a script. My guess though is that this detail is not worth worrying about, and the advantages of getting most of these numbers moved via scripting well outweighs the advantages of one format style over the other for such single line "Notes". Chavey 16:52, 16 June 2017 (UTC)
An empty "ul" list is invalid in HTML, but it too can be removed programmatically. Ahasuerus 18:13, 16 June 2017 (UTC)
There were only a couple of hundred records where the OCLC line ended with (/li) (in brackets), so I just converted them all. There are about 720 records where the OCLC line ends in (br), which is doable by hand. So if you wanted to simplify your developing script, you could assume that the lines we're looking for all end in (/a) with the possible addition of a single period and/or additional white space. Of course there are a ton of records that have the OCLC line without a URL at all, so the pattern looks like (the various starts of lines)OCLC(optional :)(optional space) number (optional period)(optional space). Chavey 04:22, 18 June 2017 (UTC)

(unindent) What are you looking for exactly for the FantLab report? As of now, it is empt but I keep finding links in the Russian works. I can find them out and move them easily enough but it seems like the report does not look for the expected format... for example - it has a standard link to Annie 19:33, 20 June 2017 (EDT)

The original logic was limited in scope in order to avoid false positives. As it turns out, it missed a whole class of FantLab links. I have tweaked the logic and added the ability to "ignore" publication records. The new data will become available tomorrow morning. Thanks for raising the red flag! Ahasuerus 21:50, 20 June 2017 (EDT)
I knew the number was too low but figured you moved them with a script or something - the FantLab links are pretty standardized for the most part. Then started finding the ones today and figured I will ask. Thanks! Annie 21:56, 20 June 2017 (EDT)

(unindent) A question about the Publications with direct Amazon links in Notes: most of the still standing links are because a direct link to an Amazon page (using ISBN or ASIN) or LookInside (based on ISBN or ASIN). Do we want to ignore these? Or do we want to clean the Amazon links? All valid ISBNs and ASINs are usually already extracted; it is just that we have an additional link needed. The direct links (when to can easily be replaced with the ASIN template (ASIN is the 10 character ISBN when it exists) but we have links to,, and all those LookInsides. Thoughts? Annie 23:37, 25 June 2017 (EDT)

If an explicit link can be replaced with a template, then I would suggest using that template. If it can't be replaced, e.g. if the note links to Wayback Machine snapshots of Amazon pages, then I don't think we have any choice but to "ignore" the affected record. Ahasuerus 00:40, 26 June 2017 (EDT)
That's what I started doing and then stopped to post - because that left a lot of direct links over there. What would reduce them is a LookInside template (it is always<ASIN> unless if it is to another Amazon but we cannot help that). Does it make sense to add it? Or even a series of LookInside to the big amazons? We have ~350 of those links (searching for "gp/reader/" in Notes). Annie 00:51, 26 June 2017 (EDT)

Post-submission bug fixed

The post-submission bug introduced about an hour ago has been squashed. Sorry about that! Ahasuerus 01:30, 14 June 2017 (UTC)

British Library ISBN links

British Library ISBN links (accessible under "Other Sites") should be working again. Ahasuerus 22:58, 14 June 2017 (UTC)

Andrew M. Greeley: Nongenre?

I am pretty sure most of what's on Andrew M. Greeley's page is non-genre-- can anyone comment? To be more specific, my rather cursory search of Goodreads suggests that the Family Saga, Father Blackie Ryan, Passover, Time Between the Stars, and World of Maggie Ward series don't belong, and none of the nonfiction does. --Vasha 00:07, 15 June 2017 (UTC)

It is kinda funny - I was looking for something else earlier today and stumbled on one of his non-fictions - and submitted a "non-genre" update because it is clearly not genre. Looking at the full list now, none of this non-fiction sound even remotely spec-fic... And reading the note on SFE3, clearly most of his fiction is not genre either... Annie 00:33, 15 June 2017 (UTC)
Ready to exterminate... (in the afternoon).Hauck 06:14, 15 June 2017 (UTC)
If memory serves, an editor looked into this author's output a while back and discovered that a number of seemingly non-genre works contained genre elements (perhaps psychic abilities?) We may want to do additional digging before zapping his series. Ahasuerus 14:20, 15 June 2017 (UTC)
So far, I've only found Angels of September to be vaguely supernatural, the rest being mysteries. Note that at the present time there was only one publication that was verified (out of nearly a hundred) which is quite telling. Hauck 15:04, 15 June 2017 (UTC)

<sarcasm> Adventures in Gaming </sarcasm>

My brain has shut down for the night and so maybe someone else can figure out how best to do this. contains a story titled

<sarcasm> Adventures in Gaming </sarcasm>

You can't see the <sarcasm> in the title at but it's visible if you click "edit title".

The easy fix is to change the title to

&lt;sarcasm&gt; Adventures in Gaming &lt;/sarcasm&gt;

The problem is - someone searching for that title is more likely to use <sarcasm> than &lt;sarcasm&gt;

Is there a way to do this that would create the least astonishment? --Marc Kupper|talk 08:41, 15 June 2017 (UTC)

There is a patch in the works to auto-convert reserved HTML characters to their HTML-encoded representation at data entry time. The only exception will be our Note fields which will allow a limited subset of HTML. The patch will also modify the search logic to change entered reserved HTML characters to their HTML-encoded form. Ahasuerus 14:26, 15 June 2017 (UTC)
Thank you. Are you planning on using the decimal &#nnn; form or will you use the hex and/or named entities? I might as well fix the <sarcasm/> title but also want to normalize it to the form you'll use. --Marc Kupper|talk 18:25, 15 June 2017 (UTC)
The ISFDB software has been using entities like "& l t ;" all along, so that's what I plan to use in the patch that I have been working on (with Marty's help.) Ahasuerus 23:19, 15 June 2017 (UTC)

Lots of publication updates

During the evening I moved several hundred OCLC numbers from Notes into External IDs. Barely made a dent in the backlog, but for some editors this may have dumped a large number of "Changed Primary Verifications" into your report by that name, for fairly trivial changes. My apologies for that, but it seems inevitable as we try to move these IDs. Chavey 10:27, 15 June 2017 (UTC)

I am doing the same on the ASINs list and someone is working steadily on the British Library list. I dropped notes to the PVs that I see often (just so they do not freak out when they login) but your note is a good idea. Annie 17:37, 15 June 2017 (UTC)

Title merge bug fix

The following bug has been fixed:

  • Merging a SHORTFICTION title with a "length" value with a non-SHORTFICTION title can cause the resulting title to have a "length" value even if the title type is not SHORTFICTION.

Ahasuerus 18:13, 15 June 2017 (UTC)

"Editing Tools" change

The "Editing Tools" section of the navigation bar is no longer displayed if there are no editing options available. Moderators will not be affected by this change because the "Editing Tools" sections contains the main "Moderator" link. Ahasuerus 21:22, 15 June 2017 (UTC)

979 ISBNs

"979" ISBN-13s no longer display a bogus derived ISBN-10 on Publication pages. Ahasuerus 19:50, 16 June 2017 (UTC)

"Publication Format" --> "Binding"

All data entry forms have been changed to use "Binding" instead of "Pub. Format". This should synchronize the terminology used on bibliographic and data entry pages. Ahasuerus 20:26, 16 June 2017 (UTC)

Karel Čapek

We currently have the canonical name of Karel Capek without the hachek over the C in his name. We have his name in what I believe is the proper and more frequently appearing form as a pseudonym, Karel Čapek. I believe that most of the titles that currently have the author as "Karel Capek" are actually incorrectly entered. These were undoubtedly entered before we fully or easily supported other alphabets. I'd like to clean this up, but wanted to ensure there was no objection first since this will involve several verified publications. Thus, I'm doing a community notification because of the number of publications involve. My intent is to convert all titles to "Karel Čapek" and then make new variants for "Karel Capek" only for those few cases where it is truly appropriate. If there are any objections, please post them here. Thanks. --Ron ~ RtraceTalk 23:56, 16 June 2017 (UTC)

I agree that "Čapek" should be the canonical name just like "Stanisław Lem" is the canonical name.
As far as the process goes, I believe it would be safe to change the author of all of our Czech titles/publications to "Čapek". On the other hand, other languages may require additional digging. Some clearly use the original spelling, e.g. this German pub or this English pub. Others don't. For this reason I think it would be safer to convert his titles and pubs individually. Keeping title/pub author pairs in sync can be a pain when doing mass changes.
Another thing about our Čapek's bibliography page is that we list many stories which, if memory serves, are not SF. I read a number of his books some decades ago and most of his fiction wasn't SF. Ahasuerus 01:20, 17 June 2017 (UTC)
Coming late as ever, but -- you are absolutely right; the overwhelming majority of the Tales from Two Pockets are nonfantastic, crime or straight psychology. Also, the "stories" include some theatre plays (like in the Romanian omnibus) -- what to do about those, anyway? move/merge them to chapbooks? -- and the most recent, from the end of Toward the Radical Centre: A Karel Capek Reader are just as obviously essays.
I'll try to do something about those as time and capacity allows; I think the Reader is available in the city library, as well as as (sic!) a PDF. But to do the work properly (I like to annotate), I might have to get hold of the collected works (which should be available online, legally, at but in an unfrendly format) with a thorough bibliography... --JVjr 18:56, 19 September 2017 (EDT)

New Moderator nomination

It's been a while since there has been a new Mod. I think it's time. I would like to nominate the one and only [Annie]. She's indefatigable, conscientious, actually listens, and seems to understand the database extremely well. Participates [more than I do] avidly and intelligently. A great long-term addition. I'll leave the usual set-up process to Ahasuerus. Cheers all! --~ Bill, Bluesman 01:25, 18 June 2017 (UTC)

(Bureaucrat notes: See Moderator Qualifications. The nominee has accepted. Ahasuerus 14:55, 18 June 2017 (UTC))


  1. Support, as nominator. ~ Bill, Bluesman 01:25, 18 June 2017 (UTC)
  2. Support. I think Annie would make an excellent addition to the moderators. She has been working on lots of parts of the system, which I think is crucial for a moderator, and is well-skilled in many parts of it. Chavey 04:15, 18 June 2017 (UTC)
  3. Support. C'est parti ! Hauck 05:52, 18 June 2017 (UTC)
  4. Support. Absolutely! PeteYoung 06:21, 18 June 2017 (UTC)
  5. Support. Agree! Should we nominate Nihonjoe as well? --Willem 08:14, 18 June 2017 (UTC)
  6. Support. I've been working with a bunch of Brits lately, and "brilliant!" immediately leaps to mind. Willem, I like your suggestion, too. --MartyD 11:04, 18 June 2017 (UTC)
  7. Support. Highly agree. -- JLaTondre (talk) 12:25, 18 June 2017 (UTC)
  8. Support. Agree, Annie's climb up the learning curve is the fastest that I have ever seen. —The preceding unsigned comment was added by Rkihara (talkcontribs) .
  9. Support. With over 20,000 submissions to her name, Annie is certainly familiar with the process. I believe she meets the rest of the requirements as well. Just don't hesitate to ask if you come across anything remotely unclear! :-) Ahasuerus 15:16, 18 June 2017 (UTC)
  10. Support. A natural! Stonecreek 15:36, 18 June 2017 (UTC)
  11. Support. I think she will do a fine job. ···日本穣 · 投稿 · Talk to Nihonjoe 05:23, 19 June 2017 (UTC)
  12. Support though I'd been wondering who Anniemod in the new-submissions queue at times is and assumed she was already a moderator. --Marc Kupper|talk 07:44, 20 June 2017 (UTC)




The nomination was a success; the moderator flag has been set on the account. Congratulations! Ahasuerus 01:18, 23 June 2017 (EDT)

Submission view changes

The web pages which display submission data have been modified. They now show the date/time when the submission was create/approved/rejected as well as the name of the reviewing moderator. Ahasuerus 16:31, 18 June 2017 (UTC)

Along similar lines, all pages that show lists of "Recent Edits", "My Recent Edits", "Recent Rejections", etc have been modified to display more data about each submission. In addition, each page displays the current ISFDB server time to provide a frame of reference for the displayed information. Ahasuerus 20:44, 18 June 2017 (UTC)
"New Submissions" has been updated as well. Ahasuerus 19:26, 19 June 2017 (UTC)
While you are playing with those pages - can we have a "Are you sure you want to cancel?" popup on the "Pending" list's Cancel links? I won't even try to count how often I will open the page to see where I left it off and press cancel on one of the items (especially when working from one of the touchscreens or with the laptop pad... Usually it is something easy to redo but one of those days, I will wipe out something that is not that easy to reproduce... :)Annie 21:12, 19 June 2017 (UTC)
We can certainly create a Feature Request, but I am not sure it would help much. People tend to become accustomed to pop-ups and dismiss them without reading. Still, it would be better than nothing. Ahasuerus 22:33, 19 June 2017 (UTC)
And then it is their own fault that they canceled it. :) But just because I was in the vicinity of the request with a mouse and was distracted, I had to redo some requests. May be just me being clumsy - but the thing is just too annoying when it happens - and it is not an operation everyone needs to do too often - so it won't have much of a delaying factor for most editors. So thought I should mention it :) Annie 23:03, 19 June 2017 (UTC)
Personally I would find such a pop-up really annoying. And on the rare occasion that I cancel something I didn't intend to, I can just go to the Rejected Submissions, open the record of the one I just cancelled, and copy it to a new submission. Vasha 23:06, 19 June 2017 (UTC)

Translations - display changes

As per FR 1059, the way translations are labeled on Summary Bibliography and Series pages has been changed. The label, which used to say "Variant Title", now says "Variant" or "Translation" as appropriate. Here is what Heinlein's Future History" looks like now.

I believe that the new labeling system is more clear, but there is room for improvement. For example, we currently display variants and translations together, e.g.:

  • Variant: Life-Line (1939) [as by Robert Heinlein]
  • Translation: Ligne de vie [French] (1967)
  • Translation: Lebenslinie [German] (1972)
  • Variant: Lifeline (1977) [as by Robert Heinlein]
  • Translation: Levenslijn [Dutch] (1979)

My thinking is that we may want to display variants first and translations second:

  • Variant: Life-Line (1939) [as by Robert Heinlein]
  • Variant: Lifeline (1977) [as by Robert Heinlein]
  • Translation: Ligne de vie [French] (1967)
  • Translation: Lebenslinie [German] (1972)
  • Translation: Levenslijn [Dutch] (1979)
  • etc

Once we do that, we can consider other improvements. Ahasuerus 23:09, 19 June 2017 (UTC)

I like the idea of having them grouped together with variants before translations. I would make a case that we should also group the translations into a certain language (when more than one is done) together as well. That would make it easier to spot that you have a few translation in a language and to be extra careful which one you are importing. Annie 23:14, 19 June 2017 (UTC)
I think there are advantages and disadvantages to grouping either chronologically or by language. We'll also need to consider how this will affect the display of variants and translations on the Title page. Ahasuerus 01:27, 20 June 2017 (UTC)
I like Annie's suggestion. Chavey 02:12, 20 June 2017 (UTC)
What should we do with our Title pages then? Would you be in favor of ordering their "Translations" sections by language and then by year? Ahasuerus 03:08, 20 June 2017 (UTC)
Either that or by year of first translation into the language and then add the rest of the ones into the same language. So if the first ever translation is into French, French goes first but instead of followed by First English, you get all French first. Annie 03:11, 20 June 2017 (UTC)
I prefer the first solution with timeline (chronologically)--Wolfram.winkler 02:55, 22 June 2017 (EDT)

Variants and disambiguated authors

I am looking at this and I am confused on what is the idea here. Shouldn't we just change the author name inside of the two publications to the (I) variant of the name and be done with it? Yes, they did not have the (I) in the actual publication but that is the one case where we do not record as written in the book because of operational reasons. Am I missing something here or does it need fixing (change author inside of the pubs and delete this unneeded variant)? Annie 17:48, 20 June 2017 (UTC)

You are correct. Fixed. -- JLaTondre (talk) 14:41, 20 June 2017 (EDT)
Thanks! :) Annie 14:45, 20 June 2017 (EDT)

ISBN template

While moving FantLab links, I am also adding the additional ISBNs where applicable. Can we have a template so they are easier to spot and run a report on when we eventually have the ability to add multiple ISBNs? Annie 22:00, 20 June 2017 (EDT)

Good point, will do! Ahasuerus 22:46, 20 June 2017 (EDT)
Thanks! And talking about multiple ISBNs and Publishers - in the Russian case they are (almost?) always paired (an ISBN per publisher). I am yet to find one that does not match that rule - but it is not impossible. I know we cannot do a 2 parameter template so I will use the notes to pair properly but I hope that when we have the feature one day, we will be able to pair them as well when applicable. Or I am overthinking again... Annie 23:13, 20 June 2017 (EDT)
OK, {{ISBN}} has been added.
Re: publisher-specific ISBNs, I am not sure even Russian publishers are consistent. For example, take this Russian translation of The Lord of the Rings. The FantLab page lists two publishers, one additional imprint, four ISBNs and two printings. Ahasuerus 12:40, 21 June 2017 (EDT)
Thanks! For the publisher-specific ISBNs: АСТ Москва is a separate publisher even if it is part of the АСТ group and adds its own ISBN when published. So we have 3 publishers here:). I am seeing the АСТ Москва (5-9713-2149-8), the АСТ (5-17-018845-5) and the Terra Fantastica (5-7921-0618-5) ISBNs. The 4th one is a Belorussian one - I need to do some more digging but I think I had seen that before - Belorussian distribution only ISBN from АСТ - will be back with more details as soon as I find them again (or find my notes wherever they may be). So yeah, they are consistent (mostly :) ) and even if we cannot pair all of them, some pairing is possible. Annie 15:49, 21 June 2017 (EDT)
Plus it looks like that last ISBN could have been only on the additional printing (being ISBN-13) and even then 2006 is stretching it so I wonder if it is even valid for that book or if it was a distribution only one added later and one that was on whatever copy was looked at for the entry. In which case it does not belong on the record really... so the consistency holds. Annie 16:14, 21 June 2017 (EDT)
Well, ultimately we will want to allow multiple publishers per publication. At that point we will have to decide whether we want to associate ISBNs with individual publishers. In certain cases it may be feasible (see above), but in many other cases it may not be, e.g. when a jointly published book has only one ISBN.
More importantly, we will first have to decide how we want to reorganize our publisher records, including:
  • imprints vs. publishers
  • publishers becoming imprints after an acquisition
  • imprints getting spun off as separate publishers
  • "stated published" vs. "regularized publisher"
  • multiple forms of the publisher name used in the same publication
etc. Ahasuerus 17:49, 21 June 2017 (EDT)
Understood :) I was just adding that to the mental notes of the thinking - I don't like having derivable information just sitting because we did not structure properly for it. That's all. Being elbows deep in the Russian mess of ISBNs make me think of all kind of weird things about easier way to figure out what is what. In the long run, it won't really matter that much - it was just my curiosity a few years ago to untangle the mess that made me aware of how the Russian ISBNs work at all. And I have a suspicion that each country may have an oddity or two hiding under the hood. Annie 18:01, 21 June 2017 (EDT)

Apex publisher name?

Apex Book Company is the same company as Apex Publications and as far as I can see the two names are used interchangeably. Does anyone know if there's actually some reason to use either one or the other? --Vasha 11:20, 21 June 2017 (EDT)

I cannot see a reason to have them separate unless if we want to keep the magazine publisher from the book publisher (and then a lot of books need moving). Annie 16:18, 21 June 2017 (EDT)
I think Apex Publications is more general (FWIW I have two Apex books and they both say Publications throughout); but do we ever do alternate/pseudonym publisher names? ABC redirected to AP? --Vasha 18:24, 21 June 2017 (EDT)
In the system now? Only soft-linked via notes. Or in some cases the notes contain the alternative names. We chose one for "the name" and that is it in this system. IMHO opening the door for pseudonyms will end up with abolishing the regularization and will make a mess with mergers and splits and what's not... Annie 18:33, 21 June 2017 (EDT)
I think we will need a single comprehensive solution to the "publisher problem". It will have to address the issues of multiple publishers, imprints and name regularization at the same time. We'll need to come up with a solid design first. Ahasuerus 18:36, 21 June 2017 (EDT)
I tried that Verified Publishing Names thing for a while but it never gained traction. I've drifted towards recording less detail and not being as picky about inconsistencies in the details. If a bunch of people PV a publication as a Bantam/Spectra but I only see "Bantam" for example I'm likely to let it slide. Maybe I'll add a pub note but that means notifying people and creating work to fetch and recheck a pub for an area that the consensus does not pay attention to.
I also discovered that as we record more details that it increases the chance of error, inconsistencies, and subjective calls. I assume it also increases the verification load on later PVers as they all presumably need to check and compare a detail that's not part of what we normally cover. The only exception to this is that my personal PV process for collections and anthologies is to double check the spelling of the titles, author names in the Contents, copyright acknowledgments, and the start of each work. I'll document inconsistencies on the theory that it would explain why different sources have slightly different data about a topic.
I believe this was an area where the publication wiki pages were useful as we could record extra details that were not part of the primary verification record. I did not care that the wiki part was not moderated as the edit history is available and it was never something that was formally part of the PVed information. Someone doing a later PV could chose to ignore the link to the wiki page. As long as the main record matched we had the data that most of us care about.
I don't have any Apex publications and so can't answer the original thread topic. --Marc Kupper|talk 20:06, 21 June 2017 (EDT)
That Verified Publishing Names thing looks like a lot of work-- I can understand why no one devoted themself to it for long enough to get it to a usable state. --Vasha 21:11, 21 June 2017 (EDT)
I agree, it was a lot of work. Something that could have simplified the workload would be to include scans of the title page, copyright page, and other parts of the publications that seemed relevant. That also would have allowed for future projects such as confirming the copyright date, or other things.
On the Apex issue - There are Amazon Look Insides available for the 2015 and 2016 books. All of the Apex Book Company publications are by "Apex Publications, LLC." per the copyright pages meaning the proposed merge is safe. The title pages have "An Apex Publications Book" over "Lexington, Kentucky". We are missing 978-1-937009-31-1 and 978-1-937009-35-9 on ISFDB.
978-1-937009-31-1 For Exposure hardcover - look inside is of tp
978-1-937009-35-9 The Apex Book of World SF: Volume 2 - Apex Publications
For Exposure: 978-1-937009-30-4 - Apex Publications
King of the Bastards 978-1-937009-32-8 - title page and copyright page not available. Back cover has Apex Publications
The Apex Book of World SF 978-1-937009-36-6 - Apex Publications
The Apex Book of World SF: 4 978-1-937009-33-5 - Apex Publications
The Apex Book of World SF 3 978-1-937009-34-2 - Apex Publications
Best of Apex Magazine: Volume 1 978-1-937009-37-3 - Apex Publications
978-1-937009-38-2 - 404 not found on Amazon
978-1-937009-39-1 - 404 not found on Amazon
978-1-937009-40-7 - 404 not found on Amazon
978-1-937009-41-6 - 404 not found on Amazon
978-1-937009-42-5 - 404 not found on Amazon
978-1-937009-43-4 - 404 not found on Amazon
978-1-937009-44-3 - 404 not found on Amazon
978-1-937009-45-2 - 404 not found on Amazon
978-1-937009-46-1 - 404 not found on Amazon
978-1-937009-47-0 - 404 not found on Amazon
978-1-937009-48-9 - 404 not found on Amazon
978-1-937009-49-8 - 404 not found on Amazon
978-1-937009-50-4 - 404 not found on Amazon - Ugly As Sin on ISFDB
978-1-937009-51-3 - 404 not found on Amazon - The Wicked on ISFDB
978-1-937009-52-2 - 404 not found on Amazon
978-1-937009-53-1 - 404 not found on Amazon - Beautiful Sorrows on ISFDB
978-1-937009-54-0 - 404 not found on Amazon - Greener Pastures on ISFDB
978-1-937009-55-9 - 404 not found on Amazon
978-1-937009-56-8 - 404 not found on Amazon
978-1-937009-57-7 - 404 not found on Amazon
978-1-937009-58-6 - 404 not found on Amazon
978-1-937009-59-5 - 404 not found on Amazon
--Marc Kupper|talk 06:00, 22 June 2017 (EDT)
Wow, thanks for checking all that. I've put a request for the merge on the moderator page.
You have mistakes in those ISBNs you couldn't find. Ugly as Sin is 978-1-937009-50-2; The Wicked is 978-1-937009-51-9; Beautiful Sorrows is 978-1-937009-53-3; etc. (wrong final digits) --Vasha 07:31, 22 June 2017 (EDT)
Thanks for catching that. It turns out I had bugs in the code for a web page that lets me enter an ISBN as 978-1-937009-5?-?. It's supposed to generate a list of the ISBNs with each linking to It makes it easy to scan a range of ISBNs. It's been working well for ISBN10s but had bugs for ISBN13s. Here's an updated list formatted as a table. Non-linked titles are pubs we don't have. All of them seem to be specfict and we already have the authors.
978-1-937009-00-7 1937009009 The Tower of the Forgotten Sara M. Harvey
978-1-937009-01-4 1937009017 Escape from Zombie City: A One Way Out Novel Ray Wallace
978-1-937009-02-1 1937009025 The Fields Ty Schwamberger, Jonathan Maberry
978-1-937009-03-8 1937009033 Disintegration Visions J.M. McDermott
978-1-937009-04-5 1937009041 On ISFDB but not Amazon
978-1-937009-05-2 193700905X The Apex Book of World SF 2 Lavie Tidhar
978-1-937009-06-9 1937009068 Appalachian Undead Eugene Johnson
978-1-937009-07-6 1937009076 Dark Faith: Invocations Maurice Broaddus, Jerry Gordon
978-1-937009-08-3 1937009084 Seasons of Insanity Gill Ainsworth, Frank W. Haubold
978-1-937009-09-0 1937009092 Like Death Tim Waggoner
978-1-937009-10-6 1937009106 The Lost Level Brian Keene
978-1-937009-11-3 1937009114 The Book of Apex Catherynne M. Valente
978-1-937009-12-0 1937009122 What Makes You Die Tom Piccirilli
978-1-937009-13-7 1937009130 Machine Jennifer Pelland
978-1-937009-14-4 1937009149 Desper Hollow Elizabeth Massie
978-1-937009-15-1 1937009157 Plow the Bones (Apex Voices) (Volume 1) Douglas F. Warrick
978-1-937009-16-8 1937009165
978-1-937009-17-5 1937009173 I Can Transform You (Apex Voices) (Volume 2) Maurice Broaddus, Matt Forbeck
978-1-937009-18-2 1937009181 Appalachian Undead Jonathan Maberry, John Skipp, Gary A. Braunbeck, Jason Sizemore, Eugene Johnson
978-1-937009-19-9 193700919X Glitter & Mayhem Lynne M. Thomas, John Klima, Michael Damian Thomas, Amber Benson
978-1-937009-20-5 1937009203 The Book of Apex Lynne M. Thomas
978-1-937009-21-2 1937009211 Maze J.M. McDermott
978-1-937009-22-9 193700922X Mountain Dead Geoffrey Girard, Sara M Harvey, Lesley Conner, K. Allen Wood, Jason Sizemore, Eugene Johnson
978-1-937009-23-6 1937009238 Midnight Mari Adkins
978-1-937009-24-3 1937009246 The Apex Book of World SF 3 Benjanun Sriduangkaew, Lavie Tidhar
978-1-937009-25-0 1937009254
978-1-937009-26-7 1937009262 War Stories: New Military Science Fiction Karin Lowachee, Andrew Liptak, Jaym Gates
978-1-937009-27-4 1937009270 Severance Chris Bucholz
978-1-937009-28-1 1937009289 Sing Me Your Scars (Apex Voices) (Volume 3) Damien Angelica Walters
978-1-937009-29-8 1937009297 Rosewater Tade Thompson
978-1-937009-30-4 1937009300 For Exposure: The Life and Times of a Small Press Publisher Jason Sizemore
978-1-937009-31-1 1937009319 For Exposure: The Life and Times of a Small Press Publisher Jason B Sizemore
978-1-937009-32-8 1937009327 King of the Bastards Brian Keene, Steven L. Shrewsbury
978-1-937009-33-5 1937009335 The Apex Book of World SF: Volume 4 (Apex World of Speculative Fiction) Usman T. Malik, Mahvesh Murad, Lavie Tidhar
978-1-937009-34-2 1937009343 The Apex Book of World SF: Volume 3 (Apex World of Speculative Fiction) Karin Tidbeck, Lavie Tidhar
978-1-937009-35-9 1937009351 The Apex Book of World SF: Volume 2 (Apex World of Speculative Fiction) Lauren Beukes, Lavie Tidhar
978-1-937009-36-6 193700936X The Apex Book of World SF: Volume 1 (Apex World of Speculative Fiction) S.P. Somtow, Lavie Tidhar
978-1-937009-37-3 1937009378 Best of Apex Magazine: Volume 1 Ursula Vernon, Lesley Conner, Jason Sizemore
978-1-937009-38-0 1937009386 Freeze/Thaw Chris Bucholz
978-1-937009-39-7 1937009394 Apex Magazine 2015 Sampler Jason Sizemore
978-1-937009-40-3 1937009408 The Kraken Sea E. Catherine Tobler
978-1-937009-41-0 1937009416 first communions Geoffrey Girard
978-1-937009-42-7 1937009424 Glitch Rain Alex Livingston
978-1-937009-43-4 1937009432 Stay Crazy Erica L. Satifka
978-1-937009-44-1 1937009440 Upside Down: Inverted Tropes in Storytelling Monica Valentinelli, Gates Jaym
978-1-937009-45-8 1937009459 Kentucky Kaiju Shawn Pryor, Justin Stewart, Tressina Bowling
978-1-937009-46-5 1937009467 Upside Down: Inverted Tropes in Storytelling Valentinelli Monica, Gates Jaym
978-1-937009-47-2 1937009475 Yours to Tell: Dialogues on the Art & Practice of Writing Steve Rasnic Tem, Melanie Tem
978-1-937009-48-9 1937009483 Apex Magazine SFFH: Issue 0, Winter 2017 Lavie Tidhar, Douglas F. Warrick, A. Merc Rustad, K. T. Bryski, Jason Sizemore
978-1-937009-49-6 1937009491
978-1-937009-50-2 1937009505 Ugly As Sin James Newman
978-1-937009-51-9 1937009513 The Wicked James Newman
978-1-937009-52-6 1937009521 Mars Girls Mary Turzillo
978-1-937009-53-3 193700953X Beautiful Sorrows Mercedes M. Yardley seems non-genre though
978-1-937009-54-0 1937009548 Greener Pastures Michael Wehunt, Michael Bukowski
978-1-937009-55-7 1937009556 Shine Your Light On Me Lee Thompson
978-1-937009-56-4 1937009564
978-1-937009-57-1 1937009572
978-1-937009-58-8 1937009580
978-1-937009-59-5 1937009599
978-1-937009-60-1 1937009602
978-1-937009-61-8 1937009610
978-1-937009-62-5 1937009629
978-1-937009-63-2 1937009637
978-1-937009-64-9 1937009645
978-1-937009-65-6 1937009653
978-1-937009-66-3 1937009661
978-1-937009-67-0 193700967X
978-1-937009-68-7 1937009688
978-1-937009-69-4 1937009696
978-1-937009-70-0 193700970X
978-1-937009-71-7 1937009718
978-1-937009-72-4 1937009726
978-1-937009-73-1 1937009734
978-1-937009-74-8 1937009742
978-1-937009-75-5 1937009750
978-1-937009-76-2 1937009769
978-1-937009-77-9 1937009777
978-1-937009-78-6 1937009785
978-1-937009-79-3 1937009793
978-1-937009-80-9 1937009807
978-1-937009-81-6 1937009815
978-1-937009-82-3 1937009823
978-1-937009-83-0 1937009831
978-1-937009-84-7 193700984X
978-1-937009-85-4 1937009858
978-1-937009-86-1 1937009866
978-1-937009-87-8 1937009874
978-1-937009-88-5 1937009882
978-1-937009-89-2 1937009890
978-1-937009-90-8 1937009904
978-1-937009-91-5 1937009912
978-1-937009-92-2 1937009920
978-1-937009-93-9 1937009939
978-1-937009-94-6 1937009947
978-1-937009-95-3 1937009955
978-1-937009-96-0 1937009963
978-1-937009-97-7 1937009971
978-1-937009-98-4 193700998X
978-1-937009-99-1 1937009998 Let's Play White Chesya Burke
Beautiful Sorrows seems non-genre. --Marc Kupper|talk 16:32, 22 June 2017 (EDT)
Kentucky Kaiju is a graphic novel; I added the other missing ones. --Vasha 18:41, 22 June 2017 (EDT)

Vote changes

As per FR 1049, the way votes are displayed on Title pages has been changed. Previously, a title had to have at least 5 votes before the average was displayed. With the new system the average rating is displayed regardless of how many votes have been cast. The number of votes is displayed in parentheses.

For example, the title page for Nine Princes in Amber now says:

  • User Rating: 9.22 (36 votes)

The part of the FR which requests to "calculate an additional "Combined User Rating" across the main parent title and its variant titles" hasn't been implemented yet. Ahasuerus 18:58, 21 June 2017 (EDT)

I am not seeing votes for titles I voted on (although I do see it for Nine Princes in Amber). For example Hunted says "This title has no votes" even though my vote is there. Do you see Hunted differently? --Vasha 21:20, 21 June 2017 (EDT)
Oops! It's a display problem which affects titles that have fewer than 5 votes recorded prior to the change. It's just a display issue, so no data has been lost. Sorry about that! I'll fix it tomorrow morning. Ahasuerus 21:55, 21 June 2017 (EDT)
Fixed. Ahasuerus 14:45, 22 June 2017 (EDT)
The Title page has been modified to display the currently logged-in user's vote (if present.) Ahasuerus 16:48, 22 June 2017 (EDT)
The part of the FR which asked to "calculate an additional "Combined User Rating" across the main parent title and its variant titles" has been implemented. For example, the Title page for The Hobbit, or There and Back Again now says:
  • User Rating: 7.52 (21 votes) Including variants and translations: 7.60 (42 votes)
Ahasuerus 18:30, 22 June 2017 (EDT)

Cloning and External IDs

A new chechbox has been added to the intermediate cloning page. It asks the editor whether to re-use external IDs. By default the box is not checked. Ahasuerus 22:45, 22 June 2017 (EDT)

"Other [records] with the same name"

A while back we enhanced the Summary Bibliography page to display a note if there were other authors with the same name, e.g. Andrew Smith (V). The software has been further modified to display similar warnings when viewing disambiguated series, publication series and publisher pages. Ahasuerus 17:26, 23 June 2017 (EDT)

Is there a reason placement of the series one is not consistent with the other three? -- JLaTondre (talk) 17:34, 23 June 2017 (EDT)
Yes, indeed! It's known as "human error" :-) , now fixed. Ahasuerus 17:59, 23 June 2017 (EDT)
They look at the same place for all 3 elements for me? Annie 17:57, 23 June 2017 (EDT)
That's because I fixed the Series page 3 minutes ago :) Ahasuerus 17:59, 23 June 2017 (EDT)
Ha, testing done before you even managed to announce it. Annie 18:22, 23 June 2017 (EDT)

New cleanup report: Titles with mismatched parentheses

A new cleanup report, "Titles with mismatched parentheses", has been deployed. The data will become available tomorrow morning. Of the 232 suspect titles that I expect it to find about 20% will be false positives, so the report lets moderators "ignore" records. Ahasuerus 19:50, 23 June 2017 (EDT)

Down to 2 - which may need an ignore but I am still to talk to the verifiers. Annie 19:23, 25 June 2017 (EDT)
Thanks! Ahasuerus 11:52, 26 June 2017 (EDT)
The verifier fixed those 2 remaining ones so we are all done with Titles with mismatched parentheses legacy issues. Anything showing from now on will be newly entered data. Annie 15:29, 4 July 2017 (EDT)

Award pages and parent titles

As per FR 1058, the following pages have been enhanced to display parent titles:

  • The Award Bibliography page, e.g. see this page
  • The Award Year page, e.g. see this page -- note how "The Dark Forest", "Folding Beijing" and "Hollowgirl" are displayed

Ahasuerus 18:21, 24 June 2017 (EDT)

I like it! ···日本穣 · 投稿 · Talk to Nihonjoe 20:06, 24 June 2017 (EDT)
Very nice! Annie 22:16, 24 June 2017 (EDT)

ASIN IDs - Display changes

The way ASIN IDs are displayed has been changed. In addition to the 6 Amazon stores that were already supported (Canada, France, Germany, Japan, UK, US), we now link to the Amazon stores in Australia, Brazil, China, Spain, India, Italy, Mexico and the Netherlands. The Chinese store seems to be missing a fair number of ASINs, but the ones that are available seem to use the same ASINs as the rest, e.g. see this publication. Please note that the "Other Sites" section of the navigation bar on the left hasn't been changed (yet.) Ahasuerus 15:38, 25 June 2017 (EDT)

I still wonder if all the external IDs should be moved out from the top panel - and it their own one or in the Secondary verifications one. This way the top will remain for what the book is and it won't look too crowded when we have no notes (or single lines). Annie 17:25, 25 June 2017 (EDT)
There are certain advantages to displaying External IDs immediately below the Note field. For example, if the Note field says "Data from BNF", then having the BNF number right below it makes it easier to figure out what it's talking about (and follow the link if needed.) Ahasuerus 19:15, 25 June 2017 (EDT)
Oh, I know - but when they take over visually, the notes are becoming invisible. Annie 19:17, 25 June 2017 (EDT)
Well, how about we move them below the "Upload new cover scan" line? It would hopefully separate the Note field and the ID area. Ahasuerus 20:49, 25 June 2017 (EDT)
That may work - it will push these under the image (or low enough so it does not crowd the Notes. Annie 20:52, 25 June 2017 (EDT)
Let's see what other editors think. In the meantime Amazon has made certain changes that affect Fixer (effective almost immediately), so I am mega busy trying to mitigate the fallout. Ahasuerus 13:19, 26 June 2017 (EDT)
Not urgent at all -- good luck dealing with the Amazon changes! I will stop bugging you and go fix some ASINs :) Annie 13:23, 26 June 2017 (EDT)

A new cleanup report: ISBN-less e-pubs without ASINs

A new cleanup report, "ISBN-less e-pubs without an ASIN", has been deployed. I expect it to find around 3,200 publication records when it runs overnight. Moderators will be able to "ignore" records.

Once the data has been cleaned up, Fixer will be able to submit ISBN-less publications without having to worry about creating duplicates. Ahasuerus 19:09, 25 June 2017 (EDT)

POD numbers

Do we want to add in a field for tracking POD numbers (print on demand numbers, they are under the barcodes on the last page inside POD books)? Since POD is becoming a bigger and bigger thing, it might become a useful tool. Usually, they have the POD number, a print date, and a print location. ···日本穣 · 投稿 · Talk to Nihonjoe 02:17, 28 June 2017 (EDT)

Oh, and the POD numbers sometimes have letters in them, so they would have to be alphanumeric instead of just numeric. ···日本穣 · 投稿 · Talk to Nihonjoe 02:18, 28 June 2017 (EDT)
Is there a Web page explaining how these numbers are assigned, what the standard is, etc? Ahasuerus 15:57, 29 June 2017 (EDT)
Not that I know of. All the POD books I own have the same basic info (what I described above) on the last page inside the book. It is very consistent. I can take pictures of several if you want. Maybe some others here know of something. ···日本穣 · 投稿 · Talk to Nihonjoe 17:25, 29 June 2017 (EDT)
Most of the printers print there an identification for the data file being used. I am not sure if Batch number is not incorporated as well. Annie 19:27, 29 June 2017 (EDT)
Yup. I thought it might be another useful way of tracking them. We might be able to see patters after a certain number have been entered (not sure what that certain number would be). ···日本穣 · 投稿 · Talk to Nihonjoe 20:18, 29 June 2017 (EDT)
We will only see a pattern if we have multiple copies from the same book and know which ones are from the same batch (based on the date maybe). I don't think that it will any useful information to the db but I may be wrong. I certainly would not like to copy a long alphanumerical string anytime I am entering a book. :( Annie 20:35, 29 June 2017 (EDT)
It would only be with POD books, not every time. And we already copy a number of different long strings: LCCN, ISBN, and so on. ···日本穣 · 投稿 · Talk to Nihonjoe 02:09, 30 June 2017 (EDT)
I agree with Annie. While I don't have that many POD books I'm not inclined at all to be hand-entering and double checking "LaVergne, TN USA" over "04 December 2009" over "165914LV00005B/13/P" from Youthanasia. If the information was available via barcode or data matrix then I would consider scanning it. I am entering the POD date in ISFDB's date field in the hopes that others also check the POD date to start gathering data on if there are many publications with the same date of if a few, or even one, publication is printed each time someone orders a copy.
@Nihonjoe, an LCCN is much shorter and can also be verified on the spot by clicking on the resulting link. If it links to the correct record on the LOC site then you know you have the right value. An ISBN has a checksum meaning the editor gets near immediate feedback if there's an error and can also instantly click to verify that it leads to the expected record on many web sites. The POD value is 19 characters in the book I pulled off the shelf that can't be verified other than by manual inspection. It's also has long string with no delimiters, for example with "165914LV00005B/13/P" I had to be extra careful with that long run of zeroes.
Earlier this week I ran across an anthology I want to buy. While checking to see who had copies for sale I noticed there were eight AbeBooks sellers with all of them stating they were selling new copies printed on demand. It made me wonder if all eight had their own POD presses or how that worked. --Marc Kupper 02:51, 30 June 2017 (EDT)
Well, all of the POD books have the POD number with an attached barcode for it on the lower left of the last page, and the date and location of printing on the lower right. All of them are that way. Either way, it seems people don't care for this idea, so we'll leave it at that. ···日本穣 · 投稿 · Talk to Nihonjoe 14:40, 30 June 2017 (EDT)
But you do care. And that is important. So one option may be to use the wiki (under your account or in general on the wiki) - create a POD/PUB page and create a table with publinks and the POD numbers? Maybe something will emerge from that. This way if one day it becomes trackable/important, moving the data will be possible. Annie 14:51, 30 June 2017 (EDT)
How about this? Any suggestions on columns? ···日本穣 · 投稿 · Talk to Nihonjoe 18:36, 30 June 2017 (EDT)
None of my POD books looked like yours - I have some where the code is on the left and the ISBN is on the right, I have some where the code is on the left and the date/place is in he middle and I have some where the whole thing is in the middle lower part of the page. So really starts sounding like each printer having its own ideas? But I am getting my library in order these days so I will see if I can find more of the books and see if I can spot some pattern... Annie 14:51, 30 June 2017 (EDT)
One of the lessons that we have learned over the years is that it's important to have a solid understanding of real life practices and usage before we try to support them in the software. After all, our database is an attempt to model the world of books and needs to follow it closely. On the other hand, if we wait too long before we add software support, we end up with a lot of "stuff" in notes which takes forever to convert once the software has been updated. The balance can be tricky. Ahasuerus 11:01, 1 July 2017 (EDT)
I have a lot of POD books (not just genre - I have a soft spot for small and micro-publishers and most of them had switched to POD in the last years). And it seems to be different - sometimes I will get a book printed months ago, sometimes Amazon will deliver me a book that had been printed on the day I ordered. It is great for remembering when I bought the book but it will be meaningless for a bibliography. It seems like there are printers out there where you can just order a copy and get it in a few days (depending on the ordering agent) as long as they already have the file. Doesn't Lulu also operate in a similar way - they print and mail the book when you order it? Annie 02:59, 30 June 2017 (EDT)

Bad HTML in publication notes

A while back we added a cleanup report to identify and correct notes with empty HREFs. This report has been now expanded to look for HREF values without quotes. Although some versions of HTML allow "naked" HREF values under certain circumstances, they will become invalid in all notes Real Soon Now (tm) as we continue to tighten security.

The report has been renamed and moved to the Notes/Synopsis area of the cleanup menu. The data, a few hundred publication records, will become available tomorrow morning. Ahasuerus 15:41, 29 June 2017 (EDT)

P.S. A new cleanup report, "Mismatched HTML tags in Publication Notes", has been deployed. It looks for HTML tags without a closing tag and vice versa: "b" and "/b", "i" and "/i", etc. Approximately 830 pubs are currently affected. Ahasuerus 23:09, 29 June 2017 (EDT)

I just did 50 of those "Mismatched HTML tags". The vast majority of the problems, 44 of them (88%), were mismatched ul - /ul tags. Surprisingly, the second most frequent error (3 of the remaining 6) were editors who preferred to add the (unnecessary) /li tag with their li tags, but then mistyped the "/li", e.g. as "/l" or as "/i". Chavey 00:21, 1 July 2017 (EDT)
I am also seeing mismatched pairs of "i" or "/i" and a quote on the other side. But yeah - the missing "ul" or one too many "/ul" are the most common.
On the other hand most of the HREFs with no quotes in the other report are for OCLC or LCCN links, a smattering a links to a Portuguese library (same editor I would presume - most of them not verified) and a few links to artist's sites and Amazon.Annie 01:07, 1 July 2017 (EDT)
HTML errors are hard to notice when moderating submissions. Things will get a bit better once we restrict human-entered HTML to a subset of valid tags. We'll also want to add moderator warnings about mismatched tags and other HTML weirdness. Ahasuerus 09:51, 1 July 2017 (EDT)
I was working on the mismatched tags starting with titles that began with "The ..." (so that Annie and I weren't fighting over the same sets of books :-). I made fairly good progress, and all of the remaining books alphabetically from "The Best .. " to "The Golden .." share the same issue: They used tables that included styles or other settings within the td or tr specifiers. (And all in ways that, IMHO, were appropriate.) None of those 6 titles actually has any mismatched html tags. So eventually, you will want to add the ability to recognize a wider variety of td and tr tags, or else add the inevitable "ignore" buttons. I suspect the later. Chavey 12:17, 1 July 2017 (EDT)
I see. The report was looking for things like "<table>"/"</table>" mismatches and found notes with snippets like "<table align=center>".
The bad news is that, as we discussed a few weeks ago, we will discontinue support for all user-entered HTML attributes (except "href") in the near future. They create huge security issues, including potentially having our data destroyed or, worse luck, massively corrupted.
The good news is that the impact of the security changes will be very limited. As of this morning, here is how many notes will be affected:
2 "tr" tags
19 "table" tags
1 "th" tag
19 "td" tags
If a record absolutely has to include HTML attributes, the note can be moved to the Wiki and linked from the database side. Ahasuerus 13:18, 1 July 2017 (EDT)
I was wondering if that was the case. After we correct all the easier cases of these mismatched tags, I'll go back and see if I can find other ways to deal with these tables using "easy" html. Chavey 13:32, 1 July 2017 (EDT)
The non-table ones are all cleared. The remaining 16 are there because of tables :) Annie 20:22, 2 July 2017 (EDT)

(unindent) "Invalid HREFs in Publication Notes" are all taken care of. All hrefs in the system are now decent and clothed - no more naked ones hiding in corners. :) Annie 03:12, 2 July 2017 (EDT)

Thanks! Ahasuerus 08:21, 2 July 2017 (EDT)
Good job! I did a bunch yesterday and today, and was very pleased to see them all finished while I was out dancing :-) Thanks much. Chavey 02:18, 3 July 2017 (EDT)

2017-06-30 downtime at 4pm

The server will be unavailable between 4pm and 4:03pm server time. Ahasuerus 15:50, 30 June 2017 (EDT)

Oops! I meant 4pm. Ahasuerus 15:56, 30 June 2017 (EDT)
Everything should be back up. Approximately 3,300 OCLC IDs have been moved from notes to the External ID field. More to follow. Ahasuerus 16:02, 30 June 2017 (EDT)

New cleanup report: Mismatched OCLC URLs in Publication Notes

While working on moving OCLC links to the External IDs field, I noticed that some links were internally inconsistent, i.e. the displayed OCLC record number and the URL did not match. I ended up creating a new cleanup report which will become available tomorrow morning. Some mismatches will require additional TLC and/or judicial application of the OCLC template. I expect the new report to find fewer than 200 records. Ahasuerus 18:37, 30 June 2017 (EDT)

There are some of those in the other kinds of links as well - a few in the FantLab ones for example while I was clearing them and I've seen a few in ASIN and the LCCN links(and not the 62-21 format vs the 6200021 format which are normal). Annie 18:44, 30 June 2017 (EDT)
Originally I planned to auto-move LCCNs and OCLC IDs at the same time. However, LCCNs have additional quirks, including the hyphens that you mentioned, so they will be handled separately. One monstrosity at a time... Ahasuerus 10:35, 1 July 2017 (EDT)
I constructed a macro that was pretty good at moving the OCLC numbers into external identifiers (but still, one at a time), but it was the hyphens in the LCCN numbers that made me despair of doing that as a macro. Well, that and the "sometimes a hyphen, sometimes not". Chavey 12:12, 1 July 2017 (EDT)
If you are on Windows, AutoHotkey is very useful for semi-automating moving the external ids. You can bind the appropriate key presses (ex. copy selection, select line, delete, tab, select id type, tab, paste) to a single key. Doesn't matter what format the LCCN is, just need to select the 'correct' one. As for LCCN formats, I would prefer we don't try to standardize them, but keep allowing entry of the format shown in the publication. But that is a very minor preference. -- JLaTondre (talk) 08:33, 2 July 2017 (EDT)
I do not think that we are standardizing - I know I am not. What we are talking about is the case where the URL has the hyphen and the string does not and the far more often vice versa - which is a mismatch for automation. In such cases when moving, you need to make sure they are indeed just different formats as opposed to a typo in one of them( in which case you need to find out which is the correct one to leave in place). I go for the hyphen format in such cases or whichever is the correct ifone of them is a typo. But it is up to the moving editor preference when both are presented. Annie 18:21, 2 July 2017 (EDT)

(unindent) Those are done. It was mix of a few situations:

  • Different numbers on purpose from 2009/2010 (if I remember correctly, WorldCat was doing some reindexing back then - I've seen this kind of weird double links before). The one that points to the entry exactly is now left in place
  • Transposed numbers in one of the two numbers (don't people use copy/paste?)
  • Missing or additional number at the start or end of one of the two numbers.
  • Non-number string in the text part or additional parameters in the link part
  • Non-finished editing of a template (12345678 was a very popular one amongst the ones I cleaned
  • Two different numbers that have nothing to do with each other or with any OCLC that belongs to that book.

Now we have a funny python error in the empty report page. Did you forget to check of an array or a result is empty or something? :) Annie 17:12, 6 July 2017 (EDT)

Yup, all fixed now. Thanks for working on these! Ahasuerus 18:42, 6 July 2017 (EDT)

Title missing publications

This title is missing publications. Anyone know where it's pubs went? ···日本穣 · 投稿 · Talk to Nihonjoe 19:49, 30 June 2017 (EDT)

Most likely, it was ejected from the publication it was initially added in (because someone decided they do not need to record this) and as we do not have a report to find the orphan essays, it was not spotted before. It needs a deletion. :) Annie 19:52, 30 June 2017 (EDT)
Deleted. -- JLaTondre (talk) 09:17, 2 July 2017 (EDT)

"Publications with Invalid Prices"

The cleanup report "Publications with Invalid Prices" has been enhanced to include prices without a currency symbol. Ahasuerus 16:26, 1 July 2017 (EDT)

Proposal: modifying capitalization help

See discussion at R&S. --Vasha 11:05, 2 July 2017 (EDT)

OCLC URLs in Notes

As we discussed a few weeks ago, I have been trying to create a script to move OCLC numbers from publications notes to the external ID field. Unfortunately, I have found a number of odd permutations which make it hard to perform an automatic conversion. For example, consider this 1985 publication. The two lines that contain OCLC URLs are followed by additional lines related to the referenced OCLC records. The way my script is currently written, it would move the OCLC numbers to the External ID field and delete them from the Note field, which would create a mess. There are other patterns that make it difficult to automate the process without jeopardizing the integrity of the data.

On the plus side, our indefatigable editors have manually migrated close to 10,000 OCLC numbers to the new field over the last month+. At the rate we are going, the conversion process may be done by early 2018. Ahasuerus 16:53, 2 July 2017 (EDT)

If OCLC, or WorldCat (in all kinds of combinations of small and capital letters) are mentioned anywhere in the note except on the OCLC line itself, then it should not be touched, I agree. But we still have a lot of them where it is only on a line on its own, never referenced outside of the line... I know that text analysis of the full field to check if it is free to lift is time consuming but that will move a lot of OCLC numbers automatically Annie 18:29, 2 July 2017 (EDT)
Well, let's consider the publication that I linked above. I will copy and paste the relevant section of the Note field to make it easier:
How would the parsing logic know not to touch these 2 OCLC numbers even though there are no other references to OCLC or WorldCat next to them? A human would immediately realize that the lines that follow are related, but it's really hard for a computer program to make that type of determination. Ahasuerus 21:40, 2 July 2017 (EDT)
This record have a third OCLC and the word WorldCat under these two - which will mean "nope, human needed". And these 2 messages are faulty anyway - one of them is a bit wrong (they are the same number, different text). I can see the problem though - single OCLC but text on the next line for it.
Here is a new idea - OCLC as a last line on the notes? Annie 22:00, 2 July 2017 (EDT)
That's a good point. There are at least 14,661 notes with an OCLC link as the last line (except for the optional "/ul".) I'll see if I can take advantage of that. Ahasuerus 22:17, 2 July 2017 (EDT)
There are also lots of records where the last 2 lines are: OCLC link; then LCCN link. Those should also be easy to recognize and move. Chavey 02:27, 3 July 2017 (EDT)
It looks like any "pure OCLC link" line followed by a line that starts with a recognized pattern should be doable. Ahasuerus 10:19, 3 July 2017 (EDT)
... and then I find a publication note where the last line consists of an OCLC link and the next-to-last line reads 'OCLC <.....> Table of contents from <carriage return>". Sigh... Ahasuerus 14:15, 3 July 2017 (EDT)

Titles without Pubs report

Can we have this extended to include essays(for now) so we can start clearing those out of the DB? The ones that need to stay for one reason or another will need to get notes to explain so and ignored but there are a lot that need just old fashioned deleting. Thanks! Annie 18:33, 2 July 2017 (EDT)

Done. The data will become available tomorrow morning. Ahasuerus 21:51, 2 July 2017 (EDT)
Cleaned up, it was mostly essays by a few authors that seemed to have had a special dispense (granted by who, I don't know) like Shirley Jackson and Arlan Andrews (and also some quite unknown authors that probably used our system to beef up their bibliography), some leftovers from varianting, some purely electronic texts, some mismatched types (ESSAY for SHORTFICTION), some wrongly split NONFICTION books and some duplicates (under or not slightly different names). We can now try to clean the POEMs. Hauck 03:31, 3 July 2017 (EDT)
Great! POEMs will become available tomorrow morning. Ahasuerus 10:11, 3 July 2017 (EDT)
Done, note a concentration on very few authors (Hardy -see below-, Payack -a special dispense like Hardy?-, Baudelaire -problem with pseudonyming-, Poe -some working or "generic" titles or variants-) and a few leftovers. Ready to go to the next level (SHORTFICTION), perhaps in small batches as I won't try to guess the numbers involved. Hauck 03:58, 5 July 2017 (EDT)
Excellent! Actually, SHORTFICTION isn't too bad, only 1,300 records. I have tweaked the report and they will pop up tomorrow morning. Ahasuerus 13:54, 5 July 2017 (EDT)
A lot of the pub-less stories got nuked when I was assigning languages - because they were not in a publication, they did not get automatic language through any of the usual means so they got in the manual inspection and got deleted there. Thus the much lower number that one would expect (same happened with the poems and essays by the way).Annie 16:13, 5 July 2017 (EDT)

BNF cleanup report tweaked

The cleanup report "Publications with direct BNF links in Notes" has been tweaked to ignore links to Gallica and other non-catalogue parts of the BNF Web site. Ahasuerus 16:13, 3 July 2017 (EDT)

K. S. Hardy pub-less poems question

A huge amount of the poems in K. S. Hardy's account a pub-less (example: this one). What is our policy on that - do we allow them (if they were in printed books or ebooks and not solely online)? I can see that going both ways - if someone bothers to add the non-genre magazines/newspapers they were in, they are eligible (poems are fiction after all and we allow fiction from non-genre sources) but without the magazines, it looks weird. So what do we want to do:

  • Delete the lot of them
  • Add Magazine entries for the magazines/newspapers that contained them and add them as the only content
  • Leave them as they are. (my least preferred but if that is what we want to do, then I am fine with it).

Thanks for any opinions. PS: There are couple more poets like that (and I suspect we will find a few short fiction writers like that as well). Annie 15:40, 4 July 2017 (EDT)

If someone had added it and did not bother to add a note where they appeared, they are getting deleted - if someone cares enough, they need to write notes. But someone did here - thus the question. Annie 15:52, 4 July 2017 (EDT)
IIRC they were added by the author (or its representative, here). I'm going to delete them as our policy (or so it seems to me) was to add non-eligible (either because of their genre or their place of publication) texts at author's level and not as "proper" titles. For example, even if I've written a few essays about SF in non-genre publications, I won't try to include them in the db. I'm also wary of a recent tendency of "aspiring" poets (this is strictly without sarcasm) to massively enter their works in the db (note the length of this page.Hauck 02:39, 5 July 2017 (EDT)
Essays are clear. But if we have a short story in Nature (a non-genre magazine), we catalog these. As we will do any SF story in any non-genre magazine of any type. So I was wondering what we are doing for poems. I have no issue with deleting the lot of them. Annie 02:50, 5 July 2017 (EDT)
Yes, but in the case of Nature there is (in theory) a publication record that goes with the text, it's not the case here IMHO either because the poems are not spec-fic or the publication itself is not eligible. To go beyond this case, the underlying question is "Are POEMs to be treated like SHORTFICTIONs?". This could easily make a few pages at R&S. (For the record, I'd say "no" only on practical grounds as I fear that there are a looooooooot of poems in looooot of "obscure" supports waiting to be entered). Hauck 02:57, 5 July 2017 (EDT)
Another point is that I found the practice of some "self-centered" contributors to enter only their own texts (either as a stub as in this case or as the sole content of the publication like here as I've seen numerous times) instead of taking the time to enter the whole publication quite contrary to our spirit and aims. In Hardy's case, it's quite interesting as some titles were entered as stubs and were (before I deleted them) given as being published in this magazine or that one, BUT the contributor simply didn't take the pain to add the issue. This is just lazyness or self-publicity but it's not acceptable bibliographic work. Hauck 03:02, 5 July 2017 (EDT)
Maybe they did not know that they should because a boatload of those were accepted like that. But then - people will be people, there will always be people with different agenda in a community project... When I end up with one of those incomplete ones, I just edit and add the rest if I can find it. Maybe that will show a new contributor what is the expected behavior. And even if it does not, we do not have a half-way entered publication.
PS: Agree on poems - unless if they are in a genre publication, they are out in my book. Annie 03:31, 5 July 2017 (EDT)

Add AWards From the SFPA

The Science Fiction and Fantasy Poetry association holds an annual contest as well as offering yearly awards for speculative poetry published the previous year: the Rhysling Awards for individual poems, the Dwarf Stars Award for short-short poems, and the Elgin Awards for genre poetry books and chapbooks.

I would like to have the SFPA award, Dwarf Stars and Elgin added to the Awards list —The preceding unsigned comment was added by Akua (talkcontribs) . on 23:05, 4 July 2017 (EDT)

The Dwarf Stars Award and the Elgin Awards look solid and I would have no problem adding them. However, I am not sure what "the SFPA award" means in this context. We already support the Rhysling Award and I am unaware of any other SFPA-administered awarwds. Could you please clarify? Ahasuerus 20:32, 6 July 2017 (EDT)
The SFPA holds an annual contest for sf/f poetry. "The 2017 SFPA speculative poetry contest is open to all poets, including non-SFPA-members. Prizes will be awarded for best poem in 3 categories: Dwarf (poems 1–10 lines [prose poems 0–100 words]); Short (11–49 lines [prose poems 101–499 words]); Long (50 lines and more [prose 500 words and up]). Line count does not include title or stanza breaks. All sub-genres of speculative poetry allowed in any form. Entries will be read blind"Akua 22:10, 6 July 2017 (EDT)Akua
OK, I have added new award types for Dwarf Stars Award and Elgin Award.
As far as the "SFPA Poetry Contest" goes, I am not sure it counts as an award in the regular sense of the word. Its Web page says that it "is intended to raise funds for SFPA, as well as to draw more attention to speculative poets and reward writers of good speculative poems". All poems must be unpublished and and it costs $2 to enter a poem. Ahasuerus 15:21, 10 July 2017 (EDT)

Stephen Crane and the supernatural

On this page, a scholar answers a question about which of Stephen Crane's stories have supernatural elements by listing just three, one of which he considers to have merely "overtones of the supernatural." Do you think this is sufficient evidence to remove Crane's other stories from the database? --Vasha 18:46, 5 July 2017 (EDT)

Ashley and Contento's The Supernatural Index lists 10 stories. I wouldn't recommend deleting any of those. --Ron ~ RtraceTalk 21:08, 5 July 2017 (EDT)
The thing is, Ashley & Contento tend to indiscriminately list the complete contents of a book of "Tales of Terror and the Supernatural" -- because they can't read everything, that's fair enough. I would recommend in that case adding a synopsis and a nongenre mark and a note saying that Ashley & Contento list it even though it's not supernatural. Vasha 21:24, 5 July 2017 (EDT)
If someone reads them and write synopsis for each (or find some) and determines that they are non-genre, a flag and an explanation will be a good idea. I would say that without a reading, it is just one source against another - we need a real human :) Deleting them will cause issues downstream - someone somewhere will decide to add them again. So staying in sync with Ashley & Contento plus notes is probably the best way to go. Annie 21:48, 5 July 2017 (EDT)
Sure, that's fine. Vasha 22:20, 5 July 2017 (EDT)

Stephen Crane wrote a number of parodies of ghost stories in which an apparently supernatural phenomenon turns out to be something ridiculously mundane -- for example "A Ghoul's Accountant" and "The Black Dog." I believe these should be marked non-genre, right? --Vasha 21:28, 6 July 2017 (EDT)

Linking to Smashwords images

Please note that we can now link to Smashwords images. The syntax is similar to what we do for SFE3-hosted images -- see Template:Image Host Sites for details. Ahasuerus 18:42, 7 July 2017 (EDT)

Cool. ···日本穣 · 投稿 · Talk to Nihonjoe 18:50, 7 July 2017 (EDT)

The Whole Science Fiction Data Base 8 question

Hello, I've seen a reference to this: 'The Whole Science Fiction Data Base 8' stating that it provides more detail about a specific book ( Can anyone please tell me what this 'Data Base 8' is and how to access it?

Thanks in advance!—The preceding unsigned comment was added by Voodoomailman (talkcontribs) .

Answered the duplicate post in the Help Desk. --Ron ~ RtraceTalk 22:42, 8 July 2017 (EDT)

LCCN double formats, PV'd publications and the current cleanup project

LCCN have two valid formats: year-number and year(some number of zeroes)number. Both are valid, both can be used for search and linking. In the current LCCN cleanup project, my understanding is that we are moving "as is" - if the editors preferred the 91-38127 format to be visible, this is what we put in the new External ID field - we are not trying to standardize at all (by using the 91038127 format always in the External ID and leaving the 91-38127 in the notes); if they preferred 91038127, that is what we are putting in the field. Especially on PV'd publications. There are a few cases really:

  • Link is pointing to 91038127, the visible part is 91-38127. The new field should be 91-38127 (link ends up in the same place; the visual part is preserved).
  • No link at all, just a string 91-38127, same as the above (actually in this case, I would say that using the 91038127 format is unreasonable.
  • If the visible part is already 91038127, we follow the same principle.

Leaving the link in the notes clutters them - if it is already moved down, why keep the note at all unless if it is tied to more details and needs to stay for clarification? Any opinions on that? Annie 18:58, 10 July 2017 (EDT)

I'm all for removing the note. A difference between LCCNs and the other external identifiers is that LCCNs are usually printed in the books. As an editor/verifier, I'd prefer to have the number formatted as it appears in the book. Assuming editors are likely to have entered them that way, preserving any existing visual formatting appeals to me. --MartyD 21:16, 10 July 2017 (EDT)
Me too (tm) Ahasuerus 21:56, 10 July 2017 (EDT)
There are cases (e.g. The Clocks of Iraz) where a LCCN number is quoted in a book that has no bearing to the present records of the Library of Congress, thus offering no valid link upon entering. I think all links should be checked, and the actual number must be used for linking, as was the case with the old method of entering LCCN numbers in the notes.--Dirk P Broer 21:33, 10 July 2017 (EDT)
We are not talking about the cases where there is a difference between the printed value and the actual one in more than format. Yes, all links need checking - and if it is plainly wrong, the note should explain the situation. But if the LCCN is printed as 62-123, I think this is what needs to be in the External Identifier field - as long as it is the valid one. I strongly disagree that we should enforce 91038127 in the link in such cases - this will make it appear as the value as well as the link now and that is the difference from those old links.
We may have an alternative option - have the external indentifier link be changed when creating the link - doing the conversion on our side (so it emulates the current most prevalent links). But this will be an overkill - LOC already does that. Annie 21:41, 10 July 2017 (EDT)

Similar series titles

This series has a notice at the top that another series has the same name. However, their names are not the same. In fact, the second series is a subseries of the first. Is there a way to suppress these notices on specific series? ···日本穣 · 投稿 · Talk to Nihonjoe 13:39, 12 July 2017 (EDT)

I submitted a series name change for the second one that may fix this. ···日本穣 · 投稿 · Talk to Nihonjoe 13:43, 12 July 2017 (EDT)
Yep - instead of brackets (), use another separator in the series name. They have the notice as the same because the brackets syntax is considered disambiguation. The change helped (different names outside of any brackets). All good nowAnnie 13:46, 12 July 2017 (EDT)
Cool. Any way to fix the sorting in the main series? It currently has them listed 2, 1, 3, 4 instead of 1, 2, 3, 4. ···日本穣 · 投稿 · Talk to Nihonjoe 16:40, 12 July 2017 (EDT)
The problem is that there are two different elements here - the books directly into the series (2) and the subseries entries (the rest). They sort separately and I am not aware of any way to sort them together - it sorts first the books into the series, then the subseries. So... one option is to put the single one that is in position 2 now into its own "main" series and put that main series as #2. Or you can wait to see if someone else has a better idea. Annie 16:49, 12 July 2017 (EDT)
Guess we should wait for Ahasuerus to wander by. :) ···日本穣 · 投稿 · Talk to Nihonjoe 18:08, 12 July 2017 (EDT)
Hmm...found an art book for the series, so I submitted that and changed the series to レア・ガルフォース to match the art book. That should fix it once approved and the new series is placed as #2 in the main series. ···日本穣 · 投稿 · Talk to Nihonjoe 18:17, 12 July 2017 (EDT)
Approved everything you had in the queue - and added the numbering (I think). let me know if this is what we were trying to achieve :) :) Annie 18:24, 12 July 2017 (EDT)
Looks good. Thanks. ···日本穣 · 投稿 · Talk to Nihonjoe 18:28, 12 July 2017 (EDT)

Here's a fun one to sort

Somehow, things got mixed up with this volume:

Volume one should have the following contents:

Instead, the first entry is for this omnibus (which isn't actually a real book). I'm not sure how it happened, and I'm not sure how to fix it since there's an omnibus within an omnibus. ···日本穣 · 投稿 · Talk to Nihonjoe 18:42, 12 July 2017 (EDT)

Use 'Remove Titles From This Pub' and remove the 'bad' omnibus . Once approved, you can delete the omnibus record (since it doesn't seem to be a valid publication) and import the missing novel's title record into the 'good' omnibus. -- JLaTondre (talk) 18:50, 12 July 2017 (EDT)
Submitted ···日本穣 · 投稿 · Talk to Nihonjoe 19:05, 12 July 2017 (EDT)
That's the wrong one -- you want to kick out the 銀河無責任時代 omnibus, not the one that is the same name as the publication :) Annie 19:07, 12 July 2017 (EDT)
Fixed. ···日本穣 · 投稿 · Talk to Nihonjoe 19:11, 12 July 2017 (EDT)
here is your stray title. Do you want to delete it? Seems like that series is missing its #1 now? Annie 19:13, 12 July 2017 (EDT)
Submitted the deletion and the addition of the stray title (I knew about it, since that's how I noticed something weird was up). ···日本穣 · 投稿 · Talk to Nihonjoe 19:14, 12 July 2017 (EDT)
Yeah, you were moving faster than I was typing above :) All looks good now I think? Annie 19:15, 12 July 2017 (EDT)
Almost. I submitted a sort order fix, too, for both in the series. ···日本穣 · 投稿 · Talk to Nihonjoe 19:16, 12 July 2017 (EDT)
Got them less than a minute after you sent them - probably while you were typing. :) Now all should look fine. Annie 19:25, 12 July 2017 (EDT)
Thanks for helping to sort it out. :) ···日本穣 · 投稿 · Talk to Nihonjoe 19:52, 12 July 2017 (EDT)
I remember the anime, but I had no idea that the (very) irresponsible Captain Tylor was so popular! Ahasuerus 20:25, 12 July 2017 (EDT)
And there's more, too. I still have 4 more novel series to add to the pile. ···日本穣 · 投稿 · Talk to Nihonjoe 20:47, 12 July 2017 (EDT)
Also, I wasn't aware there were this many. I knew the anime was based on some novels, but I'd never looked into it before. ···日本穣 · 投稿 · Talk to Nihonjoe 14:52, 13 July 2017 (EDT)
You never know what you may find once you start digging. Back when I started working on the database, I had no idea that so many Kenneth Bulmer books were only available in German (at the time.) Ahasuerus 16:00, 13 July 2017 (EDT)

Jonathan Strange & Mr Norrell

Based on the following links ( WorldCat, LOC, Locus) and other references, the correct canonical title for this title should be "Jonathan Strange & Mr Norrell". The author's intent based on interviews is to use the punctuation of the time, and so the title should not include a period after 'Mr'. I can correct the publication entries for which I am PV, and suggest modifications to other PVers, but I think we should correct the Title Name as well. What's the recommended procedure for this?

WorldCat LOC Locus

—The preceding unsigned comment was added by Taweiss (talkcontribs) . 19:46, 12 July 2017 (EDT)

Changed. Stonecreek 23:42, 12 July 2017 (EDT)

Nominating user Nihonjoe for moderator

See Moderator Qualifications#Becoming a moderator for the nomination process.

I would like to nominate user Nihonjoe (talkcontribs) for moderator. He has been with us for a number of years now, mainly working on Japanese entries and the Wiki side of the database. His talkpage shows excelent communication skills, and hardly any problems with his submissions. I think he meets the Moderator Qualifications and he has accepted the nomination. His language skills are an enormous asset.


  1. Support, as nominator. --Willem 02:10, 17 July 2017 (EDT)
  2. Support, sorry to be late to the party, and discussion below notwithstanding. I don't share those concerns. I've found him to be a good communicator and his submissions and understanding of ISFDB policies and procedures to be solid. I also think beefing up the moderator corps' Asian language capabilities would be an added benefit. --MartyD 21:11, 20 July 2017 (EDT)
  3. Support, Nihonjoe is careful and all of his submissions that I have moderated have been uniformly good.Rkihara 23:06, 21 July 2017 (EDT)
  4. Support, with a small note - I had never seen him submitting an English language item (except when it was needed for a variant for a Japanese one or something like that.) Which does not disqualify him (and I had not been around long enough to know the complete history) but the conventions around capitalization, names normalizations, same-named publishers and so on can be a bit tricky (a lot trickier than they are in any of the other languages) and need a lot of attention when moderating. But the easiest way to learn those is either to edit a lot in that language or start helping people do them properly - and the first one is unrealistic for someone whose interests are in other languages. So when/if he becomes a moderator, I hope he will be careful around English-language submissions. :) Annie 13:06, 22 July 2017 (EDT)
  5. Support, I'm about 65/35 split [positive/neutral] and only because I have not moderated enough submissions for Nihonjoe. I think most were Variants, so the original submissions were already done [while I'm comfortable with french, German, Spanish or Italian I have no experience with cyrillic or pictographic languages]. But that's my lack, not his. I went through past pages of commentary and I will say there is a definite involvement at all levels. That's important. I don't feel the 'lack' of submissions is crucial [I certainly had far fewer when I became a Moderator, but the database was much simpler then - I actually knew what I didn't know]. What one submits also doesn't need to cover everything [I've never done one involving an Award - so maybe both my and Willem's flags should be lowered to half-mast]. I agree with Marty that more language coverage is a plus at the Mod level. And I haven't seen where Nihonjoe has been the slightest bit reluctant to ask if there's something he's unsure of. That's how I learned. --~ Bill, Bluesman 15:29, 22 July 2017 (EDT)
  6. Support, despite a moderate amount of submissions, what I moderated was usually the result of serious, careful work. It being often Asian-centered is no problem, on the contrary : competence in this field seems to be an asset for this db. Linguist 08:16, 23 July 2017 (EDT).


  1. Oppose Not for now, as Nihonjoe has (IMHO of course) not enough contributions (7.000). In the present state of the db, 10.000 contributions is a bare minimum (15.000 being the correct number to have learned most of the ropes), the last two accepted moderators were above the 20.000 mark. There is also, at this time, no shortage of moderators. Hauck 02:18, 17 July 2017 (EDT)

Comments/ Neutral

  1. Neutral. Since I was asked to comment, I will post this versus not responding. I would prefer that someone becoming a moderator be familiar enough with the tools to handle this situation (discussion two topics above). However, I don't oppose the nomination. Nihonjoe's work has been good & I highly doubt he will cause any damage. If others are fine with him picking the rest up 'on the job', I'm comfortable that he will remain conscientious and so don't have an objection. -- JLaTondre (talk) 21:03, 23 July 2017 (EDT)

  • Regarding Hauck's comment, there are a couple of moderators who have fewer contributions than me per the list: Albinoflea (~2800), Tpi (~4000). Another has just a few more than me: Kpulliam (~8500). So these seemingly-arbitrary "bare minimum" or "correct number" of contributions seems to be just that: arbitrary. As for the other argument (no shortage), there is no mention anywhere that I can find of a cap on the number of them (or anything similar). ···日本穣 · 投稿 · Talk to Nihonjoe 12:06, 17 July 2017 (EDT)
What don't you understand in the acronym IMHO? Of course it's arbitrary, it's my opinion. As for Albinoflea, I've exactly had the same position (not enough contributions), as for tpi he's with us since 2008 and was made moderator more than seven years ago in different times. On the relationship level, this immediate outburst just confirms my opinion that you're not ready for the job. Hauck 12:50, 17 July 2017 (EDT)
I apologize if you feel my comments were an "outburst", but I wasn't aware I wasn't allowed to make comments. If that's the case, I'll certainly refrain. I understand the IMHO acronym, too, as I've been online since the glory days of IRC way back in 1990. I'm fine if you don't think I'm ready for the job. I was merely pointing out that the numbers you gave seemed arbitrary (to me, even though I didn't use "IMHO" when I made the comment). You are certainly welcome to have your own opinion about anything, and I wasn't trying to imply that you couldn't. Apologies if it sounded that way. ···日本穣 · 投稿 · Talk to Nihonjoe 14:46, 17 July 2017 (EDT)
A little bit of background. Back when ISFDB 2.0 was open for alpha testing in 2006, all testers were automatically made moderators because we were still testing the software. Once the beta phase started in December 2006, many original testers slowly became inactive. Eventually their moderator flags were removed. Some even asked to have them removed because they felt they were no longer on top of things.
Over the following few years a more formal process -- formal requirements, nominations, voting, etc -- was developed. We also added a "demoderatorization" policy for inactive moderators.
As far as the number of contributions goes, I think it's a useful gauge, although necessarily a rough one. As the size and complexity of the database grows, moderators need to be aware of more and more areas which didn't exist in 2006: awards, publication series, external IDs, cloning, cleanup reports, links to third party web pages, various title flags, languages, chapboooks, user preferences, etc. It's possible for an active editor to be very familiar with one part of the database editing process and be a rookie in other areas.
When reviewing the edits made by a potential moderator, I find it useful to look not only at the total number of edits, but also at the breakdown by submission type, which can be found here. In this case, I see that Nihonjoe has created 432 New Award submissions, which suggests that he is probably familiar with awards. Similarly, he has created 318 Title Merge submissions and 765 Make Variant Title submissions. On the other hand, he has created fewer than 10 Title Unmerge submissions, a tricky area.
One thing that we have found useful in the past is identifying the areas which a prospective moderator hasn't been exposed to and having a more experienced moderator work with him to make sure that he or she becomes familiar with them before we turn the moderator flag on. Perhaps it's something that we could consider in this case. Ahasuerus 13:54, 17 July 2017 (EDT)
I'm sure there are far more areas where I haven't done much, too. I don't think a moderator has to be intimately familiar with every area before becoming one, but rather willing to learn and willing to ask questions if they don't know what should be done in a given area. ···日本穣 · 投稿 · Talk to Nihonjoe 14:46, 17 July 2017 (EDT)
True, but there are times when, as the saying goes, "we don't know what we don't know". There have been times when I thought that I knew what the policy was and made edits based on my understanding only to find out later that my understanding was incorrect or out of date. On the other hand, as the system grows, it becomes harder to keep track of all of our features and policies; if we required perfection, we would be left with no moderators. Kind of a tricky balance... Ahasuerus 14:57, 17 July 2017 (EDT)
However, if that's the general consensus, I'm happy to live with it. I wasn't looking to be made one, but Willem said he thought I could do the job, so I indicated that I was fine if he wanted to nominate me. Whatever the decision here, I plan to continue contributing as I can (which is all I've ever done). ···日本穣 · 投稿 · Talk to Nihonjoe 14:46, 17 July 2017 (EDT)
I find that this too is a matter of balance. On the one hand, we obviously do not want new editors to be able to approve their own edits -- even the most well-meaning newbie is liable to make mistakes early on. On the other hand, having moderators approve every trivial change made by a veteran editor is not a very good way to spend out limited resources either: it makes both the veteran editor and the approving moderators less productive. Making qualified editors moderators is in everyone's interests, we just need to make sure that they have been properly "seasoned" :-) Ahasuerus 15:09, 17 July 2017 (EDT)
If people want to work with me in specific areas in which they think I need more experience, I'm open to it. I tend to just work on whatever interests me (though I do have a very large list of projects which seems to never get smaller). So, feel free to pass the salt and pepper and make sure I'm seasoned. ;) ···日本穣 · 投稿 · Talk to Nihonjoe 15:13, 17 July 2017 (EDT)
I think this discussion is going nowhere. Sorry to put you through this Nihonjoe. Perhaps my moderator flag should be removed because I don't do anything in the awards section? --Willem 15:22, 17 July 2017 (EDT)
No worries. I've been through far worse in similar discussions. This one barely registers. I'm fine it continuing if people want to continue it. Either way. ···日本穣 · 投稿 · Talk to Nihonjoe 16:03, 17 July 2017 (EDT)

Request for additional comments

It's been 5 days since the nomination. Normally (and as per the standard process), it would be enough to determine if we have consensus. However, only 4 votes have been cast so far: 3 in favor and 1 against. I worked with Nihonjoe last year, when he was learning the ropes, but I have processed only 14 of his submissions in 2017, so I hesitate to cast a vote.

I have compiled a list of moderators who have approved more than 50 of Nihonjoe's submissions in 2017 and who haven't voted. I will ask them to chime in based on their experience. Unfortunately, one of them is on vacation. Ahasuerus 11:02, 22 July 2017 (EDT)


The final tally is 6 in favor, 1 opposed and 1 neutral, so I will go ahead and set the moderator flag on Nihonjoe's account. Based on the comments about his limited exposure to non-Asian publications, I will encourage Nihonjoe to concentrate on self-approvals while he gets more experience in other areas. Ahasuerus 16:07, 24 July 2017 (EDT)

I'll keep Hauck's comments in mind. As for non-Asian contributions, I've submitted quite a few. More recently, I've been focusing on a few long Japanese series (one of the series had around 160 volumes in it!), so most of my recent contributions have been related to those. Thanks to everyone who participated. ···日本穣 · 投稿 · Talk to Nihonjoe 19:56, 24 July 2017 (EDT)
Congratulations from me also... don't be afraid to ask questions, and the beers are in the fridge. :) PeteYoung 06:24, 25 July 2017 (EDT)
As long as some of them are root beers since I don't drink. :) ···日本穣 · 投稿 · Talk to Nihonjoe 10:48, 25 July 2017 (EDT)

Content field for omnibus variants

The way we variant ombinuses is that we do not variant if the content is not exactly the same. At the same time, if you want to see the content of a variant, you need to have it added on the variant itself. Is there a reason why this field is not treated as the Series and SeriesNum fields and to be greyed out on the variant and inherited from the parent? Annie 18:10, 19 July 2017 (EDT)

Let me think about it... How about this scenario: the parent is [/1,2] while the variant is [/1,2+ss]? Is it a legitimate parent-variant pair or should their VT link be broken? Ahasuerus 20:56, 19 July 2017 (EDT)
Well, I do not like to variant different content - if it is a variant, it should be the same content (abridged, translated and whatsnot make a variant, split volumes (these I still do not like but oh well); different fiction elements altogether make new title). Otherwise where do we draw the line? At 1 story, 2 stories? 1 novel? If we take that wiggling to the extreme, we should just variant all omnibuses that contain the same 2 novels - after all the fact that one of them have 1 more novel and the other 2 more novellas and the third 4 more novellas is ignored if we say that 1-2 and 1-2+ss is the same. Or if we decide we will accept that, we need to draw the line somewhere firmly (and to keep in mind that if 1-2+ss is varianted under 1-2 but that will also mean that 1-2+2ss needs to go under 1-2+ss which will mean under 1-2 as well.
Of course if the community disagrees, I will just shut up and go do the variants the way we agree we should (and I had never seen your hypothetical situation so never thought on that - I was just cleaning series from variants and saw it on the omnibus and decided to ask. Annie 21:11, 19 July 2017 (EDT)
I only variant those with the exact same content. Otherwise, it's a different publication. ···日本穣 · 投稿 · Talk to Nihonjoe 00:08, 20 July 2017 (EDT)
I do this on a case by case basis (that is a really precise definition!). When the "bulk" is kept, especially in translation, I make the variant, for example in Ahasuerus' scenario for a translated work, I'd make the variant. As for the content of the field proper, I rarely use it as some title series are quite fluid and what is true at one moment may not be so in the future. Hauck 02:09, 20 July 2017 (EDT)

Award year listings not working for Gemmell Award

Examples: 2016, 2015, 2010. If you view them by category, it works fine: Legend, Morningstar, Ravenheart. ···日本穣 · 投稿 · Talk to Nihonjoe 15:54, 20 July 2017 (EDT)

It is the dates - someone had added them as 2010-06-00 (for example - the one I looked at) and not as 2010-00-00. I fixed one so now one shows under 2010. I will wait for Ahasuerus to see this and confirm before I fix the rest. Annie 16:11, 20 July 2017 (EDT)
Just looked at the help page: "Enter the award year using the YYYY or the YYYY-00-00 format" :) So at least the help page is correct. Maybe we need a check added to make sure that the page does not allow the month and date to be added. Annie 16:14, 20 July 2017 (EDT)
Probably easiest would be to have it be a YYYY-only field and add the "-00-00" on the back end if needed to make the database happy. ···日本穣 · 投稿 · Talk to Nihonjoe 16:45, 20 July 2017 (EDT)
It will do it if you just add 2017 :) The field accepts two formats: YYYY and YYYY-00-00 and stays valid. Because it is called year, I guess the assumption is that people will just add the year so no additional checks were done. I think the validation is the same as a date field but the help page for them is different. Annie 16:48, 20 July 2017 (EDT)
PS: That is why I said that I want to wait for Ahasuerus to chime in :) Annie 16:50, 20 July 2017 (EDT)
Yup. Another idea: just have the database strip out "-MM-DD" for the awards and replace them with "-00-00" when submitted. Might be quicker than changing the form and such. ···日本穣 · 投稿 · Talk to Nihonjoe 18:20, 20 July 2017 (EDT)
Thanks for identifying the problem! I am kind of under the weather at the moment, but I will take a closer look as soon as I get better. Ahasuerus 20:33, 20 July 2017 (EDT)
Get better! This can wait. Annie 20:41, 20 July 2017 (EDT)
Yup, definitely hope you feel better soon. ···日本穣 · 投稿 · Talk to Nihonjoe 01:42, 21 July 2017 (EDT)
Thanks, folks, I am feeling better now. Not at 100% yet, but "mostly functional" (or at least that's what Fixer tells me.)
I have changed the affected Web page to use the award year only, so values like 2010-06-00 should no longer be a problem. While I was at it, I fixed a couple of other bugs. They only affected processing improperly formed award URLs, so fixing them should have no impact on most users. Please let me know if you run into any issues. Ahasuerus 15:44, 21 July 2017 (EDT)
So what is the policy now - dates or year only? If we officially allow dates, we need the help page changed. I know that it will work for dates now but that does not make it correct according to the current rules - thus me asking. Annie 15:50, 21 July 2017 (EDT)
I guess the first question that we need to answer is whether we plan to support awards that are not given annually. For example, Writers and Illustrators of the Future gives its awards quarterly. It's really a contest for amateur writers rather than an award:
but there may be other awards that are given more than once a year. If we were to decide to add them, we would need to redesign our software because it currently assumes that all awards are annual. (Gaps are allowed, of course.)
For now, I think we should keep the current policy, i.e. "YYYY or YYYY-00-00", for award dates. Once we clean up the data (there are 90 award records with months), I can add another layer of validation to prevent editors from entering YYYY-MM-00 and YYYY-MM-DD values. Ahasuerus 16:10, 21 July 2017 (EDT)
I agree. :) I've corrected any that I was approving last few days. Can you get a list of the affected ones so we can fix them? Annie 16:25, 21 July 2017 (EDT)
I have coded and deployed a cleanup report to find them. The data will become available tomorrow morning. I'll go ahead and create an FR to add that extra layer of validation that we discussed earlier. Ahasuerus 17:53, 21 July 2017 (EDT)
There is another example of 2 sets of awards in one year with the 2010 Grand Prix de l'Imaginaire. My assumption is that because their eligibility year was drifting, they awarded a second set of awards in 2010 to harmonize the eligibility year with the calendar year. I originally tried adding those with different months, but quickly discovered the problem seen with the Gemmells. I ended up adding both sets under 2010 with notes on the award record explaining which set it belonged to. I also added an explanation on the award type record. It might be a nice idea to have a note field for the award year where these sorts of things could be explained closer to where they appear. We could also note eligibility periods and time and place of the award ceremony. Just an idea. --Ron ~ RtraceTalk 18:52, 21 July 2017 (EDT)
An interesting example; thanks. I agree that there would be value to adding an "award year note". Unfortunately, our current database layout doesn't lend itself to this addition. Notes are associated with records. At this time the only awards-related records are "awards", "award types" and "award categories". We would need to create a new record type -- something like "award year" or "award ceremony" -- in order to add notes.
After sleeping on it, I think it may be a viable idea. We already need to create an award type and an award category before we enter an award. If we were to add a new record type for "award years/ceremonies", it would be similar to award categories and would be available as a drop-down list on New/Edit Award pages. You would have to define an "award year" before you could enter any awards for it and there would be a note explaining where and when the awards were announced and/or presented. This functionality would also allow us to support non-annual awards like "2015 3rd quarter". Food for thought... Ahasuerus 09:28, 22 July 2017 (EDT)

(unindent) I have created FR 1086 to document what has been discussed here. It's not a high priority; I expect that we will discuss it when it's time to talk about Roadmap 2018 . Ahasuerus 12:45, 24 July 2017 (EDT)

Sounds good. Thanks! ···日本穣 · 投稿 · Talk to Nihonjoe 13:29, 24 July 2017 (EDT)

"Note to Moderator" field name change

Hello, I find Annie's idea to use the "Note to Moderator" field to "notify" PVs really great. I'm starting to use it intensively as does Annie (of course) and Bill. This may be the long-searched-for way to simplify our notification procedures as this field is accessible anytime in the "My Changed Primary Verifications" link. Of course this way of doing things should be just used for "simple" updates or additions (The process for more problematic cases should stay as it is via the Talk Pages). To make things even clearer, perhaps should the field be renamed along the lines of "Note to Moderator / Notification to PVs" (or something else). Any thoughts on the matter? Hauck 02:56, 21 July 2017 (EDT)

One of the issues is that if there is more than one update in a day or since the last time you opened your Changed verifications, the previous notes get lost - each pub is on the list just once (I believe so anyway - maybe adding all updates to the list there is the way to go?). They are in the submission itself but not in the quick link under "Changed verifications" so finding them is not that easy). When I know I am multi-editing (typos usually), I make sure I repeat both but if I do not know about the change (or it was long enough ago to forget), the lost ones are getting lost. And yes - anything that is not moving fields around, format issues and so on should still be following the old process.
Nope, multiple edits show up on the list, and each change is different. I can't say if Mod notes stay as the ones I just checked on my list didn't have any. --~ Bill, Bluesman 15:42, 22 July 2017 (EDT)
Another concern is that these are supposed to be just for the moderators (so people may add personal information there) - even though they were never really hidden. A rename of the field just for the publications (how often do we get personal information on changed publications? Authors edits - maybe, but on publications, not that much. Plus if we do see one, a rejection WILL keep it safe (I am pretty sure that only moderators and the editor that submitted the original have access to rejected submissions or at least it is not that easy to find them) and the moderator can reapply the change and add a proper note) Annie 03:26, 21 July 2017 (EDT)
So split in two fields? Hauck 03:32, 21 July 2017 (EDT)
Also an option :) But we have editors that put notes to the moderators into the Notes or Synopsis fields (and notes for the edition into the moderators notes) :) So technically, no matter what is done, we will have things that are just in the wrong place anyway. And I am really not that concerned about these on publications - most updates on verified publications won't have personal data. It's just that we tell people this field is safe - so we need changes in that. As long as people know that the field is visible, we should be fine with one field I think.
But I do agree, it is very useful for cleanup operations (and cuts down on the "what did she change now?" questions I hope) which is where I use it mostly (bigger changes get the note AND a post on the Talk page). As long as the multiple edits issue can be resolved, that may be a very effective tool indeed (not that there will be a lot of multiple edits in the same publications very often but corner cases tend to snarl designs and it is a valid case). :) Annie 03:47, 21 July 2017 (EDT)
Extremely useful, for all minor changes. I've actually re-submitted sometimes because I forgot to add a note [by rejecting the non-note version only one shows up on the other PVs page[s] ]. If we make it more visible/renamed/whatever more people will use it. --~ Bill, Bluesman 15:42, 22 July 2017 (EDT)
I seem to have missed the original suggestion / discussion. Searching on "Note to Moderator" turned up nothing and I'm not looking through everything Annie has submitted. Any chance of a link? Thanks. Doug H 11:46, 22 July 2017 (EDT)
I think that the seed of the idea is here. Hauck 12:43, 22 July 2017 (EDT)
There was never a discussion per se. I used to use the moderator notes a lot before I was a moderator to remind myself what I did or what else needs to still be done or to show the moderators what I am changing for a small change (it is very annoying to try to figure out the change in a 6 lines of notes when visually you cannot spot it:) ). Then I realized that I can see the moderator note when I click on the changed verification or in an item in Recent Edits (before I was a moderator) so figured that it won't hurt to add there a summary of what I am changing during my cleanup -- too small to notify (moving lccn or adding a missing tag or changing to a template or adding a transliteration or changing a capital/small letter in a story - that kind of thing). So that is the backstory - a side unintended effect of a feature basically that is very useful on its own. Annie 12:53, 22 July 2017 (EDT)

(unindent) It looks like there are three overlapping issues here:

  • Add a new field to allow entering "Change Summaries". We may want to add this field to all types of submissions, not just EditPubs, since the functionality can be useful in a variety of cases.
  • "Create a history of changes to primary-verified publications by storing a snapshot of the way each verified pub looked like right before it was changed" as per Roadmap 2017. Unlike the proposed "Change Summary" field, the data will be saved automatically at the time the submission is approved. Even if the submitter doesn't enter anything in the "Change Summary" field, the data will be captured. The display software will be able to show two versions of the record side by side with differences highlighted.
  • Ensure the privacy of the "Note to Moderator" field since it can contain private information like e-mail addresses.

If this functionality is deemed higher priority than other outstanding items listed on the Roadmap 2017 page, I can reshuffle the development queue accordingly. Ahasuerus 16:28, 23 July 2017 (EDT)

I do not think that the first two are opposites - if you just change a letter somewhere in the notes (or add a tag), finding it even when you have both versions takes a lot of time and effort. So having a change summary in that case is very useful. Annie 14:13, 24 July 2017 (EDT)
I agree that they are complementary. Changing the software to highlight the differences will help, but a Change Summary field can do a lot more to explain the reasons for the change. Ahasuerus 16:23, 24 July 2017 (EDT)
Something else to think about: Changes in contained titles. We have the publications changes but if something is changed on the titles level (being it the title itself or content), there is no notification. And these are the ones that can really mess things up -- changing a story to a novella when you know it is not true or merging the translations of two stories when they need to stay separate. Any plans to start adding those to the changed primary verifications as well (and/or make them more available)? Annie 16:39, 24 July 2017 (EDT)
As I recall, I briefly considered the issue of EditTitle submissions affecting verified pubs when I was implementing the Changed Primary Verifications report. I agree that it would be desirable to add EditTitle submissions to this report. Title merges may be trickier; I'll have to think about them. Ahasuerus 17:55, 24 July 2017 (EDT)
I think that just highlighting that there had been a merge will be enough if someone wants to look into it - with a surviving note (because the current Recent Edit entry does not show the moderator notes for Merges and Deletes I think. Merges are mostly important in two cases: pseudonyms and translations. In most other cases, they are usually because someone added a book and typed the content instead of importing. On the other hand, this will make the changes page very busy so having two separate levels (show pubs only and show all titles) may be desirable. Just thinking aloud here - may be overthinking some of that. Annie 18:36, 24 July 2017 (EDT)
However - in both cases, we need to think about multiple changes - either by adding ALL changes at the Changed Primary Verifications or by making a list of "here are all changes, click away". Otherwise we are running the risk of me changing substantially the format (fixing links for example) and then someone updating a small to a big letter - leaving both my summary and changed record away from the eyes of the PV. Annie 14:13, 24 July 2017 (EDT)
As Bluesman pointed out earlier, the Changed Primary Verifications page already displays all changes, including multiple changes per day. If you have come across cases where a change to a verified publication didn't appear on the primary verifier's Changed Primary Verifications page, it was a bug and I'd love to hear the gory details! Ahasuerus 16:27, 24 July 2017 (EDT)
I need to remember what the issue was. Maybe if one of the "changers" was the PV? Will post back when I remember what that was all about (and because of that my mind somehow decided that we never have the duplicates. Looking at my own list, we obviously do so apparently I am a bit confused. Sorry! Annie 16:39, 24 July 2017 (EDT)
I think that in the meantime, using the moderator note field is a useful, albeit imperfect workaround. Annie 14:13, 24 July 2017 (EDT) covers

Apparently, 4 years or so ago, the site owner gave permission to use their covers over here. What is the correct process here:

  • Ask the user to send an email from the domain (to verify that it is them?)
  • Just accept it as stated
  • Something else?

And if that is enough, can we have it added? I am holding a submission from them with another link and I explained the situation to them.

PS: And it is a good site to have permission to link to. Annie 12:30, 21 July 2017 (EDT)

I think the permission posted on our Wiki should be sufficient in this case. I have updated the software to credit the site. Approve away! Ahasuerus 13:14, 21 July 2017 (EDT)
Done! Can you add it to the help page as well? Thanks! (over here :) Annie 13:44, 21 July 2017 (EDT)
Added and alphabetized. Ahasuerus 14:35, 21 July 2017 (EDT)

Our UK-based spies are reporting that "UK Demon Internet" ( has undergone a number of changes over the last few months. As a result, many non-business users, including some SF people, have moved their Web pages elsewhere. We have a couple dozen author records which link to Demonic web pages, so it would be good to double check that our data is still valid. Calling for volunteers! Ahasuerus 13:24, 21 July 2017 (EDT)

I will see what I can do today. Annie 14:59, 21 July 2017 (EDT)
The only ones that need attention are the 11 that used to be on (this is the list. The rest are either verified (they are still where we think they are) or updated now (for the ones that had moved). All of these are invalid - but I want to spend some time to try to find where the sites went (if anywhere) and a quick look did not find them. Annie 15:40, 21 July 2017 (EDT)


I've tweaked the layout of Template:PublicationFields:ExternalIDs in order to make it more readable and helpful. I added some descriptions so people will have a better idea about each one. If those familiar with the different IDs can add information on how to find it when viewing a record on each site, that will make the page even more helpful. Also, if there is a Bulgarian name for the SFBG, that should be added to the entry in the table. Thanks! ···日本穣 · 投稿 · Talk to Nihonjoe 15:09, 22 July 2017 (EDT)

No Bulgarian name really - it is an abbreviation of Science Fiction - Bulgaria :) Known as SFBG - both on the site and in any conversation or site that cares what a site calls itself. Added some notes. I like the new format :) Annie 15:36, 22 July 2017 (EDT)
Glad you like it. ···日本穣 · 投稿 · Talk to Nihonjoe 18:45, 23 July 2017 (EDT)

Search by External Identiier issue

Can we have this search strip leading and trailing spaces? See this? for and example. Annie 20:23, 26 July 2017 (EDT)

Done. Ahasuerus 21:35, 26 July 2017 (EDT)
Thanks! Annie 13:36, 27 July 2017 (EDT)

Duplicate Finder and COVERART

The logic behind the three Duplicate Finder pages has been tweaked. COVERART titles are no longer considered potential duplicates of non-COVERART titles. Ahasuerus 20:09, 30 July 2017 (EDT)

2017 Dragon Award nomination withdrawals

I noticed that Scalzi's and Littlewood's nominations are listed as withdrawn on the award page. However, Dragon Con refused to remove them, so it's still possible they may win. I think these two entries should be changed to reflect that they haven't been removed from the ballot. Thoughts? ···日本穣 · 投稿 · Talk to Nihonjoe 15:19, 9 August 2017 (EDT)

The first thing that comes to mind is that this is still an ongoing process. Scalzi first announced the nomination, then announced that he was withdrawing. Now the organizers have announced that they will not honor the withdrawing authors' requests. Since things are changing rapidly, we may want to wait and see what happens.
That said, it raises a larger issue. There is nothing preventing an organization from presenting an award to an author who doesn't want it. Consider the Golden Raspberry Awards given to the worst movie/actor/etc of the year. I am sure many awardees would love to be able to refuse them. We don't have anything similar at this time, but there have been complicated situations related to declined nominations/awards. For example, in 1986 Judy-Lynn del Rey was given a posthumous Nebula in the "Best Professional Editor" category. It was refused by Lester del Rey, her husband. We currently list it as a win and explain the situation in notes. Should Scalzi or Littlewood win, we could use a similar approach. Ultimately, we should probably separate "withdrawn" and its siblings from the "award level", but it will require non-trivial software changes. Ahasuerus 18:03, 9 August 2017 (EDT)
I like that idea. I also think that unless the sponsoring organization actually removes them (regardless of which award it is), they shouldn't be listed as "withdrawn". I think it will cause less confusion that way. We could just add a note to each of them that they had declined the nomination but that Dragon Con refused to remove them from the ballot. ···日本穣 · 投稿 · Talk to Nihonjoe 19:20, 9 August 2017 (EDT)
2017-08-14 update. As per this post, the award committee has decided to honor author requests and removed The Hidden People by Alison Littlewood and The Obelisk Gate by N. K. Jemisin from the ballot. They will "re-issue ballots to those people who voted for these two books". In the meantime, John Scalzi has withdrawn his withdrawal and will remain on the ballot.
Our award records have been updated to reflect the current state of affairs. Everything should be in sync assuming no other surprises. Ahasuerus 18:38, 14 August 2017 (EDT)

Finding month of publication

I was searching for the true month of publication for a book where the pre-release data that we displayed had the paperback being published prior to the hardcover that stated it was a first edition. I discovered that is generally able to give months of publication, back at least to 1976. A random sampling search of that site on a couple dozen books that we had months for showed very high agreement with their publication months. I found two disagreements: (1) One that claimed it was published a month later than we had listed (possibly due to Amazon selling books a little earlier than the official release date); and (2) One book where we had a January publication date and they had a December. This later case may be because Amazon defaults to Jan. 1 when they don't actually have a release date, and possibly because might default to a December date if they don't know (but I'm not sure of this). This site was fairly week on non-English books, but when they had such data it still agreed with us. Unless someone knows otherwise, this would appear to be a useful site for such bibliographic data. Chavey 09:39, 10 August 2017 (EDT)

Do they explain where they get their data from? Ahasuerus 11:21, 10 August 2017 (EDT)
I can't tell for sure. I suspect this site is descended from the isbnsearch open source project, but I can't find any actual evidence for a connection. If they are connected, then the SourceForge development page describes it as "A distributed search portal of common sources of ISBN numbers". That project has been inactive for 2 years (and the principle force behind it hasn't been involved in any other of his earlier projects, so likely either gafiated or passed on). The source code for that project is available, so you could always take a look. There is also an online discussion about accessing that site via Perl, which looks pretty straight forward. Chavey 03:14, 11 August 2017 (EDT)
The date they list for that pub is the same date as Amazon UK shows. Given that the search results show links to buy the book, the odds are they are pulling the data from their various affiliate sites (Amazon, etc.) and aggregating the results. -- JLaTondre (talk) 08:21, 12 August 2017 (EDT)
That's consistent with the statement on the SourceForge page that said they were aggregating from several other sources. If their description is accurate, though, they are pulling from more than just the Amazon pages. Chavey 09:45, 15 August 2017 (EDT)

Is it necessary to distinguish Nynorsk from Bokmål?

Currently the language options here only include "Norwegian." However, there are two written Norwegian languages, Nynorsk and Bokmål, and publishers generally specify which one their book is in on their book listing page (unfortunately, Worldcat does not). Since we only have 130 titles in "Norwegian," it should be possible to figure out which of the two languages most of them are, shouldn't it? --Vasha 21:14, 10 August 2017 (EDT)

Language belongs to titles though - not to publications. Are the two language different enough to require the split of the titles? And when titles are translated, it generally says Norwegian. Trying to split them may not be as trivial as you think. Maybe we can use a note or a new template. Annie 21:31, 10 August 2017 (EDT)
The question of what constitutes a language and what constitutes a dialect is a complex one. Linguists joke that the main difference is the presence of an army and a navy :-) For this reason we use what is listed by ISO 639-2 and MARC21. They are not identical, but they are described by the Library of Congress (which maintains the two lists) as "compatible". We do not use ISO 639-3, which adds hundreds of additional language/dialect/sub-dialect codes.
In this particular case things appear to be a bit more complicated than usual. There are three MARC21/ISO639-2 entries for Norwegian:
  • "Norwegian", coded as "nor", also used for Bokmål, Dano-Norwegian, Riksmål and the Trøndersk sub-dialects
  • "Norwegian (Bokmål)", coded as "nob", also used for Bokmål, Dano-Norwegian, and Riksmål
  • "Norwegian (Nynorsk)", coded "nno", also used for Landsmaal (Landsmål) and Nynorsk
I have managed to find which of the two it is for all of the Norwegian titles but four. --Vasha 19:48, 20 August 2017 (EDT)
It would appear that we should keep "Norwegian" and add "Norwegian (Bokmål)" as well as "Norwegian (Nynorsk)" to be used when we know which version was employed. We may want to consult User:Linguist to make sure that the summary above is correct. Ahasuerus 23:09, 10 August 2017 (EDT)
Wikipedia writes "After the latest Bokmål reforms, the difference between Bokmål and Riksmål have diminished and they are now comparable to American and British English differences." That's the impression that I've had from my Norwegian friends. So if we do distinguish between Bokmål and Riksmål, should we begin to distinguish between books in American English from those in British English? Iberian Spanish from South American Spanish? Canadian French from European French? High German from Bavarian German? Chavey 03:27, 11 August 2017 (EDT)
"Language vs. dialect" can be a thorny area, which is exactly why we rely on a better qualified third party to make the decision for us. We generously let them figure out whether "Zacatlán-Ahuacatlán-Tepetzintla Nahuatl" and "Tetelcingo Nahuatl" merit separate language codes :-) Ahasuerus 10:37, 11 August 2017 (EDT)
I would tend to go along with Ahasuerus here, and think it would be useful to have extra "Norwegian (Bokmål)" and "Norwegian (Nynorsk)" at our disposal in case a pub is specifically ascribed to one or the other. But on the other hand, it wouldn't be a great sin to keep using "Norwegian" in other cases — or even all cases — as we have been doing so far (and it would save a bit of trouble). In other words, I don't think it's essential, as indeed the difference is not enormous today, but if some finicky editors are prepared to go into these details, why not ? :o) Linguist 05:09, 11 August 2017 (EDT).
Well, if they're that similar, then it certainly isn't essential. The reason I asked in the first place is because the book I just added, Knots by Gunnhild Øyehaug, had a reviewer talking about "the original Nynorsk" in a way that made it sound like a difference. --Vasha 13:22, 11 August 2017 (EDT)
One thing to keep in mind is that we need to support not only "living" languages, but also "historic", "extinct", "ancient" and "constructed" languages like Old English, Latin and Esperanto. If the differences between "Norwegian (Bokmål)" and "Norwegian (Nynorsk)" were sufficiently important to merit separate ISO codes in the past, we probably need to be able to capture them. Ahasuerus 14:00, 11 August 2017 (EDT)
And the publisher did specify Nynorsk. But I think no harm in ignoring that, myself... To be sure, German publishers distinguish books translated "aus dem Englischen" and "aus dem Amerikanischen." Eh, matter of taste... --Vasha 13:22, 11 August 2017 (EDT)
FWIW, some English language publishers include notes along the lines of "This book uses British English spelling". Ahasuerus 14:00, 11 August 2017 (EDT)
That level of detail strikes me as being worth a title note, but not a specification of a different language. My Norwegian friend (born and raised in England; spent all of his professional life in Norway) compares Nynorsk to King James English, but there remain certain literary journals that require submissions to be written in that form. Chavey 00:28, 13 August 2017 (EDT)
The nice thing about relying on a "trusted third party", in this case ISO 639-2, is that it lets us sidestep various linguistic controversies like Serbian-vs.-Croatian-vs.-Serbo-Croatian. If we were to make an exception, no matter how well-meaning, we could be asked to make further exceptions in the future, thus raising the very issues which we tried to avoid in the first place. Ahasuerus 17:01, 13 August 2017 (EDT)

(unindent) "Norwegian (Bokmål)" and "Norwegian (Nynorsk)" have been added. Unfortunately, "Bokmål" is currently displayed as "Bokmal" for technical reasons, at least for now. Ahasuerus 17:09, 19 August 2017 (EDT)

I have managed to find which of the two it is for all of the Norwegian titles but four. --Vasha 19:50, 20 August 2017 (EDT)
Great, thanks! 20:27, 20 August 2017 (EDT)

New Wiki page: current genre publishers

While searching for new releases in short fiction, I started taking notes on publishers, and have now posted my notes here: Guide to Genre Publishers. It includes all those I know of who've published something in 2017, with the exception of those (many!) who only publish a single author. There's 150 at the moment; naturally, I must have missed quite a few. At the bottom is a rather haphazard selection of publishers who've brought out some speculative fiction this year but don't specialize in it.

I hope you may find this useful as a resource for keeping up with what's new. I hope you will expand on it, and feel free to improve the design. --Vasha 16:35, 13 August 2017 (EDT)

New genre publishers only or reprint ones as well? Because I am missing a couple from the latter. :) Annie 16:49, 13 August 2017 (EDT)
It's supposed to be new publications of all sorts, including reprints. Please do add more! --Vasha 16:54, 13 August 2017 (EDT)
Looks quite useful. Perhaps we can automate all or at least some parts of the compilation/maintenance process. Running a query against the database for 2017 publications, I get the following list of the 20 most prolific publishers:
| publisher_name                        |           count |
| Tor                                   |             169 |
| Turtleback Books                      |             163 |
| Escape Artists, Inc.                  |             137 |
| Titan Books                           |             115 |
| Gollancz                              |             113 |
| HarperCollins                         |             102 |
| Stone Arch Books                      |              99 |
| Orchard Books (UK)                    |              96 |
| Little, Brown Books for Young Readers |              86 |
|                               |              86 |
| Bloomsbury USA Children's Books       |              79 |
| Baen                                  |              77 |
| Orbit (US)                            |              75 |
| Gateway / Orion                       |              73 |
| J. H. Sweet                           |              73 |
| Aladdin                               |              72 |
| Angry Robot                           |              70 |
| Orbit                                 |              69 |
| Disney/Hyperion Books                 |              68 |
| Scholastic Paperbacks                 |              67 |
A little tweaking can further improve the query logic, e.g. by eliminating single-author publishers and displaying each publisher's Note field. We can then turn this query into an overnight report and add it to the ISFDB Statistics and Top Lists page. The only thing that may require additional work will be linking the new page to Amazon search pages. (At the very least we will need to strip all disambiguators like "(US)", "UK", etc.) We will also want to display multiple publisher- and store-specific Amazon links: one for all US/UK/CA/etc books, one for all 2016/2017/etc books, and so on. Ahasuerus 12:34, 14 August 2017 (EDT)
I've added the ones in that table (except for Turtleback, Escape Artists, and J. H. Sweet which are all irrelevant for various reasons). But I'm afraid I don't understand what you mean about automating updates. It will almost always take some hand-work to find the correct links. --Vasha 00:51, 15 August 2017 (EDT)
Well, it depends. Some publisher names are unique and unlikely to cause issues across Web sites. For example, 47North can be used "raw" in searches. It can also be used in WorldCat keyword searches. On the other hand, Orbit (US) can only be used in Amazon searches after stripping the parenthetical disambiguator and it's completely useless on the WorldCat side. Then there are imprints like Orchard Books / Scholastic which clearly require manual massaging. Finally, there are publishers like The Black Coast Press, which Amazon lists as something completely different.
Perhaps we could add a new field to publisher records, something like "Name used by external sites". We could then use this name (if provided) to build search URLs for Amazon US/UK/etc and potentially other sites.
It also occurs to me that broadly similar functionality could be useful when working with other records. For example, if we could link Summary Bibliography pages to WorldCat/Amazon/etc (and optionally limit them to a single year), it could benefit both active editors and occasional users. Ahasuerus 11:55, 15 August 2017 (EDT)

(unindent) BTW, here are the next 30 "most prolific publishers of 2017":

| DAW Books                             |              66 |
| HarperTeen                            |              66 |
| Yen On                                |              66 |
| Del Rey                               |              64 |
| Sky Pony Press                        |              57 |
| Scholastic                            |              55 |
| Scholastic Press                      |              54 |
| WordFire Press                        |              53 |
| Saga Press                            |              52 |
| Ace Books                             |              49 |
| Wildside Press                        |              49 |
| Brilliance Audio                      |              48 |
| 47North                               |              46 |
| Square Fish                           |              46 |
| Random House                          |              44 |
| Heinrich Bauer                        |              44 |
| Simon Pulse                           |              42 |
| Bloomsbury Children's Books (UK)      |              42 |
| Head of Zeus                          |              42 |
| Egmont Books Ltd                      |              39 |
| Houghton Mifflin Harcourt             |              39 |
| Scholastic UK                         |              38 |
| Heyne                                 |              38 |
| Candlewick Press                      |              38 |
| Games Workshop                        |              38 |
| Amulet Books                          |              38 |
| Alfred A. Knopf                       |              37 |
| Hodder & Stoughton                    |              36 |
| Delacorte Press                       |              36 |
| Berkley Books                         |              36 |

Ahasuerus 12:02, 15 August 2017 (EDT)

Here's the ones that are different on that page so far, which is the tip of the iceberg, since Amazon is rarely consistent.

ISFDB Amazon
Ace Books Ace
The Alchemy Press Alchemy Press
Alfred A. Knopf Knopf
Apex Publications Apex Publications OR Apex Book Company
Atheneum Books for Young Readers Atheneum
Avon Avon Books
Berkley Books Berkley
Bizarro Pulp Press Bizarro Pulp Press - JournalStone
Bloomsbury Children's Books Bloomsbury Childrens
Covet / Entangled Publishing Entangled: Covet
Crown Publishers Crown
Disney/Hyperion Books Disney-Hyperion
Egmont Books Egmont OR Egmont Books
Emily Bestler Books Atria/Emily Bestler Books OR Emily Bestler Books
Faber and Faber Faber & Faber
Farrar, Straus and Giroux Farrar, Straus & Giroux
G. P. Putnam's Sons G.P. Putnam's Sons
Harlequin MIRA MIRA OR Mira Books
Henry Holt Henry Holt and Co. (BYR)
Indigo BHC Press/Indigo
Laksa Media Laksa Media Groups Inc.
Little, Brown Little, Brown and Company
Penguin Books Penguin OR Penguin Books OR Penguin Classics
The Penguin Press Penguin Press
Pocket Star Books Pocket Star
Select Otherworld / Entangled Publishing Entangled: Select Otherworld
SMP Swerve Swerve
Talos Press Talos
Vintage Books Vintage
Weidenfeld & Nicolson W&N

Vasha 11:46, 17 August 2017 (EDT)

"Resist and Refuse" Magazine

Does the magazine Resist and Refuse sound like something that could be included here? It's a political magazine by weird fiction/horror writers. There will be fiction and the description says "Each story is to some degree related to the general themes explored in the zine and most have a dark and/or weird tone." Having a weird "tone" isn't exactly the same as being weird fiction... I think, personally, I would not include it in the database until someone reads it and says whether any of the stories are speculative. --Vasha 21:14, 17 August 2017 (EDT)

It looks like a non-genre magazine, although I expect that, given participants like Cody Goodfellow, some of the content may be at least borderline SF. I agree that we should wait for the first issue to appear. Ahasuerus 22:14, 17 August 2017 (EDT)

Add an editor to series entries

It has been suggested that series entries should have a field for "series editor". I think this is a good idea as there are a fair number of series that have editors who are different than those for the titles included in them. Thoughts? ···日本穣 · 投稿 · Talk to Nihonjoe 16:59, 18 August 2017 (EDT)

How would it work if a series had different "series editors" at different points in time? Would it require splitting the series? Ahasuerus 18:15, 18 August 2017 (EDT)
Maybe have an optional text field where the volume numbers or years can be entered? I'd prefer not to split the series, though perhaps it could be turned into two sub-series within a container series, the way we do with books in series within series. ···日本穣 · 投稿 · Talk to Nihonjoe 18:29, 18 August 2017 (EDT)
And about the cases where there are more than 2 editors in different times - a new one every 5 years or one doing the series for 30 days and then followed by a few with a few each year - what qualifies? Or when there are multiple ones on the series level and they are different each year? Where will be the boundary?
And what about translation editors for example - in some series, they do more work on the translated series than the series editor does. Who will make the decision who is eligible to be added to the field? The idea sounds interesting on the surfaces but opens the door way too much for a subjective decision on what is really a series editor and what does not... Annie 19:23, 18 August 2017 (EDT)
I don't think it's terribly ambiguous. If they are credited as "series editor" in the book, that's who you put in the field. If they are credited as something else, you don't list them as a series editor. We can make a specific list of who would be included. ···日本穣 · 投稿 · Talk to Nihonjoe 19:28, 18 August 2017 (EDT)
And what happens when the series editors change every year or every few years? We list all of them? Enforcing a rule based on a number will make it very unfair. Annie 19:37, 18 August 2017 (EDT)
Based on what number? The only number I mentioned was to perhaps have a field where volume numbers could be listed next to each series editor (like we do for omnibus volumes). How does that limit it or prevent series editors from changing? It would display something like this:
  • Series: Amazing Series Example
  • Series Editor: Editor Name (1-6), Editor Name (7-10), Editor Name (11), Editor Name (12-37)
Pretty straight forward. ···日本穣 · 投稿 · Talk to Nihonjoe 20:22, 18 August 2017 (EDT)
  • Series Editor: Editor Name (1-6), Editor Name2 (1-2), Editor Name3 (2, 14), Editor Name4 (3-4) and so on... Not all series editors are straight forward. Or single editors. Annie 20:31, 18 August 2017 (EDT)
Yes. I'm not seeing an issue. It still lists all of them. You could even do it by years if that was applicable:
  • Series Editor: Editor Name (1983-1987), Editor Name (1986-1999), Editor Name (1997), Editor Name (1999-present)
···日本穣 · 投稿 · Talk to Nihonjoe 20:34, 18 August 2017 (EDT)
If we allow multiple formats, the editor's page will become inconsistent (they will have one note for a series with (1-3), one with years and third one with titles for example (because the series was never numbered for example). I do like the idea in general but I do not think that just adding the field and modifying the author pages to show it is as trivial or so without problems... And sooner or later, we will get a "but I want to go on the editor page to see which exact books they were a series editor for" request (as opposed to the just show the list of series and needing to click down to it.
And I do not think that making half a feature (there on the series page, but not visible on the author page) will be a good idea. Once we start talking about showing that on the author page, we have to deal with pseudonyms (do we link to the pseudonym used or to the canonical (and what of the person changes their name - do we add two(+) records)? How do we show that on the author page - combined or as two(+) records?
If that feature is ever implemented, I think it needs to tie to translators and other types of people that we do not record now both on the pub and on the series level (editors for collections for example. Or for novels) as opposed to just doing something for one case. And we need to make sure that we do not just overload the page with links (longer series volumes will end up way down the page when we finish listing the series editors for some special cases. And Publication series also need editors assigned if we are going there. Annie 21:00, 18 August 2017 (EDT)
Opposed to this move. All this steming from (IMHO) an ego problem, an "uncredited" editor that wants absolutely to be listed (he tried this before under another pseudonym). Apart from this dishonest attitude, the concept of "series editor" is way too much nebulous to be used.Hauck 06:06, 19 August 2017 (EDT)
I'm not sure it's terribly nebulous when the front cover or title page indicates a "Series Editor". If it's not credited, then we wouldn't include it. Not nebulous at all. ···日本穣 · 投稿 · Talk to Nihonjoe 01:40, 20 August 2017 (EDT)
Sure, when you'll find the mention of a "Series Editor" in another language than english (you know that we also cover those, I suppose), just call me, I'm waiting. Hauck 04:02, 20 August 2017 (EDT)
Why are you being so snarky? You're acting like I'm some clueless idiot who needs your benevolent wisdom (at least that's the way it reads to me). Regardless of whether the editor who initially wanted this was acting in his own self interest, it's a good suggestions that should be addressed seriously. And yes, I know we have other languages here than English (that should be pretty obvious to anyone who has seen my contributions, of which you have seen and approved many). I've seen series editors for Japanese anthology series, too. You even mention a French one below, so that's at least three languages with clear examples. I'm fine with waiting to implement it with other contributor roles, but please don't treat me like a moron. ···日本穣 · 投稿 · Talk to Nihonjoe 12:46, 20 August 2017 (EDT)
Unfortunately, tone can be difficult to convey on the internet :-( In addition, sarcasm is treated differently in different cultures (and even sub-cultures), which increases the likelihood of misunderstanding. Once you add various linguistic nuances, even between different versions of English (e.g. "When would you like me to knock you up tomorrow?" or "let's table this plan"), to the mix, misunderstandings start to pile up and it can be difficult to deescalate the resulting conflict. Ahasuerus 13:59, 20 August 2017 (EDT)

(after sleeping on it) Based on the fact that "series editors" can change over time, I think that the proposed field would have to be associated with Title records rather than Series records. It would also have to be a multiply occurring field (similar to authors, transliterations, Web pages, etc) in order to support scenarios with multiple "series editors" per volume. Finally, it would need to support pseudonyms.

Although quite time-consuming, it would be possible to implement. However, I agree with Annie that it would be better to address this functionality as part of a larger project to add support for "contributor roles" like translator, "adapter", "cover design", etc. Ahasuerus 08:54, 20 August 2017 (EDT)

We may even be forced to locate this "role" at publication level. In France, we have the "directeur de collection" ("Publication Series Manager/Editor") role that may bear some likeness to the proposed Series Editor. The person at this post is of course usually changing with time. So let's suppose that a precise title is first published under series editor X's reign, then reprinted (with a new cover to distinguish things) at a time when series editor Y is at the helm. Who shall get the credit if it's located at title level? X because he likely bought it, Y because he choosed to reprint it? Both? Hauck 09:26, 20 August 2017 (EDT)
I would say we credit it however it's credited in the work. If both are listed, then list both. If only one is listed, then list only the one credited. ···日本穣 · 投稿 · Talk to Nihonjoe 12:48, 20 August 2017 (EDT)
I think the first question that we will need to address is which "roles" will be associated with title records and which ones will be associated with publication records. For example, I am reasonably sure that we will want to associate "translators" with title records. On the other hand, "cover designers" will be most likely associated with publication records: multiple publication records can use the same COVERART title, which will be arranged differently by different cover designers.
However, what about "single author book editors"? A number of publishers, including major publishers like Tor, have been know to list the name of the person who edited the first edition of a book. Should we associate this "role" with title records or with publication records? Ahasuerus 14:11, 20 August 2017 (EDT)
Maybe we should make a list of all the different roles people have suggested over the years first. I think it will be easier to consider which records to attach them to if we have a good list to work from. Here's a quick list:
  • Series editor
  • Translator
  • Cover designer
  • Text/Book designer
  • Adapted by
Anything else? ···日本穣 · 投稿 · Talk to Nihonjoe 17:10, 20 August 2017 (EDT)
The original discussion of roles is here. A lot has changed since 2009, so take it with a grain of salt, but I think that some of the ideas are still relevant. Ahasuerus 20:32, 20 August 2017 (EDT)

List of roles

Role Level
Adapted by Title?
Back Cover Artist Publication
Collection Editor Title, Publication
Cover Designer Publication
Cover Photographer Publication
Editor Title, Publication?
Interior Layout Designer Publication
Narrator/Reader Publication
Series Creator Series, Publication Series
Series Editor Series, Publication series, other?
Translator Publication

I also like MHHutchins' comment at one of the previous discussions:

I don't think authors (in our current use of the word which includes artists and editors) should be assigned roles, but that roles should be assigned to authors. There's a difference there. Let me explain. Robert Silverberg should have one record for all his credits. His relationship to the publication is defined by the role we assign him. The system automatically assigns him the role of editor when the pub type is ANTHOLOGY, as it assigns him the role of author when the pub type is NOVEL. If he were credited with an INTERIORART credit, we would assume him to be the artist, even though there is no such credit explicitly assigned by the system. I would suggest a entry form that allows a drop-down menu to assign pre-defined roles then allow the editor to add another credit (as it does now for cover art). That way we can assign as many roles to a publication as are necessary. The system will combine these credits into one "author" summary page, but only display the assigned role on the publication record. To bring up a recent example: Robert Conquest edited The Robert Sheckley Omnibus, but there is no credit on his summary page. Once role assignment has been implemented, it will. The question is: how will it be displayed? Perhaps under a newly created Book Editor category. Which begs another question: how will this is differentiated from books listed under Anthologies which at the moment assumes the author to be the book editor? MHHutchins 23:13, 17 June 2009 (UTC)

I think his suggestion is a good one: making a drop-down where we can assign different roles to every person attached to a particular title. This allows quite a bit of flexibility in doing things like adding an editor to each title in an anthology (if they each had different ones, which I've seen occasionally).

Feel free to edit the table to add other possibilities (or suggest them and I'll add them). ···日本穣 · 投稿 · Talk to Nihonjoe 22:27, 27 August 2017 (EDT)

The following is IMHO. First, we're going way too far in the details. Presently, the photographer of a photographic cover is credited as cover artist, why create another category (and so why not "cover aquarelist", "cover photoshoper" etc...), this kind of list is nearly infinite (I've seen pure genre books -NESFA?- where the font designer is credited) and can be easily expanded many times (proofreader, typesetter, packager, publication series designer, inker...). Second, for the huge majority of the records most of these new roles will be left void and are irrelevant to a bibliographical enterprise or simply unknowable. Third, there is as usual the problem of definition and the managing of the resulting conflicts, "Series creator" is a perfect example especially when applied to media-related work (who's the creator of Alien, Van Vogt?) and its wonderful world of lawsuits. Fourth there is the problem of levels and its possible impacts on the display. There are some publication series in french that have seven known series' editors, should the PS be split? On what grounds? (i.e. When is a new Series's Editor to be considered as leaving its mark on it? When he enters the company? After a fixed delay?). In general, I'm afraid that we're blindly forging ahead into minute details at a time when more bibliographically central points are still not adressed like the implementation of the printing rank. To be more constructive (It's quite a rare sight), I'd keep for now: Editor (either for COLLECTION or NOVEL records), Translator and a more general Designer. Hauck 02:45, 28 August 2017 (EDT)
To answer some of your comments:
  • Two: If they are credited in the book (or in a reliable source), they would be credited here. If not, then they wouldn't be credited. There's no guessing involved. Most books credit at least some of the suggested roles, so it's not really relevant that a given book wouldn't use all of them.
  • Three: Unless Van Vogt was credited somewhere as the creator of Alien (which he's not, as far as I I've seen, even though there are a lot of similarities). As with #2, unless they are specifically credited either in the work or in a reliable source, they would not receive credit here. Just like we do it now.
  • Four: We can discuss how any new roles would display (that's part of why this discussion is here). As for your example of the French series with seven series editors, I think you're making it more complicated than it needs to be. I don't think a series needs to be split just because a new editor takes the helm (or joins the previous editor). The new series editor would be recorded when he is credited in the work or in a reliable source. Just like we do it now.
Regarding the rest, what do you mean by "printing rank"? I haven't heard that term before.
Thanks for your comments. ···日本穣 · 投稿 · Talk to Nihonjoe 17:48, 28 August 2017 (EDT)
It's the last item listed in the Roadmap 2017 section:
  • Add a "printing rank" field to order multi-reprinted titles without pub dates
Basically the printing number. Ahasuerus 18:20, 28 August 2017 (EDT)
Aha. I was wondering if that's that it meant. I like that idea, and that might be easier to implement since it's a standalone feature (unless you include the printing year part from more recent publications, as well as the printer information that some publishers include on that line). Thanks for the clarification. ···日本穣 · 投稿 · Talk to Nihonjoe 18:43, 28 August 2017 (EDT)
Yes, it should be relatively easy to implement compared to roles. We'll have to decide whether the new field will be strictly numeric (1, 2, 3, etc) or whether we'll allow arbitrary text. Making it numeric would make it easy to sort things. Ahasuerus 20:00, 28 August 2017 (EDT)

(unindent) Going back to roles, I believe translators will have to be associated with titles rather than publications. A single pub may contain two dozen titles translated by 10 different people. The same logic would apply to narrators/readers if we decide to implement them.

As far as "series creators" and "series editors" go, I believe that they would need to be associated with titles rather than series. In addition, you can have a standalone title, e.g. a game novelization, with 3 "creators" and 2 authors. If we decide to implement these roles, we'll have to think it through. Ahasuerus 20:11, 28 August 2017 (EDT)

New cleanup report - Potential Duplicate E-book Publications

A new cleanup, "Potential Duplicate E-book Publications", has been coded and deployed. All editors will be able to see the data while moderators will be able to "ignore" records. The data will become available tomorrow morning. Ahasuerus 17:25, 18 August 2017 (EDT)

Note that I'm leaving "as-they-are" the pairs of records where one is given as a Kindle book and the other a Nook book. Hauck 10:56, 28 August 2017 (EDT)

De Nederlandse Bibliografie publication number added as a supported external ID

De Nederlandse Bibliografie publication numbers have been added as supported external IDs. Ahasuerus 16:14, 20 August 2017 (EDT)

And now we have a cleanup report to go with it. The data will become available tomorrow morning. Ahasuerus 16:17, 31 August 2017 (EDT)
Good. Thank you! --Willem 16:34, 31 August 2017 (EDT)
Sure thing! To complete the circle, I have created a "PPN" linking template. Ahasuerus 21:17, 31 August 2017 (EDT)

Award to add: Australian Shadows Award

See AHWA website. --Vasha 17:56, 21 August 2017 (EDT)

Looks pretty solid. Any objections? Ahasuerus 18:42, 21 August 2017 (EDT)
Without objection, so ordered. They have changed their structure a few times since 2005, so I had to create 9 categories to accommodate them. They should be ready for your data entry pleasure! Ahasuerus 19:15, 24 August 2017 (EDT)

(unindent) A correction is needed. As this post shows, the long fiction award began to be called the Paul Haines Award in award year 2013, not 2011. (Haines himself had won it for 2011, before he died.) Therefore, this category needs to be divided into two: "Long Fiction (2011-2012)" and "Paul Haines Award for Long Fiction (2013-)." Everything else can stay as it is.

Note that, following AHWA practice, the years stated for the awards are the years of publication (the award itself is given out the following year). Thus, the 2013 Paul Haines award (the first one) was given in the spring of 2014. --Vasha 12:42, 26 August 2017 (EDT)

OK, I have split the "Long Fiction" category into 2 and added the winners. Ahasuerus 14:09, 26 August 2017 (EDT)
Upon reflection, I am not sure that splitting 2011-2012 and 2013- is the best approach. We have a number of categories whose names have changed over time yet we don't split them. For example, the SFWA Grand Master Award, which we currently list as a Nebula category. Its name was changed to "Damon Knight Memorial Grand Master Award" in 2002, but we have only one category for 1975-2017. Ahasuerus 15:41, 26 August 2017 (EDT)
True. Would make perfect sense just to call it "Long Fiction" with a note. --Vasha 15:55, 26 August 2017 (EDT)
OK, done. Ahasuerus 10:35, 27 August 2017 (EDT)
Good... and the data entry is complete --Vasha 10:50, 27 August 2017 (EDT)
Excellent! Ahasuerus 12:36, 27 August 2017 (EDT)

Nonexistent book

I think I have found a record for a book that doesn't actually exist -- this edition of Samurai and Other Stories by William Meikle. That ISBN is only on and a few other random booksellers; and there is an interview with the publisher in December 2013 where they say that Samurai and Other Stories is forthcoming. Any objections to deleting that publication and changing the title date? --Vasha 16:12, 24 August 2017 (EDT)

I've fixed it. It was likely just a typo. The Kindle edition is listed as January 21, 2014, and the TP version is January 5, 2014. ···日本穣 · 投稿 · Talk to Nihonjoe 16:38, 24 August 2017 (EDT)
One interesting thing: 978-0-9922182-3-2 brings up the Kindle edition, not a trade paperback edition, on Amazon. ···日本穣 · 投稿 · Talk to Nihonjoe 16:40, 24 August 2017 (EDT)

Canonical name question: Geoff Brown/G. N. Braun

The author whose canonical name is currently G. N. Braun has equal numbers of records, published simultaneously, as G. N. Braun (fiction) and Geoff Brown (editorial work). The canonical name should be the one that's more familiar, and normally that would be the one used for fiction, but not in this case. As Braun, he's respected but neither very famous nor very prolific. As Brown, he's editor of magazines and anthologies (twice award-nominated), past president of the Australasian Horror Writers' Association, columnist for a major fan site and author of numerous articles, and publisher of Cohesion Press. I have found five interviews with him using the name Brown, one using Braun. So I propose changing the canonical name. Thoughts? --Vasha 09:35, 27 August 2017 (EDT)

Since he writes of himself as G. N. Braun on his own website, I think it would be somewhat confusing to change the canonical name. Stonecreek 04:02, 28 August 2017 (EDT)

OCLC as External Identifier vs. OCLC as Secondary Verification

Now that we have the External Identifier field in place which covers OCLC numbers, is OCLC as Secondary Verification still necessary? I feel like they're redundant now, and that we could probably remove the Secondary Verification option as long as we run a list of all the records that have OCLC 2V's that haven't had their OCLC numbers added as External IDs. (Which isn't likely to be a quick process, but is one we could start.) Albinoflea 16:09, 27 August 2017 (EDT)

It would appear that this will need to be a three-step process:
  • Move all OCLC numbers from notes to the External ID field
  • Ensure that all OCLC-verified publications have at least one OCLC External ID
  • Re-evaluate to see if we would be losing any other functionality by removing OCLC verifications
Does this sound about right? Ahasuerus 12:47, 28 August 2017 (EDT)
I'm not sure why we would remove OCLC verifications. My assumption is that when we do an OCLC verification, we make sure that all of the data we collect that is in that record has been pulled into our record (is that naive of me?). So the OCLC verification would, at least in my (possibly imaginary) world indicated that this data extraction has been done, while just having it as an external ID would mean that we have located the record, but not necessarily extracted the data. Chavey
Plus the verification allows to set a value as N/A do the next editor do not lose time looking for a record. It also contains a date (so it is clear when something was valid so one knows to check for a new record or not). The two things serve different purposes in my mind. Annie 23:50, 28 August 2017 (EDT)
Since WorldCat is not a finite resource, in my mind the N/A would only be a temporary state of affairs... in that sense I feel like its not really like the other resources listed in the secondary verification section, which are primarily bibliographic resources fixed in time. WorldCat is sort of like ISFDB in the sense that it's simultaneously adding new, current material while infilling with older material, especially non-English editions, so if there isn't a record for it today not that's not an indicator that there might not be one at some future date.
While I haven't been here long enough to know the original raison d'être for the Secondary verifications, my sense was that they were a way of asserting the existence of a volume when no primary verification was forthcoming based on the collections of the editors at any given time... basically along the lines of "We've never seen one but it's in Currey so we assume it exists", with the relevant data being extracted into ISFDB until a PV or more detailed source becomes available.
WorldCat entries are not terribly detailed by ISFDB standards, and the Using Worldcat data help page is really just a long cautionary tale. I primarily use it just to grab the OCLC accessioning number so that it can be linked to, which also gives a nice way of locating a copy of an edition in a library. Otherwise I occasionally use it for creating entries to prevent dangling reviews in older publications where the Amazon data is lacking or suspect... but in that sense they're really just stub entries. Most of the time I see an OCLC number and it's already attached to a fairly healthy entry in the DB. And we don't have an Amazon entry for Secondary Verifications, even though we're probably more reliant on Amazon than we are on WorldCat by now.
I would miss the date stamp and user info that Annie mentions, because we don't get that from regular edits and sometimes it can help identify who created or edited a record. In that sense if a WorldCat 2V is the only verification on record, perhaps they are valuable, but if there are multiple verifications WorldCat is probably the weakest, so there isn't likely to be much data coming from the WorldCat record aside from the ID number anyhow.
But that's just my sense from how I use it and how I've seen it used. I'm interested to hear how others are using it. Chavey and Annie's take is interesting because I just assumed that we would want parity... if there was a WorldCat 2V we should have a External ID and vice versa, which is why I felt they were redundant, but if people are using them for two different purposes then that is definitely not the case. Albinoflea 01:15, 29 August 2017 (EDT)
OCLC records are *very* helpful if we don't have a PV (page numbers, anthology contents, and a few other things). If we have a PV on the book, the OCLC has limited usefulness, although I use it for two other purposes: (1) I'll often grab the OCLC plot synopsis to add as our title summary; and (2) It can often be used to identify other editions of that book that exist, and add them as well. Chavey 01:25, 31 August 2017 (EDT)

A. Bertram Chandler Rim Worlds Books

I recently had reason to revisit the Chandler page to enter some new acquisitions. I would like to reorganize the "Rim Worlds" section, and the John Grimes Rimworld section in particular. First, the numbered pubs are in no particular order, neither in order of publication or in order of their point in Grimes' career. But if that were all, I wouldn't bother making changes. The situation that bothered me was that not all of the pubs under the John Grimes heading involve John Grimes. The first two novels (items 1 and 3) are about a character named Derek Carver. The second item, a collection, also features characters other than Grimes (although he appears in one as a minor character). There are others that are non-Grimes as well. And it should be "Rim Worlds", not "Rim World". Any thoughts? Bob 13:07, 28 August 2017 (EDT)

I only read a subset of the John Grimes stories and it was years ago, so I hesitate to give advice. However, this list of Grimes works maintained by David Kelleher looks useful. His site also has a master list of Chandler's works.
My personal recollection is that some stories featuring Grimes were either peripheral or set in what appeared to be alternative continuities, which may present bibliographic challenges. Ahasuerus 12:49, 29 August 2017 (EDT)
That material looks useful; thanks! Bob 16:34, 29 August 2017 (EDT)

Suppression of reference titles

I have been thinking about our last round of discussions re: how we display nonfiction titles in publications. After experimenting with the software on the development server I concluded that we probably have a larger problem.

Back when the "ISFDB 2.0" software was designed and implemented, we wanted the Publication page to distinguish between each pub's "reference title", i.e. the main title in the pub, and all other titles. For this reason we displayed the former on the "Title Reference" line in the metadata section while displaying all other titles in the "Contents" section.

A few years later things began to change. First we realized that some reference titles, notably novels, could have page numbers associated with them. If we didn't display them in NOVEL publications which contained additional titles, the Contents section would look incomplete. For example, consider what this pub with its 4 introductions and 1 afterword would look like if we didn't display the novel title. And so we changed the software to display NOVEL reference titles twice -- once on the "Title Reference" line and then again in the Contents section -- but only for pubs with additional titles. It was inconsistent and duplicative, but we figured it was a relatively uncommon occurrence.

Then we began adding information to the Contents section. We added series data, variant data, pseudonym data, translation information and the juvenile/non-genre/etc flags. However, we didn't change the data on the "Title Reference" line, which is now out of sync with what's displayed for Contents titles.

Consider this NOVEL publication. Its "Title reference" line reads "Fanfaren der Freiheit". If the pub contained an afterword, the same title would also be displayed in the Contents section as follows:

  • Fanfaren der Freiheit • [Interplanetary Relations Bureau • 1] • novel by Lloyd Biggle, Jr. (trans. of The Still, Small Voice of Trumpets 1968) [as by Lloyd Biggle]

One heck of a difference.

Also, back when we implemented non-genre/juvenile/etc flags we tweaked the "Type" line of the metadata section to display the reference title's flags. It's really a kludge since the flags are part of the title record and not of the publication record, so they don't belong in the metadata section.

I should add that this is not a novel-specific issue. It also affects other types of reference titles: collections, anthologies, omnibuses, etc.

I can think of two ways of addressing this problem. The first approach is to change the software behind the display of the "Title Reference" line to display the same information as what we display in the Contents section. It would synchronize the two sections, but we would be left with the issue of data duplication for novels and the kludgy flag display. Also, my experiments on the development server suggest that adding more information to the "Title Reference" makes it too long/unwieldy if there is a cover scan.

Alternatively, we could remove the "Title Reference" line from the metadata section and simply display the reference title in the Content section along with the rest of the pub's titles. I think it's a more comprehensive solution and will work better in the long run. However, it will change the appearance of the Publication page, especially the Contents section of NOVEL pubs, so I'd like to solicit feedback before I do anything. So, what do you think? Ahasuerus 20:39, 5 September 2017 (EDT)

That would certainly look good for novels and for monograph NONFICTION. Indeed, I very much like the idea of always displaying the reference title in the contents section in such cases where it is also a regular title. But are you suggesting also displaying container titles in the contents section? A much more awkward prospect. --Vasha 22:01, 5 September 2017 (EDT)
Well, let's use this collection as an example. The "Title Reference" line currently reads "Der Tag ist nicht mehr fern". If the display software leveraged the logic used in the Contents section of the page, the text would change to:
  • Der Tag ist nicht mehr fern • by Lloyd Biggle, Jr. (trans. of A Galaxy of Strangers 1976) [as by Lloyd Biggle]
Would that be OK?
Also, it occurs to me that we don't have to change everything in one fell swoop. We could start by taking some of the following steps:
  • Always display novels in the Contents section
  • Rename "Title Reference" to something like "Container Title"
  • Do not display the "Container Title" line for novels
  • Change the display logic for other container title types to what's shown for "Der Tag ist nicht mehr fern"
We could then take a break and re-evaluate in a few days. Ahasuerus 23:54, 5 September 2017 (EDT)
So, any objections to the last 4 bullet points? Ahasuerus 18:12, 13 September 2017 (EDT)
No, let's give it a try! Stonecreek 23:47, 13 September 2017 (EDT)
The 4 listed changes have been applied. Let's see if the new layout looks better than the old one. Ahasuerus 17:58, 15 September 2017 (EDT)
I really like how this looks. My only issue is that "Container Title" sounds like obscure jargon; but at the moment I can't think what to call it instead. At least it's obvious what it is, no matter what we call it. --Vasha 19:03, 15 September 2017 (EDT)
A better term would be nice. We can always use more better English! :-) Ahasuerus 21:06, 15 September 2017 (EDT)

Victor Valla

Jonebergquist has submitted Victor Valla as the cover artist for multiple publications. I have the original edits on hold as there was no source or notification to the primary verifiers. After posting on the user's talk page, the user made subsequent edits (also on hold), providing a moderator's note that gave the source as:

" also and my Facebook conversations with Mr Valla"

As this impacts multiple verifiers, I'm posting this centralized notice here and pointing active verifiers to it.

If anyone has concerns with this sourcing, please post here. Otherwise, I will accept the edits tomorrow and add an appropriate statement to the publication notes. Thanks. -- JLaTondre (talk) 21:57, 6 September 2017 (EDT)

Not a fan of blogs as sources, but the data seems solid. --~ Bill, Bluesman 12:37, 7 September 2017 (EDT)
(edit conflict) The Cthulhu Files post seems thoroughly researched, so I would accept it. Here's an archived version of the page, if you wish to use it: archive. ···日本穣 · 投稿 · Talk to Nihonjoe 12:40, 7 September 2017 (EDT)
Edits have been accepted & source note added. -- JLaTondre (talk) 21:02, 7 September 2017 (EDT)


Just returned from a vacation provided by my medical insurance, although I'm still having some problems. My brother picked up my mail and as soon as I can get to it, I'll list the contents of these magazines, unless somebody gets to them first. MLB 21:55, 8 September 2017 (EDT)

Good to have you back! Just take your time! Christian Stonecreek 00:39, 10 September 2017 (EDT)
Glad you made it through! Ahasuerus 21:03, 10 September 2017 (EDT)

Introductions to Lem's Dzienniki gwiazdowe / The Star Diaries / Sterntagebücher

There seem in fact to exist three different introductions (see here for the initial discussion):

  • the introduction from the first edition 1957
  • a new introduction plus an introduction for the expanded edition from the second and third editions 1966/1971.

They appear to have become mixed up for the various publications. It seems very likely that the 1957 introduction in fact wasn't published in English, as these editions feature expanded versions and omit the original 26th voyage.

In addition, it will likely be the case that (at least the two later ones for the expansions) are either credited to 'Professor A. S. Tarantoga' or 'Prof. A. S. Tarantoga' (which would be a pseudonym for Lem).

As this issue impacts multiple verifiers, I'm posting this centralized notice here and will pointing some active verifiers to it. Stonecreek 08:36, 9 September 2017 (EDT)

I'm not sure what the 'issue' is? These 'Introductions' are clearly in-universe, thus whatever 'signature' is attached to them can't be considered a pseudonym of the author. If this were a novel they wouldn't even get separate content status. I think it's a stretch even in a collection. --~ Bill, Bluesman 11:55, 9 September 2017 (EDT)
I agree with Bluesman about the authorship. In universe fictional essays don't get a fictional author. As for the dating, my edition states "Original edition: Dzienniki Gwiazdowe, published by Czytelnik, Warsaw, 1971", so I assume both introductions were first published in that publication. --Willem 15:57, 9 September 2017 (EDT)
Thanks for the comments, guys. I still think that we could use the signature as source for the credit (especially since one of the introductions refers to the absurdity of Lem being the author of the stories), but I also can live with that.
I'll variant the varianting of the introductions. There may exist four different introductions from the years 1954, 1966 (we have an edition for that year but without contents), 1971 (reference to LEM and two now shortfictions) & 1971 again? (for the expanded edition). Thanks, Christian Stonecreek 00:37, 10 September 2017 (EDT)

Old ISFDB backups

Now that additional Google Drive space has been donated to our project, I am in the process of making select pre-2017-06 backups available once again. The ISFDB Downloads page should be updated over the next day or two. Ahasuerus 13:30, 12 September 2017 (EDT)

MySQL 4.0/5.1-compatible versions of the existing monthly backups have been uploaded for 2006-2017. If someone needs a particular weekly backup file which hasn't been uploaded, please let me know and I will make it available. Ahasuerus 16:46, 14 September 2017 (EDT)

Date order inconsistency

I just noticed that if a title or publication does not have an exact date, only a year, it is listed after exact dates in the publication table on the title page, but before exact dates in title lists. That ought to be made consistent. --Vasha 10:51, 13 September 2017 (EDT)

Originally records with 0000-00-00 dates were displayed first. There was a feature request to make them sort last. It was implemented in 2010. Somewhat later (I can't find the FR right now), the display logic was streamlined to display records with "00" months and days last: 2014-00-00 appears after 2014-01-00, 2015-01-00 appears after 2015-01-12, etc. Ahasuerus 11:35, 13 September 2017 (EDT)
P.S. Of course, these things are never cast in stone. Changes can be reverted (completely or partially) if we determine that they do more harm than good. Ahasuerus 11:50, 13 September 2017 (EDT)
I don't really care whether they are before or after, it's just that they should be the same in publication summaries and title summaries. So change titles to "after"? --Vasha 11:52, 13 September 2017 (EDT)
Sorry, I didn't realize that there was a discrepancy. Could you please post a couple of sample URLs? Ahasuerus 13:49, 13 September 2017 (EDT)
To take the example of the author I was just working on: notice that in the list of publications for "The Afterlife Is What You Leave Behind", 2011-00-00 comes after 2011-07-26; however, in Ben Loory's chronological summary page, "Wings" (2010-00-00) comes before "The Poet" (2010-01-12). --Vasha 14:01, 13 September 2017 (EDT)
Thanks! I'll take a closer look once I am feeling better. Ahasuerus 14:41, 13 September 2017 (EDT)
Actually, it's not as involved as I suspected. Here are the 4 places where we try to account for various 0000-00-00 and YYYY-00 permutations when sorting records:
In other words, we have 3 "YYYY-00-00 before YYYY-MM-DD" scenarios and 1 "after" scenario. It would be easy to make things consistent, we just need to decide which option is better. Opinions? Ahasuerus 14:59, 13 September 2017 (EDT)
Actually, I like before better... and that's the easier option too... But I agree that 0000-00-00 should continue to be at the end of the list. --Vasha 15:17, 13 September 2017 (EDT)
I agree with Vasha. ···日本穣 · 投稿 · Talk to Nihonjoe 18:21, 13 September 2017 (EDT)
I think I prefer the "before" option as well. Ahasuerus 20:11, 13 September 2017 (EDT)
I can live with both. Stonecreek 23:50, 13 September 2017 (EDT)
I would say that exact dates should be first but I can live with either. Annie 10:17, 14 September 2017 (EDT)

(unindent) OK, the change has been made. Please note that it also affected the "Publisher Year" page. For example, "1978-00-00" is now displayed before "1978-10-00" on this page. Ahasuerus 16:05, 14 September 2017 (EDT)


Our friends and co-conspirators at SFE3 have added an article about the UK magazine Punch. We already have a couple of issues on file, but the SFE3 article includes additional information, notably a 1960 series of SF parodies written by well-known authors. In addition, Advanced Title Search for "Title Note contains Punch" finds a number of additional SF stories published in Punch (like this one.) Calling for volunteers. Ahasuerus 16:55, 14 September 2017 (EDT)

I'll work on it. -- JLaTondre (talk) 18:41, 14 September 2017 (EDT)
Thanks! Ahasuerus 19:02, 14 September 2017 (EDT)
I have entered magazines for:
  1. Existing genre stories which had first published in Punch notes containing the full issue date.
  2. New stories from the SFE3 entry.
There are some additional genre stores stating first published in Punch, but without the full issue date. I tried searching for more information, but didn't find anything. I couldn't find an obvious Punch bibliography, etc. to make it easy. If anyone wants to try their Google-fu to try and find dates for those issues, that would be a help. -- JLaTondre (talk) 19:52, 14 September 2017 (EDT)

Invalid birth places

This report indicates that this author entry has an invalid birthplace. Following the instructions at Template:AuthorFields:BirthPlace, I changed it from "Burford, Oxfordshire, England" to "Burford, Oxfordshire, England, British Empire", but it's still showing up in the report. Since that location was in the British Empire during at the time of his birth (1753), I'm not sure what it needs to be changed to. Ideas? ···日本穣 · 投稿 · Talk to Nihonjoe 19:14, 15 September 2017 (EDT)

For England, the software recognizes the following combinations as valid:
  • Kingdom of England (before 1707-05-01)
  • England, Kingdom of Great Britain (1707-1800)
  • England, UK (1801 or later)
I thought the template had sections for England and Scotland, but apparently not. Ahasuerus 20:47, 15 September 2017 (EDT)
The template has been updated. Ahasuerus 20:50, 15 September 2017 (EDT)
Okay, thanks. ···日本穣 · 投稿 · Talk to Nihonjoe 20:56, 15 September 2017 (EDT)

Date issue in one-line title summaries

While we're discussing display issues, here's something to consider. Currently, regular titles have a top-level record that combines the canonical title, the canonical name of the author, and the date of first publication (or of first book publication in the case of serials). Variants and translations have their date set to the first time that title+author combination appeared (or that's how it's supposed to be done in theory; it's applied extremely inconsistently). Here's how some variants appear when displayed in tables of contents. The first is the canonical name + canonical title; the second is a variant title (and pseudonym also); the third is a canonical title + pseudonym.

1. Record # 1197 | Title: Triplanetary | Author: Edward E. Smith | Date: 1948-00-00

  • Triplanetary • [Lensman • 1] • (1948) • novel by Edward E. Smith

2. Record # 893797 | Title: Triplanetary: A Tale of Cosmic Adventure | Author: Edward E. Smith, Ph.D. | Date: 1954-00-00

  • Triplanetary: A Tale of Cosmic Adventure • [Lensman • 1] • novel by Edward E. Smith (variant of Triplanetary 1948) [as by Edward E. Smith, Ph.D.]

3. Record # 1281436 | Title: Triplanetary | Author: E. E. Smith, Ph.D. | Date: 1965-08-00

  • Triplanetary • [Lensman • 1] • novel by Edward E. Smith [as by E. E. Smith, Ph.D.]

I hope you can see the problem: in the third instance, the date of original publication does not appear. You will be on a publication page when you're looking at that; if you want to know the original date of the work, you click on the contents item to go to the title page, and that takes you to a variant title with a variant date, and you then have to click again and go up to the top-level title to find the original date. What's needed, then, is a revision of the display so as to include the original publication date in all three of those combinations.

It's actually not simple to do that. The current logic has the year of the variant displayed following the title only for the top-level title, as "Triplanetary • (1948)", and that's a good thing; it would be confusing if the year of the variant was always displayed, because we'd see something like "Triplanetary • (1965) • novel by Edward E. Smith [as by E. E. Smith, Ph.D.]" and it wouldn't be clear that that wasn't the date of first publication. So instead, in scenario #2, the date is placed in the "variant of... ", thus: "(variant of Triplanetary 1948)". But that leaves nowhere for the date to go if there isn't a "variant of...", that is, in scenario #3.

I've been thinking about this for a while, and I'd like to submit, as a shy proposal, a way of rewriting these table-of-contents summaries. How about VariantTitle • (OriginalYear) • length by CanonicalAuthor (variant of CanonicalTitle) [as by Pseudonym], thus:

1. Triplanetary • [Lensman • 1] • (1948) • novel by Edward E. Smith
2. Triplanetary: A Tale of Cosmic Adventure • [Lensman • 1] • (1948) • novel by Edward E. Smith (variant of Triplanetary) [as by Edward E. Smith, Ph.D.]
3. Triplanetary • [Lensman • 1] • (1948) • novel by Edward E. Smith [as by E. E. Smith, Ph.D.]

There's only one problem with this: translations. I would prefer to have the date of the first appearance of that translation displayed next to it rather than the date of original publication. Here's how translations currently look, using the example of a translation that first appeared in 1961:

  • Die grauen Herrscher • [Lensman • 4] • novel by Edward E. Smith (trans. of Gray Lensman 1951)

(That is to say, the translation title is being treated the way a variant title would be currently, except for the notation "trans. of..." instead of "variant of...".) I would like to see translations either continue to look like that, or look like so:

  • Die grauen Herrscher • [Lensman • 4] • (1961) • novel by Edward E. Smith (trans. of Gray Lensman 1951)

However, we do already treat translations differently from variants for some purposes, so shouldn't it be possible to display the summary line differently for them also? Even if it's not possible, and we had to display translations as "Die grauen Herrscher • [Lensman • 4] • (1951) • novel by Edward E. Smith (trans. of Gray Lensman)" to be uniform with variant titles, I would still prefer that to the current state of scenario #3.

I hope this all makes sense! --Vasha 22:41, 15 September 2017 (EDT)

Interesting points, thanks. I'll need to spend some time playing with the data to internalize the permutations... Ahasuerus 12:09, 16 September 2017 (EDT)
Here's another thought; how about displaying both the variant date and the original days after three title, something like this: "Triplanetary: A Tale of Cosmic Adventure • (1954 / 1948)" "Die grauen Herrscher • (1961 / 1951)"
Here are a few more permutations.
1. Publication Record # 633461
Current: Spiegel, the Kitten • novelette by Gottfried Keller (trans. of Spiegel, das Kätzchen. Ein Märchen 1855)
Proposed: Spiegel, the Kitten • (1946 / 1855) • novelette by Gottfried Keller (trans. of Spiegel, das Kätzchen. Ein Märchen)
Note: Currently no variant date displayed because the variant title date (1946-00-00) is the same as the publication date
2. Publication Record # 421763
Current: Cultos da carga da Ilha do Beijo Picante • novelette by Rhys Hughes (trans. of The Cargo Cults of Salty Kiss Island 2017)
Proposed: Cultos da carga da Ilha do Beijo Picante • (2007 / 2017) • novelette by Rhys Hughes (trans. of The Cargo Cults of Salty Kiss Island)
Note: Translation published before original
3. Publication Record # 458637
Current: Fantastic Mr Fox • (1970) • novella by Roald Dahl (variant of Fantastic Mr. Fox)
Proposed: Fantastic Mr Fox • (1970) • novella by Roald Dahl (variant of Fantastic Mr. Fox)
Note: Variant year same as original year.
--Vasha 14:34, 16 September 2017 (EDT)

Poe canonical titles

Currently Poe's stories have their title of first appearance as their canonical title, even though often that title is very rarely used (in a number of cases, such as "Morella" or most of the Tales of the Folio Club, we have no publications using it in the database). What do you folks think about changing to the most common variant as canonical instead? --Vasha 15:24, 18 September 2017 (EDT)

I think we should leave it as it is: It would lead to some chaotics and in theory, these publications should eventually be added. Stonecreek 23:59, 18 September 2017 (EDT)
I agree with Christian. Another wrinkle to consider: If we're dating the variants based on when they first appeared, having something other than the original as the parent will cause everything to appear with an inappropriate date. --MartyD 07:36, 19 September 2017 (EDT)
There are actually quite a lot of cases in the database where the canonical title is the most familiar, widely used variant rather than the oldest one. In that case, the top-level record is given the date of the first publication. The combination of canonical title+canonical author name+oldest date may never have actually appeared in a publication, but it provides a standardized top-level title record. That said, I'm not actually very enthusiastic about creating such records for Poe. I agree with you two that it might disrupt work that people have done. --Vasha 10:59, 19 September 2017 (EDT)

Documenting Verne translations.

Someone has done a wonderful job of documenting the various translations of "Voyage au centre de la Terre" here. Unfortunately, this has not (yet) been extended to his other novels. I'm prepared to take a shot at this, but wanted to discuss the organization of the material first.

The translation information is laid out chronologically. If you examine the first title for a translation (e.g. the 1877 translation), the first appearance and translator are provided, as well as the text that characterizes that translation. Subsequent titles under the translation (e.g. 1965 retitled translation) repeat the first appearance and translator, but not the text.

For someone entering a publication where the translation/translator are not identified (as is frequently the case), it is not clear which title should be used, or when a new one should be generated. It requires opening each of the primary translation titles to locate matching text. Then, if the titles do not match, checking in the notes for the date of the earliest match for the title. Only then can the title be selected and a publication be added. This assumes that you do not have an earlier publication, for which you would have to updated the notes (as the dates are not generated).

Given that we do not have ways to identify translators or translations, or to make variants of variants, I was hoping we could still provide a better approach for people entering publications. I would like to make the identifying text more obvious by moving it up to the notes. For two different translations this might work, but here there are twelve. (The thirteenth is for unidentified translations - meaning someone did not have (or use) the text to differentiate the translation.) The amount of text would push the publications down by a screen or two. Is there some way to handle longer title notes? Is there a simple way to refer or even link to the titles without relying on the date? Doug H 15:07, 21 September 2017 (EDT)

Personal tools