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 · 22

---

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)

Following up on this, it looks like Template:PublicationFields:Pages is already transcluded to Help:Screen:NewNovel, Help:Screen:NewPub, and Help:Screen:EditPub, but it is not transcluded to Template:PubContentFields:Page. Should we transclude it there, too? I don't think it needs to be transcluded to Help:How to determine the value for the "Pages" field in a book, and there is already a link from Template:PublicationFields:Pages (at the bottom) to Help:How to determine the value for the "Pages" field in a book. ···日本穣 · 投稿 · Talk to Nihonjoe 14:29, 22 March 2024 (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 ~ Rtrace^Talk 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 https://www.thehugoawards.org/category/voter-packet/ and en.chengduworldcon.com/help/1. 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 ~ Rtrace^Talk 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)

I went ahead and took a stab at the 2024 packet which can be seen here. I don't know if any of our editors with a better knowledge of Chinese than I are Hugo voters and have access to the packet, but the excerpts of the books edited by 姚海军 (Yao Haijun) are in a locked down PDF format that wouldn't allow me to copy and paste into Google translate. These are the Chinese titles beginning with 万物之主 (Herr aller Dinge). It is possible that the cover art is credited in the excerpts, but if it is, I was unable to determine that and thus added the covers as uncredited. Also, now that we have an example of how this can look, we can discuss if there is anything folks would like to see differently. I'll wait a while before I try to tackle another one of these. They are a little labor intensive to enter. --Ron ~ Rtrace^Talk 21:09, 10 June 2024 (EDT)

Thanks for working on this publication record, Ron. "A little labor intensive" indeed... Ahasuerus (talk) 22:22, 10 June 2024 (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, https://www.isfdb.org/cgi-bin/view_submission.cgi?5796089, 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:

• "Untitled" -- note the capitalization -- e.g. this story or this poem
• "untitled" -- all lowercase -- e.g. this story or this poem
• "Untitled" or "untitled" followed by the first few words in the body of the work in parentheses, e.g. Untitled ("1.6: These texts are a book about the people and their Gods ...") (SHORTFICTION) or untitled ("A gate in rubble") (POEM)
• "Untitled" or "untitled" followed by a short description of the work, e.g. "(Untitled Congratulations to Isaac Asimov)"
• The same as immediately above except disambiguated, e.g. "(untitled editorial) (ERB-dom, June 1973)"
• "[Untitled]" or "[untitled]", e.g. this story or this poem
• "(Untitled)" or "(untitled"), e.g. this story or this poem
• The same as immediately above except disambiguated, e.g. (untitled) (Twisted #4, Summer 1987) and (untitled) (Twisted #4, Summer 1987) [2\]

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 archive.org 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? Archive.org 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 Archive.org 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. annas-archive.org. 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 https://annas-archive.org 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 ~ Rtrace^Talk 09:36, 29 January 2024 (EST)

It looks like the consensus is that archive.org links are OK to add. By default, archive.org 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 archive.org 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, Archive.org, 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 Archive.org 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 annas-archive.org, oceanofpdf.com, 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, archive.org, 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)

As per the discussion immediately below, the wording of the FR has been changed to "Create a blacklist of disallowed third party URL patterns". Ahasuerus (talk) 08:53, 4 April 2024 (EDT)

Luminist's PDF files

Reviewing the above discussion, and until the FR is implemented, I note that we agreed there was consensus for adding links to archive.org. I'm seeing new edits to add links to pdfs hosted by wasabisys.com. This seems a different kettle of fish. Do we have consensus on whether links to downloadable pdfs from this site should be allowed? --Ron ~ Rtrace^Talk 17:49, 28 March 2024 (EDT)

I haven't seen wasabisys.com, which redirects to wasabi.com, before. Based on this FAQ it appears to let anyone upload and store arbitrary amounts of data. Kind of like Google Drive or Amazon's S3, right? Ahasuerus (talk) 19:34, 28 March 2024 (EDT)

This is the edit that gave me pause. I believe Username refers to the as "luminist" links. The ones he has added all appear to be served from the wasabisys domain. The question would be whether wasabisys has any safeguards to prevent copyright violation, or are they a site that will host files for bad actors. I stopped approving the addition of any links to scans of books under copyright when this topic was raised. I resumed approving links to archive.org once we had consensus to include those, but am hesitant to approve others if we haven't agreed that they are acceptable. --Ron ~ Rtrace^Talk 20:07, 28 March 2024 (EDT)

They're from a site at Luminist.org, the URL's contain the word luminist, and the guy who runs the site calls himself Luminist, https://isfdb.org/wiki/index.php/User_talk:Luminist. Also, links from when the site still used Adobe document links are to be found in many PV Analog records on ISFDB, [2], added a few years ago by Dave888 and approved by...RTrace. I did add the Naked Storm one, though. --Username (talk) 20:24, 28 March 2024 (EDT)

Oh, right, I remember it now. Luminist has apparently moved all (?) of his PDF files to wasabisys.com -- see our Web Page Search results. I expect that it may be a more cost-effective solution for small operators since sites like wasabisys.com and backblaze.com host files relatively cheaply, in the $6-7 per month per terabyte range.

This presents a problem from our perspective since the solution proposed above was to create a "blacklist" of sites which are known to violate copyright: Anna's Archive, oceanofpfd.com, etc. With an aggregate site like Wasabisys, Backblaze or even Google Drive, there may be no easy way of telling who the owner of the linked files is. It makes the "blacklist" approach unworkable for this type of cases. Still useful in other cases, but not as comprehensive as I hoped it would be.

I note that all Wasabisys.com links start with "*wasabisys.com/luminist/", so it may be something to pursue, although it wouldn't help with files hosted by Google Drive since it doesn't have that kind of convenient URL structure.

Going back to the Luminist situation, he hosts a variety of PDF files. There appear to be three separate types of scenarios:

• Scans of books that are no longer under copyright protection, e.g. A Trip to Venus (1897) or The Altar of the Legion (1926), which were published before 1929 and are therefore in public domain in the US.
• Scans of books published between 1929 and 1963. Their copyright status is often unclear since they only enjoy copyright protection if copyright has been renewed, which is rare for genre books like Zip-Zip Goes to Venus (1958). Project Gutenberg and some other sites look for copyright renewal notices in The Catalog of Copyright Entries before making their files publicly available, but Luminist doesn't seem to do it.
• Scans of books published after 1963 and therefore still under copyright protection. Luminist justifies it 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.” By accessing files linked to this site you are agreeing to abide by these restrictions. If you do not agree, do not download. If any copyright owner objects to our inclusion of their material on this web site, please do not harass our hosting providers; just contact us with the pertinent information. We will remove contested content promptly upon receipt of legitimate requests. Readers who wish to obtain a permanent copy of any item are encouraged to acquire one from a bookseller of their choice.

This is presumably based on Chapter 1, section 107 of Title 17, "Limitations on exclusive rights: Fair use". Perhaps it may be argued that some relatively obscure books like The Tsaddik of the Seven Wonders (1971), which hasn't been reprinted since 1981, are only of interest to researchers. However, Luminist also has scans of books that have been recently reprinted, e.g. The Secret of Barnabas Collins, which has had multiple editions since 2019.

I don't think we are (or should be) in a position to decide which post-1963 books fall under "fair use" and which ones do not. "Fair use" cases are settled by the courts on a case by case basis; we don't have the knowledge or the resources to deal with its complexities. Template:TitleFields:WebPage currently allows:

• legally posted versions of the title's text

but doesn't define "legally posted". My current thinking is that we could clarify it to disallow "texts known to be under copyright protection and made available without the copyright owner's permission". The clause "known to be under copyright protection" would exclude everything from 1964 on.

If we decide to do this, then it would be easy to create a cleanup report to look for PDF files associated with post-1963 publication records.

Thoughts? Ahasuerus (talk) 13:55, 31 March 2024 (EDT)

That sounds workable, at least for moderation and as a guide to editors for what is allowed. How hard would it be to add a yellow warning (for both editors and moderators) for this? Not a big hurry for that, but it would make things easier, assuming that others, if any, agree with handling Luminist and Wasabisys in this manner. --Ron ~ Rtrace^Talk 16:46, 31 March 2024 (EDT)

It would be a simple task. The process of adding new warnings has been much more straightforward since the "yellow warning" system was revamped in 2023. Ahasuerus (talk) 17:09, 31 March 2024 (EDT)

Date-based warning sounds like a handy reminder.... @Ahasuerus, you could think about a pattern-based approach to blacklist/whitelist, instead of relying strictly on domains. E.g., for the Luminist example on Wasabi, a pattern might be: *.wasabisys.com/luminist/* (or whatever pattern-specification syntax appeals to you -- regex, SQL, ...). Since the pattern itself would not be created by ISFDB end-users, but rather "internally", it doesn't really matter what the pattern syntax would be, as long as we can explain it in plain English. --MartyD (talk) 17:34, 31 March 2024 (EDT)

A good point. We'll just have to change the name of the menu option and the text of the associated yellow warning from "Blacklisted domains" to something like "Blacklisted Web page URL segments". Ahasuerus (talk) 18:06, 31 March 2024 (EDT)

(unindent) I still have a number of held edits adding Luminist scans for post 1963 published works. I also just held a new edit for a scan of a 1976 publication hosted by fantlab. Have we reached a consensus on this? My sense was that we were leaning against allowing these scans. Thanks. --Ron ~ Rtrace^Talk 11:12, 18 July 2024 (EDT)

The FantLab link is an interesting example. "A Song for All Women" by Robert E. Howard was first published in 1976. FantLab servers are presumably located in Russia, so they are governed by Russian laws just like the Australian version of Project Gutenberg is governed by Australian laws. I know little about the current state of Russian copyright laws except that they were apparently rewritten in the 1990s and then again in the 2000s. Wikipedia says:

• Among the true novelties introduced by the new legislation in the area of copyrights were a publication right (a copyright granted to the publisher of a previously unpublished, uncopyrighted work with a period of 25 years from the publication) [snip] A subtle change concerned the calculation of the copyright term for posthumously published works, which began newly from the disclosure instead of from the publication. (See above for the difference.) For a work that was disclosed during the author's lifetime, the copyright term of 70 years thus runs from the year the author died (or was rehabilitated, if the rehabilitation was posthumous), even if the work is published only later.

This is rather convoluted and not something that we can realistically sort out.

I guess there are two ways to approach these types of scenarios. The first one is to say "We can't be sure what the copyright laws that apply to the servers hosting the text say about this particular case, so we won't be linking to it." The second one is to say "FantLab.ru, like Project Gutenberg, is an established bibliographic resource and we will trust it to enforce copyright restrictions in the jurisdiction that it is based in, so linking to their files is OK".

Re: Luminist-hosted files, I think the majority opinion was that we shouldn't be linking to post-1963 works made available in the US without getting permission from the copyright owner. We should probably create a separate R&S section to discuss specific changes to the Help text that started this discussion. Something like:

• 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. For texts published after 1963, US-hosted versions need to be authorized by the copyright owner.

Ahasuerus (talk) 16:45, 19 July 2024 (EDT)

I'm going to go ahead and reject the "recent" Luminist edits that I've been holding. For the Fantlab edit, I'll leave it on hold in the hopes that someone else will chime in. I will mention that my sense is that Fantalab does not usually provide complete texts of titles. I usually see covers and title pages there. I agree about not wanting to parse Russian copyright law. I'm not a lawyer, and certainly not a Russian lawyer. Thanks. --Ron ~ Rtrace^Talk 18:52, 23 July 2024 (EDT)

Before you go ahead and reject the links I took the trouble to find you may want to suggest an alternative, possibly notes for each saying the book is available at Luminist.org without actually providing a clickable link, so people still know where they can find them. Some of them are very hard to get and/or expensive these days. --Username (talk) 19:01, 23 July 2024 (EDT)

Definition of "advertising", and so on

Moving this here from the discussion on the moderator noticeboard, and putting it as a subheading under the conduct policy discussion above since it's pretty closely related.

Help:Contents/Purpose#What the ISFDB Wiki is not states that "The Wiki is a support tool for the ISFDB, and should not be used for anything that is not appropriate for that purpose." The Conduct Policy is also connected to this discussion. I think it boils down to the first paragraph of the policy: "Policy intent: Anything that helps make the ISFDB a more useful and more reliable bibliographic tool is encouraged. Anything that hinders this process is discouraged." (emphasis added)

In my opinion, political or religious content (regardless of which set of politics the content espouses or supports) on user pages and elsewhere on ISFDB that is not directly related to the stated purpose of "[making] the ISFDB a more useful and more reliable bibliographic tool" should be not only discouraged, but removed. ISFDB is not the place for making political or religious statements. There are plenty of other places online to do that.

Perhaps adding something like the following to the Conduct Policy would be good.

"The Wiki is a support tool for the ISFDB, and should not be used for anything that is not appropriate for that purpose. This includes posting—anywhere on the wiki—political, religious, or other content that is not directly related to making ISFDB a more useful and more reliable bibliographic tool. Such content is not allowed and may be removed by an Administrator on sight. Appeals regarding removed material can be made on the Moderator Noticeboard."

Thoughts? ···日本穣 · 投稿 · Talk to Nihonjoe 16:35, 31 May 2024 (EDT)

I do support this intended move including the addition to the Conduct Policy. Christian Stonecreek (talk) 02:08, 3 June 2024 (EDT)

I like the idea, but we may want to clarify what we mean by "other content that is not directly related to making ISFDB a more useful and more reliable bibliographic tool". User pages in particular often contain information that is not "directly related to making ISFDB a more useful and more reliable bibliographic tool" yet is perfectly innocuous. For example, User:Alvonruff/ReadingLists lists all SF books that he has read since 2006.

In general, I suspect that it would be fine to have non-SF content on User pages as long as it is about the user and as long as it is informational as opposed to promotional. For example, if L. Neil Smith -- who was an SF author as well as a musician and a political activist -- were still alive and wanted to contribute to the ISFDB project, I don't think it would be objectionable if he added a list of links to his other Web sites/pages to his User page, but I don't think we would want his User page to become an advertising platform for his numerous social and political ideas or a place to sell his books. Defining this distinction in a way that could be fairly applied across the board could be admittedly tricky. Ahasuerus (talk) 21:54, 3 June 2024 (EDT)

I can see that. The main thing I don't want to see is divisive content such as political stuff (of any stripe), pushing of religious beliefs or anti-beliefs, and stuff like that. It will definitely be tricky to be explicit enough without painting ourselves into a corner. ···日本穣 · 投稿 · Talk to Nihonjoe 15:09, 4 June 2024 (EDT)

(unindent) How about something like this?

"The Wiki is a support tool for the ISFDB, and should not be used for anything that is not appropriate for that purpose. This includes posting—anywhere on the wiki—political, religious, or other content (images, essays, promotional material, and similar content) that is not directly related to making ISFDB a more useful and more reliable bibliographic tool. Users may post a brief link section if they wish to link to other websites they run, as long as the sites do not contain illegal content. Content deemed inappropriate toward ISFDB goals may be removed from any wiki page by an Administrator or Moderator. Appeals regarding removed material can be made on the Moderator Noticeboard."

Thoughts in this version? ···日本穣 · 投稿 · Talk to Nihonjoe 15:36, 23 July 2024 (EDT)

One thing that comes to mind is that the term "illegal" is harder to define these days. Different countries have different laws, including copyright laws, which is why Project Gutenberg makes different texts available in different countries. Similarly, obscenity laws vary from jurisdiction to jurisdiction and can be subject to interpretation. US laws are a well known example since they use "contemporary community standards", "patently offensive way" and "serious literary, artistic, political, or scientific value" as parts of the standard obscenity test. Some books that we catalog, e.g. Samuel R. Delany's Hogg, may well run afoul of some interpretations of obscenity laws in some jurisdictions. Ahasuerus (talk) 18:23, 24 July 2024 (EDT)

Clarifying "legally posted"

As we discussed in January-March 2024, Template:TitleFields:WebPage currently allows links to:

• legally posted versions of the title's text

but it doesn't define "legally posted".

My current thinking is that we could clarify the template to disallow:

• texts known to be under copyright protection and made available without the copyright owner's permission

This wording would disallow links to copyrighted texts in cases where the hosting site doesn't have the copyright owner's permission and claims "fair use" or "libraries and archives" exception under Title 17 of the United States Code. As mentioned previously, I don't think we are (or should be) in a position to decide which copyrighted texts fall under "fair use" and which ones do not. "Fair use" cases are settled by the courts on a case by case basis; we don't have the knowledge or the resources to deal with its complexities. Moreover, it's possible -- although in our case, perhaps, unlikely -- for links to copyright-infringing sites to be considered "contributory infringement". To quote this Cornell Law School Web page:

• One who knowingly induces, causes or materially contributes to copyright infringement, by another but who has not committed or participated in the infringing acts themselves, may be held liable as a contributory infringer if they had knowledge, or reason to know, of the infringement. See, e.g., Metro-Goldwyn-Mayer Studios Inc. v. Grokster, Ltd., 545 U.S. 913 (2005); Sony Corp. v. Universal City Studios, Inc., 464 U.S. 417 (1984).

Whether knowingly linking to an infringing Web page constitutes "materially contributing to copyright infringement" is something that may be debated in court, but I don't think it's an area that we want to touch.

For US cases, the clause "known to be under copyright protection" would apply differently to 3 different periods:

• Texts of books that are no longer under copyright protection, e.g. A Trip to Venus (1897) or The Altar of the Legion (1926), which were published before 1929 and are therefore in public domain in the US, would be "linkable".
• Texts of books published between 1929 and 1963. Their copyright status is often unclear since they only enjoy copyright protection if copyright has been renewed. Since copyright renewal was rare for 1930s-early 1960s genre books like Zip-Zip Goes to Venus (1958), we can probably allow links as long as there is no evidence that copyright has been renewed. If such evidence is found, as was the case with H. Beam Piper's Space Viking the other day, we can document it and remove offending links.
• Texts of books published after 1963 and therefore still under copyright protection would not be "linkable".

If we decide to do this, then it would be easy to create a cleanup report to look for PDF files associated with post-1963 publication records.

Thoughts? Ahasuerus (talk) 22:10, 22 April 2024 (EDT)

I think this sounds reasonable. This table from Wikimedia Commons may be useful, too, as it walks through how to determine if something is in the public domain in the States, or still under copyright protection. ···日本穣 · 投稿 · Talk to Nihonjoe 13:33, 23 April 2024 (EDT)

Looks useful, thanks. Including the body of the linked table (with WikiMedia-spedific templates removed/expanded) for reference purposes:

US copyrights for works first published in US, excluding audio works

Year of first publication Note: publication is not creation	Copyright duration
Before [current year-95] During [current year-95]–1963: without notice, or with notice but not renewed within 28 years of first publication During 1964–77: without notice From 1978 to March 1, 1989: without notice and without registration within 5 years of first publication	Work has entered US public domain
During [current year-95]–1963: with notice and renewed During 1964–77: with notice</translate>	Copyrighted for 95 years after first publication
From 1978 to March 1, 1989: pre-1978 creation with notice, or without notice but registered within 5 years of first publication From March 2, 1989 to 2002: pre-1978 creation	If author is known, copyrighted until the later of either 70 years pma (post mortem auctoris or "after the author's death") or Dec 31, 2047 If author is unknown or corporate authorship, the earlier of 95 years after first publication or 120 years after creation, but not earlier than Dec 31, 2047
From 1978 to March 1, 1989: post-1977 creation with notice, or without notice but registered within 5 years of first publication From March 2, 1989 to 2002: post-1977 creation Unpublished before 2003 (i.e. first published after 2002)	If author is known, copyrighted for 70 years pma (post mortem auctoris or "after the author's death") If author is unknown or corporate authorship, the earlier of 95 years after first publication or 120 years after creation

Template:PD-US contains a summary of the public domain logic, but the page is specifically talking about images. The logic could be pulled out of that into a central page that is then referenced for images and text as applicable. -- JLaTondre (talk) 08:10, 25 April 2024 (EDT)

I'm not certain as to the scope being proposed here. While I agree with avoiding links to Luminist and other dodgy sites, as proposed, this seems to also be prohibiting links to the Internet Archive. I would object to that. While the Internet Archive lost a lawsuit over their "National Emergency Library" (where they dropped the one viewer at a time restrictions during the COVID pandemic), their regular library so far has survived challenges. I do not see a reason to ban links to their texts. -- JLaTondre (talk) 08:10, 25 April 2024 (EDT)

Thanks for the reminder about Archive.org and its lending program. I agree that we don't want to disallow links to archive.org. As I wrote in the discussion linked above:

• It looks like the consensus is that archive.org links are OK to add. By default, archive.org 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 archive.org remains a legitimate organization and complies with relevant court orders, linking to its Web pages shouldn't be an issue for us.

We may want to clarify the proposed Help language to make sure that we don't accidentally exclude legally operating "internet-based lending libraries" like archive.org. Ahasuerus (talk) 13:41, 25 April 2024 (EDT)

Luminist is not a dodgy site; they have no torrents, membership fees for faster downloads, or anything like Anna's Archive or the defunct Ocean of PDF or any similar sites have and the person running the site even has a board on our site under the name of Luminist. They offer individual PDFs (years ago they used other formats but converted most to PDF a while back) that are linked offline at, I assume, a free or cheap site to save money. They've been around for many years and are perfectly legitimate. Anna's Archive, in particular, is a very useful site to research from for sites like ISFDB because they have a huge number of books not on The Internet Archive (an almost full run of Pulphouse: The Hardback Magazine, for example) but links are not appropriate because they offer those shady things I mentioned above. I've added the bulk of the 300+ Luminist links here over the last few years and I don't recall anyone ever leaving a message asking for any of them to be removed. What should be done is the same thing as the Archive does; if any individual or publisher asks for something to be taken down, do it; if not, leave it. --Username (talk) 09:06, 25 April 2024 (EDT)

Based on what I have seen on the internet, there is a spectrum of questionable -- or "dodgy" -- Web sites that make copyrighted material publicly available without copyright holders' permission. On the one end of the spectrum we have sites that openly state that they do not respect copyright and do everything in their power to evade law enforcement.

On the other end of the spectrum we have sites that come up with legal theories justifying their actions. In the Luminist case the justification is 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.” By accessing files linked to this site you are agreeing to abide by these restrictions. If you do not agree, do not download. If any copyright owner objects to our inclusion of their material on this web site, please do not harass our hosting providers; just contact us with the pertinent information. We will remove contested content promptly upon receipt of legitimate requests.

As I pointed out in January 2024, 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.

To what extent this and similar interpretations of the copyright law would be upheld by the courts is not something that we can guess at with any degree of certainty. Even if a Web site cites the correct "fair use" language of Section 107 of the Copyright Act, we are not (and shouldn't be) in a position to decide which copyrighted texts fall under "fair use" and which ones do not. "Fair use" cases are settled by the courts on a case by case basis; we don't have the knowledge or the resources needed to deal with its complexities. That's why the current proposal would exclude them.

Another thing to note is that the Luminist site makes 3 different types of texts available:

• All pre-1929 (pre-1930 come January 1, 2025 etc) texts that it hosts are always OK to link to because they are out of copyright.
• Under the current proposal, a 1929-1963 text would be also OK to link to unless it happens to be one of the rare (under 15% based on some estimates that I have seen) cases where copyright has been renewed and we know about it, e.g. H. Beam Piper's Space Viking (1963).
• All 1964-2024 texts are under copyright and links to them would be disallowed under the proposed Help language.

Ahasuerus (talk) 15:43, 25 April 2024 (EDT)

If nobody here is clear then nobody elsewhere is, either, because nobody to my knowledge has ever asked for anything from Luminist to be taken down. If they do for individual works because they think they can squeeze more money out of a new edition, take those links down; if not, leave them. They're here for research on this site to fix or add information, not for profit. Any real collector would want a physical copy, not some scanned electronic version. --Username (talk) 16:13, 25 April 2024 (EDT)

Sorry, I've tried to post this 3 or 4 times and keep getting edit conflicts. It may no longer make sense, but I'm out of time.

I'd suggest broadening the proposed bullet to be something like

• works under copyright protection and made available in contravention of applicable copyright law

We need to worry about more than pirated texts, and "without the copyright owner's permission" is both narrow and difficult to determine. Plus, of course, there is the problem of which laws govern the work in question.

The rest of the discussion seems awfully complicated to me, and I still worry about anything that can be construed as ISFDB's having responsibility to determine whether something is "legally" posted. I feel we are over-thinking it. I do not like the suggested date-specific criteria at all (US-specific, doesn't allow for truly legal postings of copyrighted material, for two things). I think ISFDB's position should be that if a court with appropriate jurisdiction: (a) decides the availability is ok, then it is "legal" and (b) decides the availability is not ok, then it is "illegal". That could be widened a bit at both ends to allow for temporary orders in one direction or the other (e.g., an injunction blocking specific posting(s) or an entire site's content provision, or the denial of a motion to grant such an injunction, until an open case is decided). Beyond that, I believe the ISFDB should have a policy of allowing links, with some process for case-by-case exceptions (e.g., the copyright holder or their representative asks to have the link removed) and perhaps one bigger exception where if multiple copyright holders make the same complaint/request involving the same source the ISFDB might decide to prohibit links to anything hosted by the same entity. IANAL, but we're not producing, hosting, or delivering the content, merely documenting where it is available on the internet. If ISFDB's having a link constitutes copyright infringement or abetting copyright infringement, then that would be the case for every search engine as well. --MartyD (talk) 16:24, 25 April 2024 (EDT)

Accounting for Different Jurisdictions

Different jurisdictions have different copyright laws, which often means that a text may be under copyright in one jurisdiction and out of copyright in another jurisdiction. The US/Australian versions of Project Gutenberg are probably the best known example of the resulting divergence. For this reason I would like to change the proposed Help language from:

• texts known to be under copyright protection and made available without the copyright owner's permission

to:

• texts known to be under copyright protection in the jurisdiction where the third party Web site is hosted if it is made available without the copyright owner's permission

(This change doesn't account for the "internet lending libraries" exception discussed above since it will need additional consideration.) Ahasuerus (talk) 16:04, 25 April 2024 (EDT)

Looks good to me. ···日本穣 · 投稿 · Talk to Nihonjoe 13:09, 21 June 2024 (EDT)

Audio Book Publishers for Reissues

Ordinarily when entering a publisher I first what is listed on the title page. For Audio Books, I feel that the introduction to the book is the equivalent or if the publisher is not listed there, they may be mentioned in the end credits. In either case, I would take the publisher as stated in the actual audio. However, I'm not sure whether this is best when audio books are reissued by a new publisher using the same recording as the earlier publisher. I recently entered title this title which I recently purchased. The end credits of the book state that the book is published by BBC Audiobooks. However, Audible claims that it is published by Audible Studios and provide a cover with their "Only from Audible" banner. The also list the release date as 2007-10-18. I did some research using archive.org's wayback machine and was ableo to find Audible's listing from 2014. Where they list the publisher as "AudioGO Ltd." (AudioGO acquired an 85% stake in BBC Audiobooks in 2010 per Wikipedia). Additionally the cover at that time bore the BBC Audio logo. The 2018-02-18 archive of the listing now has the publisher as "Audible Studios" but still has the BBC logo on the cover. The 2015-04-20 archive has an altered cover removing the BBC logo and lists the publisher as Audible Studios. Finally, the 2020-04-07 archive alters the cover to include the Audible banner. No publisher is listed in this archive, but the publisher is still listed as Audible Studios in the current listing. I created records for each version of the cover that I could find. However, I maintained the publisher as BBC Audiobooks, as that is what is stated within the recording. After I had done this, I did notice a copyright statement included in the metadata after I had imported this into iTunes reading "Public Domain (P)2014 Audible, Inc.". My theory is that even when an audio book is reissued by a new publisher, they do not bother re-record the publisher credit and I know I have seen other examples of this. My question is that in these situations, should we go to a source outside of the recording to determine the publisher? Should we consider the stated publisher in sale listings? Should we consider logos in the cover? Despite how I entered these, I'm thinking I should consider the stated publisher in the Audible listings which would necessitate adding another record for the AudioGO publication. What do others think? Thanks. --Ron ~ Rtrace^Talk 17:24, 19 June 2024 (EDT)

To go along with this, I'm finding that Amazon has become more unreliable in listing the publisher correctly. It's become fairly common for them to list the current owner of a publishing house/imprint for a recording that was published before the actual audiobook publisher was acquired by the current imprint owner. The publisher listed on Audible.com in the summary data tends to be that same as the one listed on Amazon.com but fairly often the copyright data shown in the same Audible listing contains the actual publisher. A lot of times, the logos on the cover will also match the correct publisher. Phil (talk) 17:46, 19 June 2024 (EDT)

We have a similar issue happening with POD printed editions of older books - a new publisher slaps new covers (sometimes just a back cover so they can put an ISBN/EAN) and then reprints the original book with copyright page and all in tact. The better ones at least add their own title page. At least one I caught last week was much worse (no link, sorry - if I remember which one it was, will post it): not only they had not changed a thing on the cover but even Amazon was reporting the publisher as the long defunct original one. At this point, I think all we can do is due diligence and document what we are seeing and where (in the case above I removed the publisher from the publisher field as the book is definitely not theirs...)

For the case above, I will use the external source to credit the publisher and add a note that it is the same recording (publishing credit and all) as the older one. Not much more we can do I guess. Annie (talk) 15:49, 23 July 2024 (EDT)

Movie shots

Did not find in NewPub Help: if the cover image of a book (flagged novelization or not) is a frame from a cinema movie, should we put the film director or the photography director as author, or leave it blank? thanks! --Fantagufo (talk) 14:28, 3 July 2024 (EDT)

Leave blank. Add a note into the publication notes that it is a still from the movie. We do not create COVERART records unless we have an actual artist. Annie (talk) 14:35, 3 July 2024 (EDT)

thanks Annie - that's exactly what I did until now... will continue. --Fantagufo (talk) 16:35, 3 July 2024 (EDT)

Updating the help to cover stills as cover images

So, in order to explicitly address the point above (since this seems to come up periodically), I suggest we update Help:Screen:NewPub#Cover Art by adding the following directly under the "Cover Art" header (not as a bullet point, but as a paragraph immediately following the header):

• If the cover art is a film, television, video, or video game still, do not add it as cover art. Instead, include a note in the Notes field stating something along the lines of "The cover art is a still from the film."

Thoughts? ···日本穣 · 投稿 · Talk to Nihonjoe 19:04, 5 July 2024 (EDT)

what about this:

• If the cover art is a film, television, video, or video game still, do not add it as cover art and leave the cover arts fields empty. Instead, include a note in the Notes field stating something along the lines of "The cover art is a still from the film [Title of the film], [directed by XXX] and/or [year of the film release]" or something otherwise useful to distinguish films with the same title, remakes, etc. --Fantagufo (talk) 19:14, 5 July 2024 (EDT)

Sure, that's fine. ···日本穣 · 投稿 · Talk to Nihonjoe 18:53, 11 July 2024 (EDT)

We may want to change "useful to distinguish films with the same title, remakes, etc." to "useful to distinguish media works with the same title, remakes, etc."

Re: "leave the cover arts fields empty", it would work OK for pubs with a single COVERART record, but some pubs -- e.g. dos-a-dos books -- have two covers. It's possible for one of them to be a still and the other one to be real cover art. I would suggest limiting the Help text to "do not add it as cover art."

Also, just a note that the proposed change would be made to Template:PublicationFields:CoverArt, which is linked from Help:Screen:NewPub, Help:Screen:NewNovel and Help:Screen:EditPub. Ahasuerus (talk) 20:39, 11 July 2024 (EDT)

So something like this?

• If the cover art is one or more stills from a film, television, video, or video game, do not add it as cover art. Instead, document it in the Notes field. Wording such as "The cover art is a still from the MEDIATYPE MEDIANAME (YEAR)." may be used to avoid confusion between similarly-named works, remakes, and so on.

How does that look? ···日本穣 · 投稿 · Talk to Nihonjoe 13:28, 12 July 2024 (EDT)

Looks reasonable to me. Ahasuerus (talk) 18:10, 12 July 2024 (EDT)

Are there objections to the proposed Help change? Ahasuerus (talk) 18:09, 20 July 2024 (EDT)

Not an objection per se, but sometimes a cover is a photograph taken on set and although it looks like a still, it's not. One of the Midwich Cuckoo covers comes to mind. I think there's another one, but I can't remember where it is, where the back cover only has the film credits (I think it was the Ladd Company) but one of our editors had found the photograph credit on the artist/photographer's website. How do we want to cover these? - I'm sure there are already credited works in the db. Apologies for the late contribution. Kev. --BanjoKev (talk) 14:38, 21 July 2024 (EDT)

An interesting point. So something like:

• If the cover art is one or more still frames from a film, television, video, or video game, do not add it as cover art. Instead, document it in the Notes field. Wording such as "The cover art is a still frame from the MEDIATYPE MEDIANAME (YEAR)." may be used to avoid confusion between similarly-named works, remakes, and so on. Note that "still frames" are not the same as "production stills", which are photos taken during film/TV episode production; the latter should be recorded according to the regular rules listed above.

Ahasuerus (talk) 16:29, 21 July 2024 (EDT)

That nails it for me, beautifully! Thanks. Kev. --BanjoKev (talk) 16:48, 21 July 2024 (EDT)

Looks good to me. ···日本穣 · 投稿 · Talk to Nihonjoe 15:22, 23 July 2024 (EDT)

Still frames and Production stills -- Outcome

Template:PublicationFields:CoverArt has been updated with the text proposed above. "Additional Notes on Cover Artists" has been split into 3 bullet points for readability. Rules and standards changelog has been updated. Ahasuerus (talk) 11:35, 24 July 2024 (EDT)

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

Last year at the end of August, a discussion of the use of captions for Interior Art titles was begun here but somehow was just dropped without any resolution. Could we please continue/restart that discussion?

Here's the existing text in the Regular Titles Help:

"Artwork. Interior art should have the same title as the fiction or essay it is associated with. If it is independent of other content, and has no apparent title or caption, give it the title of the publication in which it appears, disambiguating if necessary. [Note: occasionally a work may actually be titled "Untitled" which can correctly be given as the title of the work in the ISFDB record.] Cover art should have the same title as the title of the publication that it is associated with. Artwork on the back cover of a publication is treated as interior art. For works with multiple illustrations per story (usually a magazine or anthology) where the illustrations are not individually named, use the format "Story Title" for the first illustration, then "Story Title [2]", "Story Title [3]", and so on."

I see some issues with the existing wording.

• At the very least, there's an ambiguity between the first two sentences. The second sentence implies that at title or caption can be used while the first sentence says it must be the name of the publication.
• I'm not sure why the Cover art sentence is included here since Cover Art has its own separate help section.
• I think the sentence "Artwork on the back cover of a publication is treated as interior art." should be moved to the Cover Art Help to make it more visible.

My opinion: We should be using artwork captions for titles. This includes using titles that are on the artwork for cases such as maps. Phil (talk) 15:29, 10 July 2024 (EDT)

The cover art line is probably there as the help is also linked from the Title record field of the Title editor window as well as from the Publication editor. When editing a title record, it could be a cover art title record. However, none of the other help windows displayed from the Title editor seem to mention cover specific items. It might be better to drop cover art items from the Title help and instead add a statement that says "for COVERART see the CoverArt help". -- JLaTondre (talk) 21:06, 11 July 2024 (EDT)

Maybe something like this?

"Artwork. Interior art (which includes artwork on the back of a publication) should use the title or caption included with the artwork, or (if no title or caption is included) have the same title as the fiction or essay it is associated with, disambiguating as necessary. If it is independent of other content, and has no apparent title or caption, use the title of the publication in which it appears, disambiguating if necessary. [Note: occasionally a work may actually be titled "Untitled" which can correctly be given as the title of the work in the ISFDB record.] Cover art should have the same title as the title of the publication that it is associated with. Artwork on the back cover of a publication is treated as interior art. To disambiguate stories with multiple untitled or uncaptioned illustrations (usually a magazine or anthology) where the illustrations are not individually named, use the format "Story Title" for the first illustration, then "Story Title [2]", "Story Title [3]", and so on."

Here's a cleaner version that may be easier to read:

"Artwork. Interior art (which includes artwork on the back of a publication) should use the title or caption included with the artwork, or (if no title or caption is included) have the same title as the fiction or essay it is associated with, disambiguating as necessary. If it is independent of other content, and has no apparent title or caption, use the title of the publication in which it appears, disambiguating if necessary. To disambiguate stories with multiple untitled or uncaptioned illustrations (usually in a magazine or anthology), use the format "Story Title" for the first illustration, then "Story Title [2]", "Story Title [3]", and so on."

How does that look? It's even shorter than the original. ···日本穣 · 投稿 · Talk to Nihonjoe 13:19, 12 July 2024 (EDT)

I think that's very clear. I like it. So the next logical question: If we are using the title or caption, do we then make variants for the same title/caption that is different from the oldest version and disambiguate the variant title names? Although I can't point to a specific example, I know there are maps used in some series that have the same title and artist but which have minor differences and often different copyright dates depending on when they were published. I think it could look something like "Map of A" for the oldest and then possibly "Map of A (1990)" for a later variant. Phil (talk) 17:28, 12 July 2024 (EDT)

I think that would work. ···日本穣 · 投稿 · Talk to Nihonjoe 17:37, 12 July 2024 (EDT)

(unindent) I believe the reason that the 2023 discussion was inconclusive is that it tried to cover a number of issues, which made it difficult to reach consensus. With that in mind, we may want to start with the low-hanging fruit that we can all agree on. JLaTondre's proposal above:

It might be better to drop cover art items from the Title help and instead add a statement that says "for COVERART see the CoverArt help".

seems like a good start.

Re: the core issue of deciding how INTERIORART titles should be entered, I think we should start by separating:

1. what the current version of Help says explicitly
2. What the current version of Help implies
3. what the current de facto data entry standard is
4. what we want the new standard to be

Re: "what Help currently says", I see at least two separate Help templates that currently cover this topic. The "Artwork" paragraph quoted at the beginning of this discussion comes from Template:TitleFields:Title. In addition, Template:TitleFields:TitleType also describes how INTERIORART titles should be entered:

• In this case there is only one illustration for the story, but if there were multiple illustrations a single title would still suffice. [snip] the title is the title of the content item being illustrated. [snip] If the illustration has a separate title or caption, document in the illustration's Notes field.
• If an illustration is independent of other content, and has no apparent title or caption, give it the title of the publication in which it appears, disambiguating if necessary.
• Maps. [snip] The format for titling maps is "Title of Work (map)", for example: Asylum (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).
• Cartoons. [snip] The title should be "Cartoon: " followed by the caption, in the original case, between quotation marks. If there is no caption the words "no caption" should be used without quotation marks.

So the first question that we should probably answer is whether we want to consolidate the Help rules for INTERIORART titles -- which are currently split between two templates -- in one place in order to avoid duplication and future divergence. Ahasuerus (talk) 16:07, 14 July 2024 (EDT)

At the very least, we need to have one single set of Help rules for INTERIORART titles. I had frankly missed seeing the Help in Template:TitleFields:TitleType which give an editor a lot more entry options than those in Template:TitleFields:Title; had I known about them, I would have been using them, particularly for map titles. I think the consolidated Help will also need to discuss the correct use of variants for those titles. Phil (talk) 22:29, 14 July 2024 (EDT)

Content help section of new publication entry

In the Content Information section of the Help:Screen:NewPub page, the first paragraph states the following (emphasis added):

The third section of the data entry form is used to enter the publication's contents. It is in turn divided into three sub-sections. The first sub-section, "Content", includes all items in the publication, including reviews and interviews. (For novels, do not re-enter the main Novel title, which you already entered at the top of this Web page, in this section.) The second sub-section, "Reviews", and the third sub-section, "Interviews", provide additional details about the reviews and interviews. They are to be used in addition to the entries for those items in the Contents section.

I think this wording is a little confusing, because it could be interpreted to mean that reviews and interviews should be entered in both the main content section as well as their individual sections. This hasn't been the case as long as I can remember. I propose rewording it to something like this (changes marked in bold):

The third section of the data entry form is used to enter the publication's contents. It is in turn divided into three sub-sections. The first sub-section, "Content", includes all items in the publication, except reviews and interviews. (For novels, do not re-enter the main Novel title, which you already entered at the top of this Web page, in this section.) The second sub-section, "Reviews", and the third sub-section, "Interviews", are where reviews and interviews (respectively) should be entered. In cases where a review is of a non-genre work but was written by a genre author, editor, or artist who is already in the database, the review should be entered as an ESSAY in the main Contents section rather than in the Reviews section.

Thoughts? Not sure if there are special cases for interviews as I've never run into one. ···日本穣 · 投稿 · Talk to Nihonjoe 20:06, 11 July 2024 (EDT)

The last part should be something like "In cases where a review is of a work ineligible for inconclusion in the database, the review should be entered as an ESSAY in the main Contents section rather than in the Reviews section." The proposed wording would allow reviews of non-genre books by non-genre authors. If the work is eligible for inconclusion in the database (even if non-genre), then there is no issue with having it listed as a review. The issue is to avoid Authors That Exist Only Due to Reviews or people adding ineligible books because there is a review. -- JLaTondre (talk) 20:43, 11 July 2024 (EDT)

As for interviews, I've often thought we should have a rule similar to reviews such that interviews should only used for people who have written works eligible for inconclusion. If the interviewee has not written works eligible for inconclusion, then it should be entered as an essay. We have a number of entries for people who have never written a genre work (many are actors like this one). -- JLaTondre (talk) 20:54, 11 July 2024 (EDT)

So, something like this (addressing both of your comments)?

The third section of the data entry form is used to enter the publication's contents. It is in turn divided into three sub-sections. The first sub-section, "Content", includes all items in the publication, except reviews and interviews. (For novels, do not re-enter the main Novel title, which you already entered at the top of this Web page, in this section.) The second sub-section, "Reviews", and the third sub-section, "Interviews", are where reviews and interviews (respectively) should be entered. In cases where a review is of a work normally ineligible for inclusion, but the review is by an eligible genre author, editor, or artist, the review should be entered as an ESSAY in the main Contents section rather than being entered in the Reviews section. For interviews, if the interviewee has not produced eligible works, and if the interviewer is an eligible genre author, editor, or artist, the interview should be entered as an ESSAY in the main Contents section rather than being entered in the Interviews section. If both the interviewer and interviewee are ineligible for inclusion, the interview should be documented in the Notes section.

How does that look? ···日本穣 · 投稿 · Talk to Nihonjoe 12:42, 12 July 2024 (EDT)

Looks good to me. -- JLaTondre (talk) 17:57, 12 July 2024 (EDT)

(unindent) I believe that the current wording of this Help paragraph was an attempt to explain how to enter "review columns". Consider Science Fiction Adventures, September 1953, which includes 3 reviews by Damon Knight on pages 92-95. Each review is entered as a REVIEW title, but the pub also includes an ESSAY titles, "The Dissecting Table (Science Fiction Adventures, September 1953)", for the review column as a whole. That's how most magazine reviews are currently entered.

That said, I agree that the current wording is confusing and probably misleading. Once we are all on the same page re: the underlying idea, we'll want to come up with a better explanation. Ahasuerus (talk) 19:06, 12 July 2024 (EDT)

Perhaps for the review columns, have a sentence like this added to the above?

For review columns, where the reviewer is giving commentary outside of the specific reviews in the column, create an ESSAY entry by the reviewer for the column as a whole in the Contents section in addition to the appropriate REVIEW entries.

I think that captures what you mentioned. Thoughts? ···日本穣 · 投稿 · Talk to Nihonjoe 19:57, 12 July 2024 (EDT)

Help:Screen:EditPub and Help:Screen:NewPub already cover this topic. However, comparing Help:Screen:EditPub#Reviews and Help:Screen:NewPub#Reviews, I see that their introductory sections used to be identical but have diverged over time. Now one of them contains a single paragraph while the other one has two paragraphs. Each section has obsolete AND more up-to-date sentences compared to the other one.

Since the text is supposed to be identical except for the number of displayed blank REVIEW records, I suggest that we start by coming up with a single, up-to-date version of this section and putting it into a new Help template to be transcluded on Help:Screen:EditPub and Help:Screen:NewPub.

In addition, the rules that determine which reviews to include in the "Reviews" section are currently split between this section and Template:PublicationInfo:WhatToInclude. I propose that we consolidate them in order to avoid confusion.

Here is my first draft of the proposed new template followed by proposed changes to the related templates:

• This section is used to enter reviews. Note that magazine/fanzine review columns as well as individual reviews within a volume of critical essays are also entered as ESSAY types in the general content section -- see the "Regular Titles" section above for details.

• What to Enter: See the "Reviews" section of Template:PublicationInfo:WhatToInclude for guidance.

• Each review record contains five fields: page, title of the work being reviewed, first date of publication of the review, author(s) of the work being reviewed, and reviewer(s). If there are not enough review records displayed, click on the "Add Review" button at the bottom of the review section and a blank review record will appear. This can be repeated until sufficient review records are available to enter every eligible review in the publication.

Template:PublicationInfo:WhatToInclude will be updated with the following paragraph:

• Only reviews of works that are actually commented upon should be entered. If a reviewer mentions that a publisher has re-issued a work, but does not comment on the quality of the book, don't include it. Even a brief comment, such as "recommended", is enough to qualify, but without any comment the review should not be listed. For old reviews of Ace Doubles a reviewer would occasionally review one side of the book and simply mention the name of the book on the other side without comment. In these cases only enter the work that was actually reviewed. A review for a work of short fiction should be entered only if the reviewer specifically comments on that specific story. If several stories from the same collection or anthology are discussed, each briefly, recording this as a review of the collection or anthology may be preferable.

Template:TitleFields:TitleType, which currently says:

• Review columns and interviews are also entered as ESSAYs.

will have the following sentence from the current version added:

• No ESSAY title is needed for a single, standalone review (one review with no additional text) unless it is a regular feature with a recurring title.

Note that I have rearranged some parts of the Help text for readability, but I didn't touch "individual reviews within a volume of critical essays" in the first paragraph and the last sentence immediately above because they are rather convoluted. They may require additional brainstorming. Ahasuerus (talk) 12:37, 13 July 2024 (EDT)

Are there objections to the proposed consolidation of what is currently stated in Help:Screen:EditPub and Help:Screen:NewPub and putting the reconciled text above in a Help template? Ahasuerus (talk) 18:08, 20 July 2024 (EDT)

Entering REVIEWs -- intermediate results

As per the discussion above:

• The introductory sections of Help:Screen:EditPub and Help:Screen:NewPub, which diverged over time, have been reconciled and put into Template:PubContentFields:ReviewSection‎
• Review eligibility requirements have been consolidated in the "Reviews" section of Template:PublicationInfo:WhatToInclude‎
• Instructions for creating ESSAY records for "review columns" have been consolidated in Template:TitleFields:TitleType‎

Rules and standards changelog has been updated.

Review columns -- Substantive issues

Now that the affected Help pages have been de-duplicated and reconciled, we can examine what they actually say and see if we want to adjust the instructions. Post-changes, Template:TitleFields:TitleType‎ says:

• Review columns and interviews are also entered as ESSAYs; no ESSAY title is needed for a single, standalone review (one review with no additional text) unless it is a regular feature with a recurring title.

I think what it's trying to say is something like:

• While individual reviews and interviews are entered in their respective sections of this data entry form, a separate ESSAY record should be additionally created for each review/interview column in the publication. A review/interview column is defined as a collection of multiple reviews/interviews under a single heading or a regular feature with a recurring title.

In addition, Help:Screen:NewPub#Content_Information says:

• The third section of the data entry form is used to enter the publication's contents. It is in turn divided into three sub-sections. The first sub-section, "Content", includes all items in the publication, including reviews and interviews. (For novels, do not re-enter the main Novel title, which you already entered at the top of this Web page, in this section.) The second sub-section, "Reviews", and the third sub-section, "Interviews", provide additional details about the reviews and interviews. They are to be used in addition to the entries for those items in the Contents section.

Help:Screen:EditPub#Content_Information starts with:

• When you are viewing or editing an ISFDB publication record, the data in the upper part of the page is from the publication record and the data in the lower (Contents) section is composed of ISFDB title records. The Contents section is in turn divided into three parts. The first part, "Regular Titles", includes everything in the publication, including review and interview columns. The second part, "Reviews", and the third part, "Interviews", provide additional details about the reviews and interviews, but do not substitute for a title in the contents section for the columns containing those items.

Once again, some of this text is duplicative and some is obsolete since we now have 4 (as opposed to 3) sub-sections in the Contents section, the 4th one being "Cover Art". I think we can consolidate them in a single Help template and change:

• NewPub: "Content", includes all items in the publication, including reviews and interviews.
• EditPub: "Regular Titles", includes everything in the publication, including review and interview columns.

to:

• "Regular Titles", includes every title record in the publication that is not COVERART, REVIEW or INTERVIEW

The last few sentences will be changed to:

• The third sub-section, "Reviews", is used to enter review records. The fourth sub-section, "Interviews", is used to enter interview records.

We'll also want to add a line about the Covert Art section along the lines of:

• The first sub-section, "Cover Art", includes cover art records. In most case there is only one cover art record per publication, but some publications, e.g., "dos-a-dos" books, have multiple cover art records.

and re-number the sub-sections since the Covert Art sub-section is displayed first.

In addition, we may want to discuss the issue of INTERVIEW eligibility, specifically what to do about interviews with non-genre authors, but it may be better handled as a separate R&S discussion.

Thoughts? Ahasuerus (talk) 16:35, 24 July 2024 (EDT)

Use of "Untitled", "untitled", and "(untitled)" in titles

Transferred from this discussion, copied here since it's not terribly long:

Curious why you're submitting changes for "untitled" to "Untitled". Was there a discussion that said they should all be capitalized? I'm aware of a related discussion here, but it doesn't seem to have anything to do with changing the capitalization on all of these. This discussion didn't have much input, but seems most closely related to what you are submitting. Thanks for your time! ···日本穣 · 投稿 · Talk to Nihonjoe 19:06, 11 July 2024 (EDT)

Pardon the uninvited comments. Untitled content titles, excluding artwork and story introductions, are not specifically addressed in the help section. That leaves us with the regular case rules. For English, "the first word is capitalized". I would approve the submissions.

It wouldn't hurt to have a R & S discussion to clarify the help. Note this from the help on author attribution. The proper case for each is illustrated.

• 'Anonymous or uncredited works. If a work is credited to "Anonymous", then put "Anonymous" in the author field. The same applies for any obviously similar pseudonym, such as "Noname". If the work is not credited at all, use "uncredited"... '.

John Scifibones 11:29, 12 July 2024 (EDT)

The rules require the title (Title field) of a piece of work to be regularized, and according to language conventions. "Untitled" is a title and therefore is not excluded from this rule.

The rules also require the use of "uncredited" (Author field) where there is no credit at all.

The difference in treatment between the two fields is the source of a lot of confusion, it seems to me, and could be explicitly addressed in the Title help. Kev. --BanjoKev (talk) 12:23, 12 July 2024 (EDT)

I guess it boils down to whether "untitled" is considered an actual title or simply a space filler since the field is required. To me, "Untitled" means the work is actually titled "Untitled", whereas "untitled" (or "(untitled)", which I've seen on occasion) means the work doesn't have a title and that word is simply a filler or placeholder since the field requires something be entered. ···日本穣 · 投稿 · Talk to Nihonjoe 13:52, 12 July 2024 (EDT)

It so happens that I noticed this issue while running a bunch of unrelated Advanced Searches a few months ago. I realized that some "untitled" titles were capitalized and some weren't. I considered starting a Rules and Standards discussion about this issue, but decided to postpone it until we had no active R&S discussion in progress. And then I forgot all about it. Naturally, I am all in favor of discussing it on the R&S page. Ahasuerus (talk) 15:15, 12 July 2024 (EDT)

Now that you've read that, Here's my take on the situation. These can be essays, artwork, poems, and possibly even stories (though this last one is fairly rare). I've seen the following used:

• untitled (Disambiguator)
• Untitled (Disambiguator)
• (untitled) (Disambiguator)

There may be others ways it's been entered, but I think these three cover the vast majority. I think we need to settle on some guidelines to achieve some consistency in how these are entered, and then clearly document them in appropriate places.

First, we need to determine whether to treat instances of "untitled" as described above as actual titles or as placeholders. If the former, then the existing guidelines here (which are transcluded in multiple places) will apply and it should always be capitalized. If the latter, then we need to determine what (if any) guidelines should be created to indicate how placeholders such as this should be capitalized (or not).

If we go the "create new guidelines" route, I suggest using "untitled (Disambiguator)" for works that have no title (such as an untitled essay, poem, or artwork), and using "Untitled (Disambiguator)" for works that are titled "Untitled". This way, we can see at a glance which is which.

Thoughts? ···日本穣 · 投稿 · Talk to Nihonjoe 16:51, 12 July 2024 (EDT)

My preference would be capitalized when actually titled Untitled and lowercase when there is no title. For stories and artwork, that is quick visual indication of the two cases. For poems, it is still ambiguous since we allow lowercase poem titles. However we do it, there should be title notes explaining the circumstances. -- JLaTondre (talk) 18:00, 12 July 2024 (EDT)

I agree with JLaTondre's approach: "Untitled" when the title is actually spelled "Untitled" and "untitled" otherwise. The issue of disambiguation, especially as it applies to different title types like INTERIORART vs.SHORTFICTION, is a separate can of worms and may require additional thought. Ahasuerus (talk) 18:51, 12 July 2024 (EDT)

(unindent) Also, for reference purposes, here is what we currently have in the database:

Lowercase "untitled":

+--------------+----------+
| title_ttype  | count(*) |
+--------------+----------+
| INTERIORART  |      180 |
| ESSAY        |       47 |
| INTERVIEW    |        2 |
| POEM         |       59 |
| REVIEW       |        3 |
| SHORTFICTION |       15 |
| CHAPBOOK     |        2 |
+--------------+----------+

Lowercase "untitled" followed by a space, a left parenthesis and a disambiguating value:

+--------------+----------+
| title_ttype  | count(*) |
+--------------+----------+
| INTERIORART  |      529 |
| ESSAY        |      216 |
| INTERVIEW    |        7 |
| POEM         |      175 |
| SHORTFICTION |      165 |
+--------------+----------+

Uppercase "Untitled":

+--------------+----------+
| title_ttype  | count(*) |
+--------------+----------+
| INTERIORART  |      113 |
| ESSAY        |        9 |
| INTERVIEW    |        2 |
| POEM         |       85 |
| SHORTFICTION |       23 |
+--------------+----------+

Uppercase "Untitled" followed by a space, a left parenthesis and a disambiguating value:

+--------------+----------+
| title_ttype  | count(*) |
+--------------+----------+
| INTERIORART  |      277 |
| ESSAY        |      403 |
| POEM         |      190 |
| SHORTFICTION |      455 |
+--------------+----------+

Ahasuerus (talk) 18:51, 12 July 2024 (EDT)

I would also prefer "untitled" with small letters when we use it as a placeholder (with or without disambiguation) and properly capitalized Untitled if the title happens to be called that way makes it the least obtrusive and least likely to confuse and require extensive notes. It is technically against the current capitalization rules but it can be easily added as an exception if we decide to. Annie (talk) 19:02, 12 July 2024 (EDT)

For page count, square brackets are used to highlight pages that exist with content but are out of official numbering e.g. `123+[4]'; the same convention could be applied here for non-titles created by the editor producing, with possible disambiguation, something like `[untitled introduction] (Book Title)' or `[untitled biography] (Author Name)' --Fantagufo (talk) 04:26, 13 July 2024 (EDT)

It's true that we use square brackets for unnumbered pages in the "Pages" field. However, we don't use square brackets in the "Author" field when no author is credited. Template:PublicationFields:Author says:

• If the work is not credited at all, use "uncredited". If you are working from a secondary source which does not specify the author, but does not explicitly state that no author is credited in the publication, use "unknown" rather than "uncredited".

I think that, in terms of function, the "Title" field is closer to the "Author" field than to the "Pages" field and would be a better precedent for rule-making. Ahasuerus (talk) 11:40, 13 July 2024 (EDT)

Ok. But would you allow to specify the kind of text, e.g. `untitled introduction', or just `untitled' alone? --Fantagufo (talk) 15:37, 13 July 2024 (EDT)

Disambiguation of untitled works is a tricky can of worms. We have a few dozen interviews whose title starts with "Untitled interview (title of the publication that the interview was published in)". We also have 4 interviews whose title is simply "untitled". Clearly, untitled INTERVIEWs mostly use the "untitled interview (publication title)" format.

On the POEM side, the situation is somewhat different. There are 144 POEMs whose title is currently "untitled" and 650 POEMs with some kind of disambiguation, mostly of the "untitled ("First line of the poem")" variety, but also of the "untitled (title of the publication that the poem was published in)" variety. Some use "untitled poem", "untitled haiku", "Untitled [1]", etc. A couple of dozen use "(untitled)" and one uses "Poem (untitled)".

SHORTFICTION is a mix of "Untitled", "untitled (first line of the text)", "Untitled (title of the publication that the story was published in)" (mostly used for Drabblecast), "untitled fragment (first line of the fragment)", "untitled synopsis (first line of the text)", "untitled synopsis (title of the story)", etc.

There is a significant amount of variety, so my current thinking is that it may be better to address the issue of capitalization and parentheses -- i.e., "untitled" vs. "Untitled" vs. "(untitled)" -- first and discuss disambiguator-related issues like "untitled" vs. "untitled fragment" later. Ahasuerus (talk) 18:52, 13 July 2024 (EDT)

Determining "untitled", etc.

Okay, here's a proposal to standardize the use of "untitled" in its various forms:

For works that have no obvious title or name given to them within the containing publication, use the following formatting:

• If the work is INTERIORART, use the format "Story Title" for the first illustration, then "Story Title [2]", "Story Title [3]", and so on.
• If the work is SHORTFICTION, ESSAY, or POEM, use "untitled (Publication Title)" for the first such work in the publication, then "untitled [2] (Publication Title)", "untitled [3] (Publication Title)", and so on.

For works actually titled "Untitled", use "Untitled (Publication Title)" for the first such work in the publication, then "Untitled [2] (Publication Title)", "Untitled [3] (Publication Title)", and so on.

Do not use "(untitled)", "[untitled]", or any other such formatting.

How does that look? I'm not sure about using the "(Publication Title)" disambiguator on works actually titled "Untitled" since we don't disambiguate same-titled works in general, unless there are multiples by the same person within the same publication. ···日本穣 · 投稿 · Talk to Nihonjoe 15:11, 23 July 2024 (EDT)

I would prefer using the lower case untitled for works that are actually not titled. I believe that the upper case "Untitled" should be reserved for works that are credited on their title page with a literal "Untitled". I'd also like to call folks attention to this discussion where there seemed to be an agreement about using the first line of untitled poems. I know I have used the first as a disambiguator for both untitled poems and short fiction (e.g. untitled ("A spaceship landed in the park. An alien emerged...")). Additionally, I'd recommend against using publication title as part of the title. This will necessitate making variants each time the item is reprinted in a new publication, rather than a simple merge. --Ron ~ Rtrace^Talk 18:45, 23 July 2024 (EDT)

Interspersed unpaginated content

Guidance for dealing with included content on unnumbered pages before page 1 or after the last numbered page is found in bullet point four under Paper Books here. There is an additional situation that is not enumerated. What should the Pages field show when the included content is interspersed throughout the text? I can think of several, hopefully we can agree on one.

Example: The text ends on numbered page 300. There are 6 unnumbered artwork plates. The content may be 1 or 6 titles.

1. The Pages field remains 300, + a publication note re: the 6 extra pages.
2. The Pages field is 300+[6], + a publication note explicitly stating the unnumbered material is interspersed throughout the text. (Without the note, someone might remove the [+6] if they don't see a content title after page 300.)
3. The Pages field is 306, + a publication note that the Pages field includes 6 unnumbered pages interspersed throughout the text. (Without the note, 306 might be deemed a typo.)

All three have positives and negatives. I prefer #2, but am fine with any of the three. Thoughts? John Scifibones 10:39, 23 July 2024 (EDT)

I prefer #2 (and have already done some like that). ···日本穣 · 投稿 · Talk to Nihonjoe 14:39, 23 July 2024 (EDT)

Though if there are 6 unnumbered artwork plates, it's likely there are 12 additional pages (the plate for each, and the reverse page for each), so it would be "300+[12]", right? ···日本穣 · 投稿 · Talk to Nihonjoe 14:53, 23 July 2024 (EDT)

I prefer #2 as well (and a note is always encouraged when just the page numbers on the contents are not enough). Annie (talk) 14:49, 23 July 2024 (EDT)

I prefer #1. If I see "+[6]" would absolutely lead me to expect 6 unnumbered pages at the end of the book. If we're going to have a note explaining there are dispersed unnumbered paged anyway, why alter the page count to possibly confuse folks. I am unaware of any other bibliographical sources that record plates in page counts. I have encountered page counts done like #2 for plates and have removed the "+[x]" wherever I've encountered it as it is not currently supported by the standards. One additional thought, if we were to go with #2, how would we reflect a publication that included both interspersed plates and unnumbered pages following the numbered section? --Ron ~ Rtrace^Talk 15:07, 23 July 2024 (EDT)

See, that's why I dislike the OCLC style notation of pages -- I'd rather go with "if it is a piece of paper, count it and then explain what ends where" approach but that ship had sailed a long time ago. I am fine with #1 - I think that #2 is cleaner and follows the spirit of our rules better (count the pages with contents that matter and/or that are numbered) but I can see how it can be confusing if there is also material at the end of the book and in not making a distinction between the two.

OCLC will mark these as "300 + vi plates" or something like that. Maybe we need something like that - leave the number itself alone but come up with a new way to indicate plates. Of course that brings the question on how these were recorded in the older books where illustrations on plates were a lot more common than they are now in fiction books. Annie (talk) 17:39, 23 July 2024 (EDT)

I agree that using "300+[6]" to document interspersed unpaginated content would likely make our users think that there are 6 unnumbered pages at the end of the book. I would go with option #1. Ahasuerus (talk) 11:40, 24 July 2024 (EDT)

I also prefer #1. With option #2, there is already enough confusion about using square brackets in the page count. I'd agree with Ron in expecting additional pages (content) at the end. #2 and #3 are basically the same approach - they beg the question - where is the additional content at the end of the book? Better to have a straight page count and then, if one is particularly interested, one can read the notes for how the additional material is handled. Kev. --BanjoKev (talk) 12:17, 24 July 2024 (EDT)

I think #1 is the cleanest method. Phil (talk) 15:33, 24 July 2024 (EDT)

Definitely Option 1. The notations for the other two options are confusing because they are identical to completely different existing scenarios (6 unnumbered pages at the end and 306 numbered pages, respectively).

However, we should not just consider the Pages field in isolation. We should also consider what we want to do in the Start Page of the Content record. There are two situations:

A) There is one umbrella Interiorart Content record for all the artist's illustrations which are scattered throughout the book.

B) Each illustration has its own Interiorart Content record.

For A) there is no issue. Leave the Start Page empty as we currently do when the illustrations are on numbered pages.

For B) I think it is desirable to have each Interiorart Content record sorted appropriately. I can see two possibilities for achieving this. As an example, suppose an unnumbered plate with an illustration is interspersed between numbered pages 123 and 124:

i) use "|123.5" for the Start Page of the Interiorart Content record. This would appear blank but it would be sorted correctly.

ii) use "123.5" for the Start Page of the Interiorart Content record. This would display as 123.5 and would be sorted correctly.

I have encountered this situation before and, in the absence of a rule, I used option i). I slightly prefer this but have no strong objection to option ii).

Incidentally, I do not know whether the software supports non-integer values in the absence of or preceding a pipe (|). If we decide on option ii), it may require a software change. Teallach (talk) 17:41, 24 July 2024 (EDT)

The software accepts pretty much any value to the left of the "|" character. The reason is that we have no control over what publishers may decide to use as page numbers. Values ranges like B1 through B88 are not unknown.

If the page number happens to be a valid decimal number and there is no "sorting" number, then the software will sort it correctly. Ahasuerus (talk) 20:38, 25 July 2024 (EDT)

HarperCollins eBooks Version Numbers

Many, perhaps all, Ebooks published by HarperCollins have a version number on their copyright page that appears to be in the format of a date (MMDDYYYY). I'm not the first editor to notice this, see these publications. My question, is whether we can take this value as a publication date. For example, my copy of Neil Gaiman's Anansi Boys is nearly identical to this publication. However, it has a version number of "08272014". It also contains an excerpt from 2013 which indicates a publication date after the stated copyright (August 2005) or the 2009 publication that we have. A 2014 date works with the additional content. I would think dating my publication 2014-08-27 with a note that it's derived from the version number makes sense. Thoughts? Thanks. --Ron ~ Rtrace^Talk 20:28, 25 July 2024 (EDT)

As long as we do not lose the original publications once the dates change - aka let's make sure these are clones and not updates. :)

One thing that worries me a bit is how often do these change and do we really want separate record every time the publisher decides to change something and reupload the files. Do we really want 30 copies of the same ebook just because the date changed if the cover and the contents did not? I know that we do that for standard printing books but we don't for PODs (which also have dates very often). I think of these dates on the ebooks more in line with the printing dates at the back of PODs than actual publication dates. But we also do not have a good way to date ebooks which had changed something (adding new excerpts for example or even changing covers), especially now that Amazon is getting even worse with dating digital media on changes.

So we probably can use these dates to date with the provision that it should not create an exact duplicate ebook (in which case the new dates go only in the notes). Which may be harder to implement than to write but I am still mulling on all that. Annie (talk) 12:22, 26 July 2024 (EDT)

I agree that "a version number of 08272014" appears to be similar to the way POD books are versioned. As with PODs, I wouldn't create a separate publication record if the data were otherwise identical. Since Ron's copy has an additional excerpt, I would create a new pub record, use 2014-08-27 as the date and document its source in Notes. Ahasuerus (talk) 12:31, 26 July 2024 (EDT)