Rules and standards discussions

Jump to navigation Jump to search

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
Verification requests
Help with bibliographic, image credit, and other questions which require a physical check of the work in question.
• New post • Archives
Rules and standards
Discussions about the rules and standards, as well as questions about interpretation and application of those rules.
• New post • Rules changelog • Archives
Community Portal
General discussion about anything not covered by the more specialized noticeboards to the left.
• New post • Archives
Moderator noticeboard
Get the attention of moderators regarding submission questions.
• New post • Archives • Cancel submission
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.

Pages - help screens and templates

There are 5 screens of help and guidance for entering page values; NewPub, EditPub, PublicationFields, PubContentFields, How to determine the value for the "Pages" field in a book.

In the light of recent discussions I think it would be helpful if, at the top of each screen,, there could be four lines (one for each of the other four screens) which includes a link to same. At present, 3 of the screens have a link to the "How to..." page but it's right at the end. The "How to..." page has references and links to the PublicationFields template (twice) and the NewPub page. Admittedly 3 of the pages contain identical wording, but knowing of the existence of them all, whichever page one first lands on is what I'm addressing. Thanks, Kev. --BanjoKev (talk) 17:21, 12 August 2023 (EDT)

It might be good to combine all of the information from each of those pages and create one page that can be transcluded to all of those locations. That way, the information on all of them will be identical, and any changes to the one location for the information will be propagated to all of them. ···日本穣 · 投稿 · Talk to Nihonjoe 13:58, 27 September 2023 (EDT)
I think that's an excellent idea Joe. Kev. --BanjoKev (talk) 07:45, 5 October 2023 (EDT)

Does inclusion in the Hugo Award Voter Packet count as a publication?

Apologies if this is an old topic, although I think this particular case might be a new spin on it.

There are (at least) 2 Chinese stories in the Hugo Voter Packet that have English translations provided. They are in PDF and/or EPUB formats. The original Chinese stories and their publications were added to the database when the Hugo finalists were announced, so these translations would be alternate titles to existing records. (Exception: some of them are stories for the Astounding Award for Best New Writer finalists, which I didn't add anything for at the time, because it seemed too hard/nebulous.)

At least one of those translations is scheduled to be an anthology due out later this year, and another I'm 99% certain will appear in Galaxy's Edge magazine at some point, so it's not as if (some of) these translations will never get recorded in the database.

After reading ISFDB:Policy#Included, I'm still unsure as to their eligibility for inclusion here. Maybe they fall under "Convention programs, guides, etc. We definitely want any convention-published "real books", but probably not the ephemera.", but as that note is marked as "Debatable", it's not exactly helpful...

Thanks. ErsatzCulture (talk) 19:38, 20 August 2023 (EDT)

I had a discussion (beginning with the first response) with Annie last year about this. We were both leaning towards adding the Hugo packet as a publication. I had (and continue to have) other priorities that I'd rather work on. However, I would still support the Hugo packet as a single OMNIBUS publication published by the Worldcon for the year. --Ron ~ RtraceTalk 21:03, 20 August 2023 (EDT)
If my understanding is correct, "Hugo Voter Packets" are sent to all World Science Fiction Society (WSFS) members -- see and Anyone can become a WSFS member (and therefore a Hugo/Lodestar/Astounding voter) by paying $50 per year.
For most practical purposes this system is similar to book clubs, APAs and other organizations which limit circulation to their members. Since we include book club editions, fanzines, etc, it seems to make sense to include these "Hugo Voter Packets". Ahasuerus (talk) 09:55, 21 August 2023 (EDT)
Just for the record - I still think it should be eligible as an e-book omnibus. Annie (talk) 12:32, 21 August 2023 (EDT)
Thanks all, I'll try to make a start on this year's some time soon.
One follow up question: for stuff like custom submissions that contain multiple stories or essays, I think it's better to group those as new OMNIBUS, COLLECTION, ANTHOLOGY or NONFICTION titles, which then get pulled into the OMNIBUS, rather than just have all the individual SHORTFICTION, ESSAY, etc imported directly into the OMNIBUS.
e.g. this year's Best Editor (Short Form) for Sheree Renee Thomas comprises 14 PDFs, which are an issue of F&SF, a full anthology, and 12 individual stories and essays extracted from F&SF and a couple of anthologies. Rather than import those directly into the "Hugo Voter Packet" OMNIBUS publication, I propose to have a "Sheree Renee Thomas Hugo Award 2023 Voter Packet Submission" OMNIBUS containing those, which is then imported into the top level OMNIBUS. This (IMHO) keeps things more consistent and tidy with for example, the Neil Clarke submission, which is a single PDF anthology of 13 stories and an essay. Objections/thoughts? ErsatzCulture (talk) 17:29, 21 August 2023 (EDT)
I'll defer to the software experts, but I'm pretty sure that an OMNIBUS cannot contain another OMBNIBUS. --Ron ~ RtraceTalk 18:45, 21 August 2023 (EDT)
Not under the current rules, no... And I really would prefer not to change this -- we had a discussion around that when someone was adding the Baen disks - creating artificial containers that had never existed is going to look ordered on the surface but will be a pain for an end user - aka - in order to get the complete list for the packet, they will need to open multiple non-existing publications (as you will need a publication for these internal omnibuses if you want to import in them). So I'd just import all stories/articles/whatever into the single omnibus and use Notes to explain what is what (and use the numbering to keep the separate pieces next to each other). If the concern is where the award/nomination gets assigned - this is not different from when a set of books are nominated - just add it to each of the title records - for the example - she did not get nominated for an omnibus containing these works, she was nominated because of all the separate works... Although technically speaking, as it is a nomination for her and not the works, these should not get the nomination added to them anyway - but if there is something where that applies, the logic is the same. Annie (talk) 18:51, 21 August 2023 (EDT)
Ah, no problem, I'll just chuck everything in the "top-level" omnibus.
The thought of adding the award nomination to those hypothetical "fake" title records didn't actually occur to me ;-) I agree that awards to people rather than titles should be done as untitled awards. ErsatzCulture (talk) 19:01, 21 August 2023 (EDT)
If you look at the thread Ron linked above, I was wondering at the time between an overall omnibus and a series/pub series for the different pieces -- mainly due to the fact that parts of it are distributed separately. But it is a special case and a single omnibus makes more sense I think -- and makes it easier to see what is inside (plus as with all other omnibuses containing other containers (collections/anthologies), you will ultimately want to add ALL contents pieces in the top level anyway for visibility - aka for people who want to see where the story can be found - as we do not have "indirect" lists so having the fake middle ones will be mostly so you can have visual separation more than anything...). Plus if we ever change our mind, we can always create the smaller containers. Does not change the fact that we want all visible in the big omnibus anyway - which means importing all in it as well...
As for the awards note - yeah I realized it as soon as I typed it but then there may be other pieces in there for which that applies so I left it and added the last note). :) Annie (talk) 19:21, 21 August 2023 (EDT)

Interior art - do we use artwork captions in the titling?

That's one of the questions arising from this discussion about the artwork in Project Hail Mary. Clarification of the rules would be much appreciated. Thanks, Kev. --BanjoKev (talk) 20:34, 25 August 2023 (EDT)

My opinion: The spirit of artwork record titling is that, except when published as a "standalone" piece of art, artwork is subordinate to the work or publication with which it is associated. Artwork record titles generally reflect that subordination. Here is what I think is de facto practice:
  • COVERART titles should always be the same as that of the publication. (In fact, I think this is the one place we do not add disambiguation for the case of two different works of art by the same artist for different publications/editions with the same title.)
  • INTERIORART titles in a publication of, or about, artwork should record the "natural" labeling used in the publication. If works are identified by title or caption, that text should be used. If works are identified by use case, then either the canonical title with " (use case)" appended or a descriptive title should be used. For example, if a plate in publication XYZ is publication ABC's cover, title XYZ's INTERIORART record "ABC (cover)". If a COVERART record for ABC's cover is present, XYZ's INTERIORART record should be made a variant of that.
  • All other INTERIORART titles should usually be the same as that of the illustrated work, or of the containing publication if not illustrating a specific work. However, each of a publication's INTERIORART titles should be unique within the publication's contents. Where the use-the-publication-or-work's-title scheme would result in the publication's having multiple INTERIORART content records with the same title text, the titles should be disambiguated. Different disambiguation techniques are employed, depending on use case and information available.
    • If the same artist is responsible for multiple works of art that are being recorded separately, the title text for each must be made unique.
      • If the works have titles or captions, those may be used.
      • If the works have different use cases, append " (use case)" to one or more of the otherwise ambiguous records. E.g. "ABC (map)".
      • If no better differentiator is available, append " [number]" to each of the otherwise ambiguous records. E.g., "ABC [1]", "ABC [2]",...
    • If different artists are responsible for different pieces of art, the normal titling scheme is followed, with each INTERIORART record having the same title text but different Artist credits. Note that "use case" disambiguation may also be employed in this case. E.g., "ABC (maps)" by artist 1 and "ABC (illustrations)" by artist 2. If differing artist credit alone is not sufficient to produce uniquely identifiable records, then one of the disambiguation schemes should be applied first to produce the title text, then the appropriate artist credit should be assigned. E.g., "ABC [1]" by artist 1, "ABC [2]" by artist 2, "ABC [3]" by artist 1.
As I said, that is my opinion. I would also note that ISFDB's view of artwork has changed over the years. We used to treat artwork as much more of an afterthought/second-class data citizen than we do today. So, for example, you will see disambiguated-by-number records entered long ago where today we would use some more readily identifiable form of disambiguation. Or older single publication-wide records where today we would tend to use multiple records to document each of the individual works. Some of the help text may not be fully in tune with the times. --MartyD (talk) 07:20, 26 August 2023 (EDT)
re "I think this is the one place we do not add disambiguation for the case of two different works of art by the same artist for different publications/editions with the same title": Cover art is not a special case. We only disambiguate artwork titles within the same publication, not across publications. I agree with you on the remainder. -- JLaTondre (talk) 08:20, 26 August 2023 (EDT)
If I read Marty's reply correctly, what it boils down to is that for the art's title, the illustrated work's title is used with all the disambiguation cases etc, as explained above (and except for the bullet point 'If the works have titles or captions, those may be used [to make them unique]' - which I don't read in the current rules btw).
My interpretation of the rules is exactly that, ie. the title of INTERIORART is the same as the title of the work it illustrates - even though there are several examples currently in the DB where the actual INTERIORART title or caption are used as title, instead of the title of the work the art illustrates. The issue that I'm having with the current rules is that they are not very clear in explaining what title to use, hence should be rewritten to make them unambiguous - because right now, the rules do not clarify what do to in case there's artwork that has a proper title of its own. - cfr. the discussion here. I have two proposals to make the rules clearer:
* INTERIORART always get the title of the work it's illustrating. If the work does not illustrate any particular work, use the title of the publication the art appears in, or
* If INTERIORART has its own title or caption, use that title or caption. Else, use the publication's title instead
(+ the disambiguation cases laid out by Marty above, of course). Thoughts? MagicUnk (talk) 12:03, 28 August 2023 (EDT)
As has been noted by others, if the interior art has a caption, use that for the title. Otherwise, it should be using the title of the work plus a disambiguator as noted above. ···日本穣 · 投稿 · Talk to Nihonjoe 13:15, 28 August 2023 (EDT)
Yes, but that's not what the current rules say. Do we agree to amend the rules to make it clear that the caption should be used if there is one, and the title of the work in all other cases? (we may want to refine for artwork publications). Regards, MagicUnk (talk) 11:46, 29 August 2023 (EDT)
It should be optional, not a requirement. Same as it is optional to enter individual titles or leave it as one record for the entire pub. -- JLaTondre (talk) 18:25, 29 August 2023 (EDT)
Works for me. Anyone else who'd like to chime in? I'll try to come up with an update for the rules text to clarify that INTERIORART gets the title of the work it illustrates, and if there's a caption, that caption can be used instead. MagicUnk (talk) 05:10, 30 August 2023 (EDT)

(unindent) If "... INTERIORART gets the title of the work it illustrates" means the publication title, then I object. It would make my favored approach outside standards. The title record Winds of the Forelands (maps) covers all the maps used in a series. It clearly shows how the maps are credited, where they appear and is easily edited if additional volumes are published. John Scifibones 07:44, 30 August 2023 (EDT)

I agree with John. It's important to be able to use one record for the same illustrations (maps in particular) used in a series. Sometimes the illustrations don't have a caption or there are several possible captions. A grouping title can provide a container that clarifies the use of the illustrations without unnecessarily duplicating them. The approach being discussed doesn't seem to provide for the flexibility to use a grouping title. It also feels like the proposed approach could inflate the number of works attributed to a given artist. Phil (talk) 08:20, 30 August 2023 (EDT)
The rules currently state that artwork is only supposed to be titled per the title of the work (story or publication). The above is relaxing that rule to match how things generally are done. I'm fine adding an additional relaxation for "series" artwork as I agree combining maps makes sense. But if you are both objecting to any change, then you should realize your way of handling maps is not valid per the current rules. -- JLaTondre (talk) 17:11, 30 August 2023 (EDT)
The current standard for Maps - "Maps. These are considered interior art for ISFDB purposes and are typed as INTERIORART. The format for titling maps is "Title of Work (map)", for example: Brightness Reef (map). Optionally, if a map is titled you can use the stated title of the map without appending the name of the work, for example The Land of Nehwon (map)." (emphasis added) I interpret work as inclusive (publication, series, or story).
Note the wording in MagicUnk's proposal - "INTERIORART always get the title of the work it's illustrating. If the work does not illustrate any particular work, use the title of the publication the art appears in..." (emphasis added) The change from work to publication was the source of my objection.
If the original intent was for work to be synonymous with publication and story only, then I am indeed proposing a change. John Scifibones 19:05, 30 August 2023 (EDT)
Concerning illustrations (eg maps) repeated in multiple volumes (of a series), under the current rules there is always the possibility to variant titles. That will effectively tie them together - under the current rules there's no need to 'invent' a common title for use across a series.
Mind that I'm not saying that we can't change the rules, but the change John's (and Phil's) proposing requires more discussion before (if) we can accept the change and can update the rules accordingly. What do we do with INTERIORART that has
  • a caption, artwork identical, and that caption is identical across the volumes of the series --> this is an easy one; use the caption. Will need a rules change, but per the discussion above I'm fairly certain everyone's OK with adding 'if it has a caption, you have the option to use it'
  • a caption, artwork identical, but captions differ between volumes? --> since we'd make the use of the caption optional, we could decide to either use the series' title instead, or go the variant route, using the different caption titles (this latter would be my preference, as that's common practice for variant work titles anyway)
  • no caption, artwork identical, --> either use the title of the work it illustrates and variant per the other volumes, or, use the series title instead
  • combination of the above - might not be common, but can't be excluded either imo
and then I've not even touched John's example: how to write down the conditions to cover this case where there's a grouping of different maps involved, which are not identical across volumes?
Note that using the series title has its own challenges: what with series titles that change over the years? Are we going to go back and update all INTERIORART titles that were based on the old, no longer applicable, series title? What with series titles that we've "invented"? Those that are not to be found on or in the publication? Is using these "invented" titles for INTERIORART a good idea?
Lastly, we're now having two topics to discuss: "optional usage of caption", "usage of series title". What do you say, split the discussion in two sub-discussions? (splitting would allow us to update the rules to at least allow usage of captions...) MagicUnk (talk) 05:44, 31 August 2023 (EDT)
Splitting it seems reasonable. Phil (talk) 22:06, 31 August 2023 (EDT)

Numbering of pages numbered in the ToC but not numbered themselves

Please go read this discussion for background. Please keep comments here, though, since this discussion will be referred to regarding any outcome.

Here's the summary: For pages prior to the main content, we generally use the numbering found on the pages themselves (this is the same for all other content, too). In some cases, those pages don't have any numbering on the pages themselves. For those, we generally include the number of those pages in square brackets prior to the main page count. For example: "[12]+374" for a book that has 12 unnumbered pages of recordable content (maps, introductions, etc.) prior to the main content. In the case linked above, the table of contents gives Roman numerals to that content, so I used that in the numbering ("[x]+690+[3]") and included a note to that effect in the notes for the publication: "Although no roman numerals are printed on any pages, the Contents page lists Maps beginning on page viii."

The question is whether using the Roman numerals is what should be done here (and in other such cases). On the Help:Screen:NewPub page, it states "Caution: Do not use the table of contents to determine the page numbers of a publication's contents." My understanding of this is that it's meant to prevent us from using the table of contents page numbers when they disagree with the actual page numbers (basically, when the publisher forgets to update the table of contents when a change is made that affects the page numbers).

However, I don't think it should be applied in this case since it's the reverse of what I believe the intention of that rule is. In this case, the pages themselves don't have any page numbers on them. Rather, the only place the page numbers are given is in the table of contents. Because of this, there's no disagreement between the actual page numbers (since there aren't any) and the table of contents.

So, let's sort this out. Should we completely ignore page numbers in the table of contents in all cases? Are there cases (like the one described above and at that link) where we should use the information found in the table of contents? Is there something else that should be done?

Thanks for your input on this discussion. ···日本穣 · 投稿 · Talk to Nihonjoe 19:02, 17 October 2023 (EDT)

If I am reading this correctly, you are thinking that where Help:Screen:NewPub says:
  • Caution: Do not use the table of contents to determine the page numbers of a publication's contents
it was actually originally meant to say something like:
  • Caution: When a page number in the table of contents contradicts the page number in the body of the publication, use the page number in the body of the publication
Or, perhaps:
  • Caution: If a Contents item doesn't have a page number within the body of the publication but has a page number in the table of contents, enter the latter in the Page Number field and put square brackets around the value
? Ahasuerus (talk) 21:58, 17 October 2023 (EDT)
I think the intent of it was the first one, as that's how I've always seen it applied in the past. ···日本穣 · 投稿 · Talk to Nihonjoe 13:00, 18 October 2023 (EDT)
The thread title misstates the fundamental problem. At question is the proper handling of unnumbered pages before page 1 which contain indexable content. Proper determination of the Pages field in the publication metadata is the source of contention. I maintain that this situation is addressed in bullet point 3, under Pages, here. User:Nihonjoe argues that Arabic numerals are not required and Roman numerals may be used instead. I see nothing in the help which allows this. The help specifically calls for Arabic numerals. The proper entry for the page field of each content title flows directly from the publication Pages field.
If we decide that Roman numerals are appropriate, bullet points 2 and 3 will need to be completely rewritten. Of course I will support any consensus decision. John Scifibones 18:06, 18 October 2023 (EDT)
Sorry if it was confusing for you, but that wasn't my intent. Perhaps the title of this discussion isn't as clear as you would prefer, but the post itself is very clear. I was trying to be concise as really long section titles can be cumbersome.
Regarding the rest of your comment, it really depends on the definition of "unnumbered" since I'm arguing that the ToC does number the pages since it has page numbers and the pages themselves do not. We need to determine if the ToC can absolutely never be used for any page numbers, or if (as I'm arguing in this case) it can be used for those page numbers when the ToC has them but the pages do not have them and the page numbers cannot be derived from surrounding pages that do have page numbers. ···日本穣 · 投稿 · Talk to Nihonjoe 18:51, 18 October 2023 (EDT)
I think that using the numbers from the table of contents, with a note stating so, makes more sense in this case than inventing new numbers and discarding information printed in the book. I've always read this part of the help in the same way as you - it is there to define what to use when the actual book and the contents page disagree not to prohibit using the TOC when it is the only source.
With this being said, I can see the other side of the argument (for consistency sake if nothing else) - but my gut feeling is to go with what is printed in the book itself. Annie (talk) 20:08, 18 October 2023 (EDT)
(after edit conflict) I am in Annie's camp. I don't have strong feelings about this, other than I think from a database user's perspective, it would be somewhat strange to have content listed as on "[7]" when the TOC says it is on "v". My inclination is to adjust the "Caution" wording slightly to say that page numbers should be taken from the numbers printed on each content item's page, not from the TOC. Then in the "Pages without a printed page number" section add a bullet stating that if the page is given a number in the TOC, that number should be treated as if printed on the page, as long as not in conflict with numbering printed on other pages or with the number of physical pages in the publication. Something like that. That should be compatible with the other rules, page count determinations, etc. --MartyD (talk) 20:21, 18 October 2023 (EDT)
My only real problem with using a Roman numeral found only in the ToC is that if a reader were to pick up the book, look at the ToC, and try to go to that page, they couldn't find it using the page reference. No matter what, there definitely needs to be note describing the situation. More than anything, I would just like a well-stated, clear rule to apply. Phil (talk) 21:34, 18 October 2023 (EDT)
I definitely agree. Having a note in these cases is very important. Having a clear and concise guideline is as well. ···日本穣 · 投稿 · Talk to Nihonjoe 12:29, 19 October 2023 (EDT)

(unintend) Let me clarify a couple of things. We are currently discussing Unnumbered pages within a range of unnumbered pages scenarios. Help currently says:

  • If a content starts on an unnumbered page within a range of unnumbered pages, its page number should first be derived and then entered in squared brackets. The page number can be derived by counting forward from the first page of the section of unnumbered pages. For example, if a content appears on the fifth page in a range of unnumbered pages, enter "[5]".

If I understand it correctly, the proposal under consideration would add a sub-rule after the second sentence, something like:

  • If the table of contents specifies the page number where the content starts AND that page number matches the number derived by counting forward, then use the numerals (i.e. Arabic or Roman) found in the table of contents. If the page number in the table of contents doesn't match the number derived by counting forward, then use the number derived and Arabic numerals.

The caveat after the capitalized "AND" above would be presumably needed to account for situations where the page number in the table of contents doesn't match the number derived by counting forward since we all know how bad tables of contents can be (my "favorite" example is here.)

Am I reading this correctly? Also, will this affect Unnumbered pages within a range of numbered pages scenarios which are covered by a separate Help paragraph? Ahasuerus (talk) 18:12, 20 October 2023 (EDT)

Very close to an edit conflict with Ahasuerus.
Ahasuerus: Your understanding of the discussion re: Unnumbered pages within a range of unnumbered pages is correct. The situation of Unnumbered pages within a range of numbered pages has not yet been considered.
What follows below is what I had prepared to say before Ahasuerus jumped in first. :-) Teallach (talk) 18:58, 20 October 2023 (EDT)
The ISFDb rules already have a method for assigning page numbers to unnumbered pages that are not derivable by counting forwards / backwards, namely, the use of Arabic numerals in square brackets. So we don't need to resort to a secondary source for the page number. The way Pages are denoted in the ISFDb is already horrendously complicated and if we adopt the use of Roman numeral page numbers from the ToC where no number is printed on the actual page then we introduce further complications and also open other cans of worms. Examples:
1) Should the Roman numeral be enclosed in square brackets? This is currently not supported in the ISFDb rules.
2) Suppose a map is on an unnumbered page that is derivable by counting backwards (page 4, say) but the ToC lists it on page iv? What do we do? [Ahasuerus' proposed sub-rule addresses this case]
3) Suppose there is an article on an unnumbered page that is not recordable in the Contents section but the ToC lists it with a Roman numeral page number? What do we do?
If we use page numbers from the ToC then all the consequences and implications need to be considered and documented.
I am in favour of not using page numbers from the ToC where no number is printed on the actual page.
Whichever way this goes:
i) the Help notes need updating to clarify what to do
ii) a pub note definitely needs to be added to explain the discrepancy and the Help notes should state this. Teallach (talk) 18:59, 20 October 2023 (EDT)
Here are a few questions using the publication which caused me to raise this issue, The Bavarian Crisis. Pages: '[x]+690+[3]'. L-O-C '690' pages
  • Is anyone else concerned that the Pages field will differ from all secondary sources? (L-O-C in the above example). When we use bracketed Arabic numerals it's an obvious ISFDB construct.
  • Looking at my copy, viii is the only Roman numeral in the TOC. I assume [x], brackets addressed by Teallach, is a count of the total pages before page 1. This differs from how we presently deal with Roman numerals. Should the Pages field be 'viii+690+[3] or would that be another explanation in the help section?
  • I repeat for emphasis Teallach's point 3.
  • The Pages field will become impossible for a reviewer to confirm unless they own the publication or there is a scan available. John Scifibones 14:12, 21 October 2023 (EDT)
Regarding each point:
  • Our page counts already often differ from those at many secondary sources. Whether the bracketed numerals are Arabic or Roman doesn't make our way of listing page numbers any less an "obvious ISFDB construct". There are a number of things we do here which can be confusing to people outside of ISFDB (the whole CHAPBOOK thing, for example). In this case, the only reason I put the Roman numerals in brackets was because the pages themselves are not numbered, and we'd do the same thing if they were completely unnumbered (meaning no mention of page numbers in the ToC or on the pages themselves).
  • The [x] is the total number of unnumbered valid content pages, derived from counting forward and backward from the one page number mentioned in the ToC for the pre-story content. Since the pages themselves didn't have any actual page numbers on them, but the page number for one of the pages was listed in the ToC, I used that.
  • I don't really understand what Teallach means by "Suppose there is an article on an unnumbered page that is not recordable in the Contents section but the ToC lists it with a Roman numeral page number? What do we do?" If the content is not recordable, then we don't include the content, regardless of whether it appears in the ToC or not, and regardless of whether it has page numbers or not. We do include the page numbers, however (for example, if there's an "Acknowledgements" or an "About the Author", and the pages were numbered, we'd include them in the page count but wouldn't record the content as a separate title. I would also include a note explaining the situation.
  • Unless a reviewer has a copy of the publication (whether physical or a PDF or scan of the publication in question), they wouldn't be able to confirm anything anyway. Maybe I'm misunderstanding this concern, but it seems like a non-concern from how I'm reading it. ···日本穣 · 投稿 · Talk to Nihonjoe 15:24, 23 October 2023 (EDT)
This pending edit,, relates to this discussion. Is the way I entered numbers the way it's been decided they're supposed to be done? Because it does mention "179" on contents page. --Username (talk) 00:49, 24 October 2023 (EDT)
Nihonjoe: here is an example to clarify my point 3).
The text of a novel starts on a page with a printed number of 1 and finishes on a page with a printed number of 999. There are ten unnumbered pages in the book before the start of the novel. A one page "About the Author" article appears on the fifth of these pages. The ToC lists the "About the Author" article and assigns it a page number of v.
Now, we don't record the "About the Author" article in the Contents section but what do we put in the publication Pages field? The possibilities seem to be 999 or v+999 or [v]+999 Teallach (talk) 18:54, 24 October 2023 (EDT)
I'd do either v+999 or [v]+999 (depending on if we want to count the ToC assigning a page number as "numbered" or "unnumbered"), unless the "About the Author" is multiple pages, and then I'd extend the Roman numeral count accordingly. In your example, I'm assuming there is no other content, recordable or otherwise, outside of the "About the Author" section? ···日本穣 · 投稿 · Talk to Nihonjoe 19:39, 24 October 2023 (EDT)
I was not looking for a solution to the example. I just provided it to clarify my case 3 because you said you did not really understand it. At this stage of the proceedings I do not consider it appropriate to start working solutions to the three cases I raised, firstly because they will not be relevant if the consensus is that we do not use page numbers from the ToC where the pages are not numbered themselves and secondly because we risk losing focus on the main issue. The existing rules for Pages are already very complicated. If we do use page numbers from the ToC where the pages are not numbered themselves then all those cases I described (plus possibly others that I and other editors / moderators have not thought of or raised yet) will need to be discussed, agreed upon and have additional rules added to the Help Notes on Pages to deal with them. This will make the rules for Pages even more complicated. I am very much against doing this unless it is necessary because the more complicated the rules are, the easier it is for editors and moderators to make mistakes. In this situation, it is not necessary. In my opinion, it's not even desirable. If we decide to not use page numbers from the ToC where the pages are not numbered themselves then we just need to add one sentence to this effect to the Help Notes and we are done. Teallach (talk) 18:41, 25 October 2023 (EDT)
I don't think it's a major change either way. Printed page numbering directs how we record the page number and the count of pages in the block where the numbered page appears. For pages with no numbers, either we always count and always use Arabic numerals, or we allow pages to be considered numbered by proxy via the TOC first, before defaulting to the counting + Arabic numeral scheme. Use of the TOC, however, would need some kind of caveat to cover the case where a TOC is reprinted from a different format edition without adjustment and does not match the layout (similar to copyright page/printing statement handling). --MartyD (talk) 06:14, 26 October 2023 (EDT)

Other Missing Values on the Title Page

It occurs to me that the "no page number on the title page" is related to other "missing values on the title page" scenarios.

What do we do if a story or an essay doesn't have a title printed on the title page, but the information appears elsewhere within the publication, e.g. in the table of contents? Help:Screen:NewPub#Regular_Titles says:

  • For short stories, essays and poems, when working from a primary source, always take the title from the heading on the page where the work begins. The title shown in/on the table of contents, running page headers, index, front cover of the publication, secondary bibliography, or a promotional website listing is secondary.

However, what does "secondary" mean in this case? Does it mean that we can use "secondary" titles if no title is given on the title page? If so, then we should spell it out and also explain the hierarchy of "fallback scenarios", e.g. whether the version in the "running page header" should be used before the version in the table of contents.

Similarly, what do we do if a story or an essay has no author credit? In most cases we use "uncredited", but Help:Screen:NewPub#Regular_Titles allows an exception:

  • If an individual work doesn't have an author credit, which is common in single-author collections, use the form of the author's name stated on the publication's main title page.

Essays whose authors sign their names at the end -- as opposed to on the title page -- are another de facto exception since we typically enter the signed names in the "Author(s)" field.

These scenarios are similar to "missing page number" scenarios in that they provide alternative values -- sometimes documented in Help and sometimes undocumented -- that editors use to populate "Title" and "Author(s)" fields. I am thinking that we should start by clarifying the current rules and bringing then up to date before we start changing the rules for page numbers. For authors, it could be something like:

  • For Content entries, the order of locations to take author names from is:
    1. The title page if author name(s) are present
    2. The last page of the content item if signed by the author(s)
    3. For single-author collections only, the publication's main title page
    4. If none of the locations listed above list author name(s), enter "uncredited"

For titles, we will also want to clarify where the pub's main title should come from if the pub has no title page, which is increasingly common with independently published books. I have been using what's printed on the cover, but we really need to spell out what the hierarchy should be.

Once we clearly document the current de facto standard for titles and authors, it should be easier to decide what to do with page numbers. Ahasuerus (talk) 15:00, 28 October 2023 (EDT)

Don't forget TOC :). Not to start down a rat hole, but I believe we also don't follow strict order once the preferred location fails to provide a value. E.g., if TOC used one name and last page used another, and one was canonical, we'd likely use that. Anyway, it also sounds like we need to distinguish the "secondary" that is from-the-pub-but-not-in-the-official-place from "secondary" that is from-somewhere-other-than-the-pub. Perhaps "fallback" for the former? --MartyD (talk) 22:12, 30 October 2023 (EDT)
I have run a few database searches and it looks like we use the following values for works without a title:
So a lot of different scenarios, all of them revolving around the use of "untitled". I don't think we have this de facto standard documented anywhere, do we? Ahasuerus (talk) 17:34, 1 November 2023 (EDT)
I believe for poems we also sometimes use the first line, or portion thereof, in quotes (without "untitled"). I believe I have done it, and I don't recall from where I got the practice. Of course, I believe lots of things.... --MartyD (talk) 11:06, 7 November 2023 (EST)
Relating to the third item in the listing of the order of locations to take author names from: I think it would be meaningful to also add novels to this item (to use the publication's main title page), in case there are forewords, prefaces, notes worthy to add, all of which are unsigned but obviously written by the author(s) of the novel. Stonecreek (talk) 05:34, 31 October 2023 (EDT)

Kindle Vella - In or Out?

We have two previous discussions I can find (this one and this one), neither of which seemed to come to any conclusion. Do we want to include them as ebooks, or do they not count as ebooks since they can only be viewed within the Kindle app or on an actual Kindle device? Would they be considered serials? They seem to be a bit outside the norm for what we accept here. ···日本穣 · 投稿 · Talk to Nihonjoe 20:19, 18 October 2023 (EDT)

Note: I've placed this submission on hold pending the outcome of this discussion. ···日本穣 · 投稿 · Talk to Nihonjoe 20:27, 18 October 2023 (EDT)
The first linked discussion petered out when we couldn't find a way to download Vella files. As I wrote at the time:
  • With regular e-books that you purchase on Amazon, you go to "Manage Your Content and Devices", then "Digital Content", then "Books". When the desired book is displayed in the list, click "More Actions" on the right. In the pop-up list select "Download & transfer via USB" and click "Download". This will download the book as an azw3 file.
  • When you follow the same steps for a Vella serial, you get to the last step, but the "Download" button is grayed out. Instead you get a "You do not have any compatible devices registered for this content. Buy a Kindle or get the free Kindle reading app." I haven't been able to find a way around it. Ahasuerus 16:49, 9 March 2022 (EST)
You then responded with:
  • That's probably due to Vella still being in beta. I haven't been able to figure out how to do it, either. I'll keep trying different ways. ···日本穣 · 投稿 · Talk to Nihonjoe 20:05, 9 March 2022 (EST)
Any luck since then? I haven't touched Vella, so I am out of the loop. Ahasuerus (talk) 17:28, 20 October 2023 (EDT)
I haven't really tried since then. I don't like Vella myself. It's a pain to use and there's not enough there that interests me enough to make a concerted effort to try to figure it out. ···日本穣 · 投稿 · Talk to Nihonjoe
One of the books I recently added to the DB is also published on Kindle Vella. I tried in vain to find the the release dates for each chapter but gave it up as a wasted effort. If we can't get critical data like the publishing date, I'd say Out. Phil (talk) 22:02, 20 October 2023 (EDT)
Yeah, Amazon has not made it easy to figure out anything regarding Vella works. ···日本穣 · 投稿 · Talk to Nihonjoe 15:04, 23 October 2023 (EDT)
And I'd say Out as well, until the releases are collected into something which has identifying information and a release date. ···日本穣 · 投稿 · Talk to Nihonjoe 13:12, 7 November 2023 (EST)

Linking to third party Web pages -- defining "legally posted"

This Community Portal discussion got me thinking. Template:TitleFields:WebPage starts with:

  • Web Page - A field for the URL of a Web page related to this title. Examples of related Web pages include legally posted versions of the title's text [emphasis added]

Our goal when originally crafting this Help template was to make sure that we wouldn't become a hub for links to unauthorized copies of texts still under copyright protection. The Help language seemed self-explanatory at the time, but how can our editors tell whether a "version of the title's text" has been "legally posted"? For example, the main Luminist page justifies the fact that they host copyrighted works without permission as follows:

This collection may contain copyrighted material which has not been specifically authorized for our use. The copyright law of the United States (Title 17, United States Code) provides for making “fair use” copies of copyrighted materials under certain conditions, including that that the reproduction is not to be used commercially or “for any purpose other than private study, scholarship, or research.”

As I pointed out on the Community Portal, that's an odd interpretation of the copyright law:

The part of the Copyright Law that they cite -- "for any purpose other than private study, scholarship, or research" -- doesn't come from the "fair use" clause (Section 107 of the Copyright Act.) Instead it comes from Section 108, "Reproduction by libraries and archives". Section 108 is a lengthy section with a set of provisions that are completely different from the "fair use" provisions in Section 107. It's odd that the Luminist Web site cites Section 108 ("libraries and archives") language to support what they state is a Section 107 ("fair use") exception.
I should add that both Section 107 and Section 108 lawsuits can get complex and technical as we saw during Hachette v. Internet Archive in 2020-2023.

This stuff can get confusing very quickly, so I think we need a set of unambiguous rules that editors and moderators could use when deciding whether to add/approve a link to a third party-hosted text.

In addition, the fact that we currently link both to the US-based Project Gutenberg and to Project Gutenberg Australia -- which use different copyright rules and have different sets of texts available for download -- suggests that we interpret "legally posted" to mean "legally posted in the jurisdiction where the third party Web site is hosted". We may want to make it explicit in the template. Ahasuerus (talk) 20:49, 26 January 2024 (EST)

I agree that making it more clear in our documentation will be a good thing. I think we should generally avoid linking to full scans in cases where the item in question may not be in the public domain. This might mean removing some links as their track record of making sure things are in the public domain is questionable. On the other hand, they do act more like a library in that (generally) things that are not in the public domain can either be browsed on the site in a limited fashion or checked out for a specific amount of time for more lengthy review. Luminist does not do that. ···日本穣 · 投稿 · Talk to Nihonjoe 20:44, 27 January 2024 (EST)
Or how about not taking any links down unless a specific individual asks ISFDB to do that? links over the last 3 years that I've added, several thousand by now probably, are mostly still working when I happen across them later on to update info but occasionally I'll click a link and there will be that message about the upload being taken down; could be lots of reasons and probably they do get complaints now and then from Harlan Ellison types who think they own everything but most (living) authors don't care with many glad to see their works available to such a wide audience because in many cases publishers have no interest in reprinting their books. Many (most, probably) copies on are ex-library and often not in the best condition with people clearly donating them instead of tossing them in the trash because they know how hard many of the books are to find these days and they want people to be able to read them. I recently did some more MZB Sword and Sorceress edits after doing a lot of them long ago and noticed that 3 links to volumes in that series I added back then had been taken down so I removed those links since all 3 had one other copy also linked; they all had the kind of URL where it's obvious that someone uploaded the books themselves, not the typical Archive URL for books they digitized, so maybe somebody asked them to take their copies down. The issues of copyright around Marion Zimmer Bradley's works are notorious and can easily be read about online; one wishes her trustees cared less about protecting/profiting off her works and more about her (and her husband's) history re: children but that's another story. So that's my suggestion - let the Internet Archive handle requests to take certain books down, which they are clearly willing to do if someone asks them, and let ISFDB stay out of it and remain solely a research site. If anyone comes across a record with a link that's no longer working, just remove it. If you allow users of this site to decide what should be taken down you're going to create a huge mess with people taking down links to authors they don't like or links added by editors they don't like and I don't think anyone wants that. I'd still like the Moondust edit to be un-rejected if that's possible but if not at least people now know where to go if they want to read it. --Username (talk) 21:25, 27 January 2024 (EST)
Let's first try to determine if there are areas that we all agree on. I can think of two scenarios that unambiguously fall under the "legally posted" clause of Template:TitleFields:WebPage:
  • Links to texts that have been made available by the copyright holder. ("Copyright holder" is important because in certain cases it may not be the same as the author.)
  • Links to texts that are out of copyright in the jurisdiction where the linked site is located. (The qualifier is important since copyright laws are different in different countries.) We could also add links either to our Wiki pages or to third party Web page explaining how to determine whether a given text is out of copyright in common jurisdictions.
This leaves us with texts that are still under copyright in the jurisdiction where the linked Web site resides, but the site owner claims some kind of exemption, whether it "fair use", "libraries and archives" or something else. The problem here is that it's hard to tell if the claimed exemption is (a) really in compliance with the relevant laws and (b) whether the site owner accurately represents the site's position on copyright.
Apparently the legality of linking to illegally posted copyrighted material has been an area of active litigation both in the US, where "contributory copyright infringement" is illegal (but the details are complicated -- see this article for a high level overview), and in Europe (see this discussion).
A recent example of how these things can go is Anna's Archive, i.e. When it appeared about a year ago, I poked around, found literally millions of copyrighted books and articles and immediately wondered whether it was legal. More digging discovered that they apparently had two lines of defense. First, they stated that:
  • We do not host any copyrighted materials here. We are a search engine, and as such only index metadata that is already publicly available. When downloading from these external sources, we would suggest to check the laws in your jurisdiction with respect to what is allowed. We are not responsible for content hosted by others.
Second, they had a DMCA page which let copyright owners request that links be taken down.
I wasn't sure whether it would be enough to make the site legal in most jurisdictions, but I am not an expert.
Fast forward to January 2024 and we have this 2024-01-08 report:
  • On December 4, 2023, the Italian Publishers Association (AIE) filed a copyright complaint against Anna’s Archive. [snip] AIE’s complaint cites over 30 books, emphasizing that this is just a glimpse of the content distributed by Anna’s Archive to which its members hold rights. [snip]
  • With no counterclaims from the contacted parties and clear evidence of mass infringement, an order was issued to Italian ISPs to disable through a DNS block within 48 hours. Visitors to the site are now met with a blocking page in Italian.
Granted, we don't position ourselves as a "search engine for shadow libraries" the way Anna's Archive does, so we are in a somewhat different position. However, if we end up with hundreds or thousands of links to Web pages whose legality we can't easily determine, we may find ourselves in a legally questionable situation. It may be safer to simply stay away from sites of that nature. Ahasuerus (talk) 18:13, 28 January 2024 (EST)
There is a very big difference between hosting content and linking to someone else's hosted content. It is unreasonable to expect our editors and moderators to be expert enough to evaluate sites' legal claims. I think our policy should be something like: "Only links to content legally posted in the host site's jurisdiction are permitted, but the ISFDB is not qualified to make legality assessments. If ISFDB becomes aware of legal action resulting in the suspension or prohibition of a site's display of certain content, links to that site's posting of the content will be removed until the matter is resolved, or permanently, according to the circumstances." And then provide a mechanism to notify the ISFDB of host site legal issues/legal challenges to a site's posting(s). --MartyD (talk) 06:51, 29 January 2024 (EST)
There are currently a large number of edits in the queue adding links. Should these be held/skipped pending the results of this discussion? --Ron ~ RtraceTalk 09:36, 29 January 2024 (EST)
It looks like the consensus is that links are OK to add. By default, only lets you access copyrighted books' metadata, cover images and the first few pages of the text, which is similar to what Amazon's Look Inside does. You have to join their "Lending Library" program in order to be able to "check out" books. The legality of the LL program is currently under review by the courts and the last brief that I know of was filed on 2023-12-15. As long as remains a legitimate organization and complies with relevant court orders, linking to its Web pages shouldn't be an issue for us. Ahasuerus (talk) 16:15, 3 February 2024 (EST)
One other thing we could do is maintain a list of sites to which ISFDB has chosen to prohibit any content links (sort of a complement to the deep-linking-permitted list) due to concerns with the site's general compliance with applicable copyright laws. That should be clear for everyone, and the software could help enforce it. ISFDB is under no obligation to permit links, so legal precision is not necessary. There could be some transparent process for managing entries on the list (e.g., an R&S discussion with a definitive conclusion required). We could have some general guidelines for what does or does not merit being on the list. For example, we might decide that sites engaged in good-faith copyright protection and infringement dispute resolution -- e.g., such as Google Books, Internet Archive, and Project Gutenberg -- are not candidates despite any specific infringement complaints, while sites subject to multiple complaints and not obviously engaged in protection management and infringement dispute resolution -- e.g., such as the Anna's Archive example above -- are candidates. --MartyD (talk) 12:09, 29 January 2024 (EST)
I agree. Anna's Archive (and the once-popular site Ocean of PDF and all the others, many probably run by the same people under different names) pretends to be aboveboard but they're really just a dumping ground for pirated e-books and their download page is a list of shady sites, users being encouraged to become members if they want faster downloads, including the infamous LibGen that encourage bulk torrent downloads that are certainly not being used just for some light reading. Any site that has individual pages for each work,, Luminist, Galactic Journey, etc. should be acceptable. Any site which mentions bulk or torrent or anything similar is a no-no. --Username (talk) 12:27, 29 January 2024 (EST)
Speaking of which, [1], I did a search for webpages with oceanofpdf and those 2 links were added by Zapp in 2023. I think they should be removed and, if you do decide to make a blacklist, Ocean of PDF should be on it, not only because of pirating but because it's virus city and you don't want anyone clicking on a link and screwing up their computer. There's no viruses on or any of the other legit sites mentioned above. --Username (talk) 12:35, 29 January 2024 (EST)
The topic is expressly the Web Page field, but does all of this apply to recording the site or document in a Note field? ../Doug H (talk) 15:40, 29 January 2024 (EST)
I don't think different displayed fields -- Notes, Web Pages, etc -- should be treated differently for the purposes of this discussion if they link to the same third party Web sites. Notes are somewhat harder to control in the software, but that's a technical issue as opposed to a legal/policy one. Ahasuerus (talk) 16:16, 29 January 2024 (EST)

A blacklist/whitelist-based solution

After mulling it over, I think a "blacklist"-based solution would be viable or at least a good first step. It would require three components:

  • A couple of new Bureaucrat menu options to add, edit and delete blacklisted domain names like,, etc
  • A new yellow warning to be displayed when a submission tries to link to one of the blacklisted sites
  • A new nightly cleanup report to find links to blacklisted sites, which will automatically flag records once a domain is added to the blacklist

A similar whitelist of "known legitimate sites" like Project Gutenberg, Google Books,, etc would also be useful. If we implement it, we should be able to create another yellow warnings for links to domains that are not on the whitelist and may require additional digging.

Re: viruses, you are much more likely to run into them when accessing well-known illegitimate Web domains, but, unfortunately, there are no guarantees on the internet. When authors (or other people/organizations) stop paying for domain names, they become up for grabs. At that point it's anyone's guess whether they may end up in the hands of spammers, criminals, etc. Swapping this information with SFE and deleting bad links is part of what I do in the background. Ahasuerus (talk) 16:30, 29 January 2024 (EST)

I like the idea of creating a blacklist and a whitelist. I think both should require some sort of documentation supporting the addition to either list, even if that documentation is only visible to bureaucrats or admins (so that they have some sort of reference as to why a specific domain was added to one or the other). It may be good to have a "last reviewed" field, too, so we can somehow indicate when a site's inclusion on one or the other list was last reviewed (since, as you said, domain names can be picked up by someone else if the original/most recent owner chooses to not renew the domain). ···日本穣 · 投稿 · Talk to Nihonjoe 13:49, 30 January 2024 (EST)
I can see how a "Note" field would be a useful addition to the proposed table of blacklisted sites. Its contents could be made available to moderators reviewing the proposed cleanup report. Ahasuerus (talk) 14:39, 30 January 2024 (EST)
Hearing no objection, I have created FR 1590, "Create a blacklist of disallowed third party domain names". Ahasuerus (talk) 09:38, 4 February 2024 (EST)

Clarifying editor data entry rules in Help

Earlier today an ISFDB editor pointed out that Help:Screen:NewPub does not explicitly tell you what to put in the "Author" field for MAGAZINE publications. Template:PublicationFields:Author, which is transcluded in Help:Screen:NewPub, says:

  • If it is an ANTHOLOGY, multi-author OMNIBUS, or multi-author work of NONFICTION, credit the editor as the "author" of the publication.

but doesn't mention MAGAZINEs or FANZINEs. I am thinking that we should add something like:

  • For MAGAZINEs and FANZINEs, credit the issue editor as the "author" of the publication. Note that for non-genre MAGAZINEs/FANZINEs, "Editors of PERIODICAL NAME" may be used instead of some or all editor names if they are unknown or unclear or not of genre interest -- see Help:Entering non-genre periodicals for details.

How does it sound? Ahasuerus (talk) 20:10, 2 February 2024 (EST)

Sounds right to me. --MartyD (talk) 15:39, 4 February 2024 (EST)
Sounds good. ···日本穣 · 投稿 · Talk to Nihonjoe 19:39, 5 February 2024 (EST)
Yes, it does fill out a very minor hole in the rules, but it will actually be helpful in some cases. Christian Stonecreek (talk) 05:59, 6 February 2024 (EST)

Clarifying editor data entry rules in Help

Template:PublicationFields:Author has been updated with the proposed language. Thanks, folks. Ahasuerus (talk) 10:49, 9 February 2024 (EST)

Currency codes

Recently I uploaded records for an Estonian book from the Soviet Union, which cost SUR 1.40. Soviet rubles SUR were in use from 1961–1991, Russian rubles RUR were in use from 1992-1997, and now the new Russian Ruble RUB is in use since RUR was devalued to RUB at a rate of 1000 to 1.

Similarly I uploaded a Bulgarian book whose cover price said "2 лв" meaning 2 levs. But there is no single Bulgarian currency. BGJ was used 1881-1952, BGK from 1952–1962, BGL from 1962-1999, and BGN is used now since 1991.

The thing is, a currency is NOT a currency just because it has the same name. The US, Canadian, and Australian dollars are not all just dollars just because they use the word "dollar" or the dollar sign "$". Estonia na SUR, then EEK, now EUR. In that case, the names changed too (ruble > kroon > euro). In Bulgaria the word "lev" applies to BGJ and BGK and BGL and BGN, but despite the name they aren't the same currency and if our database doesn't have the correct currency for a publication then the currency field is essentially worthless apart from USD and CAD and so on.

ISFDB isn't a pricing database, but its information really must be accurate. See ISO 4217 for currency codes.

I cannot find an actual link to an actual list of Rules and Standard, but I entered BGL when I uploaded the book and one of the admins changed it to BGN, which is simply not correct. If Bulgaria were to give up the lev and take up the euro, would we change all the BGNs to EUR? No; so we should not change BGL to BGN. Evertype (talk) 11:19, 9 February 2024 (EST)

Do you have a link to the publication that was changed? Also, you can find a list of currently-supported currencies at Help:List of currency symbols. Yopu're welcome to propose additions to the list, too, if there are some we should have but which aren't on that list. ···日本穣 · 投稿 · Talk to Nihonjoe 11:39, 9 February 2024 (EST)
Here is it. And the change and explanation about why was shared on the Editor's page together with the links to the help page. Annie (talk) 11:45, 9 February 2024 (EST)
This topic has come up a number of times. The longest Rules and Standards discussions were in July 2013 and June-July 2017. Here is what I wrote about the challenges associated with using ISO codes instead of currency symbols in 2017:
  • ... the ISO standard assigns a new code when a currency is revalued, so the code for the Mexican peso changed from "MXP" to "MXN" when the peso was replaced with the "new peso" ("nuevo peso") in 1993. In 1997 the word "nuevo" was dropped, so it's now back to just "peso". However, the ISO code has remained "MXN". If we were to use ISO codes, what should an editor do when entering an undated Mexican books whose price is listed as "100 peso"? Depending on whether it was published prior to 1993 or after 1996, the correct ISO code should be either MXP or MXN, something that most of us couldn't determine without a fair amount of digging.
  • To go back to the Russian example, the ISO code for the Soviet ruble was "SUR". When the USSR was dissolved at the end of 1991, the code was retired. It was replaced with "RUR" (later "RUB" as per the discussion above) for the Russian ruble and "BYB" for the Belarusian ruble. The latter was replaced with "BYR" in 2000 and then with "BYN" in 2016.
For a bibliographic database like ISFDB to keep track of these changes over many decades and even centuries would be very time-consuming and not the best way to spend editor time.
One possible "low-hanging fruit" enhancement would be to update the mouse-over bubbles that we display for prices. They currently say things like "Lev: Bulgarian lev". We could update them to say things like "Lev: Bulgarian lev. ISO codes: BGJ in 1881-1952, BGK in 1952–1962, BGL in 1962-1999, BGN since 1991". Ahasuerus (talk) 12:40, 9 February 2024 (EST)


The other day User:Elysdir added the following paragraph to Template:TitleFields:Title:

  • Appendices. If the page where the work begins includes a phrase like "Appendix A", then include that phrase in the work's title. For example: "Appendix B: Ashima Slade and the Harbin-Y Lectures: Some Informal Remarks Toward the Modular Calculus, Part Two".

I am moving the proposed language to the Rules and Standards page to see what other editors think of it. Ahasuerus (talk) 08:02, 10 February 2024 (EST)

Thanks for moving this here! I should note that before I made that change, I did a title search on “Appendix” and found that a large majority of the appendix titles in ISFDB (in cases where there’s more than one appendix) use the format that I mentioned. (The advanced-search version of that search shows all 900+ titles.) There are only three titles in those search results that use the format “(Appendix A) Title”, and hundreds that use the format “Appendix A: Title” (or “Appendix 1: Title” or “Appendix I: Title” or “Appendix One: Title” or etc). So my writeup was an attempt to document what I was (incorrectly) assuming was an existing policy, rather than an attempt to make new policy. —Elysdir (talk) 14:46, 10 February 2024 (EST)
I prefer the use of a colon as it is better at indicating the wording is part of the title. When I see parentheses, my brain interprets it as something not part of the title but used to clarify or disambiguate. So, I support this proposed wording. ···日本穣 · 投稿 · Talk to Nihonjoe 12:06, 12 February 2024 (EST)
Yeah, I prefer using : as well - and we do use that for subtitles elsewhere so it also makes sense. And Appendices with no other title should be followed by the title of the work in brackets (we may as well throw that to complete the rule although it derives from the standard naming of essays). So "Appendix B: The making of a world" if the title is there and "Appendix B (Book title) if it just say "Appendix B". That will also make it easier to determine when there was a printed title. The corner case is when the title is printed in brackets on the page itself (which the Appendix B part is not... not sure if we want a : there or to ignore the brackets or what we want to do. Annie (talk) 13:50, 12 February 2024 (EST)
Just to make sure we are on the same page: when you wrote "brackets", did you mean "[]" (aka "square brackets") or "()" (aka "parentheses")? Ahasuerus (talk) 12:52, 13 February 2024 (EST)
Sigh. Parentheses - I meant parentheses :) I usually use square brackets for [] to make sure it is clear which ones I mean and I do not always remember that () have their own word. Annie (talk) 12:55, 16 February 2024 (EST)

(unindent) A couple of questions/clarifications.

  • The proposed language is a phrase like "Appendix A". Would this be limited to the word "Appendix" or would it also cover alternative terms like "Addendum or Appendices? Some currently use parentheses, some use colons and some say things like "Addendum to Whirligig World". Ahasuerus (talk) 14:15, 16 February 2024 (EST)
Others may well have better answers, but I thought I might as well comment here: although I didn’t say this in my proposed language, I was focused specifically on the case where the appendices have individual subtitles as well as the general title of “Appendix A”. (So my language should be updated to say that; see below.) I was assuming that when a title consists entirely of a standard book-section name, it should follow the disambiguation rule at the end of that page, in the “"Standard" titles” paragraph: “you should parenthetically append the container title (title of the novel, collection, anthology, etc) to the title of the essay, i.e. "<generic essay title> (<container title>)" in order to create a unique title”. So for cases where there’s an addendum that’s just titled “Addendum”, I would use the format “Addendum (<container title>)”, which is also what the majority of those existing cases that you linked to already use. In the rare case where addenda also have their own individual subtitles, I would use the colon format, as demonstrated by the existing item “Addendum 1: Description of Maps”. And I would expect that the title “Appendices” by itself would also be covered by the “"Standard" titles” rule: “Appendices (A Magic of Twilight)”.
So maybe another way to approach this appendix-title guidance would be to reframe it as a sub-guideline of the “"Standard" titles” guidance. At the end of the page, after the “"Standard" titles” paragraph, we could say something like this (phrasing could use some further polishing):
Standard titles with specific subtitles. If the title consists entirely of a standard title, then use the standard titles guidance above. (Examples: “Appendices (A Magic of Twilight)”; “Appendix B (A Galaxy Unknown)”; “Introduction (50 in 50)”.) But if the title starts with a label for a standard section of a book (such as “Appendix” or “Addendum”) and then is followed by an individual subtitle for that specific section, then put a colon between the book-section name and the individual title. (Examples: “Appendix: Chronology of Technic Civilization”; “Appendix B: Closures and Openings”; “Introduction: 37 Divided by 3”.)
…Note that that framing does introduce a difference from how some existing ISFDB titles currently do things: it removes the quotation marks around the individual subtitle. —Elysdir (talk) 20:39, 16 February 2024 (EST)
Perhaps I am not grasping some subtleties, but wouldn't the proposed approach be the same as what the Subtitles section of Template:TitleFields:Title currently says:
  • If the title has a subtitle, enter it, with a colon and a space used to separate the title from the subtitle. For example, the 1986 edition of George MacDonald's "Lilith" has "Lilith" on the title page, and below that, in a smaller font, "A Romance". This should be entered as "Lilith: A Romance".
? Ahasuerus (talk) 10:47, 19 February 2024 (EST)
I think it’s essentially the same guidance, yes, but applied in a different case. The subtitles guidance reads to me as being about the title and subtitle of a book, as opposed to a section. If instead of adding a new section, you would prefer to clarify the Subtitles section to say that it also applies to things like an appendix or an addendum, that would be fine with me. …My goal in all of this is to clarify to editors how they should format the titles and subtitles of appendices; I’m fine with any approach y’all want to take. (…And I apologize if I’m overstepping by participating in this discussion at all—if I should step back and just leave it to you folks to decide, let me know.) —Elysdir (talk) 15:47, 19 February 2024 (EST)
Oh, no, everyone is welcome to contribute to Rules and Standards discussions! Sometimes an outside perspective reveals that Help is unclear or that it doesn't account for a certain category of cases.
The current discussion is a good example. The first three paragraphs of Template:TitleFields:Title, as currently written, are in the following order:
  • Novels
  • Subtitles
  • Short fiction, essays and poems
The way they are ordered, it's possible to assume that the "Subtitles" paragraph only applies to novels, especially since the next 2 paragraphs (SERIALs and excerpts) have special rules for subtitles and disambiguators. However, I believe the intent was to apply the "Subtitles" rule to all other title types (that do not have explicitly stated exceptions) as well.
If my understanding is correct, then we may be able to eliminate this ambiguity by moving the "Subtitles" paragraph below the "Short fiction, essays and poems" paragraph. We should probably also move "Omnibuses, nonfiction, anthologies and collections", which is currently the 6th paragraph in this template, right below the "Novels" paragraph. That way the order would be:
  • Novels
  • Omnibuses, nonfiction, anthologies and collections
  • Short fiction, essays and poems
  • Subtitles
  • Excerpts
  • Artwork
  • Etc
The 4 paragraphs preceding the "Subtitles" paragraph would all use the same subtitle rule while the paragraphs following the "Subtitles" paragraphs would have special rules. We could also make it explicit in the language of the "Subtitles" paragraph. Would this work from your perspective? Ahasuerus (talk) 16:07, 19 February 2024 (EST)
I like this idea. ···日本穣 · 投稿 · Talk to Nihonjoe 17:35, 19 February 2024 (EST)
  • Some languages -- notably French -- use a space between a word and a trailing colon. We generally follow language-specific rules for non-English titles, so is it safe to assume that we would be using " :" as opposed to ":" for French titles? Ahasuerus (talk) 14:15, 16 February 2024 (EST)
Huh, interesting, I didn’t know that about French. Given that difference, I would expect that yes, we would use " : " instead of ": " in French titles. —Elysdir (talk) 20:39, 16 February 2024 (EST)

Appendices - Outcome

Hearing no objection, I have re-ordered the first 4 paragraphs in Template:TitleFields:Title based on the order proposed above. One sentence was split into two for readability. Ahasuerus (talk) 18:47, 28 February 2024 (EST)

Secondary source artist credit in face of credit change over time

For full background, see User_talk:MartyD#Dixie_Ray. Different editions of a Ballantine Fahrenheit 451 use the same cover art but credit the artist differently. This is what we know:

ISBN Printing Date Artist credit
0-345-25027-3[-150] 40th 1975-12-00 no credit at all
0-345-25027-3[-150] 43rd 1976-08-00 Whistlin' Dixie
0-345-27431-8 44th 1977-08-00 Whistlin' Dixie
0-345-27431-8 45th 1977-11-00 Whistl'n Dixie
0-345-27431-8 46th 1978-08-00 Whistl'n Dixie

The later "Whistl'n" is canonical. The rules do not permit us to assign "uncredited" to the 40th's cover, but they do permit us to assign an identity using the later editions' credits as a secondary source. Which later edition's credit should we use here, the non-canonical "Whistlin'" or the canonical "Whistl'n"? Likely the 41st and 42nd printings will have either no credit or "Whistlin'". I'd really like to use "uncredited" to give a complete picture of the credit's evolution. It would be a little odd to have one or more earlier editions have the canonical credit, then have some later ones with a non-canonical credit, then even later ones "revert" to canonical (when in fact they progressed to canonical). For now I have gone with canonical, but I thought I'd raise the question to see if we should standardize on something else for this scenario.

Two further hypotheticals to consider: Suppose we only had the 40th (uncredited) and then entered the 43rd ("Whistlin'") and so went back and adjusted the 40th to use that. Now the 45th ("Whistl'n") gets entered, and its credit is determined to be the canonical form. What would we want done with the 40th's (now) non-canonical credit at that point? Likewise, suppose we had the 40th, then entered the 45th ("Whistl'n") and went back and adjusted the 40th to use that. Now the 43rd gets entered. What would we want done with the 40th's credit at that point? --MartyD (talk) 07:12, 17 February 2024 (EST)

Re: "I'd really like to use "uncredited" to give a complete picture of the credit's evolution."
As per Template:PublicationFields:CoverArt here is how I think we currently credit cover artists depending on what is in the publication:
What is stated in the pub What we enter in the "Artist" field
Canonical name Canonical name
Alternate name Alternate name (VT created)
Initials Canonical name if known
Artist-specific symbol (sometimes a stylized version of the artist's initials) Canonical name if known
Signature, often illegible Canonical name if known
No explicit credit, but the artist's style is recognizable The "Artist" field is left blank; Notes updated with the name of the artist and reason for attribution
No explicit credit, but a secondary source credits the artist Canonical name; Notes updated with the source
No explicit credit, but the credit is implied, e.g. a small illustration may be reproduced as a credited INTERIORART work Can be arguably considered a "secondary bibliographic source" for our purposes and treated as such, i.e. enter the canonical name in the "Artist" field and update Notes with the source
This is a tricky decision tree diagram, which, admittedly, makes it hard to "give a complete picture of the credit's evolution". I think the underlying issue here is that it would be difficult to enter artist credits the way we enter author credits, i.e. "as stated in the pub". The main reason is that signatures, symbols and barely legible stylized initials are not something that can be easily captured as text.
That being said, I think it would be beneficial to restructure Template:PublicationFields:CoverArt as a series of bullets to make it easier for new editors to parse. Ahasuerus (talk) 11:33, 19 February 2024 (EST)
I understand the current rules call for canonical when the credit is taken from a secondary source, and that is what I did. It seems wrong to me in this case, however. For argument's sake, let's assume there is also no credit in the 41st printing and the credit in the 42nd is the alternate "Whistlin'". If all printings were recorded at the same time, we would have none/canonical -> none/canonical -> alternate -> alternate -> canonical -> canonical. If instead we entered them over time we would start with none -> none, then on discovering the "Whistlin'" we might change those to: non/canonical(1) -> none/canonical(1) -> canonical(1), with the first two citing the third as secondary source. Two printings later, we would discover "Whistl'n" and realize it should be canonical, so we'd VT the existing TITLE records and end up with: none/alternate -> none/alternate -> alternate -> alternate -> canonical(2). Someone would have to know to review all previous credits to see if they came from the publication or used the source of the now-alternate credit and in the latter case change them to the (new) canonical to match what would happen if we entered them all at the same time. --MartyD (talk) 07:05, 20 February 2024 (EST)
I think we have somewhat of a grandfathered problem here. Our rules had been pretty straightforward for a long time - secondary credits of art use the canonical name. In our digital era, I'd argue that a scan that is proved to be of a certain printing should be considered primary source for this determination (and I think we had been applying it that way). However, as a practice we had often made an exception for this rule for later (and earlier) printings and even different formats altogether (audio/ebooks/paper had gotten credits based on the other formats) - we had often imported straight from the one we do know the credit for even if it is not using the canonical name because it is (usually) a good guess that most of these will match. Thus the conundrum now for the few credits where they do not match.
We have two paths: enforce the rule as written OR come up with a language that allows us a bit of creativity: "You can use the credit as found in a later or earlier printing if data for the current printing is not available, with a mandatory note on the exact source of the name used. That includes the usage of uncredited. The same applies for other formats sharing a cover (i.e. audiobooks which have only a cover and the artist may or may not be credited on it). Using the canonical name is always allowed in the cases of unknown credit (due to lack of source information or only secondary sources information) - with an appropriate note.". Feel free to rewrite/change/argue. And if we are changing the rule, can we please make it more forceably requiring a note on the decision if you are not grabbing the name straight from the book - otherwise it is a nightmare to change a canonical name for example - I am sure we had created a lot of mistakes in the DB in the process of changing canonical names of artists simply by not knowing when a credit is a direct one and when a canonical is being used.
I am leaning towards the second option - mainly because it is somewhat of a practice anyway (in the multi-formats) and it kinda covers this case here. Annie (talk) 12:10, 20 February 2024 (EST)

Clarifying Template:PublicationFields:CoverArt

Going back to the issue of clarifying (as opposed to changing) what's currently stated in Template:PublicationFields:CoverArt, earlier this week an editor asked me if we could update the template language with what I wrote above to make the instructions more clear. Here is the proposed new language to be used when deciding what to enter in the "Artist" field:

  • If the artist's canonical name is stated in the publication, enter it
  • If the artist's alternate name is states in the publication, enter it and make sure to create a Variant Title later
  • If the cover has the artist's initials, enter the artist's canonical name if known and update the Notes field with the source of the attribution
  • If the cover has an artist-specific symbol, e.g. a stylized version of the artist's initials, enter the artist's canonical name if known and update the Notes field with the source of the attribution
  • If the cover has a recognizable signature, enter the canonical name if known and update the Notes field with the source of the attribution
  • If the publication has no explicit artist credit, but the artist's style is recognizable, leave the "Artist" field blank and update the Notes with the name of the artist and reason for attribution
  • If the publication has no explicit artist credit, but a secondary source credits the artist, enter the canonical name and update the Notes field with the source of the attribution
  • If the publication has no explicit artist credit, but the credit is implied, e.g. a small section is reproduced as a credited INTERIORART work, treat it as a "secondary bibliographic source" scenario described above: enter the canonical name and update the Notes field with the source of the attribution

Does this look right? Ahasuerus (talk) 19:01, 28 February 2024 (EST)

Looks good to me, though I'd put the two "e.g." parts in parentheses. ···日本穣 · 投稿 · Talk to Nihonjoe 12:56, 29 February 2024 (EST)
I agree on the parentheses. Annie (talk) 13:19, 29 February 2024 (EST)
Spot-checking Help:Screen:NewNovel, I see that we use "e.g." inconsistently. In roughly one third of all cases we use parentheses while in the other two thirds we do not. Different grammar guides give contradictory advice. AP Style requires the use of parentheses and a trailing comma, but Fowler's Modern English Usage does not. Ahasuerus (talk) 13:45, 29 February 2024 (EST)
I find it a lot more readable when the parentheses are there. It also simplifies the reading of the sentence for non-native speakers and we have quite a lot of them - the clearer we state things and the easier we make it for someone whose English may be shaky, the better IMO. Annie (talk) 14:18, 29 February 2024 (EST)
I would also move "If the publication has no explicit artist credit, but the artist's style is recognizable, leave the "Artist" field blank and update the Notes with the name of the artist and reason for attribution" to the bottom of the list and change it to " If the publication has no explicit artist credit and no secondary or implied credit, but the artist's style is recognizable, leave the "Artist" field blank and update the Notes with the name of the artist and reason for attribution.". Otherwise it contradicts the next 2 rules in case of a recognizable artist and secondary credit for example. Annie (talk) 13:19, 29 February 2024 (EST)
That's a very good point. Here is the updated proposed order:
  • If the artist's canonical name is stated in the publication, enter it
  • If the artist's alternate name is states in the publication, enter it and make sure to create a Variant Title later
  • If the cover has the artist's initials, enter the artist's canonical name if known and update the Notes field with the source of the attribution
  • If the cover has an artist-specific symbol, e.g. a stylized version of the artist's initials, enter the artist's canonical name if known and update the Notes field with the source of the attribution
  • If the cover has a recognizable signature, enter the canonical name if known and update the Notes field with the source of the attribution
  • If the publication has no explicit artist credit, but a secondary source credits the artist, enter the canonical name and update the Notes field with the source of the attribution
  • If the publication has no explicit artist credit, but the credit is implied, e.g. a small section is reproduced as a credited INTERIORART work, treat it as a "secondary bibliographic source" scenario described above: enter the canonical name and update the Notes field with the source of the attribution
  • If the publication has no explicit artist credit and no secondary or implied credit, but the artist's style is recognizable, leave the "Artist" field blank and update the Notes with the name of the artist and reason for attribution
Ahasuerus (talk) 13:51, 29 February 2024 (EST)