Rules and standards discussions

ISFDB Discussion Pages and Noticeboards
Before posting to this page, consider whether one of the other discussion pages or noticeboards might suit your needs better. If you're looking for help remembering a book title, check out the resources in our FAQ. Please also see our Help pages.
Help desk Questions about doing a specific task, or how to correct information when the solution is not immediately obvious. • New post • Archives	Research Assistance Help with bibliographic projects. • New post • Archives	Rules and standards Discussions about the rules and standards, as well as questions about interpretation and application of those rules. • New post • Rules changelog • Archives	Community Portal General discussion about anything not covered by the more specialized noticeboards to the left. • New post • Archives	Moderator noticeboard Get the attention of moderators regarding submission questions. • New post • Archives • Cancel submission
Roadmap: For the original discussion of Roadmap 2017 see this archived section. For the current implementation status, see What's New#Roadmap 2017.

Archive Quick Links
Archives of old Rules and standards discussions.

1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 · 9 · 10 · 11 · 12 · 13 · 14 · 15 · 16 · 17 · 18 · 19 · 20 · 21

Expanded archive listing

Rules and standards changelog

Every rule change that comes out of a discussion here should be added to the Rules and standards changelog.

Page & Spine: Fiction Showcase

Page & Spine: Fiction Showcase is a webzine that publishes stories, poems and essays on a weekly basis, and it has been doing so since 2012. The current issue (as I’m writing this) came out on 31 December 2021. However, its issues are not saved as such. Rather, the various contributions to each issue are archived by month according to the sections in which they were published.

As a general fiction and poetry site, speculative fiction was originally published alongside general fiction, romance, mysteries, and so on, with microfiction appearing in its “Crumbs” section. Starting in May 2020, however, the editor/publisher of Page & Spine, Nancy Wagner, allocated a specialized section devoted to speculative fiction (and some poems), calling it “Outta This World” (though speculative microfiction has continued to appear in the "Crumbs" section). The contributions to “Outta This World” are indexed here.

Last year, Wagner posted on the site that final submissions would be considered until 31 December 2021. As readers are reminded in the current issue, the last issue of Page & Spine will be published on 6 May 2022 and that access to the site will cease on 3 May 2023 (in other words, it will be taken offline).

I think this site merits indexing, even though the “issues” themselves are not archived. However, it’s a large site and would probably need a group effort by indexers to cover it, if it is deemed to be acceptable for indexing. There’s less than a year-and-a-half available to index it and, having checked its inclusion in the Internet Archive, I found that the latter’s coverage of Page & Spine is a bit patchy. It therefore can’t be relied upon as a substitute for the actual site itself.

If speculative material across the site as a whole is not considered eligible for indexing, I think that at the very least, the “Outta This World” section should be. Any thoughts? —The preceding unsigned comment was added by Explorer1000 (talk • contribs) .

Under the current rules, the only webzines that are eligible are:

Speculative fiction webzines, which are defined as online periodicals with distinct issues (note: online periodicals without distinct issues are not considered webzines)
Special speculative fiction issues of non-genre webzines
One time speculative fiction anthologies published on the Web

Note that general NON-genre webzines are never eligible unless they publish a complete genre issue (unlike paper and ebook 'zines which can be added as non-genre and only with their genre contents insid indexed). There are a lot of fiction out there which may be worth indexing but unless we change the rules (which is possible and very likely) or find a rule under which they fit, they are not eligible (yet). This very strict definition of what is allowed was a first step - we never expanded again but we will sooner or later - and non-genre webzines with issues is probably the next step. From what I see on this site, the site as a whole is out at the moment; the "Outta This World" can be added as "webzine with monthly issues". :) Annie 15:44, 5 January 2022 (EST)

Thanks Annie! :) I'm aware of the current rules. I just wanted to open discussion on the eligibility of this particular site, especially as it will disappear in May '23. So if the rules are changed over the next year, say, so as to allow for indexing sites such as Page & Spine, then at least indexers will have advance knowledge of Page & Spine's existence.

I think that some items already listed (or which will be listed) in this database first appeared in "Page & Spine" (which doesn't publish reprints), but this first publication data is not yet being captured for the ISFDB.--Explorer1000 16:37, 5 January 2022 (EST)

We cannot just decide that a special site is eligible just because it will disappear though - we can try to organize a strike force for one that is eligible but opening the door for "ad hoc eligibility" can lead to some very wide open doors IMO - there are a LOT of them out there. :) We can discuss the change of rules though, using that one as an example and pushing for timing :) And worst case scenario, someone can add the details in Wiki preparing for the day when it MAY be eligible. It really comes down to time and effort and resources - in a perfect world, we will allow any speculative fiction anywhere in the world (online, paper, ebook, whatever the next technology is)... but the world is not perfect and we have limited resources (editors and moderators) so we are carefully opening the doors little by little. Annie 16:43, 5 January 2022 (EST)

OK. What if any updated rule specified that where a non-genre webzine has a section/department/column, which could also serve as an archive (as in the present case), that is devoted to speculative fiction, poetry and/or essays, then only this portion of the publication would be considered eligible for indexing? This would apply a general rule so that there would be no need for ad hoc decisions about specific electronic publications. It would also mean that the bulk of what such webzines/Web sites publish would not even be examined for indexing. So, with Page & Spine, only the two years' worth of material published exclusively in the "Outta This World" section (up to May 2022) would be eligible for indexing, while all the other circa ten years' worth of material, including speculative items, would not. In other words, I'm proposing a compromise between indexing everything in the 'ideal world' scenario and where nothing is indexed at all, which is the situation at present.--Explorer1000 18:46, 5 January 2022 (EST)

Figure out wording for the change, post a proposal (here is fine - and I can see if I can come up with wording based on your explanation if you want?), get enough support, revise a few times based on corner cases and other ideas (there will be a lot of them - a lot of them based on experience here and elsewhere and things you may had not seen before (or did not realize will apply to the proposal)) and the rules get changed. Same way we added the webzines at all a few years back. Same way ebooks were allowed a few years before that (it was just "paper" at the start of the project). And again - in this case the column seems like a legitimate issue-based webzine on its own so I'd allow it under that rule if someone submits it. But we do need to revise this whole set of rules anyway so... may as well discuss what we do want to try to allow. :) Annie 19:03, 5 January 2022 (EST)

Sorry for taking a couple of weeks to get back to this.

Here is my suggested insert into the "included" list of the current "Rules of Acquisition":

Published works of speculative fiction, regardless of whether they are published within or outside the genre. "Published" is defined as follows:

paper books published by:
- professional publishers
- small presses
- print on demand (POD) publications
- self or vanity publishers
paper periodicals of the following types:
- professionally published magazines (prozines)
- semi-professionally published magazines (semi-prozines)
- paper-based fanzines
- newspapers
electronic publications of the following types:
- e-books with a unique identifier such as an ISBN, ASIN, EAN, or catalog number
- downloadable e-zines
- Internet-based publications which are downloadable as electronic files in any number of ebook formats (ePub, Mobi, PDF, etc).
- Speculative fiction webzines, which are defined as online periodicals with distinct issues (note: online periodicals without distinct issues are not ::::::considered webzines)
- Special speculative fiction issues of non-genre webzines
- Speculative fiction sections or columns of otherwise non-genre webzines
- One time speculative fiction anthologies published on the Web
- Online publications available exclusively as a Web page, but only if:
  - published by a market which makes the author eligible for SFWA membership (listed here), OR
  - shortlisted for a major award
audio books, i.e. readings, but not dramatizations

So would that be sufficient or does it need more specific information?--Explorer1000 15:41, 21 January 2022 (EST)

Excerpts with no originals

I would expect that any excerpts from material would have an entry for the original (linking is another story). I would also expect some exceptions, such as previews of books never published. However, I ran across a different sort of exception - in this case a poem from the Rubaiyat of Omar Khayyam. I recognize the title but have never read any of it. If it is speculative fiction, it should be included and the referencing work (Weird Tales) is correct. If the excerpt(s) are the only speculative content from a collection of poems, should the Rubaiyat be included but only the speculative content? Which would mean dropping the "(excerpt)" from the names. If it is not speculative, should the referencing work not have included the reference? If it is left in, should an entry for the original be created even if not speculative to support the excerpt? But if you do, would we want only the original for the excerpt to link to (eventually)? This example has been translated, so should only 'link' to the translated version - which may or may not have been specified.

My query was prompted by an Icelandic translation of the Rubaiyat I ran across. I was going to put my query as to whether the Rubaiyat was speculative in the Help Desk, but the general questions seemed to belong here. Some questions may have answers, but some may simply be fodder for later discussions on generic linking for things like translations, versions of, and such like. ../Doug H 16:52, 8 January 2022 (EST)

Let's address the general issue first. We index everything in genre magazines, including non-genre stories. By extension, it applies to non-genre excerpts. Also, title records have a "non-genre" flag which can be applied to explicitly non-genre works. I don't think the existence of a title record for an excerpt from a non-genre work should require us to create a separate record for the full non-genre title.

Re: this particular case, I read Rubaiyat of Omar Khayyam a long time ago, so I am not 100% sure, but I don't recall anything particularly speculative about it. It's a collection of pithy short poems with philosophical and religious overtones. Ahasuerus 23:41, 8 January 2022 (EST)

Currency symbols (again): full-width versions of currency characters

I was being nosey and scrolling through the recent activity page for any interesting looking edits, and noticed this pub, which had an odd-looking (to my eyes) pound sign in the price field.

On further investigation, it looks to be copied directly from the publisher's site, and is a Unicode full-width version of the regular GBP £ pound sign. (There's a similar one for USD $, and I imagine there'll ones for the Euro, yen, etc.)

This isn't a big issue, but as someone who's got some hacky code to try to determine a pub's country of original based on the price, I'm wondering whether I should update my code to handle these variant characters and/or it's reasonable to submit edits to convert these prices to the standard characters? (I've not yet run a query to see how many records might be affected, but I suspect it'll be few enough that these could be manually handled without too much effort. I dunno if the price cleanup report would/could pick them up?) ErsatzCulture 11:54, 13 January 2022 (EST)

Should use the standard symbol IMO (and there should not be a space in there either) - that will make it easier to convert if/when these fields get restructured. That's partially why we allow L, E and a few other letters to be used in case one cannot do their own symbol and convert them on Submit. :)

That record also has the wrong date (6-8-2020 is August and not June in UK and almost everywhere else that is not USA) and it is also too short IMO (Kobo concurs: 32K words for the ebook on the English edition - the Japanese word counts work differently but that's definitely too short for a novel) :)

PS: We probably need this to be catchable in our cleanup report somehow indeed... Annie 12:10, 13 January 2022 (EST)

"Fullwidth" and "halfwidth" versions of regular characters as used by Unicode are a known problem. They appear to be more common in Japan, which is why our input filters automatically convert "fullwidth" yen signs ("￥") to regular yen signs. Let me add "￡", "＄" and "￦" to the list of characters to be automatically converted to their equivalents. It won't address the larger issue of all other ASCII and Japanese characters having "fullwidth" and "halfwidth" counterparts -- like in the title of this record -- but it's a step in the right direction. Ahasuerus 14:21, 13 January 2022 (EST)

Meanwhile I fixed the field (as I was updating the format and the date anyway - both were incorrect - reversed month/date and the book is really pb-sized). Plus a note added on the title level about the length. Annie 16:17, 13 January 2022 (EST)

The software and the publication record have been fixed. Ahasuerus 16:23, 13 January 2022 (EST)

Thanks both. (I'd spotted the length issue, but not the date or format ones. Will submit the ebook once I've processed the data I've scraped.) ErsatzCulture 18:13, 13 January 2022 (EST)

Different Prices

New editor here. I have a copy of Asimov / Foundation's Edge which appears identical to this one: Foundation's Edge except that both the US and Canadian prices are different. My copy is also: "Seventeenth Printing: March 1989". Based on the Last User Activity Date, the two existing PVs appear to be inactive so I haven't asked them to check. However, the cover scan for the above record confirms the prices stated in the pub record, although I suppose it could be a scan from a different printing (seems unlikely). Does the ISFDb have a policy for this situation? Should I
1) edit and change the prices in the existing record?
2) add a note to the existing record stating some copies have different prices?
3) create a new pub record?
Teallach 13:46, 19 January 2022 (EST)

That will depend - first we need to make sure that these are really the same books. The basic policy is to make sure that we don't lose editions if we can help it.

Does your say "Printed in USA"? Del Rey in 1989 published both Canadian and US editions and occasionally the only difference was the printing place and the prices. If you truly have the same book, some more ideas:
How different are the prices? Is it possible that the copyright in what you are holding was not updated but it is actually a later printing? The prices can help figure that one out compared to other books from the same publisher

In any case, it sounds like you are holding a different book from this one so we will need a new record and possibly an update on the one we have - but let's see if we can figure out what is going on first.Annie 15:57, 19 January 2022 (EST)

My copy is as follows:
On front cover and spine: US$4.95, C$6.50
On rear cover: Cover printed in USA
On copyright page:

A Del Rey Book
Published by Ballantine Books
LCCN: 82-45450
ISBN 0-345-30898-0
Manufactured in the USA
First Ballantine Books Edition: November 1983
Seventeenth Printing: March 1989

The LCCN is unhelpful; it relates to the 1982 Doubleday hardcover.
There are no house ads in the book, so no clues from date codes or contemporaneously published titlesTeallach 18:50, 19 January 2022 (EST)

My first reaction was "$5.95 (US) sounds too high for a 1989 mass market paperback". However, Advanced Publication Search finds 6 Del Rey/Ballantine paperback with a $5.95 price. Notably, 4 of them are reprints of different Foundation books (other Asimov paperbacks published by Del Rey in 1989 were priced between $3.95 and $4.95), so I suspect that the $5.95 data that we currently have on file is correct.

Based on what we know and the primary verifiers' unavailability, I think the safest bet would be to create another publication record and document the discrepancies in Notes. Ahasuerus 20:06, 19 January 2022 (EST)

I agree - something is off here but without the other PVs, it will be hard to discover what. So clone the existing one, add your own cover and add notes explaining the double 17th printing (ask if you would like some help with wording) and maybe one day, this mystery will be solved. I do wonder if all of those 5.95 are later printings keeping the copyright page but who knows. Annie 21:09, 19 January 2022 (EST)

Help entry for Cover Art

The existing entry for Cover Art makes special mention of how to (or not to) credit the designer of the cover. There are two other special cases that might be worth noting, if we have an accepted policy - namely photographs and movie images.

I'd like to observe that some of the information regarding attribution belongs in the publication, but some (e.g. designer) belongs in any associated COVERART title record that is automatically generated and would need to be tracked down after approval. Perhaps this aspect might also be covered. As might the known problem of being unable to generate a COVERART without an artist to be able to merge/variant with other publications with the same art and how to deal with it. ../Doug H 12:27, 22 January 2022 (EST)

Help entry for External Identifiers

There is an inconsistency between the template name and example - JNB / JBN. Which is correct?:

JNB | Links to Japanese National Bibliography | using a Japanese National Bibliography number (JPNO) | {{JBN|22859001}}

../Doug H 15:47, 28 January 2022 (EST)

I have checked the software and it's "JNB". The Help text has been corrected -- thanks! Ahasuerus 16:41, 28 January 2022 (EST)

Standarizing Help guidance re: editing primary verified pubs

We have two places where we describe the process of editing primary verified publications: the ISFDB FAQ and Help:How_to_verify_data#Making_changes_to_verified_pubs. Like a number of other Help pages which cover the same topic, they provide editors with somewhat different instructions.

Earlier today I created a new Help page, Help:How to change verified publications, in order to centralize the instructions that cover this process. At the moment, it has the same text that Help:How_to_verify_data#Making_changes_to_verified_pubs had a few hours ago, a minor grammar fix aside. After comparing it with what the FAQ says, I came up with the following, more detailed and explicit, Help language:

It's not uncommon for two seemingly identical copies of a publication (book, magazine, etc) to turn out to be two different versions of the publication. What at first glance may appear to be a data entry error in an ISFDB publication record may be an indicator of a different edition/printing. It's also possible that editors with access to the same publication may interpret the ISFDB data entry standards -- as they apply to a particular situation -- differently.

For these reasons, proposed changes to a primary verified publication should be coordinated with all verifiers of the affected publication using their Wiki-based Talk pages. [Sentence added on 2022-01-29:] Note that this is in addition to entering a software-enforced explanation in the "Moderator Note" field when creating a submission. If a primary verifier has created a special Talk sub-page for notifications, please read and follow any special requests placed at the top of the verifier's Talk page.

The following are the recognized exceptions to this rule:

The verifier's Talk page states that the verifier has requested not to be notified about certain types of changes
The verifier's Talk page states that the verifier has left the project or is deceased. (Note that each primary verifier's "Last User Activity Date" is displayed in the "Primary Verifications" table displayed on the verified publication's Web page.)
[Edit 1: The following exceptions were added based on the discussion below on 2022-01-29]
You are adding one or more External IDs
You are fixing formatting, spelling and/or grammar errors in Notes
You are adding publication-specific links to Web pages
You are adding one or more transliterated titles

If one or more primary verifiers do not agree with the proposed change or if no verifiers respond within 7 days, post a note on ISFDB:Moderator noticeboard. A moderator will determine whether to make the proposed change and will inform the current verifier(s).

Note that this process should be followed by all editors, including self-approvers and moderators. If an original verifier does not agree to the proposed change, only a moderator should make changes to the publication record after notifying the verifier about the decision.

How does it look? Ahasuerus 16:52, 28 January 2022 (EST)

I like the combined version. Annie 18:08, 28 January 2022 (EST)

As is current practice, I should also explicitly include allowance to notify PVs via the note to moderators for trivial/unambiguous edits. The proposed text doesn't really allow for it, does it?. MagicUnk 03:30, 29 January 2022 (EST)

No, it doesn't. The proposed wording is more restrictive than current practice and a step backwards:

It doesn't differentiate between changes and additions. Adding data to a record is also an act of changing the record. As currently written, this would apply to even the act of adding an external id.
The second exception should be based on the Last User Activity Date on the publication record. The data is shown on the publication record. People shouldn't have to waste time checking whether a user who hasn't edited in years has an inactive template on their talk page (which are not consistently applied, especially for users who were not highly active to begin with).
It should allow for simple changes (ex. spelling fixes in the notes) or changes that bring a publication inline with updated ISFDB standards (ex. removing series name from the title) via the notification system (but require a clear description of the change in the note to moderator field).

-- JLaTondre (talk) 08:38, 29 January 2022 (EST)

The concern that I had with allowing adding (as opposed to changing) data without notifying the verifier was that adding a cover scan, a cover artist, a missed price, etc would be, technically, adding data, yet we would want the primary verifier to be notified because it could indicate a different printing.

Thinking about it, I agree that trivial/unambiguous changes/additions should not require notification beyond entering an explanation in the Moderator Note field, but the challenge is defining "trivial/unambiguous" in a way that everyone agrees with. Let's try to list everything that we want to consider "trivial/unambiguous" in the list of exceptions above:

Adding External IDs
Fixing formatting, spelling and grammar errors in Notes
Adding publication-specific links to Web pages
Adding transliterated titles

I am not sure above removing series names from publication titles. In most case it's a trivial change, but I have also seen disagreements about borderline cases.

Re: using the Last User Activity Date to determine whether to notify the verifier, the software uses the following dates to calculate it:

Date of last submission
Date of last primary verification
Date of last secondary verification
Date of last Wiki edit

It's possible for an editor to stop editing the data yet to continue to be a user of the database and to want to be notified about changes to his or her verified records -- I remember it happening a few years ago.

If we do decide to use the Last User Activity Date, we will want to come up with a threshold value, e.g. 2 years. Ahasuerus 10:25, 29 January 2022 (EST)

Re. "Adding publication-specific links to Web pages", this might overcomplicate things, but could I also suggest adding "... or fixing broken links" - typically either turning them into archive.org copies, or new URLs (e.g. when publisher's decide to upgrade their CMS and don't bother to add redirects for their old links)? That the old link is broken should be explicitly stated in the mod note. Also, in cases where the old link works, but another editor thinks they've found a better one (e.g. replacing a search results URL with a proper "canonical" URL for a specific ISBN), any such edit would still need PV approval. ErsatzCulture 10:48, 29 January 2022 (EST)

We could phrase it as "Update obsolete URLs as long as they point to the same Web text". "Obsolete" is more open-ended than "broken" because it includes scenarios like "HTTP-to-HTTPS upgrade" where HTTP links still work but shouldn't be used any more (deprecated.) Ahasuerus 11:06, 29 January 2022 (EST)

I think a key element in what is 'added' is whether it is evident from the publication or not. If I add a cover artist based on the artist web site, this would in no way invalidate or change anything the other PVer's had entered or verified. As long as I document the source of the added information, no harm, no foul. Adding one based on an initial visible on the cover (possible near an edge, clipped for some printings) would be the grey area. As for checking Last Activity Date vs. Talk Page, I'd lean to Talk Page. Just because I've entered/verified all my library and have no desire to get involved in Forum melees doesn't mean I shouldn't be notified. And updating formatting is a slippery slope, sometimes (not often) the format conveys information. Until we develop format guides, I'd say that anyone who dislikes a format badly enough to change it, should have to notify the people who found it acceptable. Besides, who else is going to admire all the work put in? ../Doug H 11:19, 29 January 2022 (EST)

Given that such notifications make up a large part of Talk Page usage, is there any way to develop a 'notification template' where you point to a publication, explain what you want and have the system post it where appropriate including your talk page with links to the relevant notifications for you to check? ../Doug H 11:24, 29 January 2022 (EST)

There needs to be a trade of benefit vs. burden. For example, how many users have not edited the database in 6 months to a year, have not edited the wiki in 6 months to a year, and still check their talk page every week? I doubt it is many if any. -- JLaTondre (talk) 11:50, 29 January 2022 (EST)

I can live with 12 or even 6 months as the threshold; my main goal with the proposed Help changes is to migrate from inherently subjective and ambiguous descriptions like "significant" to a set of objective, unambiguous rules which would be easier to understand and enforce. Ahasuerus 16:27, 29 January 2022 (EST)

They will still get the notification of change through the "My Changed Primary Verifications" when they return. We can easily fix the tiny amount of issues that might slip through the cracks. Ideally, the software should handle all this - edits that change verified data go into a pending state, verifiers are automatically notified, they get an option to ok the edit or add a comment, when all active verifiers have responded or 7 days have past, it moves to the moderator queue to handle. But until then, we should be careful of driving up the burden of editing while providing little benefit. -- JLaTondre (talk) 11:50, 29 January 2022 (EST)

Re: Notification of change. The log is not a 'notification', it's a log. Checking my log I see an entry from January 18th with changes to ContentTitle, External ID. The Note to Moderator says "Have copy in hand. Will PV. Added LCCN". Based on green highlighting I see they put the page number for the novel. I also see that it is linked to Libris-XL instead of LCCN and the editor has not been PV'd. Most of the changes I see in the log I would not expect notice for - adding external IDs, cover credits (that create COVERART records) and Notes. If I was gone for six months or more (again) sifting through the chaff for wheat is possible but not friendly. Especially when the comments are vague and I end up going back through the XML to see what was really changed. As to benefit vs. burden - I invested my time creating/verifying an entry. In those cases where you think I'm wrong (i.e. not just 'adding' information), why should your time be worth more than mine? So while I agree with the trade-off concept and would probably agree with you on 90% or more the the examples we pull up - what are we trading off? ../Doug H 16:11, 29 January 2022 (EST)

It is most certainly a notification. Moderators could do better at quality checking the changes against the moderator note (as a moderator, it would be nice to be able to edit the existing note field or add an additional note when approving), but a few bad examples do not invalidate it. You are overly optimistic if you believe that telling people to use the wiki will by itself improve the quality of the information provided. You would simply have gotten the same information on your user talk page. -- JLaTondre (talk) 16:29, 29 January 2022 (EST)

Sorry for nitpicking English, a notification is an official communication. I'm willing to go with the log is a passive communication of information and the Talk page is an active communication. I'll read your arguments for the log as being for 'passive communication' unless you say that isn't your intention. ../Doug H 23:46, 29 January 2022 (EST)

Yes, that's exactly what it should be: any explanation in the notes to moderators is 'passive communication'. As it is readily accessible, and the editor gets a warning whenever a PV'd record is updated, I don't see much difference with notifying the PV on his/her talk page - at least not for trivial, uncontested (uncontestable?) changes and/or additions. It's a different matter where there's ambiguity, or you want the PV to verify something against the physical copy - that's where the talk page is needed. So, to summarize, I am not saying that you shouldn't notify the PV, I'm saying that the notes to moderator is equally valid as the talk page in the cases laid out above. I myself am oftentimes adding prices from secondary sources, external ID's, standardizing titles by applying Dutch capitalization rules, adding/expanding notes, all of which are explained in the notes to moderators, iso on talk pages - I never (that I recall) had any issues with PV's being offended or demanding I notify them on their talk page. Just my 2 cents... Regards, MagicUnk 11:37, 30 January 2022 (EST)

So that's what the "New" message meant. It showed up when the feature was introduced and I've basically ignored it since. ../Doug H 16:25, 30 January 2022 (EST)

Re: "trivial/unambiguous". In many cases (publications from the olden times - mostly pre-21st Century) we have only years. As with the example of sourced cover artists it'd also be tedious to get a hold-on for adding months or even days to those publications, as long as they are sourced. Much of my work on German publications (but also for other languages) is doing that. Stonecreek 12:13, 29 January 2022 (EST)

(unindent)

Back to the rules. I don't think there is any difference between adding, modifying or deleting information that is based on the verified copy. IF all other information is given a source attribution, it should be easy to identify what was verified. ANY change to the verified data needs to be brought to the attention of the verifier, prior to making the change. Any other change, if the source is documented in the Notes, is basically the result of research, not evident in the copy and hence need not be brought specifically to the attention of the verifier(s). With a good definition of how to record what is in the verified copy (including notes like "Designed by:", actual publisher names, non-existence of cover signatures), the rules would only need to deal with contact and drop the 'exceptions'. ../Doug H 16:44, 30 January 2022 (EST)

Varianted Cover Art

It seems redundant to me to write a note explaining where a cover art credit came from when ISFDB tells you in the cover art field what the current art is a variant of already, as can be seen here: https://www.isfdb.org/cgi-bin/pl.cgi?420922. Just a thought. --Username 10:29, 30 January 2022 (EST)

One reason I see is that it tells why the artist is credited on a publication that does not contain the information directly. One other possibility is that varianting COVERART was not possible 10 years ago, making the note necessary. ../Doug H 14:04, 30 January 2022 (EST)

Let's consider the publication in question, i.e. the 1969 Popular Library edition of Edmond Hamilton's Outlaws of the Moon. The "Cover" field says:

Outlaws of the Moon by Johnny Bruck (variant of Perry Rhodan, #205: Der Wächter von Andromeda 1965)

while the Notes field says:

Artist credit from Perry Rhodan, #205: Der Wächter von Andromeda where art was first used.

Once we follow the link to our Perry Rhodan, #205: Der Wächter von Andromeda record, we discover that the Notes field says:

Bruck is credited at Perrypedia.

In other words, the original source of the cover art credit was Perrypedia. The attribution was then propagated to our Perry Rhodan 205 record and to our Outlaws of the Moon record. I think that stating this "chain of attribution" in the Notes fields of both publication records would be useful. Ahasuerus 15:30, 30 January 2022 (EST)

Melvyl

The link to Melvyl (the Catalog of the University of California Libraries) referenced in Sources of Bibliographic Information no longer functions and has been superseded. I've noted it as such on the wiki page with its replacement, but the template itself {{Melvyl}} needs to be dealt with. ../Doug H 12:48, 1 February 2022 (EST)

I have removed the Melvyl template and all current Wiki references to it. We never implemented it on the database side, so it's not an issue there. Thanks for the heads up! Ahasuerus 14:40, 1 February 2022 (EST)

Proposed Date help text revision

Cross-posted to Community Portal as well. I have a draft revision of the help text for publication dates available at User:MartyD/ProposedDateHelp for review. This is meant to codify/clarify existing rules/policies, not to define anything new or different. A special thanks to the early reviewers. Please comment on the discussion page there. Thanks! --MartyD 13:31, 5 February 2022 (EST)

Very nicely done. I like it. MagicUnk 07:45, 6 February 2022 (EST)

The official Template:PublicationFields:Date has been updated with the proposed text. --MartyD 12:40, 12 February 2022 (EST)

Converting a NOVEL to a CHAPBOOK (Titles with more than one publication record)

The helptext is clear enough, but I think one more step is needed. In step 2-5 the NOVEL record is converted to SHORTFICTION, effectively removing the container title from the publication. In step 6 the CHAPBOOK title is added, but because there is no container title, no language is set for the new container. I suggest to do step 6 only for the first publication, and add step 7: Set the language for the new CHAPBOOK title, and import this title in the remaining publications. Perhaps with a warning not to add other contents in step 2-5. I see no easier solution, since "the software will not allow you to add a CHAPBOOK to a publication record which contains a NOVEL title.". Thoughts? --Willem 10:41, 6 February 2022 (EST)

You cannot import it either - the software does not allow you to because the publication is still a NOVEL at that point and a chapbook cannot be added to a novel record. You need to convert each publication separately and add a separate chapbook record in each - and then merge them and adjust the language and the different flags (juvenile, non-genre and so on). Annie 16:04, 6 February 2022 (EST)

That should be tried. At that point the publication does not contain a NOVEL title anymore (it is already converted to SHORTFICTION). If the pubtype prevents the import, the helptext is wrong. --Willem 16:20, 6 February 2022 (EST)

Tried a few dozen times when I forget about that... (not lately though - see the last note here - you don't win anything even if you can do it). But sure, if you don't believe me - go and try. :) I can even give you something which needs converting to try on: this one for example. I wish it was doable - but it is one of the protections put into the software in the last year to ensure that we cannot import the wrong types into publication of a certain type. You cannot import a chapbook when the publication type is NOVEL and you still need to edit the publication to change the pub type anyway. Although even if doable, it won't reduce the number of steps -- you still need to change the publication type. I breaks even at 2 pubs. In cases where you have more than 2 publications, it actually increases the number of steps (as you are replacing Edit per title + 1 Merge + 1 edit for language and flags with Edit on the first book + edit on the title level + (Edit + Import) on each subsequent publication. Annie 16:30, 6 February 2022 (EST)

D**n, you're right. I suggest we change the helptext to "the software will not allow you to add a CHAPBOOK to a publication record of the NOVEL type." :) --Willem 16:55, 6 February 2022 (EST)

Thanks for the test subject by the way. --Willem 16:56, 6 February 2022 (EST)

Took seven edits because of the variant. And I nearly forgot to set the language! --Willem 17:03, 6 February 2022 (EST)

Why do you think I have a test subject I had not converted. :) They are a lot of work so I usually work on them in batches (with a bit of help from another editor). Did you remember to set the juvenile flag on the chapbook? And yeah - we need to change the wording. Annie 17:30, 6 February 2022 (EST)

Nope, you forgot that one :) Fixed now. I usually add all needed chapbooks, then merge them, then edit the result to add language and flags. And then variant up if needed (so the parent gets the child flag) :) Annie 17:52, 6 February 2022 (EST)

It's a labor-intensive process for sure. I expect that it would be difficult to reduce the number of steps/submissions because they all affect different records :-( Ahasuerus 18:23, 6 February 2022 (EST)

We still need to fix the help page in the "why we cannot have just one chapbook" section as Willem mentioned above - the steps are what they are. :) Annie 18:34, 6 February 2022 (EST)

Oh, sure. I just wish we could streamline the process without making the software look like the plot of a van Vogt novel... Ahasuerus 18:59, 6 February 2022 (EST)

Chapbooks with a Single Story

I've run into a problem with a story titled "The Gate That Locks the Tree", that appears in Liaden Universe Constellation: Volume 5. The story was first published in a chapbook, Gate That Locks the Tree where the story has a slightly modified title. Unfortunately, chapbooks cannot be part of a series, so the story itself is number 30 in the series "Adventures in the Liaden Universe". How can I make the story in the new collection a variant of the title in the chapbook? Bob 11:03, 8 February 2022 (EST)

Ignore the chapbook - open it and go down to the story itself. Variant the stories: this record and this record. That's it :) Annie 11:10, 8 February 2022 (EST)

You know, I tried that several times and failed. This time it worked. Thanks.Bob 19:58, 8 February 2022 (EST)

Aha! I figured it out. I was hitting "Create New Parent Title" instead of "Link to Existing Title". Duh! Bob 11:52, 9 February 2022 (EST)

Is Braille a font or separate letter system?

If we base titles (etc) from title pages, are there any special considerations for braille books to consider before I go down to verify? E.g. Sweet-blood [braille] at my local library, for ISFDB title. ../Doug H 13:01, 21 February 2022 (EST)

I would consider it a tactile font since it's simply presenting the alphabet in a different manner. Any publication that is in Braille should be listed as in whichever language it's in, and have a note indicating it's in Braille. ···日本穣 · 投稿 · Talk to Nihonjoe 14:07, 22 February 2022 (EST)

I agree with that - it is a font, not a language. Annie

Eligibility for a book - second opinion needed

I have The Original Bambi: The Story of a Life in the Forest on hold because I had been trying to figure out if it is in scope - there is nothing fantastical besides the sentient animals. But we do allow allegories ("Animal Farm" for example) and fairy tales - which gives me a bit of a pause. So... anyone with an opinion one way or another? Annie 23:01, 8 March 2022 (EST)

Our policy excludes:

Animal books for very young children, i.e. books for preschoolers which depict simple scenes from animal life featuring anthropomorphized animals

but "Bambi" is above that threshold. We include other non-fantastical books about anthropomorphic animals for children like Brian Jacques's Redwall novels (the supposedly "magical" MacGuffin turns out to be mundane), so I think Bambi and it sequel should be "in". Ahasuerus 08:39, 9 March 2022 (EST)

Redwall's animals don't behave much like animals though - while in Bambi we are essentially in a nature tale with talking animals which behave as what they are. I'll approve it though - better have one extra than miss one of ours - and we can always reverse that. Annie 13:52, 9 March 2022 (EST)

I imagine that if we read it it might be clearer... If i encounter a copy I'll give my opinion. Thanks. gzuckier 17:29, 9 March 2022 (EST)

Did you mean to transient verify it? Because that implies you have a copy at hand (or had when you verified it anyway) :) If not, you may want to remove the verification. Annie 17:46, 9 March 2022 (EST)

Vlaamse Filmpjes, magazine or publication series

I need some input / opinions after this discussion. "Vlaamse Filmpjes" is a Belgian periodical for young readers that has been around since 1930, and has some 3400 issues published. 10-15% of these are considered speculative fiction. The majority of these are 32 page staple bound phamphlets, that have one short story each. Sometimes two to four issues are combined in one publication. Normally I would enter this as a publication series of mostly chapbooks, like a lot of Dutch and German periodicals are (see Terra, Apollo, Terra Astra and many more, but to my surprise the issues we have in the database are entered as both publication series and magazine. I always thought this was forbidden, but I can't find anything in the help texts. We now have a stalemate where MagicUnk wants to keep it as a magazine, and I want to turn them into a publication series. Questions are: Is the combination magazine/publication series allowed or not, and if not, which of the two should the "Vlaamse Filmpjes" be. Comments are welcome. --Willem 16:25, 26 March 2022 (EDT)

I notice the site describes them as pocket books, despite the whole numbering and "volume" numbering schemes, so chapbook seems more appropriate than magazine (although I have no objection to either treatment). Putting a magazine into a pub series and using that to record and organize by whole number is rather clever. In my opinion, however, that combined treatment is incorrect. Issues of a magazine are like a title series in that no matter who publishes it, the series continues. --MartyD 12:00, 28 March 2022 (EDT)

It is clever but if we decide to allow it, we really need to do that properly - and well defined. As it is, I agree, the usage is incorrect - pub series are not used for magazines because the series field is already a de-facto pub series one. Annie 13:39, 28 March 2022 (EDT)

What is the distribution model and what do the Belgians consider these to be? That does not always help with the determination but it may help - that's how the Italian Uranias ended up as magazines even if the whole non-Italian world will probably consider them books... It is a thin line sometimes. I am leaning towards magazines (even though ISSN numbers are not just for magazines (they are for any periodical and some chapbooks series and even books series are know to have been published that way)) but either way works for me as long as it is one way or the other. Annie 13:39, 28 March 2022 (EDT)

If I remember correctly, Urania ended up as a magazine because serials were not allowed in other pub types. Vlaamse Filmpjes were (are?) distributed by subscription, through schools and on newsstands, and are considered to be magazines, but the other examples I mentioned are also considered to be magazines. We should have some consistency in how to handle these. --Willem 14:46, 28 March 2022 (EDT)

The problem is that depending on whom you ask, these can go either way :) So I tend to ask local readers when possible. The German editors decisions on how their magazines/whatever are entered should not be determining how we deal with the other languages (and that all can change anyway). It will be very hard to make a consistent rule (requiring columns for magazines will make 90% of the modern small print magazines anthologies for example). Not to mention the webzines. :) I'd love to have a workable rule but... Annie 14:55, 28 March 2022 (EDT)

I agree! I can live with either decision, but as they are now, they're simply wrong. By the way, Fandata (the Dutch SF database) has them as a publication series to make the titles visible on the author's pages. --Willem 15:37, 28 March 2022 (EDT)

Oh, I don't think that anyone is trying to argue that they need fixing. I am just not sure in which direction. :) As I said - I lean towards magazines but a series of chapbooks also makes sense :) The Fandata example is exactly why things can get even murkier these days - sometimes definitions in different bibliographies are based on how a certain DB/project is designed and not based on what the actual reality of the publications is. Another way to look at the differentiation is the editor situation - how important is the editor for these? Moving to chapbooks/pub series means that the editor gets recorded ONLY in the notes; keeping them as magazines leaves the credit to the editor. As these seem to be unknown for the non-PVd ones, it seems like they are not considered "someone's magazine" or something like that. Which may mean that chapbooks may be better.

One other thing: Our EDITOR really stands for "PERIODICAL" which is not a real book (Aka some annual publications which are considered proper books even if they are Periodicals...) and we have only two options under it: Fanzine and Magazine (and even that line can be blurry outside of the English sometimes). So we record things like newspapers as magazines (which is incorrect really but that's what we have) -- thus masking the whole issue a bit. So the question is not if it is a magazine really but rather is that a periodical? If it is, considering that it is not really a book, it really should be under the EDITOR I would think - which means Magazine. The hsort version of all that... pick one, document it and we can always change it if needed... Annie 16:08, 28 March 2022 (EDT)

And while making a decision - the naming rules are different -- you CANNOT have the story title in the magazine title under our rules. So that may also determine what direction you want to go. We can discuss changing the magazines naming rules but under the current rules this needs to be called "Vlaamse Filmpjes, 29 February 2008" as a magazine (with February in Dutch) (or "Vlaamse Filmpjes, #12" if we claim that the magazine is not dated) and just "De ring van de textielbaron" as a chapbook (series titles of any type leave the title of the book). The current Frankenstein of a title is invalid for either case under the current rules. Just saying. Annie 16:13, 28 March 2022 (EDT)

Just one of the things that are wrong. The editor should not be a problem i.m.o. Since only 10 to 15% of the issues are speculative fiction, it would be a non-genre magazine, with "Editors of Vlaamse Filmpjes" as editor. --Willem 16:27, 28 March 2022 (EDT)

Just pointing out the differences - because sometimes the subjective decisions are based on weird things. :) PS: Under the current rules, that is not necessarily needed about non-genre magazines anymore - it is allowed but not mandatory anymore so the magazines can keep their editors if desired. We changed that some time last year. Annie 16:36, 28 March 2022 (EDT)

I notice that, indeed, I did a couple of things wrong when entering these few Vlaamse filmpjes. Pub series, even title, can be fixed to conform to magazine entry rules, but I am still inclined to enter them all as magazines, but I'm fine with whatever the majority prefers. MagicUnk 10:35, 29 March 2022 (EDT)

(unindent) So there's a clear majority for magazine. I entered a test subject here, as I think they could be. Some thoughts: Volume/number was not printed before the 54th year of publication. I would prefer the title to be "Vlaamse Filmpjes, #whole number". The publication date was first printed around #1100. I'm not sure about the editor. I'm leaning towards "Editors of Vlaamse Filmpjes" only. Some editions have a list of editors, the early editions mention no editor at all and the names I have seen are not known in the speculative fition field. Any other suggestions? Thanks,--Willem 14:43, 6 April 2022 (EDT)

That looks reasonable to me. --MartyD 08:02, 9 April 2022 (EDT)

Seeking clarity on multi-volume works.

I have run across multiple volume works a couple of times and always seem to end up spending too much time digging through the help trying to figure out the best way to deal with them. Can the Help page be updated to explain the interplay between Title and Series (title and pub) values and what should be included in each:

When the individual publications have no individual title (e.g. Works of Jules Verne [over] Volume 1).
When the individual publications have separate titles as subtitles (e.g. Isaac Asimov's Understanding Physics [over] Volume I [over] Motion, Sound, and Heat.
When the entire set is published as a single bound publication (e.g. The above set as The History of Physics - presumably an omnibus).
When different publishers use different sub-titles (e.g. 2 vs. II, punctuation, Part vs. Volume).
Variations between publishers (e.g. The Works set was published as a 15 volume set, and also as a 10 volume set, where volumes 1-10 match in content, Mysterious Island has single volume, double volume and triple volume editions).
Does continuity matter? (Physics volumes are stand-alone, Works are stand-alone generally but some stories spanned volumes, Mysterious Island is one continuous story).

One factor to consider is that the Library of Congress seems to assign a single number to the set, and the OCLC allows both set and volumes to be done. Another is ISBN's, although all my examples pre-date ISBNs.

I'm sure there are other variations to be covered (such as those raised by The Lord of the Rings) that can be added to the discussion for consideration. Thanks for any input. ../Doug H 12:30, 23 April 2022 (EDT)

Just thinking aloud - that is one of the "we do what makes sense" area I think...

I go by the title page but with the simplified titles rule in mind. So Works of Jules Verne [over] Volume 1 will be Works of Jules Verne: Volume 1; Understanding Physics [over] Volume I [over] Motion, Sound, and Heat will be "Motion, Sound, and Heat" and series Understanding Physics (1) unless the whole book is from a larger set of volumes then Understanding Physics: Volume I: Motion, Sound, and Heat" is also possible.

Omnibus if it contains really different works; if it was really a single work that was just split for publication, I will go with NONFICTION (if it is what it is) and use the series or notes to connect.

Different subtitles make variants IMO - being it for 2 vs. II or for lacking oxford comma in some cases...

When the same title means different book (aka volume 1 of 2 vs 1 of 3), notes become mandatory.

Not sure if continuity matters when I am thinking about sets and multi-volume sets. Although I am much more likely to make sure that the boxset which contains the part of a novel in volumes is added as a set as well than I would be for a 15-volumes set of novels/series/anything (which can be usually shown via a pub series anyway).

LCCN assigns a single number, OCLC assigns a single number for the whole run of a magazine and so on - we record publications... If we record these just as sets, we will rarely see verifications on them so I prefer the practice of adding the individual volumes (and if they have a slipcase and so on, add the whole thing as well). But that is a bit murky. Annie 13:22, 29 April 2022 (EDT)

Text stories in comics floppies

A lot of the older comics books have one or more text story in them - and these stories are eligible to be added here very often. What is the current consensus on the type to use for the book: non-genre anthology (because there are other stories from other authors which are not eligible - that maks their covers eligible under the current rules) or non-genre magazine issues (especially for the long running series of comics - covers here won't be eligible)? Any opinions? Annie 13:08, 29 April 2022 (EDT)

I'd vote for non-genre magazine issues: comic books that feature more than one item would typically be that (like 'Heavy Metal' for example). Christian Stonecreek 04:24, 30 April 2022 (EDT)

I agree. That seems to me to be the best treatment. --MartyD 08:54, 30 April 2022 (EDT)

I never really considered this to be a question and had already entered several Otto Binder prose stories appearing in Captain Marvel Adventures exactly as we are discussing (i.e. non-genre magazines). --Ron ~ Rtrace^Talk 10:47, 30 April 2022 (EDT)

Heavy Metal is a magazine and not a floppy though - thus the pause/overthinking for a second. Annie 13:50, 9 May 2022 (EDT)

(unindent) Magazines it is then. :) Annie 13:50, 9 May 2022 (EDT)

Other Prices

Mod added A, C, N.Z. to a note where I listed other prices for a book, then wrote a message on my board telling me that's the way I should do it. Leaving aside the fact that there's hundreds/thousands of edits I've made adding other prices in notes without those letters, it seems repetitive and unnecessary to me to add them when I enter the actual countries in the note. If it says Canada, what's the point of a C? Of course, the main price entered in the price field needs those letters, but notes are entered as particular editors want, in a whole range of styles. I'm particular about entering them exactly as in the book itself, whether it says Canada or Can., Australia or Aust., New Zealand or N.Z.; if a book actually has an A, C or N.Z. in front of the other prices then I enter them that way, but I'm sure most people here know that doesn't often happen. Not sure what the point of my note is, since I'm going to keep entering notes in my usual way, but it was suggested I open a discussion here, so that's what I'm doing. --Username 20:10, 4 June 2022 (EDT)

I've run across updates to currency in my notes once or twice, but never someone saying it's the correct way. I don't think the currency standard should be imposed in the Notes, Specifying the price as 25d in the notes if it matches the book would be more useful for comparing editions than forcing £0.25. I've used the change in foreign prices to differentiate editions. ../Doug H 23:52, 4 June 2022 (EDT)

Suppose you have a publication with $16.99 in the price field. In the notes it says '$19.99 in Canada'. It's ambiguous whether it's 19.99 US currency or $19.99 Canadian currency. If we use the same standard in the notes as in the price field, there is no ambiguity. John Scifibones 07:21, 5 June 2022 (EDT)

As a Canadian, I see no ambiguity in '$19.99 in Canada' being in Canadian dollars. More specifically though, if there is ambiguity, by forcing the C$, you remove the ambiguity, which, if it had another interpretation, you deny later readers. If you want to put C$, go ahead, but explain your reasoning - e.g. "Additional price given as $19.99 in Canada, taken as C$19.99". ../Doug H 09:22, 5 June 2022 (EDT)

Results of a quick search

Publications with a note containing 'Canada' and 'C$' 9,815 (46%)

Publications with a note containing 'Canada' and ' $' 11,501 (54%)

Clearly there is a difference of opinion among the community, at least in usage. My intention is to discuss whether a specific policy statement regarding currency symbols in the notes in necessary. If so, what should that statement be? John Scifibones 10:46, 5 June 2022 (EDT)

Clearly there is a difference of opinion, as an unknown number of the C$ entries were mine and have been changed to include the C$. And I wish they'd stop. ../Doug H 15:31, 5 June 2022 (EDT)

One thing to keep in mind is that we have FR 1158, "Allow multiple prices per publication". It's a fairly well fleshed out Feature Request as such things go and ready for implementation. Once implemented, it will eliminate most of the ambiguity. Ahasuerus 13:22, 5 June 2022 (EDT)

P.S. Let me clarify that "ready for implementation" means that it doesn't need additional discussions for a developer to start working on it. I am currently supporting Al's server upgrade efforts while working on fixing outstanding bugs (6 fixed, 15 to go.) Once these two tasks are completed, we can regroup and re-prioritize FRs. Ahasuerus 13:40, 5 June 2022 (EDT)

Prior to seeing this discussion, it never would have occurred to me to do anything but record the other pricing information as stated in/on the book. That said, it also would not have occurred to me to question a submission containing notes stating apparently normalized prices. I guess I never thought about data normalization standards applying to free text notes. One other example to think about is people's names cited in the notes: Wouldn't we expect those to be as they appear in the book, rather than normalized to ISFDB standards? --MartyD 07:40, 6 June 2022 (EDT)

Strikethrough, superscript and other HTML outside of Notes

We currently support "limited use of HTML" in Note fields. Typically, we do not use HTML in other fields, but Help seems to be ambiguous about it. Here is what Template:PublicationFields:Title and Template:TitleFields:Title say:

Fonts. Sometimes a title will have one or more words in italics, or in boldface, or in an unusual font. The ISFDB software would permit representing these via embedded HTML. However, this would mean that user searches that did not include the HTML would fail in many cases where they ought to succeed. Therefore, do not use embedded HTML to show font changes.

The way Help is worded, embedded HTML seems to be disallowed only to the extent that it changes fonts. Help even states that the ISFDB software "permit[s] ... embedded HTML".

I suspect that this is the reason why we have the following 3 title records with embedded "strikethrough" HTML:

and one publisher record with embedded "superscript" HTML.

Using embedded HTML outside of Note fields causes a couple of problems. The first one is the issue with searching which is mentioned in the Help paragraph quoted above. The second is the fact that our software can't reliably display titles with embedded angle brackets. For example, consider the following submission. The title of the story on page 69 appears to be "Adventures in Gaming". However, if you look at the underlying XML, you will notice that the actual submitted title was "<sarcasm> Adventures in Gaming </sarcasm>". The software interprets "sarcasm" as an HTML tag, which is why it's not displayed on the main submission page.

Based on the above, I am thinking that it would be best to simply disallow all HTML outside of Note/Synopsis fields. If we do that, we will have to change just 4 records. Then I will be able to update the ISFDB software to display angle brackets as text -- as opposed to as HTML -- making it possible to display titles like "<sarcasm> Adventures in Gaming </sarcasm>" correctly. It would fix Bug 673, "HTML tags in titles", and Bug 279, "Titles with "<?" do not display correctly".

Does this make sense? Ahasuerus 19:05, 15 June 2022 (EDT)

It does make sense. Also, I believe that allowing HTML in input can be a potential security issue as javascript can be injected. I agree with your proposal. --Ron ~ Rtrace^Talk 19:31, 15 June 2022 (EDT)

Indeed, JavaScript injection can be a serious security issue. However, we have partial protection against it because we had CSP (Content Security Policy) implemented a couple of years ago. If a malicious user creates a fake submission with embedded JavaScript, CSP will prevent it from being executed in any modern browser. Ahasuerus 21:08, 15 June 2022 (EDT)

I am 100% on board with disallowing HTML for non-notes/synopses. However, are there any explicit rules or help text to cover how an editor should handle titles like the examples above? e.g. should they use appropriate Unicode special characters, or just vanilla ASCII? If the former (which is what I'd expect), could/should the transliteration field be used for such titles to have a vanilla ASCII version that - I suspect - would be more amenable to searching on?

(Also, if the Unicode option is the one chosen, then it'd probably be useful to have a help page covering how to enter special characters e.g. linking to online tools like https://unicode-table.com/en/tools/strikethrough-text/ . I'll volunteer to write such a help page.) ErsatzCulture 19:55, 15 June 2022 (EDT)

The 2 Help templates linked above have sections for "Symbols and punctuation". We should probably fold them into a single sub-template. Ahasuerus 21:03, 15 June 2022 (EDT)

Hearing no objection, I have created FR 1510, "Disallow HTML outside of Notes/Synopses". Ahasuerus 21:45, 20 June 2022 (EDT)

Outcome

Template:TitleFields:Title and Template:PublicationFields:Title have been updated to explicitly disallow all HTML outside of Notes/Synopsis fields. Rules and standards changelog has been updated. All affected database records have been updated. I am still working on updating the software to support entering titles like "<sarcasm>" -- it's a bigger can of worms than I expected it to be. Ahasuerus 22:33, 28 June 2022 (EDT)

Title spelling

There is a disagreement between me and another editor about the spelling of this title [ DAS GROSSE SPIEL]. The title is written in capitals on the cover, title page and back cover. This is not intentional but is often practiced by this publisher. In German there is no equivalent for the letter ß as a capitel, but the substitute SS. Without capitels the title of the book would be Das große Spiel. On the back cover, the publisher refers in an essay about the book the title as Das große Spiel. Is the title (not the transliterated title) of the publication finally Das große Spiel oder Das grosse Spiel. Regards Rudolf Rudam 05:22, 16 June 2022 (EDT)

Generally, we change titles in all caps to standard capitalization. Das große Spiel would be correct. ···日本穣 · 投稿 · Talk to Nihonjoe 13:07, 16 June 2022 (EDT)

Sorry, but I don't think that's stated anywhere in the rules: there the regularization rules are for case, symbols & punctuation, and some special signs (like #). Instead, the rules in this case allow and demand the appropriate use of upper & lower case for the single letters, and I don't see an allowance to interchange letters with others that aren't stated on the title page. (Note that 'case' is discussed in the help for single letters, and 'capitalization' only refers to the first letter of a word). Both spellings ('ss' and 'ß') are and were allowed for this title in German (that's so, because ß stands for ss). We also wouldn't have allowed two German versions for this author. Christian Stonecreek 13:22, 16 June 2022 (EDT)

Please also take a look at the spellings with DNB and OCLC. Christian Stonecreek 13:42, 16 June 2022 (EDT)

As you wrote, "the regularization rules are for case". That means regularizing (or standardizing) ALL CAPS titles to Normal Caps titles. Onces that title is normalized as far as case, then Das große Spiel would be correct. We'd do the same thing for any title that's in ALL CAPS. ···日本穣 · 投稿 · Talk to Nihonjoe 14:39, 16 June 2022 (EDT)

And regarding Kurd Lasswitz, I suspect that name may have been used before we could use special characters in names and had a Transliteration field for the name. Looking down the list of works, "Kurd Laßwitz" seems to be the most common usage by far. The canonical name should be changed to "Kurd Laßwitz" for this author, with "Kurd Lasswitz" as a transliteration. ···日本穣 · 投稿 · Talk to Nihonjoe 14:42, 16 June 2022 (EDT)

'Case' (the upper and the lower one) is used in the rules for single letters, 'caps' (which I take stands for 'capitalization') is used only for the first letter of a word. The both notions have different meanings. And please also refer to the added note of 'Das grosse Spiel' as stated with Deutsche Nationalbibliothek and OCLC; at least the first source should know what it does. Christian Stonecreek 00:26, 17 June 2022 (EDT)

Where does it say that "case" is used only for single letters? I think you're trying to be too restrictive in applying it. And DNB has both große (at the bottom of the listing) and grosse, so we can't use them as an example. OCLC is definitely not a reliable source of information, either. Either way we decide to go, if we normalize the capitalization, we should use "große". If we choose to leave it in all caps (which I don't think we should do), it should be "GROSSE". ···日本穣 · 投稿 · Talk to Nihonjoe 12:22, 17 June 2022 (EDT)

Disclaimer: Although my parents are German, I make no claim to any German fluency or educated background.... My naive opinion is that if both "große" and "grosse" are acceptable, even though SS is compelled in "GROSSE" because there is no capitalized equivalent for "ß", then we don't know if the intent was "große" or "grosse" and so we should default to "SS" -> "ss". On the other hand, if "no" (I know, nothing is ever absolute) native/fluent German writer would use "grosse" -- even though it is acceptable -- then it seems "SS" -> "ß" is appropriate and would be understood/expected. I'm not entirely convinced of that, though. If we consider, say "KOENIG" -> "koenig" or "könig", it seems right to insist on the "oe" even if "no" native/fluent German writer would use that instead of "ö". The difference, of course, is that because "Ö" is available, the choice of "OE" vs. "Ö" is apparently deliberate, for reasons unknown. --MartyD 18:34, 17 June 2022 (EDT)

p.s. I do think for this specific case that if the publisher uses "Das große Spiel" elsewhere on the same publication, that is sufficient evidence of intent to use "ß" not "ss". My feeling is the default should be "SS" -> "ss" with "SS" -> "ß" permitted if known to be the intent (similar to the rule for "A. B." vs "AB" in an author's name regularization). --MartyD 18:41, 17 June 2022 (EDT)

And regardless of which is used, the other should be in the transliteration field so those searching either way can find it. ···日本穣 · 投稿 · Talk to Nihonjoe 20:19, 17 June 2022 (EDT)

Some general remarks regarding "ß":

A capitalized equivalent for "ß" had been discussed for many years and was introduced in 2008 as a Unicode character. Since 2017 it's officially part of the German spelling rules. Since 2020 a capitalized "ẞ" is the preferred way to write when using all caps in text and information processing: "GROẞE" instead of "GROSSE", but both forms are still allowed. However, it had already been used decades before every now and then. See https://en.wikipedia.org/wiki/%C3%9F#Development_of_a_capital_form or the German page https://de.wikipedia.org/wiki/Gro%C3%9Fes_%C3%9F I can't remember actually having seen it in publications I own, but the Wikipedia page contains some examples.
There's a distinction is the usage of "ß" depending on the German-speaking country: Switzerland and Liechtenstein never use "ß", but always "ss". Therefore Das grosse Spiel would be correct spelling in these countries.
In Germany-German spelling, Das große Spiel is correct. Surely, Das grosse Spiel will be understood by everyone as well, but I can't image why someone would write it like that in Germany. In school, you'd very likely get that marked as wrong in a grammar test.

As for the publication in question and the rules:

I think the fact that there was no capital "ß" for a long time and "SS" was used instead for all caps words, somehow as a "workaround", should not lead to wrong spelling when normalizing titles back to regular case. I don't know why DNB uses Das grosse Spiel, but I guess that's because of some (weird, probably now outdated?) library rules. For me, being used to Germany-German spelling, Das grosse Spiel looks utterly wrong (sorry, DNB and people from Switzerland and Liechtenstein :) ). Das große Spiel should be used.

Jens Hitspacebar 07:02, 18 June 2022 (EDT)

I understand the rules as to use the title on the title page, and then to apply the appropriate lower and upper case. So the proper use of case for 'DAS GROSSE SPIEL' leads us to 'Das grosse Spiel'. Right so far?

I disagree. The help for "Title" says: "Titles should have case regularized according to language-specific rules". It doesn't say "just lowercase everything". If the uppercase version of "Das große Spiel" is "DAS GROSSE SPIEL", why shouldn't the lowercase go back to "Das große Spiel"? At least regarding grammar, "grosse" is wrong in Germany. As for DNB, see my other remark below. Jens Hitspacebar 14:19, 21 June 2022 (EDT)

By our used definitions 'Das große Spiel' and 'Das grosse Spiel' are transliterations of each other: both are used on title pages, but in this case one alternative with only Roman letters in it was used. So with große/grosse we do not have a case of case here (please allow the wordplay), which seems to be the only regularization allowed, according to the rules. The letter 'ß' is a non-Roman one, and like Nihonjoe I'd be all for allowing 'Das große Spiel' into the field for transliteration. As for now the rule says 'Populate only if the title is spelled using a non-Latin alphabet/script'.

And for DNB: Deutsche Nationalbibliothek is the reference source for German publications, and imo we should try not to deviate from it, except for obvious gross errors.

'Das grosse Spiel' to me looks also more strange than the classic 'Das große Spiel', but as you, Jens, explained, the latter looked and looks more strange to many German writing people (in fact, millions of them), and I'd say nowadays both in general are almost equal in German: to me 'Ich denke, dass' still looks more strange than 'Ich denke, daß', but today the former is the only one in use. Christian Stonecreek 11:15, 18 June 2022 (EDT)

You probably mixed up something, or misunderstood me: I didn't say that the grammatically correct "Das große Spiel […] looks more strange to many German writing people". I meant the opposite: "Das grosse Spiel" will look strange to most people in Germany. It's all quite simple actually, at least when focusing on grammar: "Das große Spiel" is correct in Germany, "Das grosse Spiel" is wrong. There may be or have been some other, special rules which are or were used by librarians and which led to using "grosse" at DNB. Maybe it's not an error and they really just lowercased "SS" to "ss" for whatever reason. Maybe this way is even correct in a librarian's mind. I don't know. But there are also many decades old pubs at DNB which use "große". DNB is not error free (I've seen numerous price errors there already). So… in the end the question is: does the phrasing "according to language-specific" in the rule mean known grammar rules, or does it mean (probably unknown) librarian rules? I'd go for correct grammar. Jens Hitspacebar 14:19ta, 21 June 2022 (EDT)

I think I understood what you meant: with 'many' I just meant the Swiss and Lichtensteinian people (is the last one right? - likely not, but anyway ...).

So, both versions are correct, "Das große Spiel" was just more common, both were then in use.

You're right: the rule doesn't say just to lower-case everything: the rules only state to use the appropriate case (I recently ran over a title with 'Basic-Programm' in it that had to be regularized to 'BASIC-Programm'). But as explained above, the transformation to "Das große Spiel" would be a transliteration, and the rules don't anywhere advise to use this, there's only the point to regularize the case! (And since the lower case of 'S' is 's' - just as the lower case of 'G' is 'g' - the regularized case leads to "Das grosse Spiel".

With language-specific rules it's meant that - as an example - "DIE ROTE KÖNIGIN" could be regularized to "Die rote Königin" (if 'rote' is just a reference to a dress) or "Die Rote Königin" (if it's a name or title, or if it's the title used for a publication in English).

Regarding DNB: to me it seems they are using the same rule for regularizing case as we do. Christian Stonecreek 02:09, 22 June 2022 (EDT)

And on grammar: We also document erroneous grammar and misspelled words. If it would have been "DAS GROSSES SPIEL" the correct case regularization to record would be "Das grosses Spiel". As you wrote, "Das grosse Spiel" seems to be wrong in Germany (but is and was allowed), it is perfectly correct in nearly all other countries, including German speaking ones.

There's also the counter-example of this publication, for which I'm pretty sure Der letzte Kuss was the preferred version, even back in 1986. Christian Stonecreek 06:44, 22 June 2022 (EDT)

Apparently, it seems that a significant majority are in favor of "Das große Spiel" as the title. I will correct it accordingly. Thanks for your in-depth contributions to the discussion. Regards Rudolf Rudam 09:05, 27 June 2022 (EDT)

Help section clarification, ellipsis

The following search results indicate a lack of understanding regarding the use of ellipses.

16,923 Titles containing an ellipsis
5,369 Titles ending with an ellipsis
1,099 Titles starting with an ellipsis
10,455 Titles with an ellipsis in the middle

Here are extracts from the relevant help sections

Bullet point Symbols and punctuation under Title in the Title help section contains the following:

1. An ellipsis should be entered as the sequence "period", "period", "period" with no spaces in between the periods. If the ellipsis is in the middle of the title, it should be entered with a space after it as well, prior to the start of the following word.

Bullet point Symbols and punctuation under Regular Titles in the Contents help section contains the following:

2. An ellipsis should be entered as the sequence "period", "period", "period" without spaces in between the periods. If the ellipsis is in the middle of the title, it should be entered with a space after it, prior to the start of the following word.

My suggested changes

Use number 2 in both locations.
Add a statement 'There is no space preceding a trailing ellipsis.'
Add a statement 'Insert a space after a leading ellipsis.'

I'm not suggesting any change to our standard, merely trying to clarify it in the help. John Scifibones 14:40, 30 June 2022 (EDT)

P.S. If my understanding is in error, that further emphasizes the need for clarification! 15:40, 30 June 2022 (EDT)

The first question that comes to mind is "Is this language-specific?" For example, English titles do not use a space between a colon and the word that precedes it, but French titles do and our rules account for that just like they account for different languages having different capitalization rules. (See Talk:Title Regularization for a discussion of different language-specific rules.)

Checking French Typographical Rules and our summary of Spanish punctuation rules, I see that French and Spanish do not have a space before an ellipsis, but what about other languages? Ahasuerus 10:13, 1 July 2022 (EDT)

The search does not seem to differentiate between "..." (three periods) and "…" (ellipsis) (U+2026 or Alt+0133). Do we care? ../Doug H 16:04, 1 July 2022 (EDT)

The numbers above are for all languages combined. Here is a breakdown for English, Spanish, French.

Position	English	Spanish	French
Ending Ellipsis	3004 Total (1750 incorrect 58%)	53 Total (35 incorrect 66%)	1072 Total (21 incorrect 2%)
Beginning Ellipsis	824 Total (146 incorrect 18%)	13 Total (6 incorrect 46%)	53 Total (27 incorrect 51%)
[space]...[space]	2024 Titles	18 Titles	23 Titles

There are 3862 German language titles. It appears that usage determines whether or not to include spaces.

Perhaps a link to language specific help can be incorporated into any change? John Scifibones 16:49, 1 July 2022 (EDT)

Brazilian real (new currency symbol)

Shouldn't Brazilian real (R$) be a recognized currency symbol? --Rosab618 15:05, 6 July 2022 (EDT)

Sounds like a good idea. We have 756 publications priced in Brazilian reals and the symbol is stable, so it should be a useful addition to the menagerie. Ahasuerus 19:18, 6 July 2022 (EDT)

Hearing no objection, I have added mouseover bubbles for Brazilian reals. Here is an example. Ahasuerus 21:39, 9 July 2022 (EDT)

Old UK prices

Template:PublicationFields:Price defines how we should enter "pre-decimilisation" prices. For older UK books, it provides the following guidance:

Even older British paperback books, and magazines, may have been priced in pennies alone, indicated by a "d" suffix. E.g. 6d is six old pence, or half a shilling, 9d is nine old pence or three-quarters of a shilling. These are entered the same way as other pre-decimal prices but using the '-' for zero shillings, e.g. -/6 and -/9 in these examples.

This takes care of prices like -/6 ("6d"), but some older UK prices also included "halfpennies", e.g. New Worlds, April 1939. Our price field currently says "-/4 1/2", which looks reasonable, but I don't think we have this format documented anywhere. Is this how we want halfpennies to be entered? If so, we can update Help and then I can adjust Publications with Invalid Prices to account for this format. Ahasuerus 14:29, 10 July 2022 (EDT)

I see that the remaining 4 price values have been adjusted and the offending spaces have been eliminated. It looks like there is no need to do anything on the software side of things. Ahasuerus 12:09, 14 July 2022 (EDT)

Are photographs interiorart?

Hello! In this discussion [1] I came to attention that photographs of authors were accepted as interiorart, although it is excluded in the rules yet. Is the current policy still valid or should it be changed? I'm in favor of the present rules. Regards Rudolf Rudam 09:36, 9 August 2022 (EDT)

The rules don't really exclude them. Template:PublicationInfo:WhatToInclude (which is transcluded to various other policy pages) has the following under "Contents included with exceptions":

"Photography: As a general rule, photographs are not indexed. But, if the photograph illustrates a work, it should be entered as INTERIORART. Author photographs are usually not indexed. This determination is left to the record's primary verifiers."

As written, it leaves it up to the verifiers if they want to include them or not. -- JLaTondre (talk) 09:52, 9 August 2022 (EDT)

It seems that the current rules should be kept the same. I still find them very vague and ambiguous. Rudolf Rudam 11:08, 11 August 2022 (EDT)

I add cover artists when the cover art is a photograph, because photography is an art, but when I write a note I always mention that it's a photograph ("Cover photograph by" or however the credit appears in the book) so people know it's not an illustration. I don't remember ever adding an author photograph as interior art, though; that's ridiculous. --Username 11:42, 11 August 2022 (EDT)

Applying awards to canonical author vs credited author

Are there explicit rules/help about which variant of a title an award win/nomination should be applied to, when the author listed by the awarding body is not the canonical author record? e.g.

Leviathan Wakes was a 2022 Locus Awards finalist. "James S. A. Corey" is listed in the official announcement, but the award record here lists the 2 real authors
Conversely, 2011 Hugo finalist Feed has the award attached to the Mira Grant pseudonym rather than Seanan McGuire.

Help:Screen:AddAward states "Author: This field will be pre-filled and not editable if you are adding or editing a title-based award. For untitled awards, find the canonical name of the person associated with the award in the ISFDB and enter it here. If the author is not in the database, then enter the name as specified in the award description. If the award was shared by multiple people, click on "Add Author" and enter as many names as needed." Perhaps I'm being overly pedantic in how I read that paragraph, but I think the sentences that mention "canonical name" and "shared by multiple people" all only apply to non-title awards, which isn't the case for the two examples I gave above. (Perhaps that section would benefit from being split into two paragraphs, and expanded for each where relevant?) ErsatzCulture 12:30, 12 August 2022 (EDT)

In absence of any opinions one way or another, I've added the 3 Dragon Award nominations that had triggered this question (Leviathan Wakes in SF, Nettle & Bone in Fantasy, Gallant in YA/MG). I've added the award entries to the pseudonym variants of the titles, and these do show up as you'd hope in the parent title and canonical author pages, so I don't see that there's a downside to adding the noms this way? (Possibly this would be wrong for gestalt entities like Adam Blade or Daisy Meadows, but I guess those aren't likely to trouble award lists...) ErsatzCulture 07:34, 16 August 2022 (EDT)

Sorry, I missed the question when it was posted. You are correct, the last three sentences of the quoted Help text only apply to untitled award. For title-based awards, the "Author" field is not editable, so it would be a moot point. We can split this paragraph in 2 if it helps make it more clear.

Re: attaching awards to VTs, yes, it's fully supported by the software. In some case it's a meaningful distinction, e.g. when the award is given to a translation, so it's the preferred way of doing things. Ahasuerus 11:04, 16 August 2022 (EDT)

I attach the award to the version it was given to (provided that it is not a typo and that version exists as a published book/part of a book) or to the canonical if the award site/announcement made up a weird title for the book. That way they are visible in both places and they are attached where one would expect them to be. Annie 18:43, 16 August 2022 (EDT)

Award inclusion policy

A recent discussion of the eligibility of the Helicon Award highlighted the fact that we don't have an "award inclusion policy". ISFDB:Policy has a lot to say about which works are included and which are excluded, but it's silent on the award eligibility issue. The only time it uses the word "award" is when it states that Web-only publications are eligible if they are "shortlisted for a major award", but the term "major award" is not defined.

With the recent proliferation of genre awards, it would be best to create an "Award Eligibility Policy" instead of grasping with the issue on a case by case basis. Let's start by listing the types of awards that we currently have on file:

Awards given by convention attendees, e.g. Hugo, Ditmar, Seiun
Awards given by bodies authorized or sponsored by conventions, e.g. Endeavour, WHC Grand Master, Utah SF, Imadjinn, Golden Duck
Awards given by professional organizations, including both awards voted on by all members and awards given by officials, committees or sponsored bodies, e.g. Nebula, Bradbury
Awards given or sponsored by fan or mixed fan-pro organizations, e.g. the British Science Fiction Award, DSFP, Prometheus
Awards given by panels of industry professionals (critics, authors and/or editors), e.g. Apollo, Kurd Lasswitz Prize
- A subset of the above authorized and/or organized by a single person as long as the panel consists of industry professionals, e.g. the early Heinlein awards, ABS-premiya
Awards given by major magazines, whether by their readers or by their editors, e.g. Locus, Deathrealm, Itogi goda, Analog
Awards sponsored by schools or colleges, e.g. Campbell Memorial, Schwartz
Awards given by the users of major online forums or of social bibliography sites, e.g. Goodreads, HOMer
Awards given by book clubs, e.g. SFBC
Awards given or sponsored by library associations, e.g. Newbery, Carnegie
Awards given based on an "open vote" online, e.g. Gemmell, Ignyte, Dragon

And here are the types of awards that we generally do not include:

Contests whose participants need to pay a fee in order to enter them
Promotional efforts by publishers or authors

Hopefully this covers everything that we currently include.

The big challenge that I see is distinguishing between "small fan/pro organizations" and "promotional efforts by publishers/authors" which may try to masquerade as legitimate "small organizations". In borderline cases the difference between the two comes down to the organizers' intent, which can be hard to discern. Defining "major magazines" may be another issue. Ahasuerus 18:55, 18 August 2022 (EDT)

Though the Dragon Awards are part of the open vote online. You don't have to have a membership for or attend Dragon Con to vote in them. ···日本穣 · 投稿 · Talk to Nihonjoe 10:38, 19 August 2022 (EDT)

Thanks, I have moved it to the last bullet point. Ahasuerus 13:33, 19 August 2022 (EDT)

It may be a bit more complicated than just the small organization/promotional effort distinction - there are also the random blogs (sometimes by 1 person, sometimes by multiples) who give their own "awards" (under that name or not). Recognizing and separating them from a legitimate small organization/club/whatever is not trivial either. I don't have a solution in that direction I am afraid - just thinking aloud. I suspect that this is where we may never find a way to codify the ROA - the small awards which can take years to get recognized as the correct category... Annie 15:36, 19 August 2022 (EDT)

I had thought about how you might define eligibility-for-inclusion rules in the other thread, but that was already more than long enough, so I didn't go into things there. My thoughts - from what I can recall a week later - were that it might be difficult to come up with a set of criteria that would be met by all awards, without being so vague/minimal as to be of no use. Rather, you might have some sort of checklist, where for every matching item, a score is incremented. This might not literally be a score that has to be met, but something where if you have, say, 10 items, and a proposed award only meets one or two of them, then it's probably not suitable. The sort of things that might be criteria could be:

1. Are the works considered for the award mostly or entirely speculative in nature? (i.e. exclude stuff like the Booker) ErsatzCulture 18:21, 22 August 2022 (EDT)

I don't think this would be a useful criterion. We support a number of awards which cover multiple (or all) genres -- Newbery, Carnegies, etc. Letting our users know which SF authors have won the Nobel Prize for Literature would be a good thing. Ahasuerus 07:31, 23 August 2022 (EDT)

OK, that's an oversight/ignorance on my part. However, is there any sort of threshold for covering general literature awards. e.g. Alan Garner's Treacle Walker is nominated - or longlisted? not sure - for this year's Booker, but it had never crossed my mind to request that award to be added, given that (I'm guessing) >90% of nominees are not genre works. ErsatzCulture 08:51, 23 August 2022 (EDT)

I don't think that should disqualify awards. 90% of Nobel Prize-winning authors are not "primarily SF authors", but it would be useful to show that Kazuo Ishiguro, Doris Lessing, Gabriel García Márquez and Selma Lagerlöf won it. Ahasuerus 08:09, 24 August 2022 (EDT)

2. Do the categories apply to works that are generally eligible for inclusion in ISFDB (i.e. not comics, film/TV etc)

(With some sort of proviso to cover the likes of Hugo, Nebula and Dragon, which have categories that are not really ISFDB-relevant, but where the focus and/or majority of categories are stuff of interest.) ErsatzCulture 18:21, 22 August 2022 (EDT)

One thing to note is that media-focused SF conventions have been more popular than book-focused SF conventions, at least in 21st century North America. If a gaming convention or an anime convention with 5,000 attendees decides to create a "Best Related Novel" award category, I don't think the fact that its primary focus is gaming/anime should affect our decision to create a new award type.

Also, awards that cover more than just SF, e.g. Goodreads, have -- for our purposes -- two types of categories: "SF" and "non-SF". We enter "everything SF" and cherry-pick non-SF categories like Mystery & Thriller, Romance and Fiction. The fact that Goodreads awards cover a lot of non-genre works that are not eligible for inclusion in ISFDB didn't prevent us from creating an award type. Ahasuerus 08:36, 23 August 2022 (EDT)

Agreed - the point I was trying get across (which I don't think I clearly stated) was that there's a difference between the likes of Hugo/Nebula/Dragon where there are media/comic/game categories that might be included here, but there are awards dedicated to those areas - e.g. Eisners for comics, Saturn for SF films - that I don't think ISFDB should cover, even if they have categories that are pretty much identical ones in the Hugo/Nebula/Dragons. If awards that fall into the latter group were to add new categories that are definitely ISFDB-relevant, then absolutely they should be added, but only those categories, with no obligation to cover their "core" categories. —The preceding unsigned comment was added by ErsatzCulture (talk • contribs) .

I see. Upon reflection, I think there are three dichotomies here:

Award types like the Hugos which primarily focus on speculative fiction (with an emphasis both on "speculative" and "fiction") vs. award types which focus on something else, whether it's non-speculative fiction like the Nobel Prize or speculative works that are not written fiction like the Saturn awards
Award categories which primarily focus on speculative fiction vs. award categories which primarily focus on something else
Awards for works of speculative fiction or directly related to speculative fiction (authors, editors, critics, publishers, etc) vs. awards which are not directly related to speculative fiction (comics, films, games, TV, etc)

As far as I know, the current data entry conventions are as follows:

For award types which primarily focus on written speculative fiction (Hugo etc), we enter all categories and all awards
For award types which do NOT primarily focus on written speculative fiction, we enter:
- All awards in categories which primarily focus on written speculative fiction, e.g. Goodreads' SF categories
- Any awards in other categories which are directly related to speculative fiction, e.g. Goodreads awards given to SF works in non-SF categories

This list of rules would probably look better as a matrix. Ahasuerus 11:41, 24 August 2022 (EDT)

3. Has the award been running for a number of years/cycles, or for new awards, are the organizers people with a track record/known quantity in the community/fandom?

4. Is there are well defined rationale/eligibility/definition for the award and/or its categories?

5. Is the award structured in a way that looks like the general understanding of an award (e.g. longlist/shortlist/nominees, from which a winner or winners are selected)?

6. Can the award be accurately represented within the ISFDB data model/web interface (e.g. I'm not sure that SPSFC or SPFBO can be, although I must confess I've never looked at them closely enough to properly understand how they work) ErsatzCulture 18:21, 22 August 2022 (EDT)

Different awards use different selection/nomination/elimination/voting mechanisms, but ultimately they produce lists of nominees/finalists and winners. That's what the SPSFC and the SPFBO do even though their mechanism may be somewhat unusual. There may be other recognizable awards out there that do not map well onto our award system, but I can't think of any at the moment. Ahasuerus 07:47, 23 August 2022 (EDT)

7. Does the award have an online presence? (If for no other reason than it's straightforward to check who the winners/nominees/etc are.)

8. Are there known/published contacts, who any questions/requests for clarifications could be directed to, should there be issues with the info that has been published?

9. Is there an award ceremony and/or physical/monetary prize?

10. Are the mechanics of how the award is run clear? (e.g. for a panel judged award, is it known who is on the panel, and who is responsible for selecting the panel? For a poll award, is it clear who the voter body is, the mechanism for how they can vote, how the votes are calculated?)

Like I said, I suspect a lot of awards would fail to tick every single one of those boxes, but I reckon most would get the majority of them. (NB: I'm thinking just in the context of current/ongoing/new awards; I think any awards from the past that might get retrospectively added to the database would be treated differently.) ErsatzCulture 18:21, 22 August 2022 (EDT)

The discussion above suggests that any "award inclusion" policy would need to have at least two parts. The first part would cover "eligibility of individual awards within genre and non-genre award types and award categories" -- see the bullet points under item 2 above where I describe what I believe to be our current de-facto policy.

The second part would cover "award type eligibility". My current thinking is that the previously posted list of currently existing award sponsor types (see the top of this section) could be converted into a policy. Something like:

"Included: Awards given, authorized or sponsored by:

Conventions
Professional organizations
Fan organizations
Mixed fan-pro organizations
Panels of industry professionals
Professional and semi-professional magazines
Book clubs
Schools and colleges
Library organizations
Major online forums and social bibliography sites

Awards given, authorized or sponsored by "exclusively online" organizations need to have credible evidence demonstrating that they have a body of active members."

Note the following language in the last sentence:

"exclusively online"
"credible evidence"
"a body of active members"

This is supposed to help us differentiate between legitimate organizations, promotional efforts, single-person "organizations" and micro-blogs. It's far from perfect, but it's the best that I have been able to come up with so far. Ahasuerus 14:13, 24 August 2022 (EDT)

Invalid, inappropriate, and non-ISBN ISBNs

While working a submission with an invalid ISBN of "1230000022319" (non-ISBN in this particular case -- serendipitously flagged because '9' would not be a valid check digit) I had the contributor tell me me that the Kobo source says this is the ISBN (it does) and that submissions for several other publication with similar ISBN numbers had been accepted. That's true, too; see this list. Note that the ones formatted with hyphens most likely passed the checksum test and were not flagged in the submissions.

I was going to tell the contributor that we do not record invalid ISBNs in the ISBN field and relegate them to the Notes field, but I find the help makes no mention of what to do with an invalid ISBN. In fact, other than the mention of ISSNs and the bit about trimming off an encoded price, it does not provide any guidance as to what values should or should not be put in this field. Should we allow/accept anything the source labels "ISBN", or should we have some restrictions? For the case here, ISBN-13s start with either 978 or 979, so we know these numbers starting with 123 are not ISBNs at all. A more common case is a number that looks like a proper ISBN but where the check digit tells us something is garbled. And then there's the case where the ISBN is completely valid but is for some other publication.

Should we have some rules in place? Should I just accept the submission as is? --MartyD 12:48, 3 September 2022 (EDT)

13-digit identifiers which start with anything that isn't "978" or "979" are not real ISBNs. Some bookstores may put them in their internal catalogs' "ISBN" fields, but that doesn't make them ISBNs for our purposes. The fact that our software accepts and formats them as if they were real ISBNs is a flaw that will need to be corrected.

Re: invalid 978 and 979 ISBNs, i.e. ISBNs whose checksum digit doesn't match the rest of the value, I believe the current standard is to enter them as stated in the book and then add a note. Ditto ISBNs that have been re-used by the publisher or otherwise misused. If Help doesn't make this process clear, then I believe we should update it to spell things out.

At one point we discussed implementing a software solution to the issue of invalid 978 and 979 ISBNs. Based on the outcome of that discussion FR 176, "Add a new field to pub records for corrected ISBNs", was created, but it hasn't been implemented yet. Ahasuerus 15:03, 3 September 2022 (EDT)

The software has been changed to make sure that all 13-digit ISBNs start with 978 or 979. Any that don't will:

generate yellow warnings on submission review pages
have a red "bad checksum" warning appear next to them on Publication pages
appear on nightly cleanup reports

As of this morning we have 95 13-digit "ISBNs" which do not start with "978" or "979". Some start with "977", which may be a typo; more research will be needed. Once we sort them out, we should be able to update Help re: 978/979 prefixes being required for 13-digit ISBNs. Ahasuerus (talk) 17:42, 11 September 2022 (EDT)

A brief follow-up on this: Someone fixed up many of the pubs originally having the false, leading "123" ISBNs, moving those to Catalog ID, so that is what I ended up doing with the submissions and one lingering pub that I found, along with adding a note about its source. We can always re-adjust later if we choose, but at least the treatment is consistent for now. --MartyD (talk) 08:27, 8 October 2022 (EDT)

Place of birth: Use of subdivisions for large cities

Template:AuthorFields:BirthPlace currently says:

Use the "City, Administrative division, Country" format

However, many existing records use "Queens, New York City" and "Brooklyn, New York City" even though Queens and Brooklyn are "boroughs" within New York City and not separate cities. Some even specify the subdivision within Queens/Brooklyn. Similarly, we have a number of author records which list the subdivision ("arrondissement") for authors born in Paris and in Tokyo.

I think a finer degree of granularity makes sense for large cities with millions of residents, although I suppose it's possible for the actual birth to take place in one subdivision and the mother's house to be in another. Still, given the fact that we already heave dozens, perhaps hundreds, of author records that record this information, we may want to add something like:

For large cities, the name of the city's district or subdivision may be entered before the name of the city, e.g. "Nakano, Tokyo, Japan".

to Template:AuthorFields:BirthPlace. Ahasuerus (talk) 21:22, 9 September 2022 (EDT)

I don't have any objection to codifying this, and I agree we should capture -- and capture consistently -- things like "Queens" and "Brooklyn". I worry about lack of guard rails, though. What's "large"? What makes something a district/subdivision that should be captured (or is ok to capture) vs. a district/subdivision that should not be captured? --MartyD (talk) 10:32, 10 September 2022 (EDT)

I don't see a need for prohibiting additional fine grained locations as part of the birthplace field. I'm assuming that we are not in danger of exceeding the size of that field in the database. So long as the data is accurate and documented, it shouldn't do any harm to place it in the field. If it is prohibited, it could be added in the general notes, but given that we have a defined field where such data can reside, I'd prefer we kept in the Birthplace field. --Ron ~ Rtrace^Talk 13:55, 11 September 2022 (EDT)

And regarding Tokyo (given in the example above), Tokyo is technically not a city, but a metropolis (which operates as a prefecture). It is made up of 23 special wards that operate as separate cities, as well as 26 additional cities, 5 towns, and 8 villages. So, if at all possible, one of those 62 municipalties should be included whenever indicating someone was born in Tokyo. The only time "Tokyo" was considered a city by itself was between 1889–1943, in which case it should be referred to as "Tokyo City". ···日本穣 · 投稿 · Talk to Nihonjoe 13:00, 12 September 2022 (EDT)

Another thing to consider is that the "City" part of "City, Administrative division, Country" is often inadequate for our purposes. We have authors who were born in towns, villages or other unincorporated areas. We even have a number of authors who were born on military bases. Sometimes they are technically part of a nearby city/town while other times they are in unincorporated territory, e.g. Elmendorf Air Force Base was unincorporated until it became a part of Anchorage in 1975. "City" doesn't do a very good job of covering the variety of possible scenarios. "Municipality" would be a better choice. Ahasuerus (talk) 10:45, 17 September 2022 (EDT)

(unindent) So, how about we start by changing "City" to "Municipality"? It seems like a basic non-controversial improvement, which may not address all of the issues raised above, but one step at a time. Ahasuerus (talk) 11:21, 19 September 2022 (EDT)

I agree that "municipality" would be more clear. ···日本穣 · 投稿 · Talk to Nihonjoe 12:53, 19 September 2022 (EDT)

If there are no objections, I will change "City" to "Municipality" tomorrow. Ahasuerus (talk) 10:55, 26 September 2022 (EDT)

Done. Ahasuerus (talk) 15:13, 27 September 2022 (EDT)

I added in a section on Tokyo, too, based on the information above. ···日本穣 · 投稿 · Talk to Nihonjoe 14:56, 11 October 2022 (EDT)

Transliteration

I was asked to fill in the transliteration field for some Swedish titles, for those miserable åäö characters. I was looking for a table for the correct mapping, but couldn't find any. I can think of a couple of ways of transliterate them, but I guess we want consistency? What is the real purpose for this? Is it for searching? --Spacecow (talk) 16:10, 17 September 2022 (EDT)

The primary reason to have "transliterated values" for names and titles is to give users a way to tell that, e.g., "蝸牛くも" can be approximated as "Katatsumuri Kumo" and "Кирилл Андреев" can be approximated as "Kirill Andreev". "åäö" and other Latin-derived letters are less confusing, but it's still nice to have them transliterated.

The secondary reason is, as you said, to facilitate searching.

Re: consistency, at one point we spent a couple of years trying to come up with a standard transliteration system. After reviewing a dizzying number of transliteration systems, alphabets and scripts, we gave up and allowed an unlimited number of transliterations for each field that supports it. That's why Help says:

If there is more than one possible Romanization, click the '+' button next to the field label and enter the other Romanized spellings of the title. You can click on the '+' button as many times as necessary.

Transliterate away! :-) Ahasuerus (talk) 17:39, 17 September 2022 (EDT)

Rules for Inclusion - can we be more specific about art?

The current policy regarding inclusion of art 'books' seems to be covered under the (3) non-fiction, presumably by allowing art works that appear on covers or interiors of published speculative fiction. It might also be covered under (4) authors above a threshold, assuming all of the covers and interior art counts as published works of or about speculative fiction. There are art books and art cards for well-known illustrators such as Frazetta. It would seem reasonable to allow calendars. Could the policy statement be tweaked to make it clearer that which such works are allowed or disallowed for artist-authors as opposed to the bias to literary-authors? An example is a Julie Bell calendar (having many cover illustrations) which does not happen to have any (known) cover art included in the collection. ../Doug H (talk) 10:03, 19 September 2022 (EDT)

Personally, I use the "can be plausibly linked to published speculative fiction" standard in ISFDB:Policy's "Included 3" when deciding which art works to submit on Fixer's behalf. There are tens of thousands of works of (more or less) speculative art that are not related to published SF; I don't think they should be included. Ahasuerus (talk) 11:19, 19 September 2022 (EDT)

So if the Julie Bell calendar had two images that were used for covers of existing ISFDB entries, it could be entered as non-fiction with the two images as interior art, each linked to the corresponding cover art? And the calendar's cover may not have been used for published speculative covers but could be included on the main entry? ../Doug H (talk) 11:48, 19 September 2022 (EDT)

Image File Size Limit

I noted that the upload page now states that the maximum file size for an uploaded image is 2 MB. I currently have my image software set to reduce images to 150 KB, which I believe was the file size limit prior to the system upgrade. Have we changed the policy to allow larger files, or, is this a setting in the Wiki software that should be adjusted to reflect the smaller size limit? Thanks. --Ron ~ Rtrace^Talk 07:28, 20 September 2022 (EDT)

It's an artifact of the MediaWiki upgrade; I don't believe we have changed our policy. Let me check the configuration files. Ahasuerus (talk) 07:58, 20 September 2022 (EDT)

Yup, the configuration files currently limit file size to 2MB. I have sent Al an email. Ahasuerus (talk) 08:15, 20 September 2022 (EDT)

The configuration settings have been changed to what they were on the old server, i.e. 200KB. Thanks for pointing out the discrepancy. Ahasuerus (talk) 11:07, 20 September 2022 (EDT)

Title Series Numbering

It's been brought to my attention that there is a cleanup report that lists duplicate title series numbers within the same series. While I understand the utility of such a report to catch errors where a title in a series is mistakenly numbered, I don't think this is necessarily incorrect in all instances. this template does not mention a prohibition against assigning the same number to multiple titles in a series. The series that is being questioned is Small Gods by Lee Moyer and Seanan McGuire. Lee does the artwork for each item with Seanan writing a story. The series main website is here. In this case, the artwork and the story are tightly coupled with the artwork providing the title of the story. I feel that both the artwork and the story are part of one series, which is how they are published. A solution that was suggested would be to split this into two series, but I feel they should be kept together. I would prefer a software solution that either allowed for an artwork title to share a series number with a fiction title. Alternatively, some cleanup reports have option to ignore items that the report finds, when such items are "false positives". Thoughts? --Ron ~ Rtrace^Talk 21:39, 27 September 2022 (EDT)

A few notes: As a user, I find it confusing to open the series and see the same title twice at the same number (even though we have the type mentioned on one of them), with the interior art being first in some cases (see number 144) and the story being first in others (I am not even sure if who is first is consistent). Especially considering how variants show up in series (with enough blank space to make sure that they are related to the title above them). So there is the user experience to be considered as well. Another option may be to put the art as a decimal number (144.1 for example) and add a note in the series explaining the numbering. That will have the added benefit of having the stories always first.

There was a request at one point to allow the sorting characters to work for series but it was determined that allowing duplicate numbers and then sorting behind the scenes will be confusing (or so I remember - in any case, the discussion went nowhere). Another option may be to implement something similar to what we have in the contents pages (moving art titles a bit to the right compared to text ones).

Either way, I really dislike the idea of allowing the ignore for this specific series unless we figure out how to solve the user experience issue. Annie (talk) 21:52, 27 September 2022 (EDT)

From a purely technical point of view, adding the ability to "ignore" titles would be easy. However, I agree with Annie that having 2 (or more) titles with the same series number is confusing. The "decimal number" solution proposed above (144.1) is the best I can think of without creating another series. Ahasuerus (talk) 22:46, 27 September 2022 (EDT)

Another alternative (which would create a new series) would be to make a "Small Gods (art)" series, and make it a subseries of the main series. An explanatory note could be added to both to explain why they are sorted that way. This would remove the confusion of two items with the same number in the series. ···日本穣 · 投稿 · Talk to Nihonjoe 12:29, 28 September 2022 (EDT)

Ignoring our database for a moment and looking at the publication of these items, would someone consider each story to be part of the series? I would say yes. Would someone consider each illustration (or icon to use the term from the website) to be part of the series? Again I'd say yes. Further, I believe that folks would consider both to be part of the same series. I would also suggest that the author and artist have assigned a single number to both story and artwork. I assert that this series, as presented, has two items for each number.

Back to the database: There is no technical reason that prevents duplicate numbers within a series. I believe that the way I have entered these items is consistent with how they were published.

Regarding the user experience for viewing the series, I believe that having a story and and icon for each number is appropriate because that is how it was published. There is value in having a consistent sort within each number and I could agree to implementing a piped sort as is done with page numbers. I would not want to introduce decimal numbers to force the sort as that already has a meaning within a series (i.e. 1.5 occurs in sequence between items 1 and 2) and that is not the case here. Artwork 1.1 does not occur between story 1 and story 2, rather, it goes with story 1. I will also note that making improving the user experience for display comes with the cost of making the user experience worse for data entry. I believe that I am the only editor to have entered these since the series was nominated for a best fanzine Hugo, which it won. I intend to keep entering these and I am aware of this discussion and can abide by what it decided here. However, if someone else were to pick up entry of these items, they will likely not understand what we come up with. Regardless, a piped sort is a compromise I can live with. I don't believe that we will see many more variants, if any. I'm only aware of the one form the World Fantasy Convention books (with two different titles). I know that Lee exhibited many of pieces of artwork at an earlier North American Discworld convention in Baltimore. I don't know if any were published in the program book. I hadn't entered it, and have been unable to remember where I filed it. The variants occur before Lee started working with Seanan on the project. --Ron ~ Rtrace^Talk 18:55, 28 September 2022 (EDT)

Re: the "piped sort" proposal, implementing it (as I currently understand it) would be non-trivial. Unlike page numbers, series numbers are currently implemented as two fields: an integer field which holds the digits to left of the decimal point and a "varchar" part which holds the digits to the right of the decimal point. Adding an optional pipe component to it and making sure that everything still works correctly would take some time.

Perhaps more importantly, while addressing the sorting issue, it would leave the issue of having multiple titles with the same series number outstanding. Looking at the way Small Gods is currently displayed, I find it more confusing that either a separate art-only sub-series or a decimal numbering scheme would be. Ahasuerus (talk) 16:28, 29 September 2022 (EDT)

I thought this discussion was in part to determine whether we want to prevent having multiple titles sharing the same number is a series. As noted above, the help template is silent on the subject. What would be the point of such a prohibition, aside from the user experience issues mentioned? I've looked for the previous discussion that Annie alluded to, but have been unable to find it. She also mentioned that "the discussion went nowhere" so I have to assume no consensus was reached. If the items in this series were not numbered, would there even be a question as to whether both belong in the same series? Excepting the single instance of artwork published prior to the current project, the artwork and story are always published together. The artwork almost always includes series number and the words "Small God". The stories use the artwork to present the title and number and I don't believe have ever been published otherwise. No argument has been made to suggest that the publishers of this series consider it to be separate series of artwork and stories or an inherent ordering of artwork after story. Unless I am convinced of that, I will continue to object to separate series or relegating one title type as occurring between the numbers of the other title type. I have an additional proposal for a compromise to fix the user experience display issue, though it may be more work. Perhaps each number in a series should be presented only once with each subsequent line in the display omitting the number. I believe that would make it more clear that the number in the series is shared between titles. A third level of sort by by title type could be added to make the ordering consistent within each number. The duplicate series numbers cleanup report is still useful to catch typos where a number is duplicated unintentionally. However, when it is intentional, as in this case, an ignore option should be added to remove it from the report. No argument has been made to suggest that the publishers of this series consider it to be separate series of artwork and stories or an inherent ordering of artwork after story. Unless I am convinced of that, I will continue to object to separate series or relegating one title type as occurring between the numbers of the other title type. --Ron ~ Rtrace^Talk 15:49, 1 October 2022 (EDT)

The DB has some constraints - like it or not, we need to work within them. Some may be easier to change, some may not be. Multi series is not trivial. If we could have two series per title, that would have been easily solvable by having an overarching series with all of them and separate art and stories series. As it is, anyone looking for just the stories will be hopelessly confused when opening the series as it is now and between the lack of order (which of the two shows first) and the duplicate numbers, the display looks sloppy and as if someone made a mistake (or a lit of them really). Maybe it does not look like that to you because you know what it represents and you came up with it but looking as someone who saw that for the first time when I was trying to clear the report, it looks extremely user unfriendly. Annie (talk) 16:35, 1 October 2022 (EDT)

But there is no database constraint here. Otherwise it would have been impossible to assign same number to two title records. We're discussing whether there should be a new policy constraint and nobody has offered a reason why there should. That leaves us with a subjective display issue for which I've offered multiple suggested compromises on how to address. If those suggestions are not feasible, then can someone else come up with a compromise that maintains the artwork and stories under the same number in the same series (as they are published) and solves the display concerns? By the way, the stories or the artwork can easily be viewed individually today by viewing the series within the author/artist bibliographies. Perhaps the addition of a note on the series page explaining that each number consists of one story and one artwork would prevent confusion. --Ron ~ Rtrace^Talk 17:42, 1 October 2022 (EDT)

So a user should know to go to a very long author page and scroll around so they can actually get to the series when we supposedly have a series page? If you cannot see how user unfriendly that is, I don’t know how to explain it better.

You suggested that use case which I don't think is a likely one. I would expect that most users would want to see everything in the series i.e. both stories and artwork without having to navigate between multiple sub-series. --Ron ~ Rtrace^Talk 22:01, 1 October 2022 (EDT)

And again - opening a series which has 2 entries on each number and the order of the two entries is not consistent looks sloppy and as if someone made a mistake - if I find that on a site, I will be looking a way to report that as a data or visualization error.

I'm not sure what you mean by "visualization error", but the data is absolutely correct. Again, we are modeling a series where each number in the series contains 1 story and 1 icon. When you look at this actual publication, do you think it is an error that they present a story with an illustration under a single number? Of course not. I'm not sure what it would be inconsistent with. We have series with gaps. We have series with a mixture of numbered and unumbered items. We have series that contain multiple different title types. We have series that have interstitial titles. Correct me if I'm wrong, but I think you believe the sloppiness is because of the lack of a sort order. I've suggested one way that things could be sorted and was advised that it is technically difficult. I subsequently suggested a different sort technique which has not been commented on. --Ron ~ Rtrace^Talk 22:01, 1 October 2022 (EDT)

When you cherry pick expressions and pull them out of context, you won't understand what they mean of course. The statement you are confused about starts with "if I find that on a site". Note the "a" and the "if". I am well aware that it is not a visualization issue on our site because the series works as designed. However, it is not designed to show duplicate numbers in a good way. And while my main concern is the lack of order, the duplicate numbers are also a problem in this display. Annie (talk) 19:01, 4 October 2022 (EDT)

Your quote is directly above my comment with complete context. I simply don't know what you mean by "visualization error". I have no opinion on whether there is a visualization error on our site because I don't understand what you mean my it.

Regarding your more specific issues with displaying duplicate numbers and the ordering, the last solution I proposed addresses both of those complaints. However rather than evaluate that proposed solution, you began a straw man argument that there are database constraints, when there clearly are not. --Ron ~ Rtrace^Talk 21:08, 4 October 2022 (EDT)

A note on the page won’t help much on this I expect.

I disagree. If it's well crafted why wouldn't it be helpful? --Ron ~ Rtrace^Talk 22:01, 1 October 2022 (EDT)

I keep checking to see if you added that note which you think will be helpful and I still do not see a note. Regardless of how this discussion ends, we will need a note anyway. It may need to be changed a bit if the structure changes but we still need it (in the very least explaining how the art and the stories are related). :) Annie (talk) 19:01, 4 October 2022 (EDT)

I had not considered adding a note until consensus is reached. However, I can add one. --Ron ~ Rtrace^Talk 21:08, 4 October 2022 (EDT)

I understand that you like the formatting you came up with but everyone who chimed into the discussion either called it confusing or proposed other options or both. That should be telling you something. Annie (talk) 19:52, 1 October 2022 (EDT)

I didn't come up with the formatting. It's simply how the series page is displayed. There are only 4 editors that have participated in this discussion. 2 have stated they find the current display confusing. 1 offered no opinion on the display but did offer a suggestion on how to address the stated concerns. While the display doesn't bother me as it does others, I agreed that it could be better. I keep offering suggestions on how to accomplish that. However, I am unwilling to alter the data model to that it doesn't match the publication. If you are familiar with the MVC (Model View Controller) approach to software design, this is a view problem and we should fix it there rather than altering the model to fit the desired view. --Ron ~ Rtrace^Talk 22:01, 1 October 2022 (EDT)

You did come up with it though - because when you chose how to enter the series, you were aware of how it is formatted on the page and decided that it works for this series. While a software solution is always a better idea in cases like this one, compromises need to be made in projects like ours where we have a single developer who only has that much time. I try to live in the real world so I am trying to find a solution that works for our DB at the moment, you seem to prefer to stay in the perfect land of unlimited resources. At this point, I am not entirely sure that you came to R&D for a discussion and options - you were looking for an approval of they way you build things. I've made my argument, nothing you had said had actually addressed the user experience issue. Annie (talk) 19:01, 4 October 2022 (EDT)

How can you possibly consider me responsible for how series are displayed. While I am a programmer, I don't know the language used in this project and have not worked on this site. I chose to enter the series in the way the series is published. Nobody has argued that the series does not consist of both stories and icons and as published, both share a number. I simply entered the series as published which is what we try to do whenever possible. There is no stated policy that prohibits multiple titles sharing the same number within a series. --Ron ~ Rtrace^Talk 21:08, 4 October 2022 (EDT)

Template:TitleFields:SeriesNum says:

If you know the order in which the titles in the series are supposed to be read, you can number them starting with 1.

My interpretation of this sentence has always been that duplicate series numbers are not allowed. After all, that's what "order" means. If multiple title records could share the same series number, they would no longer be "ordered". Consequently the cleanup report Series with Duplicate Numbers doesn't have the "ignore" functionality. Moreover, we have 46,484 series on file and, to the best of my knowledge, this is the first series that prompted a suggestion to allow duplicate series numbers, so it's a corner case at best.

Re: the "pipe" approach, there is a significant difference between page numbers and series numbers. Page numbers are arbitrary, e.g. they can be "145", "xxxix" or even "M5". They can also restart within a given pub. That's why we need a better way to order/sort them.

On the other hand, series numbers are already "ordering" numbers. It could be argued that "publication series numbers" are a better candidate for the introduction of a "true ordering" number since they are not necessarily "numbers", e.g. see Ace Double which has "numbers" like "D-277". It makes the ability to "Sort by series number" less than useful in a significant number of cases. Ahasuerus (talk) 16:46, 5 October 2022 (EDT)

I don't interpret that sentence to exclude duplicate numbers. Annie noted in a prior comment that this was discussed before and to her recollection the discussion went nowhere, so I would assume the question of whether to prohibit duplicates was unresolved. I don't see duplicates as inconstant with ordering. When duplicates occur, the order shouldn't matter. The two titles with the same number are intended to go together, as they are published that way. However, the ordering of each pair of titles in relation to all other pairs does matter and is enforced. I do agree that this situation hasn't occurred before, but it could occur again, if something else were published that way.

Just to clarify - the old discussion was for allowing piped sorting fir series, not for allowing duplicate numbers. It was mainly for series with no numbering where we may know the order anyway (and to allow the sorting of subseries and books at the same time as opposed to having all books and then all subseries). The only duplicate numbers we touched on in these discussion were when we have a book and a subseries with the same number (common scenario which is a non issue now but will be an issue if we consolidate subseries orders and books orders). I brought it up when we started talking about piped order again because it was clear back then that it is not trivial at all. Annie (talk) 19:06, 5 October 2022 (EDT)

You had previously mentioned the difficulty in implementing a piped sort with series numbers, and I offered a suggestion for a third level of sort by title type (after integer and decimal portions as you described). This would keep the ordering consistent within each series number which addresses one of the stated complaints with the view. I had also suggested suppressing the series number on all but the first instance and adding ignore in the cleanup report. I believe these changes would address all stated issues with the display with the possible exception of being able filter the view by title type. As I stated above, I don't think that is a likely use case. --Ron ~ Rtrace^Talk 18:50, 5 October 2022 (EDT)

I am not sure how easy adding extra sorting would be. Series and sub-series are handled by infinitely repeating recursive functions which pass 12 parameters -- including complex structures -- up and down the stack. The code is shared by the Bibliography page and the Series page, which behave somewhat differently. It is one of the more complex and fragile sections of the system. I'd have to take a closer look to see how much work it would take. Ahasuerus (talk) 20:05, 5 October 2022 (EDT)

It's interesting that you digressed into publication series numbers since those most certainly contain duplicate numbers, even when only numbers are used. All the same complaints about the display of title series would equally apply to the display of publication series. I'm not aware of anyone suggesting that the display of publication series be altered, much less suggesting that second printings be put into a separate publication series. --Ron ~ Rtrace^Talk 18:50, 5 October 2022 (EDT)

Pub series by definition will have repeated number because we record each printing and each format separately. The UI there supports both ordering per number and ordering per date. The fact that we have different views for both types of series had always been an indication for me that we expect them to behave differently. And part of it is what needs and does not need a secondary sorting order because of possible repetitions... Annie (talk) 19:10, 5 October 2022 (EDT)

In addition, as long as we enter non-numeric values in the "publication series number" field, they remain fundamentally different from regular series numbers. Non-numeric values like "D-227" can't be sorted numerically, so the current algorithm is an approximation. If we were to deem the ability to sort publication series pages by the "pub series number" value accurately important, we would need to capture each pub's actual order within the publication series. (For the record, I don't think it's a priority because publication series pages are sorted by date by default. The ability to sort by pub series number is secondary.)

In this respect, publication series numbers are similar to page numbers within publication records. They are both objective and we need to enter them as they appear in publications. If we want them to be sorted correctly, we need to have an additional, purely numeric, data element.

Regular series numbers are different because they are strictly numeric and because they are not "objective": a prequel to a trilogy may be entered as title #0, #0.5 or #4 depending on how different editions have treated it and on how our editors approached the ordering. Ahasuerus (talk) 19:56, 5 October 2022 (EDT)

I started this discussion at your request and have presented multiple options. I believe both of my suggestions have addressed the user experience (UX) issues that you have stated. While the first was stated to be too difficult to implement (because of existing sort code), the other has not be commented on. I believe all the other suggestions create different UX issues, whether it is to hide the constituent titles from one another, or to imply that the icons are ordered between the stories, to say nothing of the new data entry UX issues that would be introduced. I didn't seek approval for how I entered this series because it violates no stated policies. --Ron ~ Rtrace^Talk 21:08, 4 October 2022 (EDT)

Just to add a fifth voice.... The display doesn't bother me, so I had no real opinion, but I agree that doing any sort of modeling gymnastics to get different display results is a backward approach and should be avoided. And, really, the same can be said for relying on entering distorted/artificial data. I do have one broad constructive suggestion. We have some "jumble"-like issues elsewhere (author bibliographies, variants, covers, magazines) for which we have implemented some display and preference features to help. One or more similar mechanisms might help here, for example:

Display the data as multiple lists, separated by title type(s).
Provide an n-way filter restricting results displayed by title type(s), with "All" the default/current behavior.
Provide an alternate display, a la the "grid" view, with titles for a particular number grouped in that number's box, in some deterministic order.

These just for illustrating how we might think about it. I am not trying to advocate for anything. --MartyD (talk) 10:56, 4 October 2022 (EDT)

How is your first option different from having two series with an overarching parent series? The only difference is that it will require software changes while it can be achieved now by simply using nested series. :)

It may be a backward approach to try to use what we have but the DB is what it is. We are not designing a new app from scratch, neither we have unlimited development resources. We have only one developer and he can only get to as much. There are a lot of things that can be improved and done better but if we wait for that to happen, we won't get much work done. Plus considering that it is the only series that is such a jumble, implementing something which uses what we have now and then changing it when/if we have a better software solution kinda makes sense.

If the majority of users think that the series looks ok as is, then it can wait for when series have more flexibility (so far we are kinda split and usually when the community is split we tend to go with the usual practices of the site - which is not to have duplicate numbers). I still find it sloppy and as if someone did not do their data entries properly. And the series still does not have a note explaining why it looks this way. Annie (talk) 19:01, 4 October 2022 (EDT)

Webzines inclusion: Proposed extension of ROA

When we opened the ROA for all speculative fiction webzines, it was supposed to be a first step so we limited them only to speculative fiction ones and a few special cases outside of the pure speculative ones. We never expanded that definition so we have a discrepancy between paper/ebook and webzines treatment. I propose to close that gap completely. This will allow:

non-fiction genre webzines which publish interviews, reviews and articles which are mostly about and related to speculative fiction
non-genre fiction webzines which are not speculative but publish a speculative story occasionally (added per the rules for non-genre periodicals)
non-fiction non-genre webzines which publish speculative fiction related contents (reviews, interviews and so on) (added per the rules for non-genre periodicals)

That will bring the three formats in line with each other so if a publication changes from e- or print to web only, we won't lose it. We may want to cleanup the whole section (sections?) but just to kick start the process, the proposal is

REMOVE:
- Speculative fiction webzines, which are defined as online periodicals with distinct issues (note: online periodicals without distinct issues are not considered webzines)
- Special speculative fiction issues of non-genre webzines
ADD
- Webzines, which are defined as online periodicals with distinct issues (note: online periodicals without distinct issues are not considered webzines).

Yes, we can open it a bit more by dropping the issues requirements (thus allowing blogs and what's not) but that will open us to cataloging the internet so I think sticking to "if you do not publish issues, you are out" is good enough for now. We do have the precedent of "date-based" issues though so we may want to tighten the definition of an issue (so we do not end up with reviewers online adding their blogs). Alternatively, we can still keep the door closed for the non-fiction webzines and then the change will be to remove the same as above but add instead:

- Fiction webzines, which are defined as online periodicals with distinct issues (note: online periodicals without distinct issues are not considered webzines).

Thoughts? Concerns? Opposition? :) Annie (talk) 12:44, 5 October 2022 (EDT)

I support the simplification in defining webzines, as well as including allowing content as described in the three bullet points at the top. I don't see any benefit to excluding webzines that are generally non-fiction. That just complicates things. I'm all for keeping the ROA as simple as possible. ···日本穣 · 投稿 · Talk to Nihonjoe 19:40, 5 October 2022 (EDT)

I agree that simplifying and consolidating the rules would be advantageous. Having different rules for different formats is especially bad when "a publication changes from e- or print to web only [and we] lose it", as stated above.

I don't think we are quite ready to open the door to online publications which do not have distinct issues. Some Web novels are both popular and influential, e.g. Mother of Learning or Worm, which has over 11,000 (sic!) fan fics, but we would need to decide how we want to enter them first. Some have over 1,000 chapters and take many years to finish, which makes them a bad fit for the current data model. They are probably best discussed separately. (Luckily, many popular Web novels, including Mother of Learning, have been published as e-books recently.)

As far as the issue of non-fiction webzines goes, I am not really familiar with them, so I don't have a strong preference. From the consistency standpoint, it would be better to include them to make the inclusion rules uniform across all formats. Ahasuerus (talk) 20:22, 5 October 2022 (EDT)

The main group of non fiction webzines will be review sites and non-genre online magazines which publish reviews or interviews we want now and again. I don’t think that there will be too many of them (the distinct issues will exclude blogs and similar)

I suspect that there will be some conversations around the web sites of usually non-genre or non-fiction print and e-magazines - some such as Kenyan Review has a separate online only magazine named KR Online which has issues cleanly; some simply post stories/articles tied to the existing print issue (but which are not in the magazine). Most of these will fit under the issues definition - either tied to the print issue or with a totally different contents. When it does not, we can see what is coming in and change things again later to extend the definition to them specifically without opening for all the internet. Annie (talk) 21:30, 5 October 2022 (EDT)

I am an Inclusionist, so this proposal appeals to me. ROA consistency is to the benefit of both contributors and moderators. We should treat Webzines the way we treat print magazines. And despite my Inclusionist bias, I do think we should avoid random blogs, so having clear criteria that exclude them is important. The "distinct issues" bar is nice in its simplicity and is based on a publishing concept unique to Webzines/websites, so it does not introduce any inconsistency with treatment of printed material. --MartyD (talk) 06:37, 6 October 2022 (EDT)

With the growth of on-line publishing, standardizing the rules of acquisition appears desirable. However, I anticipate difficulties determining genre vs. non-genre in regard to poetry. Prose and even prose poetry are relatively straight forward. Verse, particularly, the very short forms, can be problematic. The non-genre issues I've entered are the result of a work being shortlisted for a speculative fiction poetry award. I actually prefer not entering the issue but adding a note to the title record created via the award anthology. With the suggested change, the onus shifts to us. Every work in a publication will require evaluation. All the non-genre works for above the threshold will also need to be entered. In theory, this improves the bibliographies of the above the line authors. In practice, there's potential to hurt our credibility. The current rules leave some works out that would be desirable to include. The change risks including many works which should not be. I agree with the recommended changes if we can exempt poetry webzines. This is not as silly as it appears. There are other poetry specific issues which also need be addressed. John Scifibones 10:34, 6 October 2022 (EDT)

How is that different from the current situation with paper and e-zines and books? If we are able to determine if they are genre there, we should be able to do it here the same way. If you are saying that we need to revise the rules around poetry, let's not mix that here. The proposed change (either of the two options I proposes) does not mention genre or non-genre at all, these are only in the explanation of what we will gain - we just call them webzines (or fiction webzines in the second option). Annie (talk) 10:51, 6 October 2022 (EDT)

I agree that every work is looked at under the present rules. However, there is a presumption the works are speculative in a speculative fiction periodical. Since all webzines will now be eligible, that will no longer be the case. This is the problem I foresee. John Scifibones 11:14, 6 October 2022 (EDT)

I am still not sure what the difference is -- it does not matter if it is a paper magazine, an e-zine or a webzine, we always first need to determine if it is speculative or not and for non-genre, we only add speculative contents. For periodicals, usually that is determined with the first added issue. If you are saying that the editors will now need to be as careful around adding/editing webzines as they are supposed to be when adding/editing e- and paper books and zines then that is the whole point - we are trying to stop treating webzines as the red-headed cousin of the -zines family. Allowing the webzines does not add an additional risk - if anything, they are easier than paper and e-version because you usually always have access to the webzine text and need to do less guess work on the status of the -zine.

I understand your concern that determining what is speculative in the poetry world can be harder and/or subjective but we already have the practice for that and we were always supposed to do that anyway. It has nothing to do with allowing all webzines or not - because we already allow them all if they happen to produce a Kindle or ePub or PDF (or print) version anyway. How do you solve that issue when you are entering an e-zine? It will be the same here. If you want to, open a separate discussion to discuss rules for poetry addition - but I still fail to see how allowing a new format changes anything in what we had always done. Annie (talk) 11:44, 6 October 2022 (EDT)

The change is one of scope. However, I remove the qualifier from my original statement of support. John Scifibones 12:24, 6 October 2022 (EDT)

Are you worried about us being overwhelmed by non-genre poetry submissions in webzines which moderators will need to deal with? Or simply that this will make a LOT more poetry webzines eligible, most of them eligible as non-genre periodicals? If the latter, the same is kinda true for fiction webzines - there are a lot of them out there (we don't even capture all of them in the e- and paper formats, we have no chance for all web-ones either). If we change the rules as discussed here, we can certainly keep an eye on what is coming in and if needed, open a poetry inclusion conversation - part of why we took the opening of the rules for webzines that slow to start with and did not allow the non-genre webzines at the start was exactly because we did not know what scope we are looking at. I just do not see the patterns of what we see coming in changing that radically. I know that you work a lot on poetry and that inclusion will add a lot to the scope of things you *can* add (a lot of poetry zines out there, mostly non speculative but with a speculative poem now and again) so I am really trying to figure out if we need a parallel poetry conversation somewhere. We will never be able to add everything eligible (don't even get me started on non-English titles) but this at least will allow us to capture more of what is out there... Speculative poetry is a bit hard to define - I tend to just go by how it sounds to me and if someone wants to disagree and add more from a non-genre periodical, it is up to them. :) Annie (talk) 12:54, 6 October 2022 (EDT)

Yes, to the first point. While I don't anticipate a huge influx of non-genre poetry webzines, the review process per issue will not be trivial. As you point out, we can discuss this elsewhere if needed. John Scifibones 16:16, 6 October 2022 (EDT)

Review of non-genre publications (magazines, anthologies or collections) is never trivial anyway, especially when working with a new(ish) editor. If we see an influx of these, we can always reopen the rules but short of restricting all poetry non-genre webzines, a solution is impossible. And this won't solve the issue because poetry shows up everywhere these days in webzines so short of building a separate set of rules for poetry, we are already in the subjective world anyway. I am more worried about non-English webzines than I am about the poetry ones to be honest (for similar reasons) - especially when we don't have another speaker of the language on the board. But then this is true for all submissions in exotic languages so adding the non-genre webzines won't add that much to it... all we can do is to try. :) Annie (talk) 17:06, 6 October 2022 (EDT)

(unindent) It's been two weeks and it looks like there are no objections to changing the Rules of Acquisition to bring webzines eligibility in line with paper and ebook pubs. Unless I hear otherwise, I plan to change the RoA tomorrow. Ahasuerus (talk) 12:49, 21 October 2022 (EDT)

Artwork Records for Non Published Artwork

There are a number of edits in the queue to add parent title to either COVERART or INTERIORART records to reflect an original painting that was used for the cover or illustration. Further, these records are not giving a publication date, but rather the creation date of the original work. Some edits of this sort have been approved (e.g. this title). I can't find an earlier discussion where we agreed to add such records. I was able to find this discussion from 2019. I don't believe that discussion resulted in agreement to add such titles. Personally, I find these records odd. If we are going to include them, I don't think either COVERART or INTERIORART are good choices for the title type. In the example, there is no book titled "The Garden of Earthly Delights (central panel)" for which the title record was published. If we are to include these titles, we should probably give them a new title type. I am also concerned as the title appears to be the English title for a Dutch work that from the Wikipedia article appears not have been given a title by the artist. Shouldn't that record be titled "untitiled"? Since there is no title page for the painting, how do we decide what the "published" title is? Do we ignore the name in the language of the artist and always use the English title as was done with the Bosch? I also find the use of creation date in these titles confusing since for the title types in question the date field indicates publication date for that combination of title variant, author variant and language. I don't support adding such records to the database, but if we do, there are many details that should be worked out. I think a better solution would be to give the creation date and canonical title of the painting in the notes of the canonical title for the artwork appearing in a book publication. In fact, in the example given, many of the variants are actually different details from small portions of the larger painting. It's odd to see so many variants that are completely different images. Do others have thoughts on this? --Ron ~ Rtrace^Talk 21:04, 18 October 2022 (EDT)

I agree with your thoughts in general. If we do include them, we should have a different TITLE type (maybe "ORIGINAL ARTWORK" or something?). Regarding the "The Garden of Earthly Delights", if we do decide to include them, that one shouldn't be listed as three separate works (left, right, and center panels). It's a single work as a whole. I also think the title used should be the original title in whichever language. ···日本穣 · 投稿 · Talk to Nihonjoe 13:06, 19 October 2022 (EDT)

I don't think that the original works belong in the DB at all. We catalog art only when it relates to speculative fiction or is published alongside speculative fiction or books about speculative fiction (or a book by an above threshold author). The original art is out of scope IMO. And most paintings got their titles a lot later than their creation anyway. So I think we should be using the notes for the original date and title (and other details) I think. If we decide to record them anyway, they will need to be in the language of the artist and not always in English and we will need a third art type (ORIGINALART or something) because these are not used as covers/illustrations at their assigned creation date...

PS: I think the practice started a few years ago as a compromise to stop some editors from messing up with COVERART and INTERIORART titles, dates and languages because of the "it is the same art, I am going to set it with the original title/date/language" argument. But it may have predated that particular argument - that was the first time I actually saw it happening... :) Annie (talk) 14:00, 19 October 2022 (EDT)

I'm in agreement that there is no need. Title notes are sufficient. -- JLaTondre (talk) 17:51, 19 October 2022 (EDT)

Since we're a publication driven database, I favor the following. The canonical title date is the earliest of either the oldest publication in, or eligible to be included, in the database. Variants only for canonical name or title differences. The artist's native language is ignored. All additional information in the title record's notes. John Scifibones 09:15, 20 October 2022 (EDT)

Are you proposing to ignore the language for art works altogether? So if an artwork called "2022" is published in 10 different languages, we use a single title? I am STRONGLY against masking where the title is published that way. Having a separate title per language allows you to see at a glance in what languages the artwork is used. Combining them together just because the title works in all languages makes that a very very hard query to run and makes the DB essentially unusable for that. Plus it will be inconsistent - if the characters of the title exist in both languages, the title will match; if they do not, it won't. So you will have variants for some languages and reused titles for others - making it even harder to see where something is published (and essentially impossible for someone who does not understand the DB - if they come to a title and see variants in Russian and Japanese (non-Latin languages so titles will almost always be different and thus require a variant), they won't know that there may also be German and French publications but we simply decided not to show them separately). If you did not mean that, never mind :) Annie (talk) 12:28, 20 October 2022 (EDT)

Sorry if I was misleading. The titles should be credited as they appear in each publication. Of course, the different language titles will be variants of whichever is the oldest. I'm merely stating that we don't create an artificial title in the artist's native language. John Scifibones 13:09, 20 October 2022 (EDT)

So you are in agreement that we should not have an overarching parent which represents the original work unless it was an illustration or cover to start with?

For the artist's languages - you probably should post a separate discussion so we do not muddle this one - we had been talking about it now and again, maybe time to start again? :) Although when you do, you may want to think about artists who also have non-art titles (nonfiction (art albums or actual non-fiction), essays, even fiction) - at which point do we start keeping a language for an author? Maybe what we really need is to redesign how we show art titles on author pages... That solves the artist/author conundrum. Annie (talk) 13:32, 20 October 2022 (EDT)

Yes, we agree John Scifibones 13:53, 20 October 2022 (EDT)

I'm in agreement too - no artificial original artwork parent record should be in the database. MagicUnk (talk) 14:32, 20 October 2022 (EDT)

I personally feel we've gone a bit off the deep end with regard to artwork documentation, but I disagree with the general sentiment here. I think it's useful to have a "canonical" record for collecting and grouping artwork appearing in multiple, otherwise unrelated publications. We have natural relationships/documentation where the same artwork is used to illustrate a translated work, where we cover art is reproduced as interior art, and where interior art is reproduced as cover art. But close cousins -- if not siblings -- of those are artwork not intended to illustrate a publication but later used as cover or interior art and artwork originally used to illustrate one work and used to illustrate unrelated works (as we see happen with covers in particular, from time to time). I don't see how it hurts to document something like that in a way that is easy for users of the ISFDB to find. --MartyD (talk) 15:33, 20 October 2022 (EDT)

I must admit I don't follow what you're trying to say in your last but one sentence. "artwork not intended to illustrate, but later used as cover" seems to be contradictory to me... However, as Annie said above, we're a publication database, not an art database. And, we already have a means to relate art with each other via the variant mechanism. MagicUnk (talk) 15:57, 20 October 2022 (EDT)