User talk:Ahasuerus/Archive/2016

Jump to navigation Jump to search


This author is showing up on this clean-up report, and I have no idea how to remove the listing. Please help. Thanks. Mhhutchins|talk 05:30, 3 January 2016 (UTC)

This report identifies "Non-Latin authors with Latin characters in legal names". Since Jin's working language is Japanese, the fact that our version of his legal name contained the letter "P" triggered an alert. However, it turns out that "自然の敵P" is apparently his nom de theatre rather than his legal name, so I have removed it. I will check with our Japanese-savvy editors to see if they know more about this performer/author. Thanks! Ahasuerus 14:40, 3 January 2016 (UTC)

Pavel Vešinov

I've tried but can't figure out how to remove this author from this clean-up report. Can you help? Mhhutchins|talk 17:36, 13 January 2016 (UTC)

It turns out that "Pavel Vešinov" was a German transliteration of Pavel Vezhinov's name. I have updated the author record and added his legal name, "Гугов, Никола Делчев" (Gugov, Nikola Delchev). The report should be happy now :-) Ahasuerus 21:00, 13 January 2016 (UTC)

Peter Graves cover images, secure and insecure servers at Amazon

Hi. Re ISFDB:Community Portal#Cover images from Amazon, I believe that I found it via some book reviews or bookseller site(s) --as perhaps or where an option was provided automatically to "view a larger version of this image" or something of that ilk. Upon selecting that option, click, I was served the page at; at the ShopSwap page now, I know only how to save the image location.

There were numerous google hits for Peter Graves: An Extraordinary Adventures of which I saved only a few links in notes on my user page. There I experiment with all four URL, both http and https protocols for both pathnames. [Please view this section]. Do you know whether there is any urgency for me to clean up such notes, presumably copyright violation, or is this ok to retain in user space? --Pwendt|talk 22:50, 13 January 2016 (UTC)

Answered on the Community Portal. Ahasuerus 23:57, 13 January 2016 (UTC)

It's "J" time!

This report is clean. Guess it's time to tackle the big one. Thanks. Mhhutchins|talk 18:08, 15 January 2016 (UTC)

Guess which monthly cleanup report has been running on the development server for the last 2 hours? ;-) Ahasuerus 18:26, 15 January 2016 (UTC)

Length Changes

When editing a pub, if the type of an entry is changed from shortfiction to something else (essay, poem) and then the length field is changed to "-", the length field is not actually cleared when the edit is submitted and approved. It remains as "sf". Could this be changed so a separate title edit is not required? Thanks. -- JLaTondre (talk) 21:04, 16 January 2016 (UTC)

That would be Bug 563, a relatively big can of worms. I hope to split the "storylen" field into 4 (juvenile, novelization, omnibus designation, storylen proper) instead of applying yet another layer of band-aids to the crazy code that currently handles it. Ahasuerus 21:59, 16 January 2016 (UTC)
I can understand not wanting to band-aid the code, but it is annoying to have to make a half-dozen edits instead of one. Thanks. -- JLaTondre (talk) 13:00, 17 January 2016 (UTC)
OK, I will take a closer look once I get the current patch out the door. Ahasuerus 21:44, 17 January 2016 (UTC)

Odd display

This [page] isn't displaying correctly. Have a feeling all the html in the title area is causing it but I don't know how to fix it. --~ Bill, Bluesman 21:31, 16 January 2016 (UTC)

The Note field was missing the closing "/div". The immediate problem has been fixed, but I suspect that the page would look better if we used the recently added {{BREAK}} functionality described here. Ahasuerus 21:54, 16 January 2016 (UTC)
Thanks! So putting the {{BREAK}} in front of "For example ... " would move all that 'stuff' to another page? Does one need to close a {{BREAK}}?? Mark put the note there. I don't really see the point of it being so extensive, a couple of lines would have accomplished the same thing. --~ Bill, Bluesman 22:26, 16 January 2016 (UTC)
That's right. I have added {{BREAK}} in front of "For example ... " and you can see how it looks now. Ahasuerus 22:36, 16 January 2016 (UTC)

Cleanup Report

Hello, I was wondering why Cara Bullinger stays on this report even everything seems OK to me. Usually, the lines disappear from (most) of the reports when refreshing as the errors are corrected but not in this case. Is there something that I didn't understand? Hauck 11:59, 17 January 2016 (UTC)

I caused that by reversing the pseudonym/parent relationship. (She had more credits as "Bullinger" than "O'Sullivan".) It will not appear when the report is ran again. Mhhutchins|talk 16:46, 17 January 2016 (UTC)
Thanks for the clarification, Michael! Ahasuerus 17:05, 17 January 2016 (UTC)

Authors Without a Defined Language

Re this cleanup report: Now that it's been cleaned for all authors with more than 10 titles, is there a plan about how far we should go with it? Comparing the benefits to the efforts involved, I'm pretty much convinced we've reached the end, or at least I have. Even if there are 10K authors in the database who haven't been assigned a language, what's the purpose of continuing work on the cleanup report to an ever decreasing benefit/effort ratio? If I recall correctly, the origin purpose was to get the list low enough so that titles without language can be automatically converted to English. Isn't it possible to do that for non-languaged title records created before the eligibility of non-English titles into the database? Wouldn't that be easy to do since record numbers are assigned consecutively? And by adding languages to these unassigned titles, couldn't we subsequently assign authors a language based on those results? Mhhutchins|talk 20:38, 18 January 2016 (UTC)

I too have been thinking about this issue for the last few weeks. After reading your post, I checked the software and realized that ... I am yet to tweak the four "Suspected XYZ Authors without a Language Code" to include variant titles. (VTs were originally excluded to make things more manageable.) Duh!
Once I deploy the changes and the four reports are recreated overnight, we will have additional 759+317+597+535 authors to clean up. Once that has been done, we can revisit the issue :-) Ahasuerus 01:38, 19 January 2016 (UTC)

June pubs from Fixer

Is there a reason why the three pubs to be published in June slipped through? I'm assuming these are Fixer submissions, or at least the first two are. I went back to the Jan. 14 approved submissions and weren't able to find them. Thanks for checking. Mhhutchins|talk 20:00, 19 January 2016 (UTC)

Sorry, my bad. Back when Fixer captured these ISBNs, their projected publication dates were 2015-something-or-other, but they have since been pushed back to 2016-06. Fixer double checks the publication date (as well as all other Amazon-derived fields) before creating a submission record and then I review everything prior to actual submission. I have a special semi-automated process in place to avoid these kinds of problems, but in this case I didn't follow it in order to save time. I will go ahead and move these three pubs back to Fixer's internal queues. Thanks for finding them! Ahasuerus 22:02, 19 January 2016 (UTC)
P.S. Correction: it looks like I am only responsible for The Gospel of Loki. The other two pubs were entered manually. Ahasuerus 22:15, 19 January 2016 (UTC)
P.P.S. Pubs deleted. Ahasuerus 22:21, 19 January 2016 (UTC)
I get the feeling the other two may have referred to: 2914159 and 2914225 which were Plunder of Gor pubs approved by Hauck and Mhhutchins. The title record was actually at the top of the View Top Forthcoming list before it was deleted. BTW, can we get the submission record display to show the submission date, subject, etc. (e.g., moderator) too (I can see that in the list but that is hard to link to)? Uzume 02:05, 28 January 2016 (UTC)
That's right, the deleted pubs were two different editions of Plunder of Gor. Fixer will resubmit them when we reach the standard 60-90 day window. Ahasuerus 02:12, 28 January 2016 (UTC)
Ya, I found the deletion records: 2925608, 2925609, and 2926282. But you cannot see the submission subjects, dates, submitters and moderators from those links (only in the hard to find and link to submission listings). Uzume 02:36, 28 January 2016 (UTC)
As far as I recall, I didn't create a publication record for Plunder of Gor. I believe I fixed a problem with it (maybe a Fixer submission or it was on a cleanup report?), and then realized its publication was too far in advance. That's what brought up the reason for my original post here. Mhhutchins|talk 02:37, 28 January 2016 (UTC)
Submission 2914225 documents you approving a ClonePub submission for Plunder of Gor. Uzume 02:42, 28 January 2016 (UTC)

Also, I clicked on the link "View Top Forthcoming" (which I'd never done before) and got this page. Is that an expected result... being that no forthcoming novels are of interest to ISFDB users? Mhhutchins|talk

Oh dear. The logic behind this page basically says "Find top forthcoming publications that will appear prior to 2015". Since the code was last modified in 2010, the programmer apparently assumed that 2015 was always going to be in the future. I'll give that programmer a stern talking-to next time I look in the mirror! Ahasuerus 22:02, 19 January 2016 (UTC)
After you do that, then tell him he's doing an amazing job. That's how management should handle their employees. :) Mhhutchins|talk 22:49, 19 January 2016 (UTC)
Done! (And thanks!) Ahasuerus 23:45, 19 January 2016 (UTC)
Fixed and prettified. Ahasuerus 00:21, 20 January 2016 (UTC)

Translated titles with transliterated author credit

This cleanup report discovered something interesting, probably more than it was expected to find. Some ISFDB editors are entering translated titles with author credits which are also translated (or transliterated). I changed a couple before I thought you should see the problem. I'm assuming the policy of entering the author credit for translated works in the author's canonical alphabet hasn't changed. So a work by a Russian author should be credited to his canonical name when if it is published in Japanese (see here where it wasn't.) Mhhutchins|talk 20:12, 19 January 2016 (UTC)

It's possible they may not have known the correct name. I try to determine things like that when I enter Japanese publications of Western authors, but I can't always figure out the names. Sometimes it's very hard to figure out. ···日本穣 · 投稿 · Talk to Nihonjoe 21:05, 19 January 2016 (UTC)
I should note that I didn't enter the one in question. My comments were a little vague on that point. ···日本穣 · 投稿 · Talk to Nihonjoe 01:44, 20 January 2016 (UTC)
I'm going to go ahead and change the author credit to the canonical form of their name. Thanks. Mhhutchins|talk 22:50, 19 January 2016 (UTC)

(unindent) That's right, the current policy is to romanize all author names. Some parts of the software still need to be upgraded before they can handle non-Latin names, notably the Author Directory and our Wiki-based "Bio" and "Biblio" pages. (In the past, using non-Latin names could cause even more serious problems, including Python errors, but all of them have been fixed.) At the rate we are going, the software changes that will allow us to start transitioning to native scripts/alphabets will be implemented in a few months. Ahasuerus 01:44, 20 January 2016 (UTC)

That will be nice. :) ···日本穣 · 投稿 · Talk to Nihonjoe 01:45, 20 January 2016 (UTC)
I hope the change doesn't occur until we've come to a consensus about whether a variant author credit is a true pseudonym or a variant spelling and/or alphabet, especially if it's a translated title. It would be nice to distinguish the two before we start packing the variant function into another box of which it's not able to fit. Mhhutchins|talk 02:15, 20 January 2016 (UTC)
Well, there is an FR to "Differentiate pseudonym types". The request as currently worded asks to differentiate between "name variations" like "R. A. Heinlein" and "true pseudonyms" like "Anson MacDonald", but we could add another flavor for "translated names".
However, I expect that the change will be made at the author/pseudonym level rather than at the title level and the current VT functionality will be largely unaffected. Ahasuerus 02:39, 20 January 2016 (UTC)
Not exactly. Will the titles under these "fake" pseudonyms be varianted? Given a choice, I'd prefer that they weren't. Mhhutchins|talk 04:27, 20 January 2016 (UTC)
The current plan is to continue using the VT mechanism for all types of pseudonyms. However, implementing this FR will let us adjust the way things are displayed. For example, Robert Heinlein's Summary page currently says "Used These Alternate Names:" and lists 10 "pseudonyms". Once the proposed FR has been implemented, we will have three separate lines:
  • Pseudonyms: Anson MacDonald, Caleb Saunders, John Riverside, Lyle Monroe, Simon York, Anonymous #1
  • Name variations: Robert Heinlein, R. A. Heinlein, Robert Anson Heinlein
  • Translated names: Robert Hajnlajn
Also, once we assign language codes to all titles, we will be able to change the way variants appear on Summary and Titles pages. For example, "If This Goes On —" will have three indented subsections: VTs, translations and serializations. (We'll have to figure out what to do about translated serializations.)
What kinds of changes did you have in mind? Ahasuerus 04:45, 20 January 2016 (UTC)
What is the timeline for this name variant type field? Uzume 02:34, 23 January 2016 (UTC)
The current plan is to:
  • Add "transliterated values" fields to publishers, series and titles
  • Move all Wiki-based Bio and Author pages to the database
  • Add a "transliterated name" field to authors and convert the data
  • Sort out Author and Publisher directories
  • Add translator support
  • Take a deep breath and see where we are Ahasuerus 03:00, 23 January 2016 (UTC)
I am not sure I like the idea (we do not have title variant type fields and such could be more useful to differentiate related/translated works from just ones with different titles and/or different credits) Uzume 02:34, 23 January 2016 (UTC)
The software can already determine that a VT is a translation as long as both title records have language codes assigned to them. On the other hand, there is no easy way to determine that "R. A. Heinlein" is a "name variation" as opposed to a pseudonym. For "translated names", I suppose you could pre-scan all titles and check if a name is only used by translated works, but it wouldn't be dispositive because it could be a pseudonym instead. Ahasuerus 03:09, 23 January 2016 (UTC)
Since author name records have a language field we could mark Robert Hajnlajn as Serbian and since is it also marked as a pseudonym of Robert A. Heinlein (English) the software could determine it is a translated name and decide to do something different with it. In a similar vane, all the Japanese variants listed in his NDL record could also similarly be marked as Japanese (once we have native script author name support and publications with such credits of course). I am not sure how you would want detected translated names to be handled by the software (do we render them differently in some contexts?) but the detection could work much like translated works does now. Uzume 00:39, 28 January 2016 (UTC)
as some names surely will not fit into one of those categories (and it will be hard to judge such things anyway). Uzume 02:34, 23 January 2016 (UTC)
That's an interesting point. Would you happen to have an example of a name that couldn't be categorize this way? Ahasuerus 03:09, 23 January 2016 (UTC)
Sure, let's look at: 安部, 公房. In our current system he is listed as Kobo Abe. Using the three types of names you have, how would you type each of the following: 安部公房 (native script; probably the most common way he is credited in native Japanese works), Kobo Abe (probably the way he is most often credited in Latin translations), Kōbō Abe (a better Latin transliteration showing the longer vowel sounds), Abe Kóbó (a Latin transliteration apparently also showing longer vowel sounds used on some Hungarian translations), アベ, コウボウ (a Japanese transliteration based on native misreading/alternative reading of his real name that can be directly translated into Abe, Kōbō and is likely the source of the most commonly used name in translations: Kobo Abe; researched evidence shows he may actually proscribe such usage in areas like translations), アベ, キミフサ (a Japanese transliteration of the proper pronunciation of his real name), Kimifusa Abe (a Latin transliteration of the proper pronunciation of his real name).
Once we have native script canonical author name capability "安部公房" will surely evolve into the canonical name as I am sure his works are most famous and commonplace to his native Japanese audiences. But what about Kobo Abe? Is this a translated name (it surely is not his his native)? A pseudonym (since it is most common on his translated works and he may actually subscribe to such usage)? or a variation (since it is effectively a transliteration of a misreading of his name)? Uzume 00:39, 28 January 2016 (UTC)
Some people even change their legal names to what they had as nicknames before, etc. Uzume 02:34, 23 January 2016 (UTC)
Right. Andre Norton and Simon Hawke come to mind. In addition, we have people like Murray leinster, whose legal name, "William Fitzgerald Jenkins", is currently listed as an "alternate name". If we decide to implement this FR, we'll need to decide what to do with them. Add more values to the drop-down list, perhaps? Ahasuerus 03:16, 23 January 2016 (UTC)
Those are prime examples of why typing names is a slippery slope that I do not think we want to go very far down, if at all (when you mentioned Murray Leinster, I thought of John Norman). Name usage can change over time too (even after an author is dead). Uzume 00:39, 28 January 2016 (UTC)
And what about people that translate works? If I translate and have published works between English and Japanese and we eventually get credit fields/records for such, how will my name appear within such works? Uzume 02:34, 23 January 2016 (UTC)
The intent of this FR was to differentiate between various "alternate names" which appear at the top of author pages. You seem to be raising a different issue: how will these types of "alternate names" appear on title/pub pages. Is that right? Ahasuerus 03:30, 23 January 2016 (UTC)
No my thinking was what defines these types. If I am multilingual and get works published in multiple languages (and multiple scripts) which names are pseudonyms? which are translations? which are variations? In my mind a the key defining factor in a "pseudonym" is whether an author proscribes its use (e.g., actively getting something published under a certain name) but even that is not always true (especially for very long dead authors) and changes with time. What constitutes a variation though? In many cases this is fairly clear but in many others it is not (we are going to run into cultural differences here fast). I think of all of these the translated or foreign name is the most straightforward as it is a name used on works in a different language that the original author did not actively proscribe but that adheres to the cultural and linguistic norms for that cultural (used to help with pronunciation and/or needed because of differences in scripts, etc.). Uzume 00:39, 28 January 2016 (UTC)
Are those treated as pseudonyms or translated names? What about variations of translated names? Will we have types for such too? And remember not all translated names will be Latin, e.g., John Norman has six titles (the first in his huge Gor series) translated into Japanese and I bet the publications likely do not credit him in a Latin name (I can look it up for each pub of course)—more likely it will be ジョン・ノーマン or some variation that matches his NDL record entry. When we get non-Latin names, these (among many other titles/pubs) will have to get updated (and cleanup reports to find such will be useful). This is the danger of using transliterated fields and not using a variant record system. Uzume 02:34, 23 January 2016 (UTC)
I am not sure I am parsing your questions re: translators correctly. Could you provide a (hypothetical) example? Ahasuerus 03:34, 23 January 2016 (UTC)
Well let's simplify things and take translators out of the mix. The following names are derived from his NDL record (I linked this above in one of my earlier comments) for what we have here as Robert A. Heinlein: R・ハイライン, R・A・ハインライン ; ロバート・ハインライン, ロバート・A・ハインライン (these correspond to: R. Heinlein, R. A. Heinlein, Robert Heinlein, Robert A. Heinlein). Which of those are variant names and which are translated names (or are some of them a new type: translated variants)? Should any of them be considered as a pseudonym (Did he actively participate in any of his translations and proscribe the use of any of those names for his publications; I really do not know and measuring that is not always easy; why do we need this again?)? Uzume 00:39, 28 January 2016 (UTC)

[unindent] I believe there should be only two types of variant names: a) true pseudonyms and b) alternate names, which includes misspellings, "foreign" versions, and non-pseudonymous shortenings of the canonical name. But here's where I differ in how I'd like variant titles based on author credit to be displayed: I think those titles where the author is credited to a "b" (alternate name) should be displayed with the canonically credited titles. Variant titles based on "a" credits (true pseudonyms) would have a separate title record page. In other words, the software would "suppress" the display of "b" credited titles, forcing them to be displayed with the canonical title. In the publication record, the actual credit, whether "a" or "b", would naturally have to be displayed, as they are now "[as by...]". I don't know if that's possible, but I'd sure like to see if it could be written that way. (Of course, variants based on a variant title would still be handled and displayed as they currently do.) Mhhutchins|talk 21:00, 23 January 2016 (UTC)

So you are saying you want to see author pages (and perhaps title pages) which only denote titles and pseudonyms (where applicable) and suppresses displaying any alternate namings (however that is defined/determined), right? I think this goes towards how we handle variant naming in general. Currently we have complete title and author records that, to one level of indirection anyway, can refer to/defer to a canonical record of the same type (variant titles to titles and pseudonym authors to authors). We could introduce some sort of other alternative names in a light not unlike transliterations are probably being introduced (though methinks this is potentially frought with problems much like I believe transliterations will be). They then could possibly be used differently in different situations such as you mention. If we had a more active editing system (something along the lines of what the Open Library has where as you type in the name it lets you choose from a pulldown the correct existing name) we could at entry time specify both a cited alternative and a pseudonym/canonical name for pubs. I still think this is a little dangerous as it starts to hide alternative names. Names are like words and their impact changes with time. What happens when an alternative or mispelling starts to dominate in common usage? It would complicate things for us to find and reverse then. Uzume 00:39, 28 January 2016 (UTC)

(unindent) Thanks for the feedback, folks! Originally I thought that the idea was to make the Summary page's "Used These Alternate Names" line prettier, but now it's clear that there are other rationales as well. We'll need to have a more comprehensive discussion on the Community Portal before we do anything. Ahasuerus 16:10, 1 February 2016 (UTC)


I wanted to bring your attention to a comment I made here: ISFDB:Community Portal#"Titles without Pubs" made available to non-moderators. Uzume 02:07, 23 January 2016 (UTC)

Sorry about the delay. I have been under the weather the last week and haven't been very active. I'll respond shortly. Ahasuerus 02:10, 23 January 2016 (UTC)
I was more concerned that you read it. No worries—thanks Uzume 04:47, 23 January 2016 (UTC)

Possible Typo

Here is a possible typos:

Thanks. -- JLaTondre (talk) 22:05, 23 January 2016 (UTC)

Thanks for the heads up! I finished repairing my house a few days ago. Now I just have to sort out the (tens of thousands of) displaced books and magazines... Ahasuerus 22:12, 23 January 2016 (UTC)
No, that is a definite typo—either in the misspelled title name or the pub note (where it is noted without the misspelling). It is good you got your house repairs completed. Uzume 01:52, 28 January 2016 (UTC)

Data from publisher's Web site

When "Publisher's website" is selected as the "Source of the data" on a pub edit, the resultant note is "Data from publisher's Web site". That's an odd formation (especially with the capital). Could it be changed to "website" like is used elsewhere (pub edit & moderator approval screens)? Thanks. -- JLaTondre (talk) 00:14, 27 January 2016 (UTC)

Sure. FR 846 has been created. Ahasuerus 00:18, 27 January 2016 (UTC)
Done. Ahasuerus 15:54, 12 February 2016 (UTC)

Unmerging NOVELS from OMNIBUS

Hello, I've just found that when you unmerge a NOVEL title from an OMNIBUS, the "new" title created has not the title of the solo novel but the one of the whole omnibus. This may (and has recently) caused problems. Sorry to post this here, but alas I'm not good with FRs. Hauck 15:04, 27 January 2016 (UTC)

I'll have to take a look. At one point we changed the way Unmerge handles titles and it's possible that this was a side effect of that change. Ahasuerus 19:43, 27 January 2016 (UTC)
Perhaps it would have been better to use the remove function (by removing the title from the publication record), rather than to unmerge one title from another. The latter always creates a new title which matches the previous title. That's the way it was designed, and rightfully so. Mhhutchins|talk 19:54, 27 January 2016 (UTC)
Not in this case. Go to a NOVEL which is included in an OMNIBUS and unmerge it. The obtained title is not the one of the novel but the title of the OMNIBUS. It doesn't work this way for a SHORSTORY where the resultant title is the title of the short story. Hauck 20:00, 27 January 2016 (UTC)
Unmerging titles happens at the pub level. It basically is asking for pubs to have new (split) title records with the same name and type as the existing title record. I am not sure if we have restrictions in place for this but that is why it makes no sense to unmerge and check all the pubs (we probably should not even allow unmerging on single pub titles). At least one of the pubs should retain the original existing (selected) title. If you are unmerging a novel and not getting new novel records methinks that is an issue (but make sure you select unmerge from the novel title and not the omnibus title). As a side note, even container titles like omnibuses, magazines, collections and anthologies, not actually contain anything (the pubs do all the real containment)—they just help us document the pubs containment of such things (e.g., editorial credits, etc.). Uzume 01:37, 28 January 2016 (UTC)
There's a submission in the queue which attempts to unmerge an OMNIBUS title from a NOVEL title. Again, that's not the way to go about it. And the system shouldn't allow such an attempt. Mhhutchins|talk 02:53, 28 January 2016 (UTC)
The process always unmerges titles from pubs (perhaps a better is name is pub title split). I do not understand what is wrong with unmerging a novel title from an omnibus pub like the submission you quoted (not being a moderator I cannot see the mod link but I can see the submission here: 2933692). That said, I do not understand Hauck's reasoning for wanting to do such (does the omnibus not contain that novel but another novel by the same name?). It would make sense if the translators were different (I unmerged a collection title today for that same reason), however that does not seem to be the case in this submission. All the pubs under A la poursuite des Slans and A la poursuite des Slans claim Jean Rosenthal as translator; I would merge those two variant titles if it weren't for the difference in credits (A. E. van Vogt vs. Alfred E. van Vogt). Perhaps he wants to unmerge the novel in the omnibus to provide different credits for the novel in the in the omnibus (but he has no comments specifying such and that would be weird since the omnibus is currently two novels with the same credits). Uzume 03:24, 28 January 2016 (UTC)
Perhaps should it be thinkable that this particular submission was a test. Hauck 15:27, 28 January 2016 (UTC)

Let's start again one last time and make it simple. Let's say (it's HYPOTHETICAL) that this novel have seen two different translations (requiring two different titles as per a recent ruling). Let's say that I want to separate two of them form the stack. I choose to unmerge the first one, a simple novel, that gives me this submission which will create a new NOVEL also titled A la poursuite des Slans. As this different translation has been included in an omnibus here, I choose to unmerge the "contained" novel via the usual link at title level. This will give me this that will create a new novel with the omnibus title even if I never ask to work at the omnibus' title level which is here. In a nutshell the problem is visible here. It's quite misleading (and perhaps simply wrong). I brought this to a software guy's attention because a contributor (not me) had this problem with Vance and Laumer dutch omnibuses and created wrongly titled novels. Hauck 15:27, 28 January 2016 (UTC)

That was very detailed (and useful). It seems the name of (not the type of) the newly created title record is what you find issue with. I highly doubt the name of T1468493 had any bearing on the issue and rather the name of P392968 did. However, in this case the omnibus title name and the omnibus pub name are the same so I cannot be sure. We do have many cases where pub names do not match the title name of their main containing title. This is sort of the old style name variation methodology that is still in existence—often in cases where some pubs include subtitles in their names and others do not (but there are a few other reason too; it is not often valuable to create extra title records for such things).
I would have to do more testing or consulting the code to be sure but what seems to be happening is that code is naming the new title based on the type of the original title and takes either the name of original title (for non-container types like stories) or the first pub name selected (when the original title's type is one of a number of "container" types; which is questionable if one select multiple pubs with different names). I believe the issue you are seeing is that a novel can be both a container (when published as itself) and contained (when published as part of something else). Probably our code does not take that into account and just handles some types one way and others another (i.e., novel types always get handled in the container way by taking the first pub name). We could perhaps attempt to find a pub's main container title (omnibus pubs should have omnibus titles) and check if that is the same title being unmerged and handle those differently instead of purely upon type (this way novel titles would be handled different in novel pubs vs. elsewhere). Again this would not be perfect (especially when multiple pubs are selected) but I believe it was implemented to support cases like this (I have no rationale for doing this; this is a technical discussion): suppose I want to unmerge the collection T1573587 from the collection pub P556342 so the new title becomes Jeapes Japes: Featuring Stories from Interzone, Fantasy & Science Fiction, and Aboriginal SF and not Jeapes Japes (which is most likely what I wanted). That said, if I were unmerging a story title from the same collection (unfortunately these pubs are not so populated and one reason they were on my radar) then taking the pub name would not be useful.
I personally would rather have the new title always take the name of the original title but it would not be as nice a heuristic in the case of unmerging the main container title from a pub. Uzume 20:20, 28 January 2016 (UTC)

Arthur H. Landis / Camelot

Your input would be appreciated here. Thanks. PeteYoung 05:04, 28 January 2016 (UTC)

More user email validation problems

I've forwarded to you an email from a new editor who can't post on the wiki because he hasn't received an email validation request or it's not been validated by us. (I have no idea how it is supposed to work...even when it does work.) Here is another new editor to whom I can't send an email to let them know they need to respond to questions on their talk page. I can only hold submissions in the queue for so long. Maybe they want to respond but are prevented from doing so.

I still can't comprehend how someone is able to edit the database without validating their email address, but we prevent them from editing the wiki for the same thing. I thought the reason for email validation was more than just to block spamming on the wiki. Wasn't it also to make sure that we have a way to contact editors of the database? Mhhutchins|talk 18:51, 28 January 2016 (UTC)

As I recall, the original reason was to combat spammers. Ensuring that we had a way to contact editors was a secondary reason. I agree that it's a worthwhile goal, but if the e-mail validation requirement makes it harder for new editors to contribute, it's a net minus. Ahasuerus 21:08, 28 January 2016 (UTC)

The problem with email validation must be fixed or we'll practically driving away new editors who want to contribute. Who knows how many we've already lost? Shouldn't that be a priority? Mhhutchins|talk 18:51, 28 January 2016 (UTC)

Thanks for letting me know! I have poked around and there is something strange going on with our e-mail. On the one hand, the e-mail server is working: if I use the Wiki to send a test e-mail to my main e-mail address, it arrives within seconds. On the other hand, the server fails to deliver e-mail to certain addresses. For example, checking the logs, I see that George Seff tried to get his Verizon e-mail confirmed at 9:20am, but the server received a "Service unavailable" response from Verizon. It could be a Verizon-wide issue since this page currently says "Email outage: We're sorry, but we are experiencing an email system issue. Please check back in one hour." I am not sure if it's their "current status" page, though. Alternatively, Verizon may be blocking all e-mail coming from ISFDB -- they have been known to block legitimate sites by mistake. I am not sure what's happening with JKS since the status of the e-mail that requested on 2016-01-27 was " Message accepted for delivery". Perhaps it was caught by his spam filter?
In any event, this has become a bigger issue than I realized. Let me see if I can whip up a script which will let me authenticate unverified user accounts on demand. Ideally, the functionality will be available to all moderators from the submission review page. Stay tuned... Ahasuerus 20:31, 28 January 2016 (UTC)
P.S. I have confirmed that Gmail routes ISFDB-originated e-mail messages to their users' Spam folder. What a pain... Ahasuerus 21:11, 28 January 2016 (UTC)
I am not sure having a moderator bypass mechanism is a good idea (how does a moderator manually validate an email address?). Uzume 20:38, 28 January 2016 (UTC)
A new editor who has created a meaningful submission is ipso facto not a spammer. Spammers are not interested in creating submissions, they are only interested in spamming our Wiki. All we need to do is modify the submission review page(s) to display a yellow message along the lines of "This editor hasn't been verified. If the submission looks legitimate, click here to verify the editor". Ahasuerus 21:04, 28 January 2016 (UTC)
There are other ways to reduce posting/wiki spam such as with CAPTCHAs. That would allow us to not have to keep the email requirement and unblock editor responses on their talk pages. Of course it does nothing towards requiring email validation in general like what Michael seems to want (I do not really have qualms with that just trying to offer other technical options). We could have a combination where one either has a validated email address on file or has to fill out CAPTCHAs all the time (which would work around email issue situations like we are currently experiencing; they could also be built into the DB's web editing interface too). Uzume 20:38, 28 January 2016 (UTC)
I believe at one point I looked into various ways to combat spam on the Wiki side. Many of them required a Wiki upgrade, something that Al planned to do last year but hasn't gotten around to yet. Ahasuerus 21:16, 28 January 2016 (UTC)
Yes, I wanted to help work on that too (I did spend time looking into it several times) but I never had the keys and the time to do it. Uzume 00:32, 29 January 2016 (UTC)

(unindent) Please see my post on the Moderator Noticeboard about the changes that I have made in the last few minutes. Ahasuerus 00:40, 29 January 2016 (UTC)

Magazine grid table display

Looking at this grid, is it possible for consistent display when a single issue appears in a month? When there are two or more issues in the same month, each has their own box, but when there's a single issue, it isn't given a separate box, but is allowed to take up the whole space. And that space's size is dependent above the largest space of the entire year (the month with the most issues.) In the above example, the displays of months like February 2016, January 2013, and April 2010 are inconsistent with how they would have been displayed if other issues had appeared in the same month. Thanks for taking a look. Mhhutchins|talk 22:30, 3 February 2016 (UTC)

Good point. Done! Ahasuerus 23:54, 3 February 2016 (UTC)
Wow! Faster than a speeding bullet! Now, was that Superman or Mighty Mouse??? :) Either way, you've come to save the day. Thanks. Mhhutchins|talk 00:16, 4 February 2016 (UTC)
You are welcome! I haven't made a whole lot of progress on the development side lately, what with the house repairs and the collection restoration work (not to mention Fixer), but hopefully things will improve a bit now that the house is back in order. I just have to do something about this pattern of getting sick every winter -- apparently it has something to do with getting older. Who knew?? Ahasuerus 00:36, 4 February 2016 (UTC)

Help with authors with problematic legal names

Can you clear the two authors on this list? I've tried different combinations and none of them work. AFAIK, Shahrukh Husain only writes in English. And Boris Ivanov's legal name should be in Cyrillic. (I tried moving the Latin name to the other field, but it still didn't clear him from the list.) Thanks. Mhhutchins|talk 17:59, 5 February 2016 (UTC)

Ivanov was easy. Shahrukh Husain I am not so sure about. Wikipedia and BNF claim that her name is "شاہ رخ حسین" in Urdu, but it's not clear whether it's her legal name. A number of her books state that she was born and "brought up" in Pakistan and that she lives in London, but they don't explain her citizenship situation. FWIW, according to Penguin Books India:
  • Shahrukh Husain began writing while still at school. Her books include Urdu Literature, Demons Gods and Holy Men from Indian Myths and Legends and The Virago Book of Witches. Her books for children include Focus on India as well as several books on Islam. She is currently completing a series for children called Ancient Civilizations. Shahrukh has also written several screenplays, of which Merchant Ivory's In Custody, adapted in collaboration with Anita Desai, won the President of India's Gold Medal. Shahrukh lives in London.
We may want to re-post this question on the Community Portal. Ahasuerus 19:59, 5 February 2016 (UTC)
I meant to post something about this under "Help", but I forgot. Wikipedia gives both forms of her name, so it seems we should as well. Both Urdu and English are official languages in Pakistan, hence it seems likely that either form is a correct "legal" name. According to Wikipedia on Pakistan's languages, Urdu is viewed as the "lingua franca" of Pakistan, but English is the language used "in official business, government, and legal contracts", so I suspect that "Shahrukh Husain" would be viewed as her "legal name". So I don't know if there's any place to put the Urdu version, except just in the notes. Chavey 01:37, 6 February 2016 (UTC)
I placed the Urdu name into the legal name field, and the English one in the transliterated legal name field. I'm not sure if it's correct, but it did remove her from the cleanup report. (And that's all I wanted to do in the first place.) Thanks. Mhhutchins|talk 01:55, 6 February 2016 (UTC)

Use publication's title as value for <title> element of the publication's page

Hi. At the moment, each publication page uses the same fixed value for the page's "title" element: <title>Publication Listing</title>. Since this value is usually used for the title of a browser tab it's impossible to distinguish which tab contains which publication if you have several browser tabs open. Can you change it to use the publication's title? Jens Hitspacebar 12:29, 6 February 2016 (UTC)

That's a very good point. I have adjusted both the Publication page and the Title page to display this information (more) consistently. How does it look? Ahasuerus 18:13, 6 February 2016 (UTC)
Looks great! Thanks a lot. Jens Hitspacebar 18:33, 6 February 2016 (UTC)
You are welcome! Ahasuerus 19:58, 6 February 2016 (UTC)

Lancelot Biggs

I'm going to add a cover scan to Bond's The Remarkable Exploits of Lancelot Biggs: Spaceman. I'm also going to copy the note from the first edition as to why we are listing the stories as opposed to treating the fix-up as a novel. Were it those two edits alone, I would have noted this on your other page, but I can also add the price ($1.00). Thanks. --Ron ~ RtraceTalk 02:40, 9 February 2016 (UTC)

Sounds good, thanks! Ahasuerus 02:44, 9 February 2016 (UTC)

Wrong links to help on "Publisher Editor" page

On the "Publisher Editor" page the links of the "?" help icons next to the "Publisher", "Web Page" and "Note" fields all link back to the "Publisher Editor" page itself. Aren't these links supposed to point to a page in the Wiki instead? Jens Hitspacebar 23:29, 12 February 2016 (UTC)

Looks like a bug to me. I'll take a look - thanks! Ahasuerus 23:55, 12 February 2016 (UTC)
Fixed. I have linked the main Edit Publisher Help page (rather than the four templates) since the page is fairly small. Ahasuerus 01:21, 13 February 2016 (UTC)
Looks good. However, I've discovered several other pages which have the same problem. It's just a minor problem actually, but if you click on such a help icon you'll lose everything you've so far entered in the input fields. It occurs in all places in "edit/" where an empty string (no link) is used as second value of the array which is added to the help array:
help[KEY] = [text, '']
But that's not causing the problem. The real cause is that the "printtextarea" function in "edit/" creates an empty href attribute 'href=""' for the "a" element. Browsers treat such an empty href as a link to the same page (see 4.2 in RFC 2396: "A URI reference that does not contain a URI is a reference to the current document."). I suggest to inspect the value for "help[label][1]" and completely omit the "a" element if "help[label][1]" is empty here:
print '<td class="hint" title="%s"><b>%s: </b><a tabindex="0" href="%s">' % (help[label][0], label, help[label][1])
Jens Hitspacebar 09:11, 13 February 2016 (UTC)
The problem actually exists not only in "printtextarea" but in all functions in "edit/" which use "help[label][1]" to create an "a" element. Jens Hitspacebar 09:22, 13 February 2016 (UTC)
Oh, I see. Thanks, I will clean it up! Ahasuerus 15:29, 13 February 2016 (UTC)

Galactic Diplomat

to be posted also at User talk:Biomassbob -Pwendt

The stated first edition you verified P14964 needs correction of a typo in the subtitle, namely DIplomatique => Diplomatique.

With nothing else to say I would ignore that. There is a review in Toronto Daily Star 1965-10-02 p59, "He's a daring diplomat of the 27th century" by T. Humeniuk, under a subheading that I replicate here:

Keith Laumer. Doubleday.
227 pages. $4.75.

That price must be C$4.75. If the publication were not primary verified I would correct the subtitle; link LCCN 65-11536; add the Notes list item "Review Toronto Daily Star 1965-10-02 p59 gives price C$4.75."

As I write, however, it occurs to me that it would be more useful to note the review at title record T36965. Do you agree? As I understand, that would belong in a Note because Toronto Daily Star is not a genre publication. Right? --Pwendt|talk 23:31, 18 February 2016 (UTC)

Sounds about right. Ahasuerus 00:00, 19 February 2016 (UTC)

The title record now lists one year 1965 review among five reviews in the database, published in FSF Sep 1965. I don't know when that would have been printed or what it implies about the book's publication date. --Pwendt|talk 23:31, 18 February 2016 (UTC)

It's hard to tell. The F&SF reviewer (Judith Merril) may have received an ARC while the Analog reviewer (P. Schuyler Miller) didn't. Ahasuerus 00:00, 19 February 2016 (UTC)

(I found this by automated search for 'Galactic Diplomat', which hits no other 1965 or 1966 review in any newspaper covered by the ProQuest subscription at this university library, which includes in North America alone at least 2 Toronto, 1 Boston, 1 Hartford, 2 New York, 1 Chicago, and 1 L.A. daily newspaper --those I recall hitting in automated searches.) --Pwendt|talk 23:31, 18 February 2016 (UTC)

If I understand correctly, the subtitle typo will be corrected automatically at COVERART title T1845985. --Pwendt|talk 23:36, 18 February 2016 (UTC)
Editing the title field of a publication record doesn't generate any automated changes. It doesn't correct the title field of the title reference record, and it doesn't affect the title of the COVERART record either. These changes must be done manually. Mhhutchins|talk 03:16, 19 February 2016 (UTC)
I have already fixed the typo in the COVERART titles, so we should be all set in that respect. Thanks for finding the review! Ahasuerus 00:00, 19 February 2016 (UTC)
By 1965, Doubleday had started using gutter codes to indicate the week the book was printed. This can give a pretty good estimate of the publication date. I've found that it averages about six weeks after printing. Mhhutchins|talk 03:16, 19 February 2016 (UTC)

Display of unpublished works

Would you mind if I started a post on the community page to get some feedback about the use of the highlighted display for unpublished works? I don't feel this is a rules and standards issue, and that we need feedback from as many editors as possible. (I know some editors avoid R&S discussions.) Thanks. Mhhutchins|talk 21:24, 22 February 2016 (UTC)

Oh no, please go right ahead! I'll happily implement whatever works best for the other 96%! :) Ahasuerus 21:49, 22 February 2016 (UTC)

Entering data

Hi, is there an easier way to change the biding format of a pub series instead of entering it one by one. For example in the series "Argonauta" all books have the same format (pb),is it possible to change them all in a one time operation? thanks.Wolland 14:57, 26 February 2016 (UTC)

I am afraid there is no way to do it manually, but we have a Web API that lets you create submissions automatically if you know how to code.
Alternatively, I could write a one-off script to change the data, but we tend to limit the use of one-off scripts to truly massive changes. The regular editing interface has all kinds of checks and balances intended to prevent users from messing up the data while one-off scripts have the potential to do serious damage if they are not thoroughly tested. Ahasuerus 16:11, 26 February 2016 (UTC)
This is fairly typical of pub series. I recently had to make some similar changes for some Japanese pub series (the entire series uses the same binding format). Uzume 14:00, 4 March 2016 (UTC)

Linking images

Hello, we've got now the permission to deeplink to another site (message was "As the author of this book, owner the rights, and board member of Nitchevo Factory pub., I allow ISFDB to link to the cover images hosted at, including the picture I provided the link for in this Title Page ("Sacra")."), the image concerned is for this publication, alas I don't know how to translate this in code. Can you do it? Thanks. Hauck 10:22, 27 February 2016 (UTC)

Sure thing! Give me an hour or two since I am currently working on the weekly backups... Ahasuerus 15:35, 27 February 2016 (UTC)
Thanks. Hauck 15:42, 27 February 2016 (UTC)
Done! Ahasuerus 20:23, 27 February 2016 (UTC)

"Love" by Yury Olesha

Would you be able to determine the original title of this story in order to variant it? It's on a cleanup report and without knowing the original Russian title in cyrillic I wouldn't be able to clean it. Also can you take a look at the four names on this report? I don't know how to clean them off the report and they've been sitting there for a couple of weeks. Thanks. Mhhutchins|talk 07:58, 4 March 2016 (UTC)

@Mhhutchins: My Russian is not as good as my Japanese but I believe you are looking for Любовь, first published in 1928 (which was also a part of Worlds Apart: An Anthology of Russian Science Fiction and Fantasy but the contents have not been entered here; there is one verified pub you might try asking them to enter data). I shall make a submission. Uzume 13:14, 4 March 2016 (UTC)
Looks good, thanks! Ahasuerus 15:52, 4 March 2016 (UTC)
I've had no more success with them ;-(. Hauck 08:56, 4 March 2016 (UTC)
I just cannot see those cleanup reports so I cannot address them. Uzume 13:21, 4 March 2016 (UTC)
There were 4 different problems associated with the 4 affected authors:
  • Stoyan Dukov: an "n" was accidentally appended to the end of his last name
  • A. Zaplavny: one of the "a"s was spelled using the Latin version of the letter
  • E. Zelikovich: the Legal Name field used the transliterated version of the name, which is not allowed for authors whose working language uses a non-Latin script
  • Moh. Ambri: it turns out that the modern version of the author's working language, Sundanese, supports Latin characters, so this was a false alarm
I have cleaned up the first three authors and will change our settings to allow Latin characters in Sundanese authors' legal names. Thanks for reporting the issues! Ahasuerus 15:52, 4 March 2016 (UTC)
Sundanese has been updated and the report is now empty. Ahasuerus 16:06, 4 March 2016 (UTC)
Also modified the report definition to let non-moderators access it. Ahasuerus 16:26, 4 March 2016 (UTC)

Flagging titles with variants as non-genre

A couple of titles showed up on the clean-up report which finds non-genre mismatches. They were titles for which you'd flagged the parent title, but not the variant titles. Is there a way for this to be done automatically? Thanks for checking. Mhhutchins|talk 07:33, 11 March 2016 (UTC)

That's a good point and I considered it during the design stage. Keeping the "non-genre" and "graphic format" flags automatically synchronized across parents/variants was possible, but it would have added a fair amount of complexity to the filing process. In the end I decided against it and created a cleanup report instead. I may revisit the issue next time I upgrade our filing code, something that I have been doing piecemeal for the last few years. Ahasuerus 15:27, 11 March 2016 (UTC)
It happens so seldom that it may not be worth the effort. The clean-up report should suffice. Thanks. Mhhutchins|talk 18:34, 11 March 2016 (UTC)

"Missing" Coverart

Would you please take a look at Cover: The Two Towers and Cover: The Return of the King? See the "missing" coverart (no pictures, just a "Coverart" link)? In each case, the pub has a link to an ISFDB image, but there is no image on the wiki. Did the images disappear? Or did someone get confused and link to the pub image page even if there was no image? Thanks. -- JLaTondre (talk) 23:11, 12 March 2016 (UTC)

Never mind, Bluesman deleted the image from the wiki. I'll ping him to see what he was trying to do. -- JLaTondre (talk) 23:13, 12 March 2016 (UTC)

An accent problem?

Hello. I regurlarly stumble on a specific problem with 'accented' authors. When I enter some bibliographic data, I didn't encounter any problems, but the data entered seems not to be linked at the author level (the link stays red although the corresponding page seems to exist) and it's impossible to "go back" from the bibliographic data page to the author's page (with an Author not found: message). An example I have entered a bibliographic comment for this author that points there, when validating it, I have this page but when trying to go back to the author got the following message "Author not found: Isidore Ducasse Conde de Lautréamont". I suppose that the accent is the problem. What do you think? Thanks. Hauck 16:48, 16 March 2016 (UTC)

There is a known issue linking the database and the Wiki when ISFDB records contain certain characters, including accented characters -- see Bug 30, Bug 31, Bug 232, Bug 235 and Bug 256. It's one of the reasons why we are in the process of migrating Wiki-based pages to the database :-) Ahasuerus 17:06, 16 March 2016 (UTC)
Ok, thanks. Hauck 17:17, 16 March 2016 (UTC)

Pseudonyms based on alphabets?

Is this allowed? We have dozens of Russian authors that we don't enter with their native alphabet. Why should we be doing it in a pseudonym for a German author? Mhhutchins|talk 19:19, 16 March 2016 (UTC)

Non-Latin characters are currently not allowed in author names. It won't cause the software to explode in a "flash of intolerable brilliance", as Doc Smith would have put it, but it will result in some problems with the Author Directory and Bio/Biblio pages. Ahasuerus 23:04, 16 March 2016 (UTC)
Can you explain that to the editor/moderator who created/accepted it? Thanks. Mhhutchins|talk 23:15, 16 March 2016 (UTC)
Do you happen to know who entered/approved this data? Ahasuerus 23:54, 16 March 2016 (UTC)
The submission creating the publication record, and the submission creating the pseudonym. Mhhutchins|talk 01:35, 17 March 2016 (UTC)
BTW, the associated publication record also has an issue -- the three publishers are delimited using commas instead of slashes. Ahasuerus 23:04, 16 March 2016 (UTC)
Using a slash may cause confusion with its usual purpose of separating imprint from publisher. I thought we used ampersands to indicate multiple publishers (see this, this, and this), but I supposed that issue was never settled, or documented if it were. Mhhutchins|talk 23:15, 16 March 2016 (UTC)
Hm, I don't think I recall ampersands being mentioned in this context. Wouldn't that standard apply to all SFBC / [some other publisher] permutations? Ahasuerus 23:54, 16 March 2016 (UTC)
It was discussed, but I have no idea when or where, or how to find it on the wiki. As for SFBC editions, I've been thinking lately that the whole method we use is wrong. And for the same reason I mentioned: the slash should be limited to use as "Imprint / Publisher". The "Publisher / SFBC" entry method was here when I started and I just fell into step with it. Since "SFBC" is neither the publisher nor co-publisher of the editions entered this way, it would be better to come up with a unique designation. What would you think of "Publisher (SFBC)", if I presented it to the community? Or is there a different method you'd want to suggest? Or should it stand as is? Mhhutchins|talk 01:32, 17 March 2016 (UTC)
It may be best to let it be for now. If we change the way we handle publishers in the future -- see the discussion below -- we'll have to revamp things yet again and that's a lot of work. Ahasuerus 02:08, 17 March 2016 (UTC)
I guess the long term solution will be to add support for multiple publishers per publication just like we support multiple authors per pub/title and multiple Web pages per author/title/etc. Ahasuerus 23:54, 16 March 2016 (UTC)
It's so rare that I'm not sure that the effort would be worth it, or that having it might lead to confusion and/or misuse. For example, some editors may think it's for imprints or publishers at the corporation-level. I'd say let sleeping dogs lie, since there's been absolutely no discussion (that I can recall) of having support for multiple publishers, or the matter ever being brought up. Mhhutchins|talk 01:32, 17 March 2016 (UTC)
As I recall, we had a brief discussion about a year ago and the same concerns re: imprints and publisher hierarchy were raised. I agree that they are valid concerns. I'll have to think of ways to address them before I propose anything specific.
Ideally, our software would let us record infinitely nested publisher hierarchies just like it lets us record infinitely nested series hierarchies now. Unfortunately, publishers are more complex than series -- imprints become separate publishers, publishers become imprints, etc -- which makes it harder to model in the software. Ahasuerus 02:06, 17 March 2016 (UTC)

My Errored Out Edits

Hi, i made a submission of a magazine that ends in My Errored Out Edits. Only part of the data entered was validated (see here). It didn't appears in my recent edits, and there's no moderator name atached to it. Can you please help me figure out what happens here. I don't understand if it was my fault or some kind of system bug. Thanks. Wolland 13:33, 17 March 2016 (UTC)

"Errored out" submissions are always caused by system problems, so I am sure you didn't do anything wrong. I'll try to recreate the problem on the development server. Thanks for reporting the issue! Ahasuerus 14:59, 17 March 2016 (UTC)
Just chiming in, the main cause of such errors are the server shutdowns that occur every day at 16:30 (French time) during less than 10 minutes and, in smaller durations at 14:30. Having been caught a few times, I use to stop all work on the db at 16:29 (in a few minutes ;-)).Hauck 15:21, 17 March 2016 (UTC)
That's right. The daily backups run at 9:30am server time, which is currently Eastern (US/Canada) Time.
In the meantime, I have re-approved the errored out submission on the development server and everything worked correctly, so I suspect that it was indeed a problem with the backups. I really need to change the backup process to disable editing right before the backups start... Ahasuerus 15:55, 17 March 2016 (UTC)
I guess it was me who approved of your submission, Wolland. It must have been a short time before 03:10:17 server time (when I started to edit the language for some of the magazines authors). I didn't realize that part of it was eaten alive and just wondered why it didn't show up on the table of recent edits. I'll try to do better and not approving anything around that time. Sorry for the inconvenience! Stonecreek 16:41, 17 March 2016 (UTC)
No problem! But now, should i wait to the errored out submission to be recovered, or i just fill in the missing data in the publication? Wolland 16:47, 17 March 2016 (UTC)
I am afraid the only way to get the missing data into the database is to create another submission via "Publication Edit" using the body of the errored out submission as your source. Sorry about the hassle! Ahasuerus 16:51, 17 March 2016 (UTC)
OK, it´s done, just wainting approval. Wolland 17:18, 17 March 2016 (UTC)

Forthcoming Books: clean-up report vs. front page link

I've just came upon this anomaly: This publication appears on the clean-up report, but not on the list of forthcoming books. How is that possible? FYI, I also check the forthcoming books list for such titles, so I'm not sure why adding them to the clean-up report is entirely necessary. Having them listed in the one place is nice, but something like this discrepancy is strange. Mhhutchins|talk 18:31, 20 March 2016 (UTC)

Have you had a chance to look into this? Mhhutchins|talk 04:17, 28 March 2016 (UTC)
I saw your subsequent post on the Moderator Noticeboard and thought that the issue was resolved. Did I misread it? Ahasuerus 05:43, 28 March 2016 (UTC)
Yes. That post (on the Moderator noticeboard) was asking you to read this one (on your user talk page) that contains a question. I didn't want to pose the same question again on the Moderator noticeboard post. Mhhutchins|talk 06:18, 28 March 2016 (UTC)
Oh, I see. <pokes the data with a stick> Hm, that doesn't look right. I'll take a closer look later today. Thanks for reporting it! Ahasuerus 14:04, 28 March 2016 (UTC)
It turns out that the "Forthcoming Books" page was programmed to ignore pubs without cover images. Everything should be fixed now. Thanks again for identifying the problem! Ahasuerus 21:37, 28 March 2016 (UTC)

Just a ping for you

Since you have so many conversations going on the community portal, just pinging you for this one. :) ···日本穣 · 投稿 · Talk to Nihonjoe 23:11, 22 March 2016 (UTC)

Well, there are roughly 2,600 ISBNs that I still need to process this month. Between that and the software changes and the Wiki discussions, it's kind of rough. I thought someone else has been working on the publisher changes that you requested, but it looks like most of them are still outstanding. I will take a look later tonight. Ahasuerus 00:00, 23 March 2016 (UTC)
Okay, I can understand. I just didn't want it to get lost in the shuffle. Should I pop that part of the post over to the Moderator noticeboard so someone else can do it? Thanks. ···日本穣 · 投稿 · Talk to Nihonjoe 00:13, 23 March 2016 (UTC)
That would be great! :) Ahasuerus 00:32, 23 March 2016 (UTC)
Done. ···日本穣 · 投稿 · Talk to Nihonjoe 18:44, 23 March 2016 (UTC)

Bleiler '78

For some reason [probably me missing a straight line ... ] I have a verification under Bleiler '78 [had the opportunity to buy one once but it's THE most useless source of data .....]. There seems to be no way to find which record I mistakenly verified. Can you direct me, please?? --~ Bill, Bluesman 22:49, 26 March 2016 (UTC)

The culprit is this pub. It was published in 1982, so my guess is that you meant to click "N/A" rather than "primary verification" :) Ahasuerus 23:06, 26 March 2016 (UTC)
Thanks! Just started one line too high as I do have the edition. --~ Bill, Bluesman 19:21, 27 March 2016 (UTC)

Prime Books LLC

When you get a chance, if you could update Fixer to change "Prime Books LLC" to "Prime Books", it would be appreciated. Thanks. -- JLaTondre (talk) 13:28, 27 March 2016 (UTC)

Done! Ahasuerus 15:50, 27 March 2016 (UTC)

Yana Botsman

So which of the two persons in the photograph linked here is Yana Botsman? Isn't it unusual to have such a photograph linked to an ISFDB author summary page? Mhhutchins|talk 23:35, 28 March 2016 (UTC)

According to FantLab, it's a joint photograph of Yana Botsman and her co-author, Dmitry Gordevsky. I guess we could cut it in half and use the two halves separately, but wouldn't that present copyright issues? Ahasuerus 23:46, 28 March 2016 (UTC)
How would a user know which of the two persons is the author? An English speaker, like me, has no way of knowing "Yana" is a female name. It's just strange to me to have a photograph of two persons on a page for a single author. I see you've linked the same photograph to both authors. Why not just link it to their shared pseudonym's page? That at least makes sense. Mhhutchins|talk 03:03, 29 March 2016 (UTC)
I can see how having a joint photograph could be confusing. On the other hand, displaying it on the pseudonym's page could also be problematic. What if another person uses this pseudonym or they split the way the Binder brothers did in 1939? I am not really sure how we should handle this. Ahasuerus 04:44, 31 March 2016 (UTC)

Software change for editing magazine records

I've noticed a recent change (or at least the first time I've noticed it) that allows the editor to change a NOVEL record (as in a Fixer submission) to MAGAZINE, with the ability to change the title reference record to EDITOR during a pub edit. In the past that wasn't one of the drop-down menu options, and the editor had to update the title record in a separate submission. Thanks for adding that option. Mhhutchins|talk 02:53, 29 March 2016 (UTC)

I am glad you liked the tweak! It was a side effect of the recent change to "container editing". As I wrote at the time:
  • ...editors can now select "EDITOR" in the Content section of Edit Pub. It was originally removed from the drop-down list because new editors often thought that "EDITOR" stood for "editorial". Now it is back in order to support the ability to edit container titles within Edit Pub. It should be relatively harmless because we have a nightly report that finds improperly entered EDITOR titles.
Ahasuerus 05:47, 29 March 2016 (UTC)

Publisher transliteration field

Can you look at the submissions I have on hold from Dirk Broer. It's my understanding that this field is to be used transliterating publisher names that are in a non-Latin alphabet into the Latin alphabet. Should it be used for translating a non-English language into English, which is what it looks like Dirk is doing here? That is at odds with what I thought was the purpose of the field. Thanks for looking. Mhhutchins|talk 03:59, 29 March 2016 (UTC)

At one point there was a proposal to add a separate field for "translated publisher names", but there was not enough support to implement it. The recently implemented "transliterated" fields are supposed to be limited to transliterated names. Ahasuerus 05:16, 29 March 2016 (UTC)

Suspected Duplicate Authors cleanup report

There's been no items on the list for about the week, yet it still keeps showing up on the report list as having three items. Mhhutchins|talk 19:42, 29 March 2016 (UTC)

This report takes over 24 hours to compile, so it's scheduled to run on the 17th of every month. If the "false positives" are a hassle, I can add a "Resolve" column to the table. Ahasuerus 19:58, 29 March 2016 (UTC)
How is the "resolve" option different from the "ignore" one? (Which is currently available for this report.) Mhhutchins|talk 22:20, 29 March 2016 (UTC)
"Ignoring" a record (or, in this case, a record pair) adds it to the "ignore" list for the currently displayed report. From that point on, every time the report runs, it will ignore the record.
"Resolving" a record simply removes it from the list of problem records which was compiled by the nightly process. The next time the nightly process runs, it will re-add the problem record to the list if the problem still exists. We have a few reports which let moderators "resolve" records, notably the ones that find invalid URLs in notes.
That said, now that I have thought about it some more, I don't believe that it would be feasible to add a "Resolve" column to the "Suspected Duplicate Authors" report. The reason that the overview page shows a count of 3 is that 3 author records have been deleted/merged since the report last run. There is nothing to "resolve" now, so there really isn't anything we can do until the report runs again :( Ahasuerus 22:49, 29 March 2016 (UTC)
Thanks for the explanation. Mhhutchins|talk 02:02, 30 March 2016 (UTC)

Russian publisher

Filling in some data for Meeting over Tuscarora and found [this] OCLC record. Can you find the publisher in this and add it to [this] record?? Much thanks. --~ Bill, Bluesman 23:44, 31 March 2016 (UTC)

Done! Ahasuerus 23:54, 31 March 2016 (UTC)
That's the longest publisher name I've seen. Of course, it takes a long time to say anything in Entish ... er, Russian! ;-)) --~ Bill, Bluesman 00:11, 1 April 2016 (UTC)
It's even longer if you spell out the abbreviations -- "Naval Publishing House of the People's Commissariat of the Navy of the Union of Soviet Socialist Republics". And yes, it does say "Navy/Naval" twice! :) Ahasuerus 01:37, 1 April 2016 (UTC)
The closest I could find was a publisher named «Новая Москва» (New Moscow). Just "Moskva" would be Москва. I bow to your Russian capabilities. :) ···日本穣 · 投稿 · Talk to Nihonjoe 23:58, 31 March 2016 (UTC)
Hey, we all have our areas of expertise! I am sure I would make an awful mess of our Japanese titles if I tried to maintain them on my own. Ahasuerus 01:37, 1 April 2016 (UTC)

Amazing Stories: Pulp Tales Press

Please see 36191 and 41221. I thought it was not possible to have two series with the exact same name? At first, I thought there was a typo in one of the names. But as far as I can tell, they match up exactly. Even on the edit screen. Anyhow, they need to be 'merged'. -- JLaTondre (talk) 02:55, 3 April 2016 (UTC)

Well, there are software checks that are supposed to prevent this from happening, but apparently someone got around them. Out of order approval, perhaps? The good news is that we had only one problematic series and I have already fixed it. Thanks for reporting the issue! Ahasuerus 03:18, 3 April 2016 (UTC)

Non-genre magazine with only nonfiction content by a spec-fic author who is "above the threshold"

Based on my understanding of the standards this publication wouldn't be eligible for the database. I placed a message on the talk page of the record's creator about two weeks ago, but as yet have had no response. Before taking the next step and deleting the record, would you agree or not that allowing it would be beyond the scope of the database? Thanks. Mhhutchins|talk 04:14, 4 April 2016 (UTC)

As far as I know, the current standard -- as explicated by the Help page that you linked -- is that "the non-speculative-fiction contents of such a [non-genre] magazine should not be entered".
That said, I would suggest pinging Marc again before doing anything about the record. I believe that in the long run it's more important to make sure that all of our editors (and especially moderators) are on the same page. If we aren't, then we'll end up with a lot more than one questionable pub... Ahasuerus 04:34, 4 April 2016 (UTC)
I agree with Michael ("No non-fic in non-genre"). Hauck 06:25, 4 April 2016 (UTC)

Odd Fixer submission

This submission created a NOVEL-typed record that was added to a CHAPBOOK-typed title. How was this possible? (The reason I didn't add a SHORTFICTION content was because when I accepted the submission, I had to believe it was a NOVEL, and not a CHAPBOOK. You'll also notice that the publication record doesn't have a title reference, and it was this mismatch that was brought to my attention by a clean-up report.) Mhhutchins|talk 06:55, 7 April 2016 (UTC)

That's curious. I suspect it may have been caused by the fact that the submitted pub didn't have a page count. I'll try to recreate the problem on the development server. Thanks for reporting it! Ahasuerus 18:15, 7 April 2016 (UTC)
OK, we have our answer. The reason that Fixer submitted this pub as a NOVEL was that the parent title was a NOVEL as of the time of the last backup, which is what Fixer uses for reference purposes. Unfortunately, the ISFDB software doesn't let robots check its title records remotely, so there isn't much that Fixer can do. There is a Feature Request to add this functionality, but it's relatively low priority. Ahasuerus 22:12, 7 April 2016 (UTC)
I suspected that was the reason. Thanks for confirming it. Mhhutchins|talk 23:25, 7 April 2016 (UTC)

Two series with the same name?

I'm not sure how this could happen. Perhaps something to do with the apostrophe? If so, we need to let the editor know that they're using the wrong character when entering apostrophes. Mhhutchins|talk 01:04, 8 April 2016 (UTC)

At one point there was a problem with apostrophes in series names, but it was fixed a long time ago. The other day JLaTondred reported a similar problem with "Amazing Stories: Pulp Tales Press", so the issue is not limited to apostrophes.
My best guess is that the root cause is our old friend "out of order approval", but it seems odd that it would happen twice in 3-4 days. On the other hand, I don't think the relevant part of the software has been modified lately, so it's probably not a new bug either. I'll have to experiment... Ahasuerus 01:13, 8 April 2016 (UTC)
I removed the series data from the French title, then deleted the series, then added it back...and it created a new series! (See this submission.) When I do a simple search for "tales of the king's blades" under series, I get no hits. When I do a simple serach for "tales of the king", I get this series. There's something about that apostrophe in the current series that's just not right. Mhhutchins|talk 01:47, 8 April 2016 (UTC)
Thanks for looking into this! I will investigate tomorrow morning. Ahasuerus 05:23, 8 April 2016 (UTC)
You were right, it was the apostrophe that was messing things up in this case. It wasn't a regular apostrophe; it was a Unicode character masquerading as one. There is a special module in the software that converts various Unicode apostrophes, single quotes, double quotes, etc to their standard equivalents at data entry time, but this one was not converted. "Tales of the King's Blades" is a relatively old series (in ISFDB terms), so it's likely that the character was entered before we added the Unicode conversion logic. I have changed it to a regular apostrophe to address the immediate problem. Thanks again! Ahasuerus 15:14, 8 April 2016 (UTC)
I have identified and corrected the last 2 series names that had this problem. Ahasuerus 05:32, 9 April 2016 (UTC)

Wikipedia articles and redirects

Hi. Do we have any policy regarding (EN for short) redirects as targets in the "Webpages" fields? (I know of you as a Wikipedia editor by same username and suppose that you are familiar with EN use of redirects for individual people including both Joe and Beth Krush at EN. If not familiar, there are many questions I can answer.)

Browsing the archives of Rules and standards discussions this hour, albeit not on this issue, I discovered for two of our author pages:

The latter target is Joe Krush at (EN for short), although few database users will visit that page because the redirect to joint biography is executed when one uses the link from our Joe Krush page.

It occurs to me that we should prefer to target personal redirects for two reasons. Those pages are usually the ones, if any, that will become individual biographies in the future. And some users familiar with Wikipedia will be able to visit the individual page via the message "(Redirected from Joe Krush)" that now appears at the top of the joint biography for Joe, as none now does for Beth.

Do you know whether we have a policy one way or the other? Or whether Wikipedia targets have been discussed to no consensus somewhere?

By the way, we now give --and I will fix later, not this ession-- Beth Krush's actual date of death for both Krushes as Deathdate: 2 February 2009. Joe Krush is one of the "Possibly living people" at EN. Perhaps that anomaly is related to the other issue, via some work that we do automatically or semi-automatically here. (I don't know anything about what is done by robots or semi-automated by robo-reports.)

--Pwendt|talk 01:23, 8 April 2016 (UTC)

I would support pointing at redirects in the case presented above. It makes sense. ···日本穣 · 投稿 · Talk to Nihonjoe 01:29, 8 April 2016 (UTC)
I am afraid I don't recall any discussions re: Wikipedia redirects. I think our de facto policy has been to replace links to redirects with links to actual pages, but I don't think it has been discussed, mush less documented. I would suggest asking on the Rules and standards discussions page to see if anyone else has experience with redirects. Ahasuerus 01:34, 8 April 2016 (UTC)
I would think that on occasion there could be value in linking to a redirect page, if the canonical name we used for an author were substantially different than the one used by Wikipedia. For example, if you're at our page for "B. W. Clough", it currently does a direct link to Wikipedia's "Brenda Clough" page. But if we linked to "B.W. Clough" instead, it would redirect, and add the line at the top "Redirected from B.W. Clough". That might make it clearer to the user what just happened, i.e. that the link *didn't* take them to the wrong page. This may not be the best example, because it's hard to imagine someone being confused between those 2 variations of her name. But in other cases (where our name and theirs are very different), it might be useful. Not often, but occasionally. Chavey 03:45, 8 April 2016 (UTC)
Evidently I neglected to save last hour's provision of a better example on that line of reasoning. English Wikipedia has biography Robin Hobb at EN and redirect Megan Lindholm at EN. We have pseudonym Robin Hobb one variant of canonical name Megan Lindholm A1038, which summary bibliography links directly to the Robin Hobb biography.
The Robin Hobb biography does display full name Margaret Astrid Lindholm Ogden in the lead sentence, and something some similar should always be displayed boldface in the lead section. So there may be little confusion to forestall.
Regarding the other line of reasoning (below) in this instance: the Megan Lindholm redirect [1] provides no information, except the fact that Wikipedia has undertaken such maintenance. In contrast redirect Joe Krush at EN provides some information about Krush. ... --Pwendt|talk 20:19, 12 April 2016 (UTC)
User:Chavey, I have in mind primarily EN redirects from people to joint biographies such as "Brothers Hildebrandt", "Leo and Diane Dillon", "Beth and Joe Krush". And similar cases where an author redirects to a work; a publisher (or imprint, division, etc) to a bigger or subsequent entity; a novel to a series. (The latter much less useful, if at all, where we have pages for both the series and the novel, because the Wikipedia article is a substantial match for one or the other; simply list it in Webpages there). That is, the redirect pagename and target article pagename refer to different entities.
At the moment I don't know whether we sometimes target Wikipedia article sections, which may sometimes but not routinely be an alternative to the redirect. Not routinely as, say, "Brothers Hildebrandt" at EN does not have Greg Hildebrandt and Tim Hildebrandt sections; contrast "Leopold and Loeb" at EN. --Pwendt|talk 23:00, 8 April 2016 (UTC)
I have seen a few Wikipedia links that go directly to a section within a Wikipedia page. Mostly, I've seen this happen when when somebody is entering a page for a title series, or for a specific book, and link to the section of that author's page that discusses that series or that book. IMO, such links are not particularly helpful, but they do occur. Chavey 03:12, 13 April 2016 (UTC)

... It is on my list "To Do" a related note at Rules and standards. --Pwendt|talk 20:19, 12 April 2016 (UTC)

Bio pages

Hi. Only during the last couple days I have browsed for several hours, or perhaps a few dozen, in the Wiki pages where editors talk to each other. Last hour at the very end of my day that was three sections at ISFDB:Community Portal, currently #81-83 of which the first is 81 Wiki-to-Database migration: Publication and Publication Talk pages and the last begins to cover Author and Bio pages.

Yesterday I visited Bio:Catherine Asaro, following a link in one of your contributions to some discussion or other. I wonder is that page an example of the future of the Wiki Bio? Perhaps the current content of formal [ref][/ref] references would be retained in the Wiki? Thanks in advance. --Pwendt|talk 02:00, 9 April 2016 (UTC)

I think there are a couple of different issues here.
The page that you linked raises the issue of attribution and sourcing. When entering bibliographic data, our goal is to document where the data comes from -- primary verification, secondary sources, private communication from the author, etc. Since Bio pages currently reside within the Wiki and since the Wiki doesn't impose the same level of moderator oversight that regular submissions do, attribution/sourcing (or lack thereof) can be a problem.
The second issue is the larger issue of migrating our Wiki-resident data to the database proper. As I wrote on the Community Portal a couple of weeks ago:
Let me first summarize the reasons behind the Wiki-to-database migration project. We have discussed them many times over the years, but it may be useful to have all of them listed here to refresh our collective memory and to get new editors up to speed:
  • The ISFDB database and the ISFDB Wiki are effectively two databases which can (and do) get out of sync over time. As per the summary posted below, more than a third (!) of Wiki-based Publisher pages are now orphans and not visible from the database side.
  • The bidirectional links between the database side and the Wiki side are based on what we used to call "lexical match", i.e. the two names/titles being the same. In our version of the Wiki software, the match doesn't work when certain characters are present, notably accented characters. This has become a bigger problem as our data has expanded to cover non-English works and authors.
  • There is no moderator oversight when editing the Wiki, which often results in incorrectly entered data. For example, consider this publication-specific Wiki page. The data is good, but it is related to the Birthright title rather than to one of its publications.
  • With the exception of ISFDB-hosted images, Wiki-based data is not included in our publicly accessible backups. If something major happens to and its maintainers -- and we all know that many useful Web-based resources have disappeared over the years -- the site can be recreated using the latest backup file. On the other hand, the Wiki data will be gone.
  • Bibliographic data is split between two Web pages, which can prevent our users from getting a full picture.
We have always known about these limitations, but early on we were forced to use the Wiki as a crutch because the core ISFDB software didn't support many features that we wanted. Over the last few years we have implemented many of these features, including Publication Series, an Award Directory, a Publisher Directory, Series/Publisher/etc notes and so on. We have also added built-in cleanup reports which have superseded Wiki-based cleanup projects.
Because of these enhancements we no longer need to use the Wiki as a crutch and can go back to using it for its original purpose, i.e. allowing editors to communicate. Of course, editors will remain free to create and maintain Wiki pages for research purposes, but Wiki pages will no longer be automatically linked based on the notorious "lexical match". Ahasuerus 21:35, 21 March 2016 (UTC)
And a few days later:
(unindent) I haven't forgotten about this issue. I am currently experimenting with a few different options and will post once I have the results. Ahasuerus 20:11, 30 March 2016 (UTC)

"Random House Books for Young Readers"

None of the hundreds of books in the database credited to this publisher are primary verified, which led me to believe it's not a stated imprint, but a marketing strategy. Random (pardon the pun) checks of more than a dozen title pages using Amazon's Look Inside, as well as OCLC records, gives the publisher as just "Random House". I will merge this publisher with Random House, and ask if Fixer can be programmed to convert all such future data received from Amazon. Thanks. Mhhutchins|talk 20:39, 10 April 2016 (UTC)

Done -- see submissions 3005971, 3005973 and 3005976. Ahasuerus 20:59, 10 April 2016 (UTC)
Yep, just moderated them. I'll proceed to merge the publishers. Thanks a lot. Mhhutchins|talk 21:04, 10 April 2016 (UTC)

Clever, that searching

I was surprised, and amused, to see that searching for "fan" found things with "фан". Very clever! --MartyD 14:00, 20 April 2016 (UTC)

Glad you liked it! However, it occurs to me that the way search results are displayed may also be somewhat confusing. Transliterated value(s) are currently displayed as mouseover text, which may be too advanced for a "naive user". Perhaps we should add a separate "Transliterated Name(s)" column to the table. Ahasuerus 14:18, 20 April 2016 (UTC)
Hmmm. I didn't notice/realize. --MartyD 02:07, 21 April 2016 (UTC)

Pubs with multiple COVERART titles

I've been working on this clean-up off and on, but not making any perceptible progress. Is it possible to make this into a public report, and then add a column for those publications which have been primary verified with links to the primary verifier? Then we can at least have the verifying editors work on their own records. Thanks for considering the possibility. Mhhutchins|talk 16:04, 2 May 2016 (UTC)

I think it's a good idea. I have created a feature request to document the proposed functionality, but it may take some time to get to because I am currently buried in Fixer's stuff. Last month he found almost 20,000 (!) new ISBNs. Ahasuerus 16:26, 2 May 2016 (UTC)

Problem with full Talk page

I am sorry to tell you , I made a mess of my Talk page, it was full you see and I tried everything but it was not to be, It seems now that I do not have a Talk page anymore. Can you help me with this, I find it a little complicated and I can not respond to messages anymore. Thank you. William 16:38, 6 May 2016 (UTC)

Have a look at it now. Hauck 16:52, 6 May 2016 (UTC)
It is back but it is full so what to do next? William 17:04, 6 May 2016 (UTC)
It looks like everything is back to normal -- don't hesitate to ask if you need help with anything else :) Ahasuerus 17:50, 6 May 2016 (UTC)
Ok that I wil do, thank you. William 21:41, 6 May 2016 (UTC)

Murky comma-versus-cedilla business in Romanian

Hello ! I have been editing Romanian SF for some time now, and have eventually run into some kind of font problem. There exist in Romanian two letters with a comma below, ș and ț, noting respectively /ʃ/ (as in shoe) and /t͡s/ (as in czar). These letters have unofficial variants with a cedilla, ş and ţ, massively used on the net because of a former lack of computer font support (see a commentary here). The main Romanian site I'm tapping from at the moment is Moshul SF, which uses both, although ș and ț have become more frequent in recent posts.
The problem is that :

1) Most of the data I have retrieved so far from Moshul SF has cedilla ş and ţ, which are incompatible with the comma-below versions. For the sake of consistency, and in order to avoid creating unnecessary (not to say non-existent) variants, I have so far replaced commas by cedillas, as this "Second Variety" outnumbered the first one by far in this db.
2) Most of the more recent data will have official comma-below signs (this is the case for all Wikipedia data, for instance), so that anyone adding such data might create false variants, pseudonyms or duplicate names.

So would there be a way to avoid this snag, for instance by making the system automatically convert "Ş / ş / Ţ / ţ" into "Ș / ș / Ț / ț", as you did once for "..." versus "…" ? This solution is used by the Romanian Wikipedia, where all cedillas are automatically converted into commas. The difficulty here is that it would have to be restricted to Romanian, as other languages (such as Turkish, for example) have a legitimate ş with cedilla, which shouldn't be changed. If nothing can be done about it, I'll just keep on changing occasional commas into cedillas. Thanks for looking into this if you have the time… Linguist 09:50, 9 May 2016 (UTC).

Unfortunately, as you noted, the software would have to know that it's dealing with a Romanian title/pub, which is not always obvious, e.g. when changing languages. However, it would be fairly easy to create a couple of cleanup reports to find any Romanian titles/pubs that use these characters. Ahasuerus 14:33, 9 May 2016 (UTC)
Yes, that looks like a good solution. The only thing is, as I pointed out, the db is mostly using cedilla characters, which I was hoping would be reverted en masse to comma-below characters to avoid any future trouble. A cleanup report would identify thousands of the damn things — the easy solution being of course homing on comma-below characters, which can't be very numerous, and changing them to unofficial cedilla ones. Linguist 14:58, 9 May 2016 (UTC).
That's a good point. I can create a script that will convert all existing Romanian titles and pubs. The only thing that I can't (easily) do is add the same kind of "on the fly" logic that automatically converts ". . ." at data entry time. Ahasuerus 15:39, 9 May 2016 (UTC)
It would be great if all these spellings could be harmonized thanks to such a script. Once there is only one kind of diacritic (preferably the comma-below one : Ș / ș / Ț / ț, since it is the official spelling), it would be fairly easy to get rid of new cedillas brought in by occasional editors, as I would convert my own to commas first. I just hope this won't give you too much trouble. Thanks a lot, anyway. Linguist 16:08, 9 May 2016 (UTC).
Well, coding a new cleanup report or a one-time data conversion script is not particularly time-consuming. Neither is adding a new field here or a new feature there. However, there are so many of them that they require prioritization, especially considering the fact that I spend at least half of my ISFDB maintaining our robot. I wish there were more hours in the day... Ahasuerus 20:09, 9 May 2016 (UTC)
FR 890 has been created. Ahasuerus 20:12, 9 May 2016 (UTC)
Thanks a lot. The FR looks good. Linguist 08:47, 10 May 2016 (UTC).


I was just thinking : shouldn't some kind of advice be given on a help page re this particular problem, inciting future editors to prefer comma-below (Ș / ș / Ț / ț) to cedilla characters (Ş / ş / Ţ / ţ) in Romanian titles and names ? I tried to find such a font-related help page, but failed miserably. Linguist 14:49, 15 May 2016 (UTC).

I don't think we have a page dealing with fonts and/or internationalization (aside from ISFDB:Foreign Language Abbreviations), but there is no reason not to create one! :) Ahasuerus 18:00, 15 May 2016 (UTC)
Right, I'll see what I can do… Linguist 08:25, 16 May 2016 (UTC).

Will you check this Russian entry

Since you know at least some Russian, will you check this to make sure I entered everything correctly? Thanks! ···日本穣 · 投稿 · Talk to Nihonjoe 20:18, 9 May 2016 (UTC)

Looks pretty good. I have added the original Polish spelling of his legal name and a link to the Polish Wikipedia, which has a detailed article. Unfortunately, "Ary" and "Sterfeld" can be (and often are) transliterated differently in different languages... Ahasuerus 20:35, 9 May 2016 (UTC)
Thanks! ···日本穣 · 投稿 · Talk to Nihonjoe 18:33, 10 May 2016 (UTC)

Languages within anthologies

I don't know whether this is a "bug" or a "feature request". As I complete entering the contents for the Polish magazine Fenix, some of the stories ended up being assigned "English", because that's my language, and others were assigned "Polish", since that's the language of the magazine. Oddly enough, this would happen with different stories within the same issue (possibly due to the working language, if any, assigned to the author).

Language assignment for NewPub submissions is straightforward and based on the language selected by the editor in the NewPub form. However, language assignment for EditPub submissions is much more involved. The software tries to determine the language of the "reference" title -- in this cases the magazine's EDITOR title -- and then uses it as the language of newly added Content titles. If the reference title has no language assigned, the software uses the editor's "Default author/title language when editing records". The author's working language is not checked.
The safest thing to do when adding contents to a non-English publication is to (temporarily) change your "default language" under My Preferences. Ahasuerus 18:03, 21 May 2016 (UTC)
I suppose so, but I have a tendency of working on multiple projects at once, so that might backfire on me. Chavey 08:24, 22 May 2016 (UTC)

I made sure that all of the (annual) titles were assigned "Polish" as their language. I then assumed that if a story thought itself to be English, that the contents listing would say something like "Dwunasty Album [English]", since the language of that story was different than the language of the title to which that publication record was assigned. That is not the case. So although I looked through all the contents listings for "English" titles, I did not find them, hence several of those stories are incorrectly labeled as being in English. It seems to me that the existence of stories in a language other than the language of the book should be noticeable (i.e. listed with its Junifer tis . But was that ever intended? (In which case this is a bug.) Or was that language listing just intended for the author bibliographies? (In which case this would be a feature request.) Chavey 06:59, 20 May 2016 (UTC)

Yes, I can see how it could be useful. FR 892 has been created. Thanks! Ahasuerus 18:15, 21 May 2016 (UTC)
Thanks much! I'll finish fixing all of the false "English" titles in Fenix in the next few days. Chavey 08:24, 22 May 2016 (UTC)

Authentification problem

Hello, I've got this (kind of) message from a contributor. Do you think that you can do something? Thanks. Hauck 17:58, 21 May 2016 (UTC)

I have set the "authenticated" flag on the account and left a message on the submitter's Talk page. Hopefully it should do the trick! Ahasuerus 18:09, 21 May 2016 (UTC)
Many thanks (for him I hope). Hauck 18:10, 21 May 2016 (UTC)


Sorry to hear about the e-mail issues! Could you please try responding to this message? I have made a manual change to the database settings to flag your account as "authenticated", so hopefully it should work. TIA! Ahasuerus 18:08, 21 May 2016 (UTC)

It worked! I take it this is the way to respond (pasting your message into a new post on your talk page). Thanks! And thanks to Hauck and JLatondre for moderating the edits I did on my first try. MOHearn 21:44, 21 May 2016 (UTC)

Dueling magazines

Do you have an suggestions about what to do with the Fixer submission I have on hold and this pub? It seems unlikely that both exist, but both are present on Amazon, and the back cover image on each shows a barcode with the corresponding ISBN. Do you have any secret information sources or insights that might help, or should I just go ahead and accept it and have two pubs? Thanks. --MartyD 13:05, 28 May 2016 (UTC)

Well, the held submission says "Second edition" in the Note field and Look Inside shows different numbers on the title page: "v291115" vs. "v081215". I am not sure what the numbers are supposed to refer to, but they suggest that there were indeed two versions of this pub.
My guess is that the publisher decided to reprint this issue, although it's odd that Amazon lists the publication date of the first edition as 2015-12-07 while the second edition is 2015-11-30. Since the dates are so close to each other, it could be a clerical error or perhaps Amazon received the second edition first.
Unfortunately, our software assumes that a magazine issue can have only one edition, so entering 2+ editions usually involves creative shoehorning... Ahasuerus 14:32, 28 May 2016 (UTC)
Oh, EXCELLENT eye! I missed those numbers. 291115 = 2015-11-29, and 081215 = 2015-12-08. Two it is. I think it'll probably come out ok, since they're conveniently "Winter 2015/16", I can date each pub accordingly, and the grid will simply have two issues with the same title, but in two different months. That's my theory, anyway. :-) We'll see! --MartyD 15:33, 28 May 2016 (UTC)
Collaborative snooping FTW!! :-) Ahasuerus 16:13, 28 May 2016 (UTC)

Other Authors With Same Name Oddity

Minor given the nature of the entry, but this entry produces some odd other authors with the same name possibilities. -- JLaTondre (talk) 12:46, 29 May 2016 (UTC)

Thanks, I'll take a look... Ahasuerus 12:53, 29 May 2016 (UTC)
Underscore is single-character wildcard. So it probably matches everyone whose names are 6 + 6. --MartyD 14:26, 29 May 2016 (UTC)
Yes, indeed! The bug has been fixed with extreme prejudice. Ahasuerus 19:22, 29 May 2016 (UTC)

Concern about possibly erroneous editions of Ballantine Lord of the Rings titles

I have flagged two editions of the first Ballantine paperback printings of The Two Towers and The Return of the King. I have the first Ballantine printing of all three books in the trilogy. The Fellowship of the Ring reads on its copyright page: "First printing: October, 1965." The Two Towers reads: "First printing: November, 1965." The Return of the King reads: "First printing: December, 1965." The edition of The Two Towers that lists first Ballantine printing as October 1965 had no verifiers. The edition of The Return of the King that lists first Ballantine printing as October 1965 had you listed as the primary verifier, but no other verifiers. If you have information that I do not, please get into the discussion. I want the records to be accurate. Thanks! Mike 00:38, 9 June 2016 (UTC)

Thanks for letting me know! Unfortunately, my Tolkien books were in the part of my collection that was damaged late last year. They, along with thousands of other books, are currently boxed and it will be months (if I am lucky) before I have access to them. I'll make a note of it and will respond once the dust settles. Thanks again! Ahasuerus 01:19, 9 June 2016 (UTC)
P.S. Checking my list of primary-verified publications, I don't see The Return of the King listed. Could it be that you meant User:Don Erikson, who verified this printing? Ahasuerus 14:45, 9 June 2016 (UTC)

Gray Lensman

The artwork for your Gray Lensman is broken as the image was deleted from the wiki. -- JLaTondre (talk) 12:50, 4 July 2016 (UTC)

Updated; thanks! Ahasuerus 14:24, 4 July 2016 (UTC)

The immortal Francis Merton Lyon

I can't get Francis Merton Lyon to go away. Maybe because there's a submission (now rejected) referring to it? --MartyD 01:46, 6 July 2016 (UTC)

NVM. I found it -- there's a review still citing that name. --MartyD 01:49, 6 July 2016 (UTC)
It's a good point, though -- we may want to change the software to make it clear why Summary pages may appear empty. Ahasuerus 02:39, 6 July 2016 (UTC)


Please see User_talk:Anniemod#Transliteration_of_.22.D1.8A.22 and comment/contribute if you deem doing so appropriate. I will also point Linguist to it. Thanks. --MartyD 01:25, 8 July 2016 (UTC)

BTW, I notice an unfortunate aspect of the transliteration support is that with separate transliterations of pub title and title record, there's even more duplicate entry/maintenance and possibility for inconsistency.... --MartyD 01:38, 8 July 2016 (UTC)
I was thinking about this issue earlier today as I was working on one of the new cleanup reports. In an ideal world, publications wouldn't have separate titles, authors or transliterated titles. Instead every publication would inherit these data elements from its "reference title".
Unfortunately, our current rules make it impossible. For example, we merge EDITOR titles, which then diverge from publication titles. Also, some of our editors support the ability to capture certain data -- mostly subtitles -- at the publication level but not at the title level. We'll have to address these issues before we can change the database schema and the software that has been built on top of it. Ahasuerus 02:31, 8 July 2016 (UTC)
Yeah, I know. You might consider another clean-up report, though, to identify mismatched pub record + title record transliterations. If pub title text == title title text and either one has transliterations, complain if the other does not or if the other's do not match. I didn't look to see what the storage is; presumably count mismatch or length mismatch would be cheap indicators, avoiding the more expensive comparisons much of the time. --MartyD 13:17, 8 July 2016 (UTC)
FR 905 has been created. Thanks! Ahasuerus 14:31, 8 July 2016 (UTC)

Poslednyaya Bitva title change

Hi, needs title cleanup (putting the Russian title as a title here, and what we have as a title as a transliteration). As you had verified it and it's not either notes or image so posting directly here... Let me know if you want me to get that handled (and any more of those I find while clearing the cleanup report) Anniemod 06:19, 12 July 2016 (UTC)

Updated, thanks! I have a number of "sequels by other hands" in my library and Russians were particularly active in this area in the 1990s, e.g. see this series note. Ahasuerus 15:42, 12 July 2016 (UTC)
I've read a couple of those - the 4th and the 5th are even translated into Bulgarian (2002 and 2003 respectively). There are quite a few of those series around - English, French, you name it :) Anniemod 16:15, 12 July 2016 (UTC)
By the way - while talking for the Russian titles, I had been fixing a lot of these: - Interior art, set with a transliterated title and Language Set to English. That means they do not show up on any report for cleanup (latin letters only so it is fine for English language) -- I had been chasing them every time I fix a title and sometimes even down to authors when I fix one. Any chance you can get a report of the pubs where interior art has a different language than the text counterparts? I suspect that only the non-latin languages are in trouble. If it will be too much of a problem, they will show up sooner or later and can be fixed. Anniemod 16:25, 12 July 2016 (UTC)
There is a moderator-only "Author/Title Language Mismatches" cleanup report which is currently limited to NOVEL titles. The associated Feature Request says:
  • The current implementation is limited to NOVEL titles. Keeping the FR open since we will eventually want to expand the code to cover other title types.
Checking the database, I see that once we change the report to look for all title types, the title type breakdown will be as follows:
  • 347 CHAPBOOK
  • 1,190 COLLECTION
  • 2,584 COVERART
  • 23 EDITOR
  • 5,699 ESSAY
  • 219 POEM
  • 575 REVIEW
That's a lot of titles to process, so we will probably want to add one new title type at a time. We will also want to make the report available to non-moderators. Thanks for reminding me of this issue! Ahasuerus 19:04, 12 July 2016 (UTC)
I do not see INTERIORART anywhere in your list though - and there are definitely some of them lying around. Unless if by miracle all are caught and taken care of. Plus... if the art of Whelan (with author language English) for example is used in a Russian book, with a Russian title (because they tend to translate titles), will the COVERART be English or Russian as language then - the name is Russian, the artist is English? Or the opposite - Russian/Bulgarian/Chinese illustrator with an English language book cover. The text ones are clear and obviously need fixing but I am not sure if that helps with the two art categories. Unless if I am missing something. Or overthinking it... Anniemod 20:05, 12 July 2016 (UTC)
I am sorry, I missed INTERIORART in my ad hoc query. The number is 5,182, which is a lot.
To answer your question, there have been a few discussions along the lines of "Do we really want to capture/display language information for COVERART/INTERIORART titles?" My personal opinion is that it is useful. For example, let's take Michael Whelan's Stormqueen! cover. See how the "Variant Titles" section lists the language of each variant? If we didn't capture language codes for COVERART titles, this information would not be available. Ahasuerus 20:27, 12 July 2016 (UTC)
Oh, I never said that we do not need them. I think it is useful - I was just saying that artist language and work language can be different. Part of the problem in the INTERIORART will be exactly what I had been fixing - language set to English by default for too many of them. Now - just thinking aloud but - why not add language on the publication level. Then anything inside can default on it - and can be adjusted if it is different). This way we do not have issues with the art (it defaults on the publication no matter when added) and we do not end up with this kind of stuff ether: ( - this Russian title in the middle there makes as much sense as snow in August in Phoenix). There will be multi language works that will need more attention but most of the works here will be single language ones. Now - I know that I do not know a lot of things about the system and I am probably missing something crucial but... as I said - just thinking aloud. Apologies if I am stepping on someone's toes. Anniemod 21:37, 12 July 2016 (UTC)
As you pointed out, multi-lingual publications do exist and we need to support them, which is why we can't associate a single language code with publication records. However, we have a compromise solution in place. When entering a new publication, editors can change the language of its constituent titles, e.g. see this Web page. Editors can also change their default language on the User Preferences page. The default language is used when the software generates New Publication and Edit Author data entry forms. Ahasuerus 22:21, 12 July 2016 (UTC)
P.S. The Russian title in Fifty Enthralling Stories of the Middle East was a data entry error, since corrected. Ahasuerus 23:00, 12 July 2016 (UTC)
Noted for the ability to change the language on constituent titles. I have the feeling that a lot of the weird data had been around for a while and just needs cleaning. Oh well - back to figuring out why some of those titles are in the funny reports (it's actually very helpful to need to dig through old posts and whatsnot to figure out things). Anniemod 23:27, 12 July 2016 (UTC)
Oh, yes. The database has been around for over 20 years and has gone through many phases. In the early years there were no moderators and anyone could change the data on the fly. Not surprisingly, that didn't work out too well. Later on, we farmed our data acquisition to an eager (bit not very bright) robot. It took hundreds of man-hours to clean up the resulting mess.
The current software, known as "ISFDB 2.0" and written in Python in 2004-2006, was a huge improvement, but, in retrospect, it was fairly rudimentary. We have been steadily enhancing it since 2009, adding support for languages, publication series and so on. There have been many detours and blind ends which took us even more man-hours to undo.
As far as translations go, we had really bizarre and counter-productive data entry rules prior to 2010. That's when we started implementing "language support" and began undoing the damage that had been done. Once we enhance the software to allow entering non-Latin author names, the bulk of the changes will be in place. Well, except for adding "translator support", which will be a separate headache. Ahasuerus 00:08, 13 July 2016 (UTC)
Well, technically there may be fast enough way to get something for authors put together - add two more fields on the Author page - "Name in the native script" and a transliteration for the same. This way at least authors like Vezhinov will have his most common name on the page. There are already fields on the page that support special scripts (the legal name - which is great but noone in Bulgaria will recognize Vezhinov's real name and he is actually in the high school schedule and everyone had read something from him), two more... I know it is just a patch but still. This will also ensure that English people browsing do not need to stare at letters they cannot recognize. Anyway - I will stop with the ideas now - it is just one of those problems that every book related site I ever somehow gravitate into tends to struggle and work on :) Lucky for me, most of my books these days are in English so short of fixing other people's records, I should not be hitting that too much for a while Anniemod 06:28, 13 July 2016 (UTC)
And yes, I suspect that is not a new idea. Just had been thinking today -that's all. Anniemod 06:29, 13 July 2016 (UTC)
The plan is to modify author names to work the way titles, publication, publishers, legal names and publication series currently work. A new, "infinitely repeating", field will be added for transliterated author names and all non-Latin names will be corrected to use the native form. The Author Directory will be adjusted to incorporate transliterated author names.
However, before we can do that, we need to optimize the code behind our Summary pages. The way it retrieves data from the database is inefficient and causes significant delays when viewing long Summary pages, e.g. see Robert Silverberg's Summary page. I was getting ready to start working on this issue when SourceForge went down 22 hours ago. (CVS seems to be back up as of 5 minutes ago, but their status page says that they are still working on restoring certain services.) Ahasuerus 17:35, 13 July 2016 (UTC)
Good luck! One request if I may? Allow the ability to mark preferred spelling in the native language - mainly so that when you open a page and you see a name in 13 scripts you do not read (or barely read - if you read Cyrillic, you may not be able to make a difference between Bulgarian and Russian spelling for example), you can at a glance find which one is the native one (Legal name does not serve properly for authors that are not using their own name). Anniemod 23:11, 13 July 2016 (UTC)
I am afraid I am not sure I understand the scenario. Are we talking about canonical names? If so, then keep in mind that an author/artist can have only one canonical name. All other name variations ("R. A. Heinlein", "Robert Anson Heinlein", etc) are treated as pseudonyms. Could you provide an example? Ahasuerus 03:13, 14 July 2016 (UTC)
Authors like Павел Вежинов - that is the name that you will see on any of his books in Bulgarian. Because it is not his legal name, the legal name field is taken by a name that has nothing to do with the name on the cover. There will also be П. Вежинов (if we get enough Eastern European Editions). When the new system is ready, you will have his English name (Pavel Vezhinov) at the top of the screen, then the legal name and transliteration and then a jumble of pseudonyms. What I am wondering if we can have eventually is to be able to mark the actual name in his own language in a way that is recognizable. He is a bad example because the name in Russian is the same - any Cyrillic language will write it the same. But not all names are the same - a lot of them get a ь somewhere in the middle. And Russian is the biggest Cyrillic language - their writing will overshadow the actual Bulgarian one when different. If not doable, the notes of the account will need to do but I was just wondering. :) Hope I made this a bit clearer? If under the new system Павел Вежинов will be the canonical name, then that's what I am asking for. But if it is Pavel Vezhinov... Anniemod 04:04, 14 July 2016 (UTC)
Once the new system is in place, the canonical name will be "Павел Вежинов". "Pavel Vezhinov", "П. Вежинов", "Paweł Weżinow", "Pavel Vežinov", "Pavel Vejinov" and all other forms of his name will be treated as pseudonyms.
Basically, our ultimate goal is to capture all bibliographic data exactly as it appears in publications. We then use variant titles, pseudonyms, transliteration fields, and other software tricks to link various data elements and create an integrated view that makes sense to our users. And we are getting close! :) Ahasuerus 04:41, 14 July 2016 (UTC)
Ah, I missed the fact that canonical name will be in the author's language. Thanks for explaining :) Anniemod 15:36, 14 July 2016 (UTC)
Thanks for fixing that one. It was driving me crazy. And thanks for the patience :) Anniemod 23:27, 12 July 2016 (UTC)
You are welcome! SourceForge went down earlier today, so I am just sitting here waiting for it to come back so that I could commit the next patch :) Ahasuerus 00:08, 13 July 2016 (UTC)
Non-latin languages are always fun, aren't they? Anniemod 20:05, 12 July 2016 (UTC)
Non-Latin languages are bad enough, but multi-script languages (like Serbian and Azerbaijani) are much worse! Ahasuerus 20:27, 12 July 2016 (UTC)
No argument. None at all. These and the languages with 6 a-s or 3 c-s or whatever. Anniemod 21:37, 12 July 2016 (UTC)

Pending Submissions

Hello, there is one old submission that is held by you. Can you have a look at it. Hauck 09:09, 14 July 2016 (UTC)

I am waiting for Google's cache to drop their version of Lyndon Hardy's Alphabetical Bibliography page. There is something odd about the way Google handles this particular page. Ahasuerus 16:09, 14 July 2016 (UTC)

There are also two submissions by a seemingly inactive contributor (Dcarson), can you do something about them (perhaps deleting them as they concern ARC that are usually not allowed). Thanks. Hauck 09:09, 14 July 2016 (UTC)

They are apparently clones of this pub record, so it should be safe to delete then. I should be able to do it later today. Ahasuerus 16:09, 14 July 2016 (UTC)
Done. Ahasuerus 20:07, 14 July 2016 (UTC)

Loss of Non-Genre = No and Graphic = No?

Take a look at the submission I have on hold. One of Anniemod's earlier submissions showed the same thing for Non-Genre and Graphic getting wiped out (others have not). I thought I'd go edit them back in, but Edit Title doesn't show them. Any idea what's up and whether I should worry about it? --MartyD 00:49, 21 July 2016 (UTC)

It's our old friend Bug 603 aka Editing REVIEWs/INTERVIEWs sets their non-genre and graphic flags to NULL. Thankfully, it's almost harmless because the way the software handles these flags, a NULL is the same as "No". I plan to get to it before the end of the year :) Ahasuerus 01:17, 21 July 2016 (UTC)
Ah. Thanks. --MartyD 01:38, 21 July 2016 (UTC)

SQL for a statistic of titles and publications for a certain language

Hi. I wanted to get statistics about titles and publications for a certain lanaguage by title/publication type and came up with these two SQLs which seems to yield reasonable results in my local database copy. I already checked summaryStatistics in and it almost does the same. But can you maybe have a short look at these SQLs and check if I missed something regarding data model and joins for a specific language?

-- German titles (not publications) by title type
select cast(t.title_ttype as char), count(*)
from titles t
where t.title_language = 26 -- German
group by t.title_ttype
order by cast(t.title_ttype as char);

-- German publications (not titles) by publication type
select cast(p.pub_ctype as char), count(*)
from pubs p, pub_content pc, titles t
where pc.title_id = t.title_id
and p.pub_id = pc.pub_id
and t.title_language = 26 -- German
group by p.pub_ctype 
order by cast(p.pub_ctype as char);

Jens Hitspacebar 11:33, 30 July 2016 (UTC)

The first query looks good, but the second one counts every German title in every publication. I think what you want is something like:
select cast(p.pub_ctype AS char), count(*)
from pubs p
where exists(
select 1 from titles t, pub_content pc
where pc.title_id = t.title_id
and p.pub_id = pc.pub_id
and t.title_language = 26)
group BY p.pub_ctype 
order BY cast(p.pub_ctype AS char)
Ahasuerus 14:46, 30 July 2016 (UTC)
Ah, I see, right. That makes a lot more sense and is indeed what I want. I just checked my second SQL again, and now the whopping number of 24590 German MAGAZINEs it yields seems a bit too high compared to the currently 33327 MAGAZINEs altogether in the database. Your SQL results in 3058 German MAGAZINEs, which makes a lot more sense. How on earth could I think initially that my second SQL's results are reasonable? :) Thanks a lot for your help! Jens Hitspacebar 15:16, 30 July 2016 (UTC)
You are welcome! I made the same mistake repeatedly when I started working on the software, so you are not alone :-) Ahasuerus 16:05, 30 July 2016 (UTC)

Swfritter inactive?

On Tuesday it will be a year since Swfritter responded to anything on his user page. It may be appropriate to add the "inactive" template to his discussion page. Chavey 05:33, 31 July 2016 (UTC)

Add away! :) Ahasuerus 14:01, 31 July 2016 (UTC)
Done. Chavey 05:37, 1 August 2016 (UTC)

Amazon Planet

Another Ace double-use of a catalog #: [1] and [2]. The difference is in the notes. Cheers! --~ Bill, Bluesman 23:14, 9 August 2016 (UTC)

Thanks! Ahasuerus 23:29, 9 August 2016 (UTC)

'Lia Johansen' series

Do you recall the source you used to name the Lia Johansen series?

Using there's:

  • Nova Hardcover in the 'Nova' series.
  • Nova Kindle in the 'Spectre War' series
  • Nova Audio with no series
  • Nova MM Paperback in the 'Spectre War' (Book 1) series
  • Archangel Hardcover in the 'Spectre War' (Book 2) series. (upcoming March 7, 2017 and so not on ISFDB)
  • Archangel Kindle in the 'Spectre War' series. (upcoming March 7, 2017 and so not on ISFDB)

I have "Nova" ISBN 978-0-7564-1082-7 MM PB published in June 2016 in hand and the only evidence of a series is an ad on page [325] for "Nova" ISBN 978-0-7564-1081-0 (the hardcover) which has "And don't miss the trilling sequel "Archangel" coming in 2017. There's no mention of a series name.

I personally think "Spectre War" is an odd name for the series as the existence of the spectres is not revealed until near the end of "Nova." However, the author’s original plan was a series of five books and it’s clear by the end of "Nova" that the continuation of the story is likely a war versus the spectres.

The problem with "Lia Johansen" as a series name is that while she's the main character in "Nova" it appears that she no longer exists by the the of the story. --Marc Kupper|talk 18:44, 19 August 2016 (UTC)

I am afraid I don't recall where the series name originally came from. It may have been something added by the editor who updated the first edition record with the Locus data. If it was my idea, then I probably reused whatever the internet showed at the time.
As far as "Spectre War" goes, I don't think it's necessarily a spoiler since the word "spectre/specter" can be used in different contexts. If I picked up a book/series called "Spectre War", I wouldn't be sure what to expect. Once specters made an appearance, I would probably say "Oh, so that's what the title/series was in reference to!" Ahasuerus 19:52, 19 August 2016 (UTC)
Thanks. I had assumed the series title was your idea based on this which seems to be the only submission that mentions 'Lia Johansen'.
I found the source for that it's a 5-book series. It's in the author description field or whatever it's called on Up to Jul 28, 2016 it had "Author of NOVA, the 1st book in a 5-book sci-fi series published by @DAWBooks." Some time between Jul 28, 2016 and Aug 11, 2016 she changed it to "Author of NOVA & ARCHANGEL (Spring 2017) the 1st 2 books in a 5-book sci-fi series published by @DAWBooks"
The author never uses the word "spectre" when tweeting or blogging and so I assume "Spectre War" was invented by a penguin.
Do you mind if I change the series title to "Spectre War"? That'll allow it to match what the penguin's are using for most of their records
Penguin URL Title Binding Date Series
9780756410810 Nova Hardcover 320 Pages 2 Jun 2015 (no series)
9780147525413 Nova Audiobook 2 Jun 2015 (no series)
9780698197688 Nova Ebook 368 Pages 2 Jun 2015 Spectre War
9780756410827 Nova MM Paperback 336 Pages 7 Jun 2016 Spectre War
9780756412937 Archangel Hardcover 320 Pages 7 Mar 2017 Spectre War
9780698197817 Archangel Ebook 320 Pages 7 Mar 2017 Spectre War
--Marc Kupper|talk 21:52, 19 August 2016 (UTC)
Please go ahead! Ahasuerus 21:59, 19 August 2016 (UTC)

Author With No Records

Shaun Michael Jooste exists without any records. I did a "Show All Titles" as well as advanced searches for Title - Author, Title - Reviewed Author, and Publication - Author. Am I missing another place I should be looking for 'dangling' records? Thanks. -- JLaTondre (talk) 16:39, 21 August 2016 (UTC)

Hm, no, that should cover everything. Let me restore the latest backup on the development server and check submission history and a few other tables. Ahasuerus 17:03, 21 August 2016 (UTC)
Found it -- see this interview. In the past, author interviews appeared at the top of the author's Summary page, but they were moved to a separate section on 2016-08-11 (patch r2016-128.) However, the way the new logic works, author interviews are only displayed if the author has written something, hence the blank page. I'll go ahead and file a bug report. Thanks for finding the problem! Ahasuerus 17:34, 21 August 2016 (UTC)

Search Oddities

Hello, is this an intentional thing or is it just an oddity of the software that handles the Latin capital/smalls properly but treats everything else as distinct characters (so cannot make the connection?)For example:

  • Hamilton and hamilton return the same search results for authors
  • Вежинов and вежинов do not.
Our database stores Latin-1 characters natively while everything else is HTML-encoded. For example, "Агоп Мелконян" is stored as:
&# 1040; &# 1075; &# 1086; &# 1087; &# 1052; &# 1077; &# 1083; &# 1082; &# 1086; &# 1085; &# 1103; &# 1085;
which causes various oddities. If you search on "&#", the system will report 1,071 matches. As you noticed, the software has no way of telling that a lowercase "ж" and an uppercase "Ж" should be treated the same way during searches.
The ultimate solution will be migrating from Latin-1 to Unicode, which is on my list of things to do. Unfortunately, it's a much hairier project than it appears at first sight, but it will need to be done at some point. Ahasuerus 20:41, 26 August 2016 (UTC)
Figured out it may be something like that but decided to ask (in case it is just an oversight) :) Anniemod 21:08, 26 August 2016 (UTC)

And while I am around, what's with the counter of "Pubs with Romanian titles with s-cedilla or t-cedilla" report - it seems to be reporting real number +10 (and I am sure it is not just someone cleaning them before I see it - it is very consistent the last few weeks - 10 or a bit over if there are real ones :) Anniemod 19:47, 26 August 2016 (UTC)

Hm, let me see... It looks like it's a display bug in the report. Publications like this one do contain s-cedilla or t-cedilla, but the display part fails. I'll see if I can find and fix the problem. Thanks for identifying it! Ahasuerus 20:41, 26 August 2016 (UTC)
Ah, so there are 10 legitimate ones that just cannot be seen? If you get get me a list I can fix them so at least it stop annoying me (if it will be easier than fixing the display?) Anniemod 21:08, 26 August 2016 (UTC)
The display code needs to be fixed anyway, so might as well do it now. I haven't done much with the software over the last few days because I have been sick, but I am feeling better now and should be able to handle a bug or two. Ahasuerus 21:10, 26 August 2016 (UTC)
Feel better. The software can wait :) Anniemod 21:12, 26 August 2016 (UTC)
No worries, my body is very good at telling me when it can and cannot do certain things :) In the meantime, the bug has been fixed. Ahasuerus 21:44, 26 August 2016 (UTC)
Thanks. And these 10 should finally be done now :) Anniemod 21:49, 26 August 2016 (UTC)
Approved, thanks! Ahasuerus 21:58, 26 August 2016 (UTC)

Field name update ?

Hello ! As I completed a merging of two author names, I noticed that one of the field names was "author_lastname Conflict:" : I know this is not a priority, but maybe it should be changed to "author_familyname Conflict:", since "Last Name" was changed to "Family Name" some time ago (just for the sake of coherence). BTW, could any moderator do the correction, or is it restricted to the Happy (Technical) Few :o) ? Thanks, Linguist 09:08, 29 August 2016 (UTC).

Well, it's certainly possible to change the name of this field in the database. However, there are two issues here.
The first one is the fact that we have 18 modules that refer to it as "author_lastname", often multiple times. All of them will have to be changed and retested to make sure that everything still works.
The second issue is that we have a feature request to change the name of this field from "Family name" to "Sorting Name". I have been thinking about it and it occurs to me that perhaps it would be better to make it "Directory Entry" instead of "Sorting Name". What do you think? Ahasuerus 14:44, 29 August 2016 (UTC)
I can't say I have a strong preference for one or the other. "Sorting name" may be simpler, though, but I'd be just as satisfied with "Directory Entry". Linguist 15:11, 29 August 2016 (UTC).
I guess "Sorting Name" it is then. At some point we will probably have multiple directories, one per script/alphabet, but we will need to implement a whole bunch of other things first. Ahasuerus 15:36, 29 August 2016 (UTC)

Two Feature Requests

I came here to ask how hard it would be to identify the problem records and the old values, but you already responded at the community portal. So instead, as a reward for all your work, I'll give you a couple of feature requests. ;-)

  1. Ability to search the recent edits listing - by submission type, by submitter, and/or by subject. That would make it much easier than paging through the history when looking for something specific.
  2. Add language to the Author's advanced search.

Thanks! -- JLaTondre (talk) 14:01, 4 September 2016 (UTC)

Adding language to the Advanced Author Search would be fairly easy -- FR 926 has been created.
The ability to search submissions by submitter and by submission type would require a new Web page or a new section in the Advanced Search, but it's not hard to do. Searching by subject would not be feasible given the current table layout since it would require scanning the whole submission table. We'd have to move the subject information to a separate field first. Certainly doable, just more involved. In any event, FR 927 has been created. Ahasuerus 14:38, 4 September 2016 (UTC)
Instead of a new page, I'd suggest adding it to the top of the recent edits page. Similar to how the wiki my contributions page works. By default, show everything & then people can add terms & re-search to filter results. -- JLaTondre (talk) 16:30, 4 September 2016 (UTC)
Good point -- the FR has been updated. Thanks! Ahasuerus 16:36, 4 September 2016 (UTC)

Cleanup report for empty "href" attribute in links?

Hi. I just came across a publication record where the submitter forgot to add the URL to the href attribute of an a tag in the NOTE field and the link therefore simply pointed back to the page itself. Like this:

<a href="">Click me. Now! DO IT!</a>

I checked my local copy of the database and found 28 publication records with this "problem". That's nothing of any significant priority, but maybe a cleanup report might help finding these records:

select pubs.pub_id, pubs.pub_title
from pubs, notes 
-- assume that multiple spaces have possibly been entered:
where lower(replace(note_note, ' ', '')) like '%<ahref=""%'
and pubs.note_id = notes.note_id;

The same might apply for notes in titles, series, etc... Jens Hitspacebar 20:45, 8 September 2016 (UTC)

Sounds like a good idea! Please go ahead and create an FR with this information. I am still sick, so I can't really do any meaningful development work at the moment, but I plan to do a cleanup pass on the recently created FRs once I get better. Ahasuerus 21:09, 8 September 2016 (UTC)
Ok. FR has been created. Get well soon! Jens Hitspacebar
Looks good, thanks! Ahasuerus 21:24, 8 September 2016 (UTC)

Publication Advanced Search

Publication Advanced Search using ISBN doesn't work for anything other than exact match. I was able to work around it by going to the publisher page and working my way through the years, but that was only easy as the publisher didn't have a lot of records. Not a big deal, but would be nice to be able to have those searched for the odd occasions. Thanks. -- JLaTondre (talk) 00:25, 13 September 2016 (UTC)

That would be Bug 169, "Advanced ISBN Search doesn't find partial matches", one of the remaining 20 bugs :) As I recall, there is special search logic for ISBNs and it was implemented differently for regular and Advanced searches. I'll take a look and see if it would be easy to fix. Ahasuerus 00:31, 13 September 2016 (UTC)
It's not a trivial fix because the logic checks 4 different ISBN versions: ISBN-10, ISBN-13, with the hyphens, without the hyphens. It doesn't look too bad, though, I'll see what I can do tomorrow. Ahasuerus 01:01, 13 September 2016 (UTC)

SFE-F --or something

Hi. The Encyclopedia of Science Fiction, 3rd edition, is online at and The Encyclopedia of Fantasy, only edition, at (At least, the URL are all displayed with those prefixes when I visit via links in the database and in my user space.)

I wonder whether the displayed linknames can be differentiated, such as SFE3 and SFE-F, or SFE-Fantasy, etc. --Pwendt|talk 02:03, 13 September 2016 (UTC)

That's a good point. We already have special logic for Wikipedia URLs, so it shouldn't be too hard to add more logic to differentiate between SFE and EoF. I have created FR 930 to document the request. Thanks! Ahasuerus 02:24, 13 September 2016 (UTC)
Done -- see the announcement on the Community Portal. Ahasuerus 14:34, 13 September 2016 (UTC)

Thanks for all your work here. Good night. --Pwendt|talk 02:03, 13 September 2016 (UTC)

We all do our part! :-) Ahasuerus 02:24, 13 September 2016 (UTC)

Search for a Subset of Australian SF Novels

I was contacted by Dr. Catriona Mills of Austlit for assistance in winnowing out novels by Australian writers in specific years. The Advanced Search parameters do not allow them to narrow things down sufficiently, e.g., English authors born in Australia. I thought it might be possible to write a script to do this, if it is not too much trouble. I also thought it would also be good to add all authors that have a link to one of the Australian biblio sites to the search parameters, like Austlit (if possible) - as I've noticed that many authors linked to those sites have no birthplace given or were born elsewhere.

I asked for the search parameters and I've pasted her response and contact info below.--Rkihara 23:20, 13 September 2016 (UTC)

"...........Yes, ideally I’m looking for novels published by Australian authors in 2000, 2001, 2010, 2011, 2015, and 2016. If the search can include now-Australian authors who were born elsewhere, that would be fantastic, but even limiting it to Australian-born authors would be an enormous assistance to me. Many thanks to both you and Ahasuerus."
Dr Catriona Mills
Senior Researcher and Indexer, AustLit
School of Communication and Arts
Room 330, Joyce Ackroyd Building (Building 37)
The University of Queensland
St Lucia, 4072
Sure, it sounds doable. I'll take a closer look tomorrow. Ahasuerus 02:38, 14 September 2016 (UTC)
I have installed a patch which should facilitate Catriona's searches -- see the Community Portal for more details. The point about Austlit is an interesting one and I'll see what I can do. It's not as simple as what I have done so far, but it may be feasible, although the resulting report may be painfully slow.
As far as searches for books by "now-Australian authors who were born elsewhere" goes, I think the best you can do is search for publications whose price starts with "A$" or whose publisher's name contains the word "Australia". In many cases the author will live in Australia or have some kind of connection to Australia, e.g. Scott Westerfeld spends about half of his time down under. However, there is nothing stopping Australian publishers from publishing books by people living elsewhere.
Also, we seem to be missing quite a few SF titles known to Austlit. There may be various ways to rectify the situation using Fixer, but it's a big can of worms. Ahasuerus 17:42, 14 September 2016 (UTC)
Thank You! I've sent notice of the changes to Dr. Mills. Maybe we could synch our data with Austlit like we did with SFE3, and generate a cleanup report? I've been thinking of updating our Australian authors once I finish with transliterations and the like. Austlit has also given us annual guest access. I have the password, but I don't want to post it on a publicly accessible forum.--Rkihara 18:11, 14 September 2016 (UTC)
OK, Advanced Title/Publication searches have been changed to support author webpages like Austlit -- see the Community Portal announcement for details.
As far as synchronizing the ISFDB database with the Austlit database -- or any other database, for that matter -- goes, it can be done automatically, but it's a huge can of worms. I can barely keep my head above water keeping us sync'ed with major US/UK releases at the cost of many hundreds of man-hours per year. I also have to limit the sync process to major retail outlets because I don't have the bandwidth to handle library data. At one point I developed the software needed to talk to libraries directly as well as to parse their publicly available catalog dumps, but there are not enough hours in the day to do the required data acquisition and cleanup work :-(
Re: manual synchronization with Austlit, we'll need to decide what the requirements are first. With SFE3 we limited the report to author URLs, which means that we still have many discrepancies at the title level. What would you like the proposed Austlit cleanup report to do? Ahasuerus 20:17, 14 September 2016 (UTC)
For building up the Australian presence, maybe something like "Prolific Authors in Austlit, Missing or Having Less than "X" Titles in the ISFDB?"--Rkihara 21:37, 14 September 2016 (UTC)
We can certainly do that, but a low number of title records in our database doesn't necessarily indicate that Austlit has additional titles. For example, consider Barbara Lamar. We have one novel and one novelette on file, which is the same as what Austlit has.
Checking the database, I see that we have almost 1,000 author records with Austlit links, so checking all of them manually would be fairly time-consuming. And, since we expect Austlit to continue adding records to their database, any cleanup that we may do in 2016 will need to be repeated in 2017, 2018, etc.
On the other hand, if we could get a list of their ISBNs tagged "science fiction", "fantasy", etc, we could write a script that would check them against what's in our database. The resulting subset of ISBNs could be made available to editors as a cleanup report. Do you think it's something that we could request? Ahasuerus 22:06, 14 September 2016 (UTC)
I can ask. I'm also going to ask Dr. Mills would like to register for a user account, which would allow her to communicate more directly with us on our forums.--Rkihara 22:27, 14 September 2016 (UTC)
Sounds like a plan! Ahasuerus 22:29, 14 September 2016 (UTC)

I received an email from Dr. Mills thanking us for the changes, posted below. In my return message I asked if she would be able to generate the file of Australian ISBNs, and encouraged her to open a user account on the ISFDB.--Rkihara 00:01, 16 September 2016 (UTC)

"This is wonderfully useful, thank you! And please thank Ahasuerus for me, as well. Please let me know if there's anything we can help you with at any stage.
Excellent! :-) Ahasuerus 01:24, 16 September 2016 (UTC)

The Ultimate Story Collection

Hello, this publication that you created (from fixer, I suppose) is an OMNIBUS without identified content (as per notes) and it appears on the corresponding cleanup report. Perhaps, in such cases, an "Ignore" option may prove interesting. Hauck 07:22, 19 September 2016 (UTC)

Thanks for looking into this! Admittedly, it's probably an omnibus. However, given the page count, it's also possible that it's a collection of 6 stories. I've changed the title type and the publication type to COLLECTION for now. Once we have more details, we can change it back. Ahasuerus 12:33, 19 September 2016 (UTC)

The Most Thrilling Science Fiction Ever Told, No. 3

Artist found for your verified. Hauck 16:47, 20 September 2016 (UTC)

Also for this other one. Hauck 16:59, 20 September 2016 (UTC)
Thanks! Ahasuerus 18:01, 20 September 2016 (UTC)

The Invalid Record URLs in Notes report

When you have a chance, can you look at the report and see if that counter "2" is stuck or if there are entries that cannot be seen? It had been stuck like that for a few days and every time when there were other items, it was increased and then back to 2 the next day... So I wonder if this is not similar to the cedilla report issues with visualization... (or somehow the counter is stuck). Thanks :) Anniemod 08:36, 23 September 2016 (UTC)

Thanks, I'll take a look... Ahasuerus 14:37, 23 September 2016 (UTC)
There are two separate issues here. The first one had to do with displaying invalid URLs in synopses. It was fixed in the patch that I installed 5 minutes ago.
The other issue has to do with publisher notes. It would appear that they are not always deleted when the associated publisher is deleted/merged (?). I'm still looking into it. Ahasuerus 15:26, 23 September 2016 (UTC)
Why should it be easy when it can be hard. Good luck hunting it. Anniemod 16:46, 23 September 2016 (UTC)
OK, I have confirmed that old notes are left behind when publishers are merged. Now to fix the software and delete the bad data. Thanks for identifying the problem! Ahasuerus 17:07, 23 September 2016 (UTC)
Do you mind taking a look at the report again? The one that you managed to get back in the light ( ) had been fixed (the two links pointing to submissions pages are now pointing to the records of the books) but it is still showing in the report. What am I missing? Anniemod 17:55, 23 September 2016 (UTC)
This is a somewhat unusual cleanup report. As you know, almost all cleanup reports are regenerated nightly. The displayed counts remain the same for the next 24 hours even if some problems are cleaned up in the meantime. At the same time, in most cases the display logic is coded to reconfirm that each flagged record is still problematic. As soon as editors start cleaning things up, what the main cleanup page says about each report and what the actual report shows get put of sync.
This particular report is an exception. The display logic doesn't recheck each flagged record's Notes. Instead it display a moderator-only "resolve" link which lets moderators remove the flagged record from the report manually. Clicking the link simply deletes the record number from the cleanup table, so if the "resolving" moderator makes a mistake and the Notes field still has a problem, the record will be re-flagged the following night.
I have now clicked the "resolve" link for the offending record, so everything is back to normal. Admittedly, the fact that some of our cleanup reports do not follow the standard pattern can be irritating, but I can assure you that the current system is vastly better than what we had 3 years ago :-) Ahasuerus 18:50, 23 September 2016 (UTC)
Aha, so next time I am clearing an issue from there, I should tell the moderator to go and click the resolve as well after approving :) I am just used to the others dropping on their own when they are fixed (and staying if they are not fully fixed) so I was wondering what is going on. That Weird empty resolve column at the end of the line in the entry probably should have told me that something is different (and not I will know that when it is there, I need a manual moderator resolve). I knew that the "2" is staying till tomorrow - all reports do it :). Thanks! Anniemod 19:21, 23 September 2016 (UTC)
Good point about the "weird empty resolve column" -- fixed now. Thanks. Ahasuerus 01:00, 24 September 2016 (UTC)

What am I missing - Canonical names?

I had been working on some of the Russian authors and had been wondering... Can't I just request the canonical name to be changed (because I cannot change it) and then go back and fix the ones that needed to have the old latin form? No double renaming (I just did a few of those), no moving of data to a new record (as I had to do with Snegov). What am I missing on why we cannot do that? Anniemod 21:06, 27 September 2016 (UTC)

You can certain do that; it was one of the methods that I described in the original patch notes. However, you will then need to go back and change all English, German etc titles and pubs to their original form. My guess is that it's somewhat more error-prone than the other method, but it's certain viable. If the vast majority of an author's titles are in Cyrillic, then it's typically the best approach. Ahasuerus 21:57, 27 September 2016 (UTC)
I saw that but then when we talked on my page, you said that the second way is safer so... making sure I am not missing something. For a lot of the Russians, there is about gazillion and seven ways to write them in Latin (each language transliterating their own way - see for example) so not all titles are influenced (5 or 6 only are with that name - which I added English to...). Anniemod 22:11, 27 September 2016 (UTC)
Oh yes, it can be quite messy. There are almost 30 major transliteration systems in NorAm/Europe alone, which was why we decided to support multiple transliterated names/titles. Ahasuerus 23:20, 27 September 2016 (UTC)
After which I figured I would come and ask. Sounds like "make your own decisions, just make sure you clean after yourself" kinda thing :) :) Anniemod 22:11, 27 September 2016 (UTC)
Pretty much! BTW, one advantage of using "(Latin)" or "(English)" is that they are easy to find if you forget to remove them. Ahasuerus 23:20, 27 September 2016 (UTC)

Links cleanup again: Linkedin

Hello, can you run a query to see how many Linkedin links on author pages use the old format(such as for example. That one needs to be now (and yes, this is the opposite of what everyone else did with links - these guys added the name (or a name anyway) to a link that was safe from weird characters...)? I've fixed a few the last few weeks and if there are a lot of them around, we should deal with them I think :) Thanks :) Anniemod 23:22, 27 September 2016 (UTC)

I am afraid I am not familiar with LinkedIn's links. Did all of their old links contain ""? If so, then this Advanced Search finds a few dozen offenders. Ahasuerus 23:27, 27 September 2016 (UTC)
Yeah, that was the standard. The one I know of anyway. :) The new one is "name" followed by the old 3 elements in reverse order (a/833/671 becomes 671833a) or just a name if a pseudonym is selected (my profile is like that). I did not even think of the Advanced Search for finding these so thanks :) All the new ones start with /in/ so any link that does not do that is invalid (I think they finally phased out all the /pub/ ones). I will work through that list and get these links in order - if any is still valid, I will leave it alone but I think all of them will need change. Anniemod 23:41, 27 September 2016 (UTC)
Apparently some forwarders still work, most do not... Probably have something to do with the custom names - or name changes. for example forwards even though the link is the old one. Will fix the ones that do not forward - and will keep an eye on the rest. Anniemod 23:58, 27 September 2016 (UTC)
Thanks! Ahasuerus 00:03, 28 September 2016 (UTC)

The Best of John Sladek

User Horzel drew my attention to the signature on the backcover of The Best of John Sladek, see here. I added the artist and changed the note. --Willem 19:33, 28 September 2016 (UTC)

Looks good, thanks! Ahasuerus 19:55, 28 September 2016 (UTC)

Australian SF Novels in Austlit

Dr. Mills at Austlit sent a file of Australian authors with ISBNs, etc., in response to our request for same. Copy of email below. Should I use the standard file upload link or send it somewhere else?--Rkihara 05:45, 29 September 2016 (UTC)

Hi, Ron
Please find attached a list drawn from the database of all science fiction novels, with ISBNs in a separate column.
The search was set up to include what we call 'international' authors (who are sometimes listed on AustLit because their work has been adapted, for example, by an Australian author), so this should only return results on Australian authors.
It also includes extracts of novels and author-series (trilogies, multi-books works, and so on), but limiting it by 'single work' will filter those out if there are less than useful.
Excellent! Please send it to my e-mail address (ahasuerus at as an attachment. Thanks! Ahasuerus 14:48, 29 September 2016 (UTC)

Titles/Publications mismatches

Wasn't there a report for that? If so, how did this one slip: If we do not have one, maybe we need one? :) Thanks! Anniemod 19:24, 29 September 2016 (UTC)

There is a Publication Title-Reference Title Mismatches report, but it's currently limited to CHAPBOOKs and OMNIBUSes. The other title types can be tricky. Remember how I said that:
  • in an ideal world all publications would have a "reference title" and inherit that title's authors instead of having a separate set of authors. Alas, early on we made certain design decisions which make it impossible?
Well, for starters we originally decided to merge EDITOR titles for each year of publication, which means that MAGAZINE pub titles and EDITOR title titles are rarely the same. Another early decision that has made things difficult was the leeway given to our editors re: adding series names and subtitles to publication titles.
The result is that there are tens of thousands of discrepancies and no easy way to create a cleanup report that wouldn't be unmanageable. I am all for revisiting these design decisions at some point -- some of them can be undone programmatically -- but I expect that it won't happen anytime soon. Ahasuerus 20:00, 29 September 2016 (UTC)
Thanks for the explanation :) That reminds me that I had been meaning to ask why the report is missing the novels. Maybe include NOVEL as one of the types it is reporting about? There is no legitimate reason for a novel to have a publication/title mismatch in those(unless if I am missing a subcase?). I understand why EDITOR should stay out of the report but novel seems safe (or safe enough anyway)? It will highlight cases like that one - the translations that need fixing. Anniemod 20:05, 29 September 2016 (UTC)
Let's use this publication as an example. The pub title is "The Crystal Shard: The Legend of Drizzt: Book IV". The title title is "The Crystal Shard". Under the current rules it's a perfectly legal scenario. There are thousands of similar mismatches. In order for a cleanup report to be useful, we would have to change the rules first, which would require a Rules and Standards discussion. Ahasuerus 20:22, 29 September 2016 (UTC)
I see - did not think about that case. Although I would make the argument that this is a new title (the same name a different language is a new title or Harry Potter and the Sorcerer's Stone is a variant title and not just both of them into a single title... But I understand that the rules allow it. Maybe we do need that discussion after all. Can you run a query on how many novels we have with mismatched titles? I am just curious. I am trying to chase down the ones that are translations that never got properly handled and trying to stumble on them by finding them while fixing the covers languages is not very effective :) Anniemod 20:48, 29 September 2016 (UTC)
Let me see... Here is the breakdown as of last Saturday morning. The first number is the total number of mismatches. The second number is the number of mismatches after eliminating subtitles and punctuation differences:
  • Novels - 8357 / 688
  • Collections - 1611 / 235
  • Anthologies - 1411 / 307
  • Nonfiction - 894 / 134
Which reminds me that the current version of the report already ignores subtitles and punctuation differences. Now that we have cleaned up CHAPBOOKs and OMNIBUSes, I need to add the other title types. Thanks for refreshing my memory! Ahasuerus 21:09, 29 September 2016 (UTC)
You are welcome :) And these do not seem too unmanageable. :) Anniemod 21:46, 29 September 2016 (UTC)

Lost in the shuffle

This seems to have been lost in the shuffle. Will you look into it? TIA. ···日本穣 · 投稿 · Talk to Nihonjoe 23:46, 29 September 2016 (UTC)

Sure! Which ones still need to be ignored? I have reviewed the ones posted on the Moderator Noticeboard, but couldn't find them on this report. Ahasuerus 23:51, 29 September 2016 (UTC)
Hmm...looks like someone did them, and then never indicated it on the topic on the moderator noticeboard. Perhaps they ran across them through another way. Thanks for checking. ···日本穣 · 投稿 · Talk to Nihonjoe 01:36, 30 September 2016 (UTC)
No problem! Ahasuerus 01:40, 30 September 2016 (UTC)

De Engelenmaker

Sorry about this one. I knew something was off when I submitted it but could not figure out what. Thanks for getting it fixed! Anniemod 16:39, 30 September 2016 (UTC)

No worries, the learning curve is steep :) Ahasuerus 16:41, 30 September 2016 (UTC)
I think it was more getting distracted what the publication was than anything else - all of the others this morning were removals of the non-English titles so I think I went with the flow. And as the message after the request is not as informative as the ones on edit for example, I did not realize that there was an issue. Anniemod 16:47, 30 September 2016 (UTC)



Can I ask you to take a look in these two records Daniel Mroź and Daniel Mroz? They are the same person and need to be pseudonymed (which I can do easily) but the canonical name is on the smaller of them - and I cannot see a way to get all that handled without a lot of edits (in proper order) - swap the canonical names via a third name and move the details (3 edits + waiting for a moderator to adjust canonical names before I can proceed if I am not missing some way) and then variant the lot under the Mroz name. Thoughts? Am I missing an easy way? And do you want to tackle it? If not and if that is indeed the only way, I can start working on it :) Anniemod 16:07, 3 October 2016 (UTC)

I am afraid there was no easy way to link the two author names. His last name is actually "Mróz" rather than "Mroź" and I have adjusted it accordingly. Also, a number of his Polish titles had been entered as by "Mroz" and had to be moved. I have created a bunch of variants and moved the bio data to the canonical name, so we are in better shape now.
Unfortunately, our software doesn't fully support author name pairs which are identical except for a single Latin-1 letter (in this case "ó") with the same collation position as its English analog. For now, I have changed the English name to "Daniel Mroz (I)". I'll examine the code to see if I can improve it to handle these cases correctly. Thanks! Ahasuerus 17:08, 3 October 2016 (UTC)
So it is not A or B, it is actually C. Marvelous. I knew I cannot get that one fixed properly on my own :) I got a few others like that fixed this weekend but that one was too... big :) I really dislike languages that go off-script and use weird letters. Thanks for getting that one handled! Anniemod 17:29, 3 October 2016 (UTC)

You are too fast with those Russian name changes

By the time I see my first submissions approved and come to get the next batch added, you had finalized them. :) Should I just leave them to you and go find something else to work on? :) Anniemod 18:17, 5 October 2016 (UTC)

Well, we are down to the last hundred, so I guess I'll just buckle down and finish them today. Thanks for handling the bulk of the outstanding entries! Ahasuerus 19:42, 5 October 2016 (UTC)
OK - I will bail out from that list then. :) Anniemod 20:00, 5 October 2016 (UTC)

Gods of Pegana

Hi, just wanted to let you know that I've been verifying first editions of Dunsany collections and making some changes to canonical story titles that affect later editions that you verified. If you have any questions or objections, leave me a note. Thanks! --Vasha77 21:59, 5 October 2016 (UTC)

No worries -- thanks for letting me know! Ahasuerus 22:01, 5 October 2016 (UTC)

Invisible Soft Return...

Hello, this Invisible Soft Return:\ shows up in the Publication Title-Reference Title Mismatches report and no matter how I look at it, it does not seem to be a mismatch. Is it the special character at the end that confuses the code? Or is there that I cannot spot? Thanks Anniemod 04:01, 8 October 2016 (UTC)

The titles are indeed identical. It looks like the culprit is the backslash, which is messing up MySQL's "LIKE". Let me see if I can figure out how to fix it... Ahasuerus 14:28, 8 October 2016 (UTC)
Fixed! Ahasuerus 16:43, 8 October 2016 (UTC)

E-Mail Demon

Hallo, have you had any luck with the resurrection? HolgerBurghardt 12:44, 9 October 2016 (UTC)

I am afraid not. We experienced a similar problem in January when the hosting company upgraded the OS and messed up file permissions. If I can't figure it out today, I'll have to call in reinforcements. Ahasuerus 15:32, 9 October 2016 (UTC)

Two cover images

Can you think of a way for me to coax a publication into having two cover image URLs, one for the front cover and one for the back cover? For now I've just put a URL for the back cover into the title record as a web page. I have a situation where the back cover of TQF #52 was used as the basis for a write-a-story-to-match-this-art contest. In TQF #55, that back cover was reproduced as interior art prior to the winning stories. So where normally I wouldn't bother, here it seems appropriate to record two covers and to have an interior art record as a variant of the second cover. I'd love to have that second cover show the right image, or have both coverart records show both images (unfortunately, it shows the only "cover" image instead). I tried semi-colon as a separator in the cover image URL, but it did not seem to understand that. I realize I could download the two illustrations and make a composite, but I'd actually rather keep them separate. Not a big deal, I just figured I'd ask if I missed something. --MartyD 13:49, 9 October 2016 (UTC)

At one point the software was supposed to support "BACKCOVERART" as a separate title type, but that approach was abandoned a long time ago.
The current rules say "Artwork on the back cover of a publication is treated as interior art". The best way to handle this particular situation (that I can think of) would be to enter it that way and create a variant with copious notes. Sorry, I wish I could be of more help... Ahasuerus 15:39, 9 October 2016 (UTC)
Wow, that was rather buried! :-) I have converted it to conform to that. Thanks for the pointer. --MartyD 15:50, 9 October 2016 (UTC)

Romanian SF Anthology Nemira '94

Hi. I have looked into the matter of Romanian SF Anthology Nemira '94 and did what I could. There are three titles I couldn't determine the publishing date of. As for the essay by Romulus Bărbulescu and George Anania entitled "Arguments for a Combat Against the Angels", it must bear a relation with the novel Lupta cu îngerul ("Combat with the Angel") which these authors started in the 1990s, but was left unfinished. The Romanian title of this 3-page essay escapes me though. Sorry I can't do more (but I will keep searching…). Linguist 10:11, 13 October 2016 (UTC).

It also appears to me that this record must contain quite a few typos, including the Bărbulescu / Barbulescu business, and the Dănuț Ungureanu / Danut Ungureaunu variation. The lack of response from MagicKey is a bit frustrating. Linguist 10:21, 13 October 2016 (UTC).
Thanks for looking into this! Unfortunately, editors come and go, which can leave things hanging... Ahasuerus 14:30, 13 October 2016 (UTC)

What Entropy Means to Me

Hi, I've found the cover artist for What Entropy Means to Me. It is Fernando Fernandes.--Dirk P Broer 17:52, 13 October 2016 (UTC)

Updated, thanks! Ahasuerus 17:55, 13 October 2016 (UTC)

"Regal Crest" publishers

This publisher looks like a possible addition to the list that Fixer pulls from (assuming that's not too hard to include). This publisher is a step above the "self-publishers", e.g. they'll create covers for you, and I think do actual editing. We had a couple of their books in the system, and I was looking at them as I was adding some books by a friend. They have multiple imprints, listed here by theme. The books we had in the system were all from Quest Books / Regal Crest, an Action-Adventure/ Murder/Mystery imprint. I haven't evaluated that imprint much further, but I noticed they had some time-travel books we did not include. I have just finished adding all of their Silver Dragon / Regal Crest books, which is their science fiction imprint. Later tonight I'll add all of the Mystic Books / Regal Crest imprint, which is their "Fantasy/Paranormal" line. A quick glance showed that at least some of their YA imprint, "YA Books / Regal Crest" will probably deserve to be in here as well. Looking at Amazon, these imprints seem to be listed by them with publishers given as "Silver Dragon Books by RC" or "Mystic Books by RCE", or similar formats.

As you may have noticed, I'm not participating very much right now other than adding some of my library, but I'll keep an eye on this thread in case you have comments on this. Chavey 17:01, 21 October 2016 (UTC)

Thanks for looking into this! Fixer is currently aware of 59 new ISBNs whose publisher name contains the string "Regal Crest". As you have already discovered, there are many variations from "Regal Crest Books" to "Regal Crest Enterprises: Yellow Rose Books". Many/most of their authors are not in our database; adding them may involve finding author Web pages and other types of TLC. I could ask Fixer to submit the 59 ISBNs that he knows about, but if you want to curate this publisher, it may be easier to post the list on your Talk page. What do you think? Ahasuerus 00:11, 22 October 2016 (UTC)
I'm willing to curate that list, so go ahead and pull those in, then I'll work on the authors, title summaries, and other details. "Yellow Rose Books" is their "Romance" imprint, so I'd want to look at suggested books from that group carefully - but of course lots of romance books do fit the genre. Fortunately, their web site gives fairly extensive plot descriptions, so I suspect we can identify those books that belong, and I'm assuming the ones Fixer pulls in will be ones listed under various appropriate categories in Amazon. Chavey 05:36, 24 October 2016 (UTC)

Birth dates

I've noticed you take a particular interest in author profiles, so I have a question. Whenever I create an author record, I like to fill it out with information. Lately, whenever I add an anthology, or a book with an introduction, or whatever, and create an author record for someone who has written very little or nothing relevant to Spec Fic, I do add a weblink to that record, and a working language; also, if known, I add the year of their birth, but not the exact day, so as not to clutter up your "born on this day" list with non-genre authors. The thing is, it only occurred to me recently to do this. Before, I did add an exact birth day for many highly irrelevant people. I would like to go back and remove those -- would that make sense to you? I didn't want to just do it and have you say, "Whoa, why are you making the database less precise?" --Vasha77 06:21, 24 October 2016 (UTC)

It sounds like there are two issues here. The first one covers the rules for entering author birth/death dates. The second one covers the rules for displaying "authors born on this day" on the main page. They are not necessarily related since we can easily change the display logic used by the main page.
As far as the first question is concerned, our practice has been to capture exact dates of birth and death if they are publicly available, as noted on the Policy page. As far as the second question is concerned, it looks like the current display algorithm is slowly but surely becoming unwieldy. We currently display 57 "born on this day" authors, which may be too many to be useful. We'll have to think of a way to cut back without losing useful information. Perhaps implementing FR 860, "Add a 'non-genre' field to Author records" will help resolve the problem. Ahasuerus 14:54, 24 October 2016 (UTC)
A non-genre field for authors would be excellent. Though the line is fuzzy. How would you count an author who wrote only non-genre fiction except for one indisputably supernatural ghost story? Or one disputably kind-of-fantastical story? But there are lots and lots of authors I'd confidently tag as non-genre. Like Hu Shih, the Chinese scholar and cultural ambassador who wrote an introduction (in the role of ambassador rather than critic) to Arthur Waley's translation of Journey to the West; or Waley himself, who's in for his preface. Anyhow, if you implement this, I volunteer to work on searching for and tagging non-genre authors. --Vasha77 16:38, 24 October 2016 (UTC)
An author "who wrote only non-genre fiction except for one indisputably supernatural ghost story" would be considered a "non-genre author" as the term is currently understood by ISFDB editors. The rule of thumb is that authors like Ernest Hemingway, Johann Wolfgang von Goethe and {{A|Leo Tolstoy who have mostly written non-SF are considered "non-genre", which means that we should only enter their speculative fiction (and related works.) Authors who have mostly written SF are considered "genre", which means that we will include most of their non-SF. As per our Rules of Acquisition:
  • [Included:] Works (both fiction and non-fiction) which are not related to speculative fiction, but were produced by authors who have otherwise published works either of or about speculative fiction over a certain threshold (see below). This includes any non-genre works published as standalone books as well as non-genre short fiction, but exclude non-fiction which was not published as a standalone book. Thus, Poul Anderson's mysteries and his non-fiction book about thermonuclear weapons will be included, but Gregory Benford's professionally published scientific articles will be excluded.
  • [Excluded:] Works that are not related to speculative fiction by authors who have not published works either of or about speculative fiction over a certain threshold. This "certain threshold" is hard to define, but we need to draw the line in a way that would exclude Winston Churchill, who published at least one work of borderline speculative fiction. The goal here is to avoid cataloging everything ever published by James Fenimore Cooper, Robert Louis Stevenson, Honoré de Balzac and other popular authors. Instead, we want to catalog their speculative fiction works only.
Once the proposed "non-genre" flag has been added, we will create a cleanup report to look for non-genre works by non-genre authors.
Ahasuerus 16:57, 24 October 2016 (UTC)
Thanks. Actually, I have mostly been entering works of criticism lately, but it is possible to draw a distinction there too. One critic that I added lately has published 40+ papers about science fiction and lists it as one of his chief research interests; another has published about 10 papers about fantasy and magical realism but doesn't list it as a research interest. The former is a genre critic, the latter is not. --Vasha77 20:00, 24 October 2016 (UTC)

The variants again

Looking at 20.000 de leghe sub mări, there is a third contributor that is not there in the main work. And because of that, if you click on the contributor Dralco, it is not listed. It is weird to have a variant with more contributors (it almost looks like two images together thus the multiple contributors)) but we do seem to handle this poorly when visualizing... Any advices on the best way to fix this kind of issues (short of a software change to show a parent-less variant when the author is assigned to it but not to the original)? :) Anniemod 01:21, 26 October 2016 (UTC)

Do we know the nature of Dralco's contribution? If he created a "derived work", then the only way to keep the credit and have the work show up on his Summary page is to break the variant relationship between the two COVERART titles. If he was credited as a "cover designer", then we'll want to move him to Notes since we don't enter designers' names in the Cover Art field.
Oh, never mind. I see that he "is only responsible for the corner vignettes". We should probably ask Linguist to stop by to help decide how we should handle this case. Ahasuerus 02:41, 26 October 2016 (UTC)
But it is not just this case -- that would be 2 I found in as many days and I am pretty sure there are more lying around. Is there any way to track them down and clear them in some way? Anniemod 17:56, 1 November 2016 (UTC)
Well, there is a "VT-Pseudonym mismatches" cleanup report that looks for VTs whose authors are not set up as pseudonyms of their parent titles' authors. The report includes additional logic to exclude false positives, which may be too aggressive. I'll need to take a closer look -- thanks for the heads-up! Ahasuerus 20:27, 1 November 2016 (UTC)

Add award category's name to its page title

Hi. At the moment the page title of all award categories is simply "Award Category". I think if there's no reason against this change it'd be great if the page would display the category's name and also use it for the TITLE tag. Moreover, adding the award's name may be a good idea as well, like this: "Award Category: AWARD CATEGORY NAME (AWARD NAME)". Jens Hitspacebar 19:11, 27 October 2016 (UTC)

That's a good point. FR 941 has been created. Thanks! Ahasuerus 19:15, 27 October 2016 (UTC)
Done. Ahasuerus 18:34, 30 October 2016 (UTC)


Hi, I know you mentioned that there are plans to add translators as part of the records at some point. Any idea how close is that in time? Months? Years? Asking so I know how to plan my own cataloging :) Anniemod 17:55, 1 November 2016 (UTC)

It's high on the list of priorities, so normally I would say "months". However, the design hasn't been finalized yet, so it's possible that something will pop up and throw a monkey wrench into the works. Right now the biggest headache is pseudonymous collaborative translations -- see this pub for an example.
The good news is that I retired over 4 years ago, which has enabled me to implement a lot of things that had been stuck in limbo. The bad news is that I have to spend a very significant portion of my ISFDB time on Fixer, our data acquisition robot, in order to keep us current. And, of course, as John W. Campbell Jr. once said, no one is getting any younger. Ahasuerus 20:37, 1 November 2016 (UTC)
How does this differ from the same two people writing a story under two pseudonyms? :) We enter as credited and pseudonym where needed, treating translators as co-authors in effect to the title level (with a flag that it is a translation). That may open the door for proper handling of abridgements by the way (translated by, abridged by are very similar concepts in the way we are handling variants here). Anything I can help with Fixer with? I will run out of housekeeping tasks sooner or later (wiki and then a few magazines are on my list) and my evenings will be wide open for other things. :) Anniemod 20:48, 1 November 2016 (UTC)
What you have described is close to the original plan as discussed when the issue first came up. However, there are additional considerations involved.
For example, translations are not necessarily VTs. Consider the "Collections" section of Philip K. Dick's Summary page. It contains 26 translated canonical titles; I am sure there are more that we haven't entered yet. Some of them were translated by multiple translators, some by a single translator.
Then there is the issue of awards -- how do we account for awards given to translations rather than to translated work? We don't have to add "translation award" support right away, but we need to make sure that the design doesn't preclude us from doing so in the future.
Another thing to consider is whether we want to implement translators separately or whether we want to add support for other "roles" -- adapters, editors of single author collections, etc -- at the same time. At the very least we need to decide whether the same design should support other "roles" in the future. And so it goes :-) Ahasuerus 21:43, 1 November 2016 (UTC)
Well, I do understand the complexities but delaying it makes the site very unfriendly to non-English users (I am sure some of the old timers will disagree but...). Anniemod 22:01, 1 November 2016 (UTC)
The reason why adding translator support had to be delayed was because we didn't have proper support for non-Latin author names. If we had added full translator support before we had support for non-Latin names, we would have ended up with additional transliterated author names and more cleanup work. Now that author names have been sorted out, we can start working on translators. Ahasuerus 22:14, 1 November 2016 (UTC)
I understand. I should have said "delaying even more" :) Anniemod 22:38, 1 November 2016 (UTC)
So why not invent a new category called "translations" instead of dumping it into variants? Anniemod 22:01, 1 November 2016 (UTC)
That's the current (still tentative) plan. Ahasuerus 22:14, 1 November 2016 (UTC)
And the current practice IS to have each translation as its own variant - while setting language to authors last few months I saw a lot of notes on "do not merge with the other translations" and the like on variants. So even if not all of them are entered, in the long run, all of them should be VTs sooner or later. Anniemod 22:01, 1 November 2016 (UTC)
I am sorry, I may not have been clear. What I meant was that the 26 translated canonical COLLECTION titles on Dick's Summary page would never become VTs. They will remain canonical titles because they are original translated collections. There are no English originals, so there is nothing to link them to. The same thing happens with anthologies and omnibuses. Nonetheless, they have translators associated with them. Ahasuerus 22:14, 1 November 2016 (UTC)
Ah, I think I may have misread you a bit :) Well - I would say that the collection itself does not have a translator - the entries inside of it do. Technically :) So why not make our lives easier and allow translators on simple titles only (novel, sf, essay, review) and not on compound (anthology, magazine, editor, collection, omnibus). For compound, populate the field automatically by getting the entries from the simple ones? This way you never have a translator on a non-variantable work (s an original always exist). Anniemod 22:29, 1 November 2016 (UTC)
That would be one way to address the issue, but it would make many translators' (e.g. Brian Stableford's) Author pages a lot shorter than they would be otherwise. We would also lose the ability to capture awards given to translators for translating "compound titles", which the software calls "container titles". And it would introduce a new distinction between container titles and non-container titles, which will require additional software changes to support, especially when converting from titles from NOVELs to COLLECTIONs. Ahasuerus 23:20, 1 November 2016 (UTC)
Why would it make it shorter - you bubble the translators from the non-compound titles, do not allow people to set them but have them automatically set when a lower level is set. So if you have 10 translators, the collection shows on all 10 pages; if you have one - it shows on 1. At the same time on the author page, it shows once (with or without the translators mentioned). And anyone converting from Collection to Novel (or omnibus) and vice versa need to do it carefully anyway. I am not saying it is easy - I am just... throwing ideas around. Anniemod 23:27, 1 November 2016 (UTC)
Oh, derived data. Well, it's certainly possible, but derived data -- unless it's something like dynamically calculated age -- tends to add a great deal of complexity. I would only recommend its use as the last resort. Ahasuerus 23:57, 1 November 2016 (UTC)
I am thinking more of a DB trigger than coding anything explicitly or real time derivations (those are a mess, I agree on this(if the DB version supports it - if not, then code based trigger on the update/delete/insert): When a translator for a non-compound work is added/removed/changed, change the value of all works that contain it; when a work is added to/removed from a container, change the container translators if needed. Run a reconcile script once a month in case somehow something did not update - no need for calculation in real time - you just show what is in the DB. Anniemod 00:04, 2 November 2016 (UTC)
Which is where the whole point of having the translator as a second author will help I think - differentiating the variants (or the translations/whatever we call it).
Awards... this kinda ties in my question earlier on variants and new authors into a variant, doesn't it? :) Allow variants to show on pages when the main work is not on the page and allow awards to be added to variants and the problem is solved, isn't it? Then pull the awards on the highest level as well with a note of the language and translator that got the award on the line and you are all set. Anniemod 22:01, 1 November 2016 (UTC)
There are two types of awards associated with translated titles. Let's use Kurd-Laßwitz-Preis as an example. One of its categories is called Bestes ausländisches Werk. Awards in this category are given for the "Best foreign-language work which was first published in German during the award year". Then there is Beste Übersetzung ins Deutsche, which is given for the "Best translation into German". As our data shows, they are usually given to different books.
The software already supports awards given to variant titles and processes them correctly when displaying award bibliographies. However, supporting awards given to translations will require additional design work. Ahasuerus 22:27, 1 November 2016 (UTC)
OK, what I am missing? How is an award to a translation different to one to a variant? Anniemod 22:38, 1 November 2016 (UTC)
An award given to a VT is associated with the VT's author. An award given to a translation is associated with the VT's translator. Let's use Roi du matin, reine du jour as an example. It won the 2010 Imaginaire award in the "Roman étranger" category. Our award record, which is linked to the VTs, appears on the author's Awards page. If the award had been given for the "best translation of 2010", it would have been given to Jean-Pierre Pugi, the translator. At this time we have no way to capture awards given to translators in a structured way, so we would have entered it as a "non-title award".
If you think about it, it makes sense. It's possible to create a brilliant translation of a book/story that, in and of itself, may not be considered award-worthy. For example, Michael Kandel's translations of Stanisław Lem's books have long been admired even by people left cold by Lem's works. Ahasuerus 23:11, 1 November 2016 (UTC)
I was missing that the difference is that it bubbles into the author page, not just the title page. However - I do wonder if we should not bubble to the author page as well anyway -- after all the story was created by the original author and most authors list those as their awards - or somewhere close by anyway :). I grew up with Lem in Bulgarian and Russian, never read him in English but I get what you are saying :) Anniemod 23:17, 1 November 2016 (UTC)
It's been my experience that the core issue is almost invariably defining a data model that accurately captures all relevant relationships that exist in the real world. Once it's in place, data capture and display issues can be relatively easily resolved. Ahasuerus 00:01, 2 November 2016 (UTC)
But trying to find the perfect model is never working either - something will be missed. So trying to get all the corner cases covered in one fix may be a bit too much to ask for. I get your point though. :) Anniemod 00:11, 2 November 2016 (UTC)
To quote Vince Lombard, "we are going to relentlessly chase perfection, knowing full well we will not catch it, because nothing is perfect. But we are going to relentlessly chase it, because in the process we will catch excellence" :-) Ahasuerus 00:28, 2 November 2016 (UTC)
Should I pseudonym this Vince Lombard into Vince Lombardi and variant the quote? (ducking for cover very fast) Anniemod 01:11, 2 November 2016 (UTC)
Ultimately, we will want to implement abridgements (or probably a "derived from" kinda thing to account for both abridgements and novels derived by stories (after all, this is the genre where that happens a lot). So why not plan now? Or if it is too complicated, get the translators in and then think for the rest. I know the urge to try to build a system that fits everything but... that's almost impossible. So let's start from the stuff we can. Because we already have a ton of data on translators. Maybe time to pull it out from the notes Anniemod 22:01, 1 November 2016 (UTC)
Abridgements and other revisions raise additional thorny issues. For starters, we'll need to decide whether we want to add support for generic "revisions" or whether we want to support certain revision types -- "expanded", "abridged", "restored", generic "revised", etc. Will these revision types be user-defined, table-driven or embedded in the software? Will we create a "one-to-one" relationship between the original title and a revised title? Or will we also support "fix-up", which means a "one-to-many" relationship? Will we capture the names of the people responsible for the revision? How will we display the linked titles on various pages? Etc, etc.
These issues have been raised a number of times over the last 10 years. And every time we put them on the back burner because it was clear that they would open a huge can of worms and that we had lower hanging fruit to take care of first. Ahasuerus 00:14, 2 November 2016 (UTC)
I understand that (and I read most of the archived Community portal pages) and I am not trying to rock the boat (too much) but for someone coming from outside, these things are... highly visible and confusing. So I am just trying to help I think. If it is not the time - then ignore it if you want - I will be back with the same in a few months most likely (I am easily distracted in the short term but I am almost impossible to distract in the long run) :) Anniemod 01:11, 2 November 2016 (UTC)
It's been a useful exchange since you have a fresh perspective. Going forward, perhaps it may be better to have these kinds of discussions on the Community Portal where more people will see them (and hopefully participate.)
Unfortunately, I have to structure my ISFDB time very carefully these days. Fixer finds 6,000-8,000 new ISBNs every month. Processing them is very time-consuming, which limits the amount of time that I can spend on development work. Whenever I get sick, I fall behind and then it gets really ugly. Ahasuerus 01:49, 2 November 2016 (UTC)
I usually do - I came here only to ask about plans for translators. Then things went on a tangent. Sorry :) PS: I was serious above - if there is a way to help with the Fixer entries, I would not mind lending you a hand. Anniemod 01:53, 2 November 2016 (UTC)
Thanks for the offer! I'd love to be able to farm out Fixer-related tasks to other editors/moderators, but there are various technical and administrative reasons which make it at best difficult and at worst impossible.
That said, if and when you find yourself with free time on your hands, there are certain things that we could do. Back in March, I made an announcement on the Moderator Noticeboard re: Fixer taking requests. The idea was that Fixer would create submissions for select publishers/authors.
It hasn't quite worked out that way, but Fixer can be used in other ways as well. At the moment Darrah Chavey is working on a list of "Regal Crest" ISBNs. The list was compiled by Fixer and then posted on Darrah's Talk page. Fixer can easily compile similar lists for other publishers/authors that you (or any other editor) may be interested in!
The only downside to this approach is that Fixer will have no way of knowing that certain ISBNs were discarded by the processing editor because they were non-genre, invalid or otherwise not worth entering into the database. Still, it's a small price to pay. Ahasuerus 02:21, 2 November 2016 (UTC)
One does not find free time, one makes it :) I was not here in March and I think I saw that when I was reading through but it went a bit over my head when I joined :) As soon as the cleanups are done and I get some magazines up to date in the DB, I will be up for getting to work on pretty much any list - so a month or so? :). May need some guidelines and have some questions but should be able to get the hang of it. Doesn't a list exist somewhere for Fixer to know what is discarded that can be updated when an editor rejects something? Anniemod 02:45, 2 November 2016 (UTC)
Fixer has an internal database which resides on the development server. Whenever an ISBN is submitted, rejected, suspended, assigned a priority, etc, the database is updated.
However, that only happens during regular Fixer processing. If I post a list of 100 ISBNs on an editor's Talk page instead, the sequence of events will be different. Suppose the editor creates submissions for 90 ISBNs and determines that the other 10 are non-genre or otherwise not applicable. When Fixer's database is updated with the latest data from the production server, Fixer will find the 90 submitted ISBNs and reconcile them with his database. The other 10 ISBNs will remain outstanding. I will need to reject them manually. Ahasuerus 11:36, 2 November 2016 (UTC)
Thanks for the explanation! Do you mind pulling some ISBNs/titles that need work on and sending them over? I can start slowly doing some work on them. Anniemod 17:23, 3 November 2016 (UTC)
Sure, I can do it any number of ways. We just need to decide what you would like to concentrate on first. There are some 500,000 ISBNs in Fixer's main database, so prioritization is key. It's been my experience that it's important to select ISBNs that the submitting editor would remain interested in, otherwise it's easy to burn out. The options are as follows:
  • Recently published e-books. I have been keeping up with e-AddPubs, but I don't have the bandwidth to process e-book NewPubs. Every month Fixer finds roughly 1,500 new e-NewPubs that require human review; the rest are auto-suspended and we don't have to worry about them. If you decide to work on them, I will need to perform preliminary prioritization, which will likely reduce the workload from 1,500 to 500 or so. (See How Fixer Works as of 2014 for details on Fixer's queues and prioritization.)
  • Books published by certain publishers. I can generate lists of the most prolific genre publishers (as far as Fixer's database is concerned), but it can be any publisher you choose. Some of the more prolific publishers are low priority, e.g. Fixer's "unprioritized" queue contains 2,296 ISBNs by Siren Publishing, a self-publishing/SF/romance/erotica outlet. But again, you can pick any publisher you want.
  • Paper books published prior to 2012-09, i.e. before I was started keeping up with Fixer's new acquisitions. The workload is approximately 600-1,200 ISBN per month, but I should be able to whittle it down to a much more manageable number before I post the list.
  • Paper books published since 2012-09. Some sources, notably Amazon, have been known to assign genre classifications retroactively. For example, a new book may be filed under "cozy mysteries" or "juvenile adventure", so Fixer will skip it. A couple of years later an Amazon employee may add "ghosts" or "vampires" to the list of tags/genres. At that point Fixer will grab it and add it to his unprioritized queue. The months between 2012-09 and 2015-10 have anywhere between 60 and 400 unprioritized ISBNs. Yet again, the actual number will be lower once I prioritize them.
So... Pick your poison! :-) Ahasuerus 20:14, 3 November 2016 (UTC)
I just did a few hundred (thousand maybe? had not counted) transliterations and at least as many Author Language updates in the last few months. I am too stubborn to get bored with data that improves a DB :) Most of my usual publishers are already in (and the ones that miss books, I will be working on anyway). Whichever of the other groups will require the least processing on your part (so you do not spend too much on it), just drop me the ISBNs and titles and I will see what I can do with them. It may not be fast (I still have some of my own things to do and I need to get around to working on my library:) ) but I can start chipping at it. IF all require the same time and you do not feel like picking a group in random, let me have on the older books pre-2012-09 for now (and if I get bored, I can just come and ask for another set temporarily). Anniemod 20:39, 3 November 2016 (UTC)
Publisher-specific ISBN lists take the least amount of time to compile. I will post a list of candidate publishers on your Talk page shortly. Ahasuerus 23:06, 3 November 2016 (UTC)
(Just out of curiosity - any usable REST API for adding publications? It may be easier to get data together in text format and program something to dump the data via an API (it is not beyond me to do something that automates even without an API when I decide I am too tired to fill the same 11 fields all the time but just asking) :)? Anniemod 20:39, 3 November 2016 (UTC)
The Web API that Fixer uses is REST-based. The "getpub" component can be used by anyone. The "submission" component requires authorization, but the process is straightforward. All I need to do is add your user name to the white list. You'll probably need to install the ISFDB software locally in order to be able to test your submissions -- see the bottom of the ISFDB Downloads page for instructions. Ahasuerus 23:06, 3 November 2016 (UTC)
Will take a look. :) Anniemod 23:25, 3 November 2016 (UTC)
BTW: Almost sounds like Fixer does need rewriting sooner or later - if for no other reason, then because Windows tend to change... and MS is not very good with backward compatibility for very old versions. What language is it written in? Anniemod 02:45, 2 November 2016 (UTC)
It's a hodgepodge. Ahasuerus 11:36, 2 November 2016 (UTC)
And while we are talking about changes in types and authors and what's not, any plans to fix the "Part 1 of 2" being a variant of the full title? When it is longer than a serialized entry somewhere (the very nice way of the French to split books for example). I kinda understand why we do not call these serializations (is 400 pages an installment?:) ) but as they are now, it is misleading and... hard to read. Maybe if you are thinking on restructuring, that also may come into play? (If you ask me, call anything that splits one work into multiple parts serialization and be done with it but maybe I am missing something). Anniemod 22:01, 1 November 2016 (UTC)
I guess the first thing to do is to decide what these "split books" really are. Are they a separate data type? Probably not: we have "split novels" as well as "split collection", "split anthologies", etc. Sometimes they have separate titles and sometimes they don't -- see the 2-part version of Legends. We'll need to brainstorm a bit and come up with a model for this odd corner of the bibliographic world. Ahasuerus 00:23, 2 November 2016 (UTC)
My feeling is still that there is no real difference between a novel serialized in 6 issues of a magazine and one that is split into two books... Yes, bigger chunks but still the same concept. But if everyone else disagrees, then so be it. But I think we need the discussion open. :) Anniemod 01:11, 2 November 2016 (UTC)

(unindent) I am glad to see that the process of thinking this through is still going on. How much has the design stage been finalized? It has occurred to me that I would like having the translations as a separate record type, similar to cover art, rather than as separate variant titles (the latter threatens to make summary bibliographies unreadable). --Vasha 14:01, 25 November 2016 (UTC)

There is a User Preference that controls whether "All", "None" or "Selected" translations are displayed. When "Selected" is chosen, the user can select a subset of supported languages to display. Please note that it applies to translations only. Canonical titles are always displayed regardless of the language. Also, non-registered users can toggle the display of translations on Summary and Series pages. Ahasuerus 15:17, 25 November 2016 (UTC)
Not what I meant... I mean, currently some stories have different translations split into separate title records, which is good, because it's definitely good to group together the instances of that translation somehow. But mightn't it perhaps be better handled like artist records than by variant titles? That way, the instances of the same text could be grouped together regardless of title and translator credit. For a rough example, a TRANSLATION canonical title record "Twenty Thousand Leagues under the Sea" (1872) by Lewis Mercier would have others varianted to it like "Twenty Thousand..." by uncredited, and "20,000..." by uncredited. And there could be a way of displaying those (optionally by user preference) on bibliography pages, such as a second set of nested lists below the variant title lists. OK, I'm just coming up with this while you've been thinking about it for years! Go ahead and tell me I'm way off base. --Vasha 16:00, 25 November 2016 (UTC)
Oh yeah, and creating TRANSLATION as a title type would mean no more having to split up existing book title records and shift publications to different title records. We'd just add or import the translation information to the publication record. Much work saved. --Vasha 16:21, 25 November 2016 (UTC)
Oh, I see. Yes, a separate title type for translation records is very much on the table. It will be needed if we want to be able to separate awards given to translations and awards given to translated works. The implementation/display details are still fuzzy, though. I don't think anyone has proposed the version that you described above yet. I'll keep it in the back of my mind -- thanks! Ahasuerus 20:35, 25 November 2016 (UTC)
Won't that be hard with short stories? There won't be a publication level entry to add the record to so if a story by Pavel Vezhinov is translated twice into English - once in an anthology and once in a magazine, there won't be a place to attach both translation records - you still will need two different SHORT FICTION variants. Unless if the idea is for Translation to replace all text types when the text is not in the original language. Which does not sound very ordered. Unless if I am misunderstanding the idea. Anniemod 19:31, 26 November 2016 (UTC)
I think we are still in the "requirements gathering" phase with translators. Over the last year I have come up with a number of possible design approaches. Every time I thought I covered every possible permutation (authors, translators, titles, languages, awards, etc) and was ready to post a complete proposal, I found a gap in the design... Ahasuerus 19:47, 26 November 2016 (UTC)

Young Readers Science Fiction Stories

As you made the Tuck secondary verification's on the Young Readers Science Fiction Stories publications, I was wondering if you could help me. Project Gutenberg has just added this text (#53456), but the contents are completely different. I assume the listed contents come from Tuck (I don't see any other sourcing)? SFE3 says this is a variant of Teen-Age Science Fiction Stories. While we don't list it as a variant, the contents match so perhaps that's were the contents come from... I assume our contents are incorrect based on the PG edition (everything else seems to match to the pub), but was wondering if there was any additional information before I replace the contents with the PG titles. Thanks. -- JLaTondre (talk) 23:04, 5 November 2016 (UTC)

Unfortunately, my Tuck, the Reginalds and the rest of my secondary bibliographic sources are currently packed, so we may need to ask someone else. However, I note that our contents match what's listed by Contento, who also states that "Teen-Age Science Fiction Stories" is the same as "Young Readers Science Fiction Stories" and "Science Fiction Stories". Since Teen-Age Science Fiction Stories and Science Fiction Stories have been primary-verified and their contents match, it seems likely that the underlying problem was Bill Contento's assumption that Young Readers Science Fiction Stories was a variant of the other two books. Ahasuerus 00:03, 6 November 2016 (UTC)
Bluesman‎ has verified "Science Fiction Stories" ("Teen-Age Science Fiction Stories" variant). I've asked him to compare the contents just to be sure they were not re-titled, etc. for the "Young Readers Science Fiction Stories" publication. However, both Teen-Age Library and Young Reader were separate publisher series for Grosset & Dunlap and they were aimed at different age ranges. These are most likely different collections. -- JLaTondre (talk) 16:20, 6 November 2016 (UTC)

Undefined title language?

I was looking at Titles by Language. When is a book language marked as "Undefined"? (there is no way to set it like that now so are these 238 books in some kind of limbo? Or is it a valid value? Thanks! Anniemod 23:34, 8 November 2016 (UTC)

There is no way to set it explicitly. My best guess is that there is an obscure sequence of submissions which can cause this value to appear in the database. I think the number has been stuck at 238 for some time, so it's possible that the bug has already been fixed. Ahasuerus 23:40, 8 November 2016 (UTC)
Any chance to pull a list of them at some point so they can get unstuck and their languages set? Anniemod 23:52, 8 November 2016 (UTC)
Sure. FR created. Ahasuerus 23:55, 8 November 2016 (UTC)

Search by title type

I tried doing an advanced search that included specifying the title type as a chapbook; the form wouldn't accept "Chapbook" but only "CHAPBOOK". I think you don't actually intend for that search to be case-sensitive, right? --Vasha77 03:07, 10 November 2016 (UTC)

Good point -- FR 945 has been created. Thanks! Ahasuerus 04:03, 10 November 2016 (UTC)
Fixed -- see the announcement on the Community Portal. Ahasuerus 21:04, 21 November 2016 (UTC)

Another search problem

Currently, a search for the french œ (as in, a search for "cœur") does not pull up oe and vice versa. Can you do anything about that? The difference between the two is a matter of typography rather than orthography, and so perhaps all the titles should be regularized to use only "oe", but a lot of people will be searching for œ anyhow. --Vasha77 21:13, 10 November 2016 (UTC)

Our software doesn't support this functionality, but I agree that it is desirable. There are other permutations where is could be helpful, e.g. the Russian "ё", which has been all but superseded by "е" lately. Unfortunately, I don't think it will be an easy fix because of the way the software works. The most straightforward solution would probably kill performance. I'll create an FR and see if I can accommodate it during the next Advanced Search rewrite. Thanks for the heads-up! Ahasuerus 21:34, 10 November 2016 (UTC)
How about the easy solution - use the transliteration fields to add the second option. That make is searchable. Anniemod 01:08, 11 November 2016 (UTC)
Anniemod is right-- treating that character the same as other non-latin-1 characters would be a good temporary fix. Maybe you can partially automate it-- find all records with œ in the title and create a transliterated version replacing that with oe, if there isn't one already. Then, if you pull up a list of all records whose language is French and whose titles contain oe, I would be happy to go through them and see which ones could be converted to œ and given a transliterated equivalent. A little checking with Advanced Search suggests that it would be a manageable number of records.
By the way, I never got an answer to this question, about how to create an author record that differs from an existing name only in Latin-1 letters with diacritics. --Vasha 18:44, 26 November 2016 (UTC)
Sorry about the delay. I don't think there is an easy way to differentiate them short of creating a "Firstname Lastname (I)" variant :( Ahasuerus 16:51, 27 November 2016 (UTC)
Those pesky software details (grumble) -- thank you for devoting yourself to wrestling with them. In that case, I will go back to the forum and reiterate the question whether the canonical name should have the diacritics given how inconsistent Anglophone publishers are about printing them. --Vasha 17:46, 27 November 2016 (UTC)

Pirates of Zan

Found a signature on the artwork for [this]. Added credit with notes. --~ Bill, Bluesman 18:15, 11 November 2016 (UTC)

Excellent, thanks! Ahasuerus 18:35, 11 November 2016 (UTC)

Bizarre Fantasy Tales 1

Hi. On your pv'd above you have the page count as 130 for both issues. The last numbered page is indeed 130 but the addition of the back cover pages would make them both 132, I believe. Also, for #1, was there a particular reason you omitted the uncredited interior art for "The 'V' Force" on p.111? I also added some notes. Thanks, Doug / Vornoff 04:46, 15 November 2016 (UTC)

I verified Bizarre Fantasy Tales 1 in 2006, shortly after ISFDB 2.0 went live. The data entry standards were still being developed at the time, which is presumably why I entered "130" instead of "132". My copy is currently boxed, so please go ahead and submit corrections based on your copy. TIA! Ahasuerus 15:14, 15 November 2016 (UTC)
Done. Doug / Vornoff 01:19, 16 November 2016 (UTC)
Thanks! Ahasuerus 01:20, 16 November 2016 (UTC)

Submissions on hold

Hi, before the change a few weeks ago that made all "My something Edits" pages look the same, it was easy to see who has a submission on hold by just looking at the page of Pending Edits (so I could ask a moderator why it is on hold). Now it does not seem to be visible anywhere (I know who has the one because I saw them asking another editor about it but am I missing a way to find who has one of my submissions on hold)? There is no indication that something is on hold at all actually (this one is on hold by JLaTondre at the moment is my suspicion based on a question I saw and the fact that noone is touching it but no indication on my page that it really is and by whom). So am I missing it or is that an unwanted side effect of the change? Thanks Anniemod 04:08, 16 November 2016 (UTC)

It's an "unintended side effect" aka "a bug" :-) I have created a Bug report -- thanks for identifying it! Ahasuerus 04:37, 16 November 2016 (UTC)
Side effect always sounds better than a bug -- when you are the one that need to fix it anyway :) Maybe the easiest fix will be to be changing UNKNOWN for a reviewer to the name of whoever has it on hold - that field is useless in that view anyway otherwise. Just thinking aloud here :) Anniemod 04:45, 16 November 2016 (UTC)

Coven 13, November 1969

Hi. In your verified above, I corrected the p.9 title to add a "-" between "were" and "wolf" as it is on the story's title page and added a note for it. Doug / Vornoff 14:36, 18 November 2016 (UTC)

Sounds good, thanks! Ahasuerus 16:15, 18 November 2016 (UTC)

Problem with URLs in author profiles

Currently adding certain URLs to author profiles gets the error message "Invalid characters in the URL. URLs must start with http, valid characters are A-Z, a-z, single quotes and -._~:/?#[]@!$&()*+,;=". But many URLs -- for example, Wikipedia links -- do contain other characters. SeeÓscar_Hahn -- etc. etc. (Also, an extra space at the end of the URL generates this error message.) --Vasha 15:57, 22 November 2016 (UTC)

That would be FR 921 Allow Unicode characters in Web page URLs and I was just about to start working on it :-) There was a time when Web standards disallowed Unicode characters in URLs (you had to use their encoded values instead), but things have changed. Ahasuerus 16:20, 22 November 2016 (UTC)
Thanks! --Vasha 16:32, 22 November 2016 (UTC)

Strange author credit

In [this] record, the story on page 262 is credited only with a Japanese[?] name, though the book itself credits Edogawa Rampo. I know we don't treat transliterated names as pseudonyms, but is that why the story completely suppresses the version of the name that's actually in the book? Is this an unintended result? Somewhere in the credit the name, as in the book, should be shown. I feel like we're losing data this way. --~ Bill, Bluesman 06:20, 25 November 2016 (UTC)

It must have been a human error during the cleanup process. As you noted, works should be entered as they are credited in the book, in this case "Edogawa Rampo", and then varianted. I have changed the credit to "Edogawa Rampo" and things look better now :) Ahasuerus 15:12, 25 November 2016 (UTC)
I didn't realize one could variant a transliterated name. Good to know! Thanks. --~ Bill, Bluesman 17:34, 25 November 2016 (UTC)
With the implementation of the final round of transliteration changes in late August, the data entry rules are the same for authors, titles, publications, publishers and publication series: we enter everything as it appears in the book/magazine and then add variants, pseudonyms and transliterations if known. For example, we have 27 (!!) romanized forms of Аркадий Стругацкий's name on file. We capture the way his name appeared in German, Italian, Polish, English, etc books and then we create variants/pseudonyms linking them to the canonical Russian name and titles. The same logic applies to Japanese, Bulgarian, Chinese, etc authors. Conversely, when Robert Heinlein's name is spelled "ロバート・A・ハインライン" in a Japanese publication, we enter "ロバート・A・ハインライン" and set up a pseudonym/variant. Ahasuerus 20:11, 25 November 2016 (UTC)

Two new wiki pages for the "Proposal to create a rule changelog"

Hi. Based on our recent discussion "Proposal to create a rule changelog" I created proposals for two new wiki pages in my wiki namespace:

Can you please have a look at them and see if they fit our needs and what we discussed? Jens Hitspacebar 13:42, 26 November 2016 (UTC)

The changelog sub-page is a good start. I have tweaked the spelling/grammar a bit and we may want to change the wording in other ways, but the concept seems right. I would change "Link to rule's wiki page" to "Help pages affected by the change" since a single rules change may affect multiple Help pages. Ahasuerus 16:55, 26 November 2016 (UTC)
Oh dear, the page had some bad spelling errors. Thanks for correcting them :) Your change proposal makes sense. Please reword anything else which sounds awkward from your native speaker's point of view. Jens Hitspacebar 17:56, 26 November 2016 (UTC)
OK, I have made a few changes to streamline the introductory paragraphs. Hopefully they make sense. Ahasuerus 18:59, 26 November 2016 (UTC)
Thanks, that's indeed better. Jens Hitspacebar 19:04, 26 November 2016 (UTC)
As far as the second template goes, I like the idea, but we may want to discuss it separately. The changelog is fairly straightforward and unlikely to prove controversial. The other template may require additional discussion, which will delay the implementation of the changelog. Ahasuerus 16:55, 26 November 2016 (UTC)
Let's start ;). I'm not aware that there is presently an accepted way to tally the results of a discussion as I've never encountered something looking like a vote here (except for the nomination of moderators). The simili-consensus method that we're seeming to use is (AFAIC) completely undemocratic (the louder or the more implicated or the biggest bully wins). Before starting to record alterings of the rules, we must devise a way to gather the opinion of the contributors of this project. Without such a validation method by all, we'll still have people (that may alas include me) that won't consider themselves bound by a few lines of discussion in an obscure thread. This is also valid for changes to the basic structures of the db that seems to be decided largely offstage of by bilateral discussions between select members. Hauck 17:19, 26 November 2016 (UTC)
Could you please provide examples of changes to the basic database structures that were implemented without a public discussion? I can't recall any, so I am wondering if I am misremembering (entirely possible since my memory is not what it used to be) or whether I considered the changes to be so minor that they didn't require a discussion (which may have been a mistake on my part.) Ahasuerus 19:04, 26 November 2016 (UTC)
IMHO we're talking about different things here, you're talking about discussions (that usually take place albeit in some unusual places like this talkpage) and I'm talking about the decisions coming from these discussions that are taken without never having been put to a vote (or pass trough a similar validating stage) by the whole community. For example I was thinking of the whole invasion of diverse "transliteration" fields and their use, remember a few days ago Bill's surprise at such changes (you'll agree that he's far from one of our lesser lights ;-)). Hauck 11:04, 27 November 2016 (UTC)
Here is what I wrote re: the addition of support for transliterations earlier this year:
  • We started discussing transliterations in March 2012. A Feature Request was created later in 2012. The challenge that we identified at the time was the existence of multiple competing transliteration standards for different languages. In November 2013 Dirk Stoeker came up with a tentative solution based on automatic generation of transliterated values, but it was not deemed desirable. In 2014 I finally realized that we didn't have to pick one transliteration standard per language, which would have been difficult to enforce for less well known languages anyway. Instead we could make transliterated titles/names a "repeating" field and allow as many transliterations as necessary. Based on that insight (which seems obvious in retrospect, but you know how that is), I added support for transliterated legal names in April 2015, which prompted a discussion of further plans to add support for transliterated versions of other fields. Similar changes were made to Publication series in January 2016. Publisher names followed in February 2016 and now we have support for transliterated titles. Each change was followed by a community discussion. The feedback has been generally positive Ahasuerus 14:48, 27 November 2016 (UTC)
I think these are two different things: one is my suggested template which shall (as a kind of "tool") simply make sure that rule changes will be applied to the documentation pages (in other words: a way to prevent the documentation to become even more out-of-date as it already is). You're talking about another thing: the way to tally the results of a discussion. I think we can handle these two things separately. Jens Hitspacebar 17:56, 26 November 2016 (UTC)
Exactly, but before talking about "rule changes" there must be a real process that lead to the validation of such changes so your "tool" will have value only backed by a real decision-making process. If not, it'll just be another list of more-or-less agreed upon individual practices. Even in your first proposition, you're using the term "New Rule", I'll just ask who made this "New Rule", who worded it, who approved it, who opposed it, who had the final say about it, etc... Hauck 18:05, 26 November 2016 (UTC)
I think it's the other way around regarding "what is needed when": the "tool" can be used no matter what the process is that results in a rule change (or in a "new rule"). Put differently: no matter if you're using the current practice for decision-making or a new still-to-be-devised practice one, after that you'll need to make sure that the result is communicated to all editors (via documentation). My "tool" is for this "after" part and can be used for both current or future practices (as long as we don't switch to a completely different software, of course). The question is: do we need to or want to discuss both things at the same time now, or isn't it sufficient to start with having the part the template is for improved (the part after the rule change has been decided)? For me it'd be ok to do the latter now only, just to get this little improvement started (as long as it is considered an improvement), and then start more general discussion you proposed separately (then this thread should definitely moved one of the other wiki pages). But if there's the desire to bundle both things in one discussion - no problem :) Jens Hitspacebar 18:52, 26 November 2016 (UTC)
The proposed changelog is supposed to be a summary of Help changes. It doesn't change the current -- admittedly ad hoc and haphazard -- process of coming up with Help changes. It just makes it easier to find them. For this reason I believe that the introduction of the changelog can be treated as a separate issue. Ahasuerus 18:58, 26 November 2016 (UTC)
That's correct. Jens Hitspacebar 19:04, 26 November 2016 (UTC)
Well, as usual, we're short-circuiting what is (for me) the real issue: the decision-making process. I've worked as a computer analyst long enough to see this classic effect where the "tool" surpasses the rules (I've written some tools like this myself).Hauck 11:04, 27 November 2016 (UTC)
@Hervé: I'm not sure if your answer refers to the proposed changelog only or to the proposed discussion template ("tool") only or to both. But again, just to make sure: please note that I proposed two separate things: one is the "tool" template (let's probably call it "rule discussion status template" for a more explanatory name). The other is the changelog which is completely independent from any decision process because it simply logs the results, similar to Development/Recent_Patches. Jens Hitspacebar 11:59, 27 November 2016 (UTC)
BTW: seeing that there's a need for additional discussion: should this thread probably be moved to the Community Portal? Jens Hitspacebar 17:56, 26 November 2016 (UTC)

(unindent) I think we should stop discussing the topic here because it's the wrong place and continue on the rules and standards page (I had thought it'd be best to ask Ahasuerus here directly about the proposals I created based on the Rules_and_standards_discussions#Proposal_to_create_a_rule_changelog discussion after there had been almost no participation or feedback there by other editors than Ahasuerus). I'll start new threads on the rules and standards page (trying to take into account what's been said here so far). Jens Hitspacebar 11:59, 27 November 2016 (UTC)

Chapbook Synopsis

Re Whitby After Dark: I copied the series, webpage, & synposis information over to the shortfiction record. However, in the chapbook title record, the synopsis field is uneditable so I cannot blank it and when I try to save the other changes, it tells me "Error: CHAPBOOKs cannot have synopsis data.". A nice catch 22. ;-) -- JLaTondre (talk) 13:45, 26 November 2016 (UTC)

Ah yes, I remember making CHAPBOOKs' series fields editable if they contained data. Apparently I didn't make the same change to the Synopsis field. I have created a Bug report -- thanks! For now the only workaround is to temporarily change the title type, update the data and then change it back. Ahasuerus 16:28, 26 November 2016 (UTC)
Fixed. Ahasuerus 00:23, 29 November 2016 (UTC)

Your Namesake

I've been entering individual titles of some of Frank C. Papé's plates for Cabell and came across this illustration. I could scan it and send it to you, if you're interested. If not, that's OK too. It's been some years since I read the novel and I have no recollection of the character. In any case, I thought you may be interested to know it exists. --Ron ~ RtraceTalk 01:25, 28 November 2016 (UTC)

Thanks, but I think I have this edition of Domnei in my collection :-) Unfortunately, I can't check at the moment since most of my books are currently in boxes. Ahasuerus 01:41, 28 November 2016 (UTC)

Fixer ISBNs

Hi, not sure if you saw my note in my Talk page - Orbit is done, so is Night Shade Books. I still have some work on Tartarus Press (mixed up ISBNs so will take a bit) and PS Publishing -- there is a lot more missing than just these ISBNs (mainly other editions of the same titles) so I am working on another plan to have all of them added so for the time being I am using the list as a base only and will be adding them slowly (and then probably keep doing that every time they publish something). So when you have a chance, I can work on another Publisher's list. Anniemod 15:54, 28 November 2016 (UTC)

Many thanks for all your hard work on publishers and on the Wiki cleanup reports! As far as additional publishers go, I can certainly convince Fixer to generate more lists. Would you like to take a stab at Marvel or one of the other comics publishers? Ahasuerus 20:21, 28 November 2016 (UTC)
Sure - whichever is the easiest to generate, I can work on it. Marvel or DC are preferred as I am better versed in their universes and it will be easier to eliminate a lot of title just by sight. A lot of rejections expected on those :) Talking about rejections - my rejected list has a few new entries if you had not picked them up yet. Anniemod 21:30, 28 November 2016 (UTC)
Great, I will start working on DC posthaste! Ahasuerus 22:32, 28 November 2016 (UTC)
Done. Have fun mass-rejecting them! :-) Ahasuerus 23:15, 28 November 2016 (UTC)
You may want to get me Marvel as well while you are at it -- this will go a lot faster than Orbit - I do not see more than 10 titles (if that) surviving and making it into the DB) - the quick glance showed me a lot of titles I know (and that need rejection outright) Anniemod 23:39, 28 November 2016 (UTC)
OK, I will start working on Marvel. It may take a bit longer than DC since Fixer's database contains 1,021 unprioritized Marvel ISBNs. Ahasuerus 23:44, 28 November 2016 (UTC)
That's fine - I have enough in the DC list for this evening :) Anniemod 23:50, 28 November 2016 (UTC)
OK, Fixer's Marvel ISBNs have been added. It's a somewhat streamlined list of 709 ISBNs, so it's possible that there will be another pass once the current crop has been taken care of. Ahasuerus 00:29, 29 November 2016 (UTC)
Thanks. :) No worries - if needed, I can do a pass through them every month (or however often it runs) for the new catch if Fixer feels like posting me the lists (for all "my" publishers - except PS Publishing and Tartarus for now). Anniemod 01:27, 29 November 2016 (UTC)
DC is done - 3 manga, 1 coloring book and everything else was comics (there were ~50 that gave me some hope that they are something else and I probably checked at least as many more even though I recognized them - just to be sure as something was bugging me... but no luck). My wish list did not fare as well - I had not realized that some of those had been published or are about to be published. All are in the rejected table so Fixer can reject from there. Next - Marvel. Maybe there are legitimate titles there Anniemod 06:32, 29 November 2016 (UTC)
Also, just as a point of reference, the Wiki cleanup project is a prerequisite for certain software changes, which will result in additional cleanup reports. Ahasuerus 20:21, 28 November 2016 (UTC)
Why am I not surprised about the more cleanup reports? :) I am doing as much as I can on the ones we have - Fanzines are done (except for the 9 that do not have any issues entered and I need to do that), Magazines are getting there (42 remaining and some of them will be as easy as adding the link to the record; some need adding issues), I made a pretty big stab at Publishers and somewhat smaller one at Publications, and Series are also a lot lower (mainly because of the Fanzines and Magazines and me chasing the forwarders). But I need a break from that once in a while (which is how I got around to finishing Orbit this weekend) - and as I seem to be the only one actively working on that, it will take a while. I hope to get them done by the end of the year but it all depends on how things go :) Anniemod 21:30, 28 November 2016 (UTC)
By the way can you look at the bottom two in this one? I cannot get to the wiki pages they seem to be pointing to - one of them forwards to the non _ starting but there is no forwarded, the other one leads nowhere because I already cleaned the one it was non-forwarding to. Anniemod 21:38, 28 November 2016 (UTC)
There was an outstanding redirect, which is gone now. However, the cleanup report still points to two Wiki pages whose titles start with an underscore. Apparently the Wiki software treats underscores in a special way, which messes things up. I don't think it's a big deal, though. Once all series and magazine pages have been migrated, I will break the "lexical match" link in the software and decommission these cleanup reports. Problem solved! :) Ahasuerus 22:31, 28 November 2016 (UTC)
Yeah - the Asimov's one. I saw that one. However they are really annoying me staying over there at the bottom :) It is pages with spaces, I know - but I cannot convince wiki to show them to me so I can mark them for deletion. Oh well... :) Anniemod 23:06, 28 November 2016 (UTC)

(unindent) And Marvel is done except for two titles which I am pretty sure do not belong here (but the other 2 from the same series are already entered so I want to do some more research) - the Rejected list is pretty long at the moment. Turned out I recognized more titles in that list than I expected; the rest were easy enough to check. At least I found 1 legitimate title that was added to the DB in Marvel unlike DC. Anniemod 06:11, 30 November 2016 (UTC)

Excellent, thanks! Ahasuerus 15:21, 30 November 2016 (UTC)

Whenever Fixer feels like it, a new list of ISBNs is needed. :) Anniemod 06:11, 30 November 2016 (UTC)

How about a slightly different angle? Here is a list of Fixer-identified ISBNs that had to be deferred for various reasons. Would you like to take a stab at it? Ahasuerus 15:21, 30 November 2016 (UTC)
They are ISBNs, they need reviewing so sure. Anything special in the reporting for these? Do you want me to update the table/remove the ones that are added/rejected or anything else like that in addition to adding them to my two lists of Done and Rejected? Anniemod 18:37, 30 November 2016 (UTC)
As long as you add any rejected ISBNs to your Rejected page, Fixer will be happy. Everything else will come from the weekly backup file. Once you are done with the current list on Fixer's User page, we can simply zap it. Thanks! Ahasuerus 19:22, 30 November 2016 (UTC)
Quick questions - when we say we are not adding books for very young readers, how do we define very young? Is preschool (3-5) in scope? Such as this? I do not think so but verifying. And what about chapter books auxch as this? Thanks Anniemod 21:59, 2 December 2016 (UTC)
I don't think we ever defined "very young readers" explicitly. Empirically, books aimed at preschool readers are rarely added unless the author is "above the threshold". I tend to assign Fixer-captured ISBNs in the 10-24 page range to Queue 2, where they die a slow and painful death. 32-40 page books vary, but many end up in Queue 2 as well. Ahasuerus 22:15, 2 December 2016 (UTC)
I have quite a lot of the two types in the last Marvel list - thus the question. I am almost ready to reject them - they are children books more than SF books in my mind... a Junior novelization is one thing; a chapter book is a bit different. Any objections? Or do you want me to put them on a separate list so you can get them into Queue 2 for now? Anniemod 22:23, 2 December 2016 (UTC)
Perhaps create a sub-page for "Preschool books" and move the offenders there? I can write another script to parse the sub-page and move their ISBNs to Queue 2. Ahasuerus 22:36, 2 December 2016 (UTC)
Sure, I will do that and stick all the chapter books and all preschool ones there. Anniemod 22:38, 2 December 2016 (UTC)
Sounds good, thanks! Ahasuerus 22:39, 2 December 2016 (UTC)
And new page created: Over here. Named Queue2 in case I find something else that can go there in other publishers (so I do not need to create a new page for the new thing). Let me know if that format works for you. Anniemod 04:15, 3 December 2016 (UTC)
Thanks! Ahasuerus 12:36, 3 December 2016 (UTC)
By the way, the list of deferred ones contain ISBNs that are already in the DB. I am just ignoring them and not adding them to any list - do you need a record of them somewhere? Anniemod 21:59, 2 December 2016 (UTC)
I assume that the editors who added these ISBNs to the database didn't realize that they were among Fixer's "deferred" ISBNs. They can be safely deleted from Fixer's User page. BTW, it's a good example of how quickly and drastically the Wiki and the database can get out of sync. Ahasuerus 22:15, 2 December 2016 (UTC)
Considering that some of those have notes "Check in 2011" or earlier, I am not that surprised to find some of them already in the DB. And some are PV-ed so they were added by someone that actually have them. :) I am doing that as a first check followed by the usual checks for the Fixer differed ones. Anniemod 22:23, 2 December 2016 (UTC)
One more question - do you need the first column of the page to be a link or can it be just the ISBN itself (I have links for the publisher links but I need to build them for the ones from the Fixer deferred list. Anniemod 23:31, 2 December 2016 (UTC)
Are we talking about Fixer's deferred ISBNs? Currently the first column contains the word "ISBN" followed by the book's ISBN. It's a MediaWiki-recognized pattern, so it automatically links to a page with further links to book sources (including Amazon), right? Ahasuerus 23:51, 2 December 2016 (UTC)
No, my rejected ISBNs list - when I am rejecting something from a publisher list, I have a ready to use URL for the first column; when I am rejecting from the deferred list, I do not. I am not sure how your parser works so I had been building Amazon UK links (easier for me to build on the fly than the .com ones) but was wondering if you need that or if I can copy the ISBN as it is on the deferred list. Anniemod 00:03, 3 December 2016 (UTC)
Oh, I see. Either way is fine. I can tweak the parser easily. Ahasuerus 00:14, 3 December 2016 (UTC)
When you have a chance, check if the format at the bottom of the list works - I am slowly making my way through the Fixer deferred list and some do need outright rejections (and if that works, I won't spend time building URLs :) Anniemod 06:31, 7 December 2016 (UTC)
Updated, thanks! Ahasuerus 15:10, 7 December 2016 (UTC)

(unindent) Hi, I am working on the Deferred list - was mostly offline for a couple of weeks but should be back and working on these shortly. Sorry for the delay. Annie 22:30, 30 December 2016 (UTC)

No worries, there is no rush! Enjoy the holidays :-) Ahasuerus 22:49, 30 December 2016 (UTC)


Hello, there is perhaps a new bug with the processing of the flags "Nongenre" and "Graphic Format". I've moderated two submissions here and there that have these flags set to "yes" without that being a deliberate choice of the contributors (see here and there). Perhaps a new default setting? Hauck 11:06, 1 December 2016 (UTC)

Wjmvanruth also mentioned the problem in this submission. I left him a note about it. I did not make a correction yet, in case having it the original way is useful. --MartyD 12:37, 1 December 2016 (UTC)
It does look like a bug introduced in the last patch. Let me take a closer look... Ahasuerus 13:14, 1 December 2016 (UTC)
NewPub bug confirmed. Working on it... Ahasuerus 13:22, 1 December 2016 (UTC)
The bug has been fixed and the affected records have been corrected. Sorry about the hassle! Ahasuerus 13:37, 1 December 2016 (UTC)

New pubs and title notes

I happened to deal with a bunch of submissions the other day from Vasha77, where he put info into the title synopsis for want of ability to provide the info as title notes. I moved the ones I noticed for him after acceptance + title creation. Tonight I found he made a bunch of title edits doing likewise on others, only a few of which were his. Looks like several people are abusing Synopsis that way on new pub submission, and the records are going uncorrected. I realize the sheer number of notes-like fields on New Pub is already daunting, but I'm wondering if empirical evidence isn't suggesting we ought to have provision for title notes entry there. --MartyD 01:02, 6 December 2016 (UTC)

Well, we already have FR 152:
  • Add a new field for Title level Notes in the New Publication forms, but only display it when it's checked in User Preferences.
My last comment reads "It's not clear whether the described functionality is still desirable. Re-opening [this FR] pending a discussion on the Community Portal." Would you like to post a proposal on the Community Portal? Ahasuerus 01:13, 6 December 2016 (UTC)
Done. --MartyD 11:35, 9 December 2016 (UTC)

ISBNs and ebooks again

Hi, can you look at one of the Fixer ISBNs and help me figure out what to do with it (and any other similar ones). ISBN search finds it as a paperback which is incorrect (the paperback has a different ISBN). Amazon finds it on keyword as an ebook (here). However - there is nothing else to help me verify that this is indeed the angry robot ebook ISBN - Look Inside is useless (which helped me on all those Orbit books I added). So what to do with that ISBN (and any others like that)? Presume that they are the ebook ones? Add the ebook with no ISBN (I found it, it is not in the DB, I might as well add it) and reject the ISBN? Something else? Thanks! Anniemod 20:53, 7 December 2016 (UTC)

Unfortunately, Amazon doesn't display ISBNs for e-books, which makes matching ISBNs and ASINs harder than it needs to be. Fixer uses Amazon's API, which does support ISBN as well as ASIN look-ups for e-books. When I ask Fixer to do a search on "ASIN:B011G4DNDE", Amazon's API returns 9780857664884, which is usually sufficient to confirm that the ASIN and the ISBN refer to the same publication. You can also do a "field-isbn" search manually using this URL. It should result in the same ASIN. Kind of a roundabout way of doing things, but it's the best we can do until Amazon changes its logic... Ahasuerus 23:08, 7 December 2016 (UTC)
That last URL will be very helpful I think - Thanks! Anniemod 23:23, 7 December 2016 (UTC)
One URL at a time :) Ahasuerus 01:11, 8 December 2016 (UTC)

"Marie Queen of Roumania"

Hi, thanks,
Re our "Marie Queen of Roumania" 115976, you approved my addition of the Library of Congress webpage--where she is authorized heading "Marie, Queen, consort of Ferdinand I, King of Romania, 1875-1938" LCCN n50-043679.

There LC lists "Marie, Queen of Romania" (only with comma) as one Variant. Its catalog records for her books imply that they credit her as, for instance, "the Queen of Roumania", LCCN: 16-8076.

I wonder, do we have a special treatment of such authors, rather than use whatever may be printed on the title pages? Not only "Marie" but also Romania/Rumania/Roumania variation across publishers is likely to be an issue.

(Later I will look for the title pages at HathiTrust and elsewhere if necessary.)

Good night. --Pwendt|talk 02:39, 9 December 2016 (UTC)

What the Library of Congress calls a "personal name" is similar to our concept of canonical names. The LoC record which you linked contains two relevant lines:
  • Personal name: Marie, Queen, consort of Ferdinand I, King of Romania, 1875-1938.
  • The dreamer of dreams / by the Queen of Roumania
The "Personal name" line serves the same purpose as our canonical name. The "Main title" line shows the name given in the book itself, in this case "the Queen of Roumania".
We don't have to worry about their "personal names". We just use what's on the title page and then pick the most common form of the author's name to serve as our canonical name. Sometimes it's the same as what the Library of Congress uses, other times it's different. For example, they list "The Murder of the U.S.A." under "Jenkins, Will, 1896-", but we know that it was written by Murray Leinster, so we have it varianted accordingly. Ahasuerus 04:53, 9 December 2016 (UTC)

Bug: "as by" in alphabetical view

Here's an odd bug: when I change an author's page to the alphabetical listing (but not summary or chronological), all the "as by [pseudonym]" show up as "with [pseudonym]". --Vasha 01:03, 11 December 2016 (UTC)

It's a side effect of Bug 491. After rewriting and optimizing the Summary page code in July-August, I was going to rewrite the Alphabetical page, but then I got sick. I have been playing catch-up ever since, but I should get to it in the foreseeable future. Ahasuerus 01:12, 11 December 2016 (UTC)

Weird submission

Hello, can you have a look at this submission there is probably something in the notes (a tag?) that prevent it from being accepted. Thanks. Hauck 15:13, 13 December 2016 (UTC)

That's right. The current version of the Note field starts with an invisible HTML comment (<!- -<li><a href="">Star Gate #13</a>←Star Gate #14→<a href="">Star Gate #15</a> - ->). The submitted version of Notes removed the closing "-->", so everything from the Note field on was treated as an HTML comment and was not displayed. I had to use "hardreject.cgi" to reject the submission and the submitter will need to re-create the submission. Ahasuerus 16:55, 13 December 2016 (UTC)

The "English" epidemic

That said sort-of in jest. Lately there have been days when there are several hundred submissions in the queue that exist for the sole purpose of changing/adding the language of a particular record/author to ENGLISH. While I would be the first to point out that what passes for said language these days is a travesty, that opinion isn't germane to the point-in-hand. Would it be possible to change every record that currently does NOT have a language assigned to ENGLISH??? While I loathe regularization at any level, this would have a large side-benefit. Because this is a largely English database, the number of records adversely affected would be a small percentage of the total. Thus, editors could/would be expending their energies correcting that percentage instead of inundating the queue with the current deluge of -/ENGLISH edits. There are days, like today, when I look at the queue and just give up. I shouldn't [supposedly there may be beer some day .....??¿¿??] but with each passing day I realize there are fewer days ahead than behind and think "What for ...??". Then I realize we old farts have NO rookies to foist this upon ......... There's true bibliography where EVERY change matters, but .... this ain't it ..... In obeisance, oh keeper of the [software] code! --~ Bill, Bluesman 00:15, 15 December 2016 (UTC)

Well, the plan has always been to add languages to titles in two phases. During the first phase we were going to assign language codes to the "low hanging fruit" manually. During the second phase we were going to assign 'English' to the rest of the titles automatically.
The first phase has been a success in that only 46.3% of our titles do not have a language code assigned. A few months ago I decided to check if we were ready for the second phase. I ran a few reports and found hundreds of authors like Philippe Druillet whose language-less titles are a mix of English and French records. Clearly we weren't quite ready for "phase 2" yet.
I think the next step is to use this cleanup report, which finds author/language mismatches, as a template and build on it. What we need is a cleanup report to find language-less titles by non-English authors. There are 18,290 of them, which may seem like a lot, but it's just 3% of the total number of language-less titles. Once they have been upgraded, we can move on to "phase 2". And there will be much rejoicing! :-) Ahasuerus 04:26, 15 December 2016 (UTC)
FR 961 has been created. Ahasuerus 04:41, 15 December 2016 (UTC)

Tag typo ...

... (as I suppose) here. Stonecreek 06:40, 17 December 2016 (UTC)

Fixed - thanks! Ahasuerus 16:47, 17 December 2016 (UTC)

Amazing Stories, December 1976

Added letter from R. H. Blackburn to your verified pub.--Rkihara 17:01, 17 December 2016 (UTC)

Thanks! Ahasuerus 18:51, 17 December 2016 (UTC)

Transliterations for reviews question

A quick question on transliterations (don't have an example to try with so wanted to check how the software behaves before asking for a change) - if a review is added and the book reviewed has a transliteration, does the review get the transliterated value as well (it takes the title)? Thanks! Annie 01:26, 18 December 2016 (UTC)

Review records do not inherit the reviewed record's transliterated titles. However, keep in mind that the Title page handles review records in a special way. If a review is linked -- which is the case in 99%+ of all cases because we have a cleanup report to find unlinked reviews -- then the review's Title page shows the reviewed record's title, including its transliterated titles. Devious, aren't we? :) Ahasuerus 01:49, 18 December 2016 (UTC)
We are not showing - but when there is a need for transliteration, it shows up on the report even when the title itself has it (so when you look at the page it looks as if it is ok (because of what we are showing) but if you edit the interview record, it is missing there and the report catches it. Very confusing until I finally figured out what we are showing exactly :) We should either suppress the INTERVIEWS from the the report or we should start copying the transliteration with us I think :) Annie 01:53, 18 December 2016 (UTC)
That's a fair point. Would you like to create an FR? It would be good practice :) Ahasuerus 02:01, 18 December 2016 (UTC)
I had not realized that I can create FRs and not just request them here. Sure, I'll go figure out how to do that. Annie 02:04, 18 December 2016 (UTC)

Request: 2016 colections/anthologies

Hi-- I'm taking a break from adding editions to Jorge Luis Borges and would like to get started on a project before awards season comes around: adding as much 2016 short fiction as possible to the DB. I've already begun putting in magazine issues; and here's my request: could you pull up a list of all the 2016 collections and anthologies that don't have contents? Thanks! --Vasha 01:42, 22 December 2016 (UTC)

Ahasuerus is on a break right now. I don't know if he can add an additional parameter of "content = none" (or whatever it would be) to the search results, but a full list of collections and anthologies with a year of 2016 is here. That could get you started. ···日本穣 · 投稿 · Talk to Nihonjoe 01:49, 22 December 2016 (UTC)
Thanks... I know it's possible to search for no contents because there's a cleanup report like that. --Vasha 01:59, 22 December 2016 (UTC)
It must require a custom inquiry, then, as that option is not available in the Advanced Search as far as I can tell. ···日本穣 · 投稿 · Talk to Nihonjoe 03:27, 22 December 2016 (UTC)
See User:Vasha77/2016 Contentless Anthologies & Collections. I generated it from this weekend's database dump. In some cases, there may be other formats with contents so you should check to see if you can import contents vs. manually entering. But in most cases, it will probably not be that easy. -- JLaTondre (talk) 00:49, 23 December 2016 (UTC)

Three More Feature Requests

Added three more feature requests:

  • 966 Indicate existing ISBN on new pubs
  • 967 Link record number in add variant submissions
  • 968 Clarify differences in moderator merge approval screen

Hopefully they are clear enough, but let me know if you have questions. -- JLaTondre (talk) 13:48, 28 December 2016 (UTC)

967 done, looking into the other ones. Ahasuerus 16:35, 28 December 2016 (UTC)
966 done, looking into 968. Ahasuerus 17:24, 28 December 2016 (UTC)
968 done. The code behind it has been a pain to maintain ever since it was written, so I went ahead and rewrote it. Ahasuerus 00:55, 29 December 2016 (UTC)
Wow, awesome! Thanks! -- JLaTondre (talk) 02:53, 29 December 2016 (UTC)
One patch at a time :) Ahasuerus 16:40, 29 December 2016 (UTC)

Dynamic Science Stories, April-May 1939

Just got the facsimile reprint of Dynamic Science Stories, April-May 1939 and in the corner of the illustration on page 80, if you look hard, you can see the signature of M. Marchioni. Do you want to add it to your record or would you like me to? Nobody wants to tear up old pulps, so I can how this signature was overlooked. MLB 02:40, 30 December 2016 (UTC)

Please go ahead and add it. TIA! Ahasuerus 02:41, 30 December 2016 (UTC)
Having no life I went over the uncredited illustration on page fifty-nine and found Wesso's scrawl in the left hand corner. I'll add that too. MLB 00:00, 31 December 2016 (UTC)
Thanks! Ahasuerus 01:07, 31 December 2016 (UTC)

Duplication of publications for translated stories

Is this a bug you've already found? In a collection like this one, which contains both an original story and a translation of it, the title page for the original displays the publication twice: for example --Vasha 00:25, 31 December 2016 (UTC)

I do not think it is a bug - the story is in the collection twice so the collection shows up twice - once for the title itself, once for the variant... Annie 00:27, 31 December 2016 (UTC)
Oh right, a translation is a variant. But is there a way to have the publication only show up once? --Vasha 00:44, 31 December 2016 (UTC)
Not at this time, but we have FR 900 -- "Title page should let users view main title's pubs only". Once implemented, it will support the functionality that you'd like to see. Ahasuerus 01:11, 31 December 2016 (UTC)
That‘s not quite what I meant. Would it make sense to have the software detect that the same publication is showing up twice and only display it once? I suppose there's some reason why that's either impossible or a bad idea. --Vasha 01:34, 31 December 2016 (UTC)
Well, it's right to show the title twice, when it's incorporated twice. As a user I would wonder where the second occurrence has gone to when it'd be deleted from the display. A similar case that occasionally happens is the reproduction of a cover image in the interior part: there the second occurrence should be (and is) also displayed. Stonecreek 05:52, 31 December 2016 (UTC)
I think for the experienced user, it is expected because we are used to it. For the average user(*), I think it is confusing. It is something that at least gets asked periodically. It would be low on the features request list, but it might be nice that instead of listing a publication multiple times, it is only listed once, but with an additional field that shows the number of times the title (including variants) is in the publication. When there is a long listing of publications, that would also make it easier to see that a given publication contains multiples.
(*) I'm assuming we are like a typical site where we have more viewers than we have active editors. -- JLaTondre (talk) 13:59, 31 December 2016 (UTC)