Rules and standards discussions

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

Archive Quick Links
Archives of old Rules and standards discussions.

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

Expanded archive listing

Rules and standards changelog

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

Old UK prices

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

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

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

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

Are photographs interiorart?

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

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

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

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

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

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

Applying awards to canonical author vs credited author

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

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

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

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

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

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

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

Award inclusion policy

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

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

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

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

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

Hopefully this covers everything that we currently include.

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

"Included: Awards given, authorized or sponsored by:

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

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

Note the following language in the last sentence:

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

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

Invalid, inappropriate, and non-ISBN ISBNs

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

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

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

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

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

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

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

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

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

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

Place of birth: Use of subdivisions for large cities

Template:AuthorFields:BirthPlace currently says:

Use the "City, Administrative division, Country" format

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

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

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

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

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

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

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

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

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

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

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

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

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

Transliteration

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

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

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

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

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

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

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

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

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

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

Image File Size Limit

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

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

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

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

Title Series Numbering

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Template:TitleFields:SeriesNum says:

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Webzines inclusion: Proposed extension of ROA

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Outcome -- Policy changed

ISFDB:Policy#Included has been changed to include all webzines. Rules and standards changelog has been updated. Ahasuerus (talk) 16:34, 22 October 2022 (EDT)

Artwork Records for Non Published Artwork

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

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

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

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

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

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

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

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

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

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

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

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

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

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

I think we've got a consensus here although with one dissent. I don't think we need to update ISFDB:Policy as it already excludes works that are unpublished. I'm going to wait for another day in case anyone else has something to add to this discussion. After that I will reject the edits that are being held. I'll also fix any records that I encounter for unpublished artwork by collapsing them to the first title for the first published instance and moving creation date and other data to the notes. --Ron ~ Rtrace^Talk 09:14, 30 October 2022 (EDT)

The non-genre covers (Again)

With more and more magazines getting reprinted, we have somewhat of a weird situation happening with non-genre fiction (pulps and other early ones) magazines:

If the cover does not illustrate our contents, it is not eligible to be added either as a cover art record or as an image in the magazine record.
When the magazine gets reprinted (print or ebook), it is entered as an anthology as per our rules and that makes both the cover art and the image eligible for addition.

The latest example: the magazines and its reprint. As this seems to happen only with fiction magazines (non-fiction ones (almost) never get reprinted the same way the early fiction magazines are getting reprinted these days), I propose to change the rules on non-genre magazines a bit - allow the covers for all non-genre fiction and predominantly fiction magazines. That will leave out Playboy's and similar covers (which was the main reason for the rule as I understand its history) but will allow us to get the pulp covers (so when they get reprinted, we do not end up in the situation above). Thoughts? Annie (talk) 14:20, 7 November 2022 (EST)

I think there are two separate and somewhat coupled policies here:

Do we add the artists name, when known, creating a COVERART title record that can be related to reprints?
Do we allow the Image URL field to be populated either from a scan hosted by us, or one of the sites that we have permission to link to?

I am in favor of allowing this data for both cases whether or not the magazine is primarily fiction. My reasoning for the first case is that if these were non-genre books instead of magazines, there would be no question of including the cover. I'm happy to support at least as far as your proposal, but would go further. For the URL field, I again think that we should allow this in all instances regardless of focus of the magazine. An image allows folks to identify the magazine in question. I can think of no reason we would want to prevent this data from being shown. This was discussed many years ago and I held the same position then. I recall that concerns were raised about storing additional images on our servers, but I also recall that later in the discussion someone (Ahasuerus?) concluded that this was not really an issue. --Ron ~ Rtrace^Talk 18:48, 7 November 2022 (EST)

That's right. Hosting additional images shouldn't create noticeable disk space issues. Ahasuerus (talk) 10:27, 10 November 2022 (EST)

Besides my sense is that the majority of the images we are likely to add would be linked to Galactic Central. --Ron ~ Rtrace^Talk 18:48, 7 November 2022 (EST)

Both are related enough to keep them together I think - we do not have a case where we allow one of them but disallow the other one (except the unknown artist case where you have an URL/image but not a coverart but it is irrelevant here really).

I was looking for a small scale gradual change but I am not opposed to dropping the difference altogether and treating all covers the same - if the publication (magazine or book) makes it to the DB, its cover is eligible both as a coverart record and as a URL/image if no other policy applies. I remember seeing a discussion from way back around risque covers (Playboy for example) being cited as the main reason not to allow the covers (I can try to look for it but I have no memories where it was - and it had been a few years since I went through the old discussions). Those don't get put on books but there are a lot of books with at least as explicit covers anyway so not sure how much that really matters. Annie (talk) 19:13, 7 November 2022 (EST)

I'm in favor of having the same rules for non-genre books and magazines. I also believe it would be easier to allow non-genre magazine covers & cover artists than trying to also restrict non-genre books covers. We already have quite a few non-genre magazine covers in the database - the differing rules is confusing to new editors (and some moderators) in my opinion. -- JLaTondre (talk) 16:34, 11 November 2022 (EST)

I agree that it's probably easier to treat all covers the same way. Additional disk space isn't really an issue and streamlining all rules would help keep things straight. Ahasuerus (talk) 22:35, 11 November 2022 (EST)

I agree. --MartyD (talk) 08:35, 12 November 2022 (EST)

It would appear that the consensus is to lift all restrictions on entering cover image URLs and cover art artists for non-genre publications, including magazines. If we go ahead with it, then Help:Entering non-genre periodicals will be affected. The current rule is:

Do not enter a cover artist, nor a cover image URL. Leave both fields blank. Exception: if the cover art illustrates the SF content, or is by a well known SF artist, enter the credit, and if an image is available, enter the URL (See ISFDB:Image linking permissions.)

This will need to be changed to something like:

Enter the cover artist if known. Enter the cover scan URL if an image is available online and we have permission to link to it (see ISFDB:Image linking permissions for details.)

Anything else? Ahasuerus (talk) 22:33, 20 November 2022 (EST)

I think that’s the only place. We may want to add “or upload it to our DB and link it from there” or something along these lines - your proposed language makes it sound like we allow only external images. Or keep the original wording in that part. Annie (talk) 13:00, 21 November 2022 (EST)

Makes sense. We can keep the current wording ("if an image is available, enter the URL") then. Ahasuerus (talk) 13:09, 21 November 2022 (EST)

Outcome

The Help language that disallowed entering cover artists for non-genre periodicals unless the cover "illustrated the SF content or was by a well known SF artist" has been removed from Help:Entering non-genre periodicals. Rules and standards changelog has been updated accordingly. Ahasuerus (talk) 08:09, 23 November 2022 (EST)

Advance Reader Copy

I'd like to take Marty's direction and ask you to have a look at this thread with particular reference to guidance in the Help, acquisition policy, and consensus opinion. Any thoughts on how things stand? Thanks, Kev. --BanjoKev (talk) 09:19, 26 January 2023 (EST)

I think they should remain out, per long-standing practice. The reason for that is that they aren't considered a final published product, even if they are being sold. They are basically proof copies created to check for errors that need fixing prior to sending the final version tot he printer. ···日本穣 · 投稿 · Talk to Nihonjoe 12:32, 26 January 2023 (EST)

Orthogonal to the "in" vs. "out" question: Given the relatively high frequency of this topic's appearing, if ARCs are indeed "out", I think it would be a good idea to add some sort of obvious statement about them to the policy. We did that for audio recordings. The policy's definition of published is a set of qualifications, not a definition, so the Included section does not help. If someone considers them published, that list would make them "in". And if someone considers them unpublished, #1 in the Excluded section can be read to allow them nonetheless, if they fall into one of the Included section's points -- for example, if they are issued by a mainstream publisher. One approach would be to add something like Manuscripts, advance reader copies (ARCs), and similar compilations produced prior to official publication for purposes of proof-reading or marketing are not considered "published", even if offered for sale by the publisher. That could be a qualifier in the Included section or an early bullet in the Excluded section. One idea, anyway. --MartyD (talk) 15:31, 26 January 2023 (EST)

I like your proposed wording. It is clear and concise. We could also add something to the ISFDB:FAQ about it, too (with a link directly to the SCOPE). ···日本穣 · 投稿 · Talk to Nihonjoe 17:55, 26 January 2023 (EST)

I have a fundamental problem with ARC's being out. These are copies that exist out there, are genre, people have them, can be bought/sold, so need to be recorded (as a separate edition) imo. MagicUnk (talk) 04:02, 27 January 2023 (EST)

Printed ARCs are not supposed to be sold. They clearly state that on every one I've ever had. They also tend to have on the cover some marketing/bookseller info to help with ordering and whatnot. eARCs generally can't be sold, and the only ones I know of that can be bought are those from Baen (and I believe the purchaser receives a copy of the ebook once it's finalized, though I could be remembering incorrectly). If we do add them, we should have some standardized way of marking them so they clearly show up as an ARC so people don't get confused when viewing the editions of the books. ···日本穣 · 投稿 · Talk to Nihonjoe 13:03, 27 January 2023 (EST)

A couple of thoughts.

First, ISFDB:Policy#Included doesn't really define the term "published". For paper books, it says "paper books published by ..." and it's silent on the topic of ebooks and online publications.

The common definition of "published" is "made available", but "available" can be ambiguous. In most cases it means "available to the public at large", but it can also mean:

available to all members of a book club
available to those who have subscribed to a limited edition
available to members of an APA (Amateur Press Association)

Wild Talent and The Time Masters by Wilson Tucker is an extreme example:

"No publisher, no place, no date [about 1953], 1953. Hardcover. First edition. Typed sheets, rectos only, of the two novels, a total of 260 pages. Professionally bound with headbands and title and author printed in gilt on spine. This special book was typed from the magazines by Stuart Hoffman, science fiction fan and editor of Index to "Unknown ," probably before the books were published in 1954. " - from Robert Gavora, Fine and Rare Books, ABAA

Is this really a "publication" or just something put together by a dedicated fan for his own use and perhaps for his friends?

Second, a number of authors make new works available to their Patreon subscribers before they are "officially" published on Amazon. For example, Glynn Stewart's Discretion was officially published on 2022-07-26 and cost $5.99 on Amazon.com, but Patreon subscribers could read it 4 weeks earlier, on 2022-06-28, and it cost them only $5.00. It's not clear whether these types of "early releases" should be considered separate publications or ARCs. At one point I proposed that we call anything with a price a "publication", but the Rules and Standards consensus was against it. Ahasuerus (talk) 14:10, 27 January 2023 (EST)

A few more random notes (after resolving an edit conflict):

Dating: we cannot date the title records based on an ARC date or any search of titles from a specific month/year becomes meaningless. Which means that we either need a new special date (7777-00-00 for example) or we need to consider allowing publications with titles in the future (which is a big no-no now).
Multiple ARCs are produced for some books - usually some of them are for specific projects, sometimes they come at different times. As such they can have different covers and more importantly contents - images and excerpts may or may not be included; other bonus material like extra stories or essays may also appear in some and not in others. Do we record them all separately? If not, how do we decide which one takes precedence and whose contents to add? Annie (talk) 14:19, 27 January 2023 (EST)

(unindent)Two advanced searches of notes fields resale and not for sale reveal some interesting anomalies. Here are a few random-picked from a search on "advance": [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]. Kev. --BanjoKev (talk) 15:59, 27 January 2023 (EST)

In my opinion ARCs and proofs should be excluded from the ISFDb. The general sense of the word "published" means an item that is made available for sale to the general public. That is not the case with ARCs and proofs although I am aware that there are rare exceptions (eg Baen eARCs, as Nihonjoe says). They are designed to be distributed freely to proofreaders and reviewers. Some of these recipients then choose to sell them but by then they are second hand goods. I have no problem with Ahasuerus' ambiguous examples because a member of the general public can join a book club / subscribe to a limited edition / etc. Regarding the Wilson Tucker extreme example, well... there will always be extreme examples.
Works available to Patreons may need to be treated separately. There was a discussion in Rules and Standards in 2021 with no clear resolution.
Whichever way this goes, I definitely agree with MartyD that ARCs and proofs should be explicitly mentioned in the Included or Excluded section of the Rules of Acquisition Policy. Teallach (talk) 19:02, 27 January 2023 (EST)

Re: "a member of the general public can join a book club / subscribe to a limited edition / etc.", it can get complicated. For example, I was thinking about entering a Russian language edition of George Orwell's 1984 earlier this week -- see filial.shpl.ru/wp-content/uploads/2020/10/Oruell-555x800.jpg . As far as I can tell, it came out in 1984 and was the first edition of the novel published within the Soviet Union. The catch is that even though the publisher, "Прогресс" (Progress), was a major Soviet company and published books in many languages, which were then sold all over the world, this particular edition was limited to trusted members of the Soviet government as indicated by the statement that you can see at the top of the cover. Was it really "published" as we define it? Ahasuerus (talk) 14:06, 28 January 2023 (EST)

I don't have any preference about this, but I suppose movies might provide a precedent and useful analogy. Pre-release screenings for cast/crew or critics are not considered the "release" of a movie (and, indeed, what gets released may be different). ARCs seem very much like those pre-release screenings. --MartyD (talk) 08:00, 28 January 2023 (EST)

I also have no preference but desire clarity (having entered one and run across multiple). Abebooks has a page delineating advanced copies, galleys, proofs & other pre-first edition books which our Help/Policy pages should cover in explaining what we mean by 'publish'. Using the general sense of "published" isn't good enough, we have our own definition for several terms (e.g. Chapbooks) and the interpretation of 'publish' is context sensitive (looking at the number of online definitions). A couple of (random) thoughts - given the 'publication' of material on the internet, does something have to be 'sold' for a 'price' to be considered published? We consider the change of advertisements (e.g. in Ace editions) to be additional 'printings' and worry about how many editiorial changes constitute a new 'edition' - so where are ARC's on this spectrum? Are they a valid 'printing'? ../Doug H (talk) 11:23, 28 January 2023 (EST)

Re: "does something have to be 'sold' for a 'price' to be considered published?", I don't think so. We have numerous pubs which have no price associated with them: webzines, fanzines, certain promotional editions like this one. Ahasuerus (talk) 16:21, 28 January 2023 (EST)

Shouldn't we be better of discarding the 'published' notion? As discussed above, there's not a definition to be found that will not spawn exceptions. Rather, shouldn't we consider any single "version" (to not have to use the word published), and record that? MagicUnk (talk) 16:27, 29 January 2023 (EST)

Well, we have a "publication date" field, which implies that a pub has been "published".

Also, if we were to use "version" instead of "publication", how would we distinguish "versions" produced by the author or the publisher internally from versions delivered to the intended audience, whatever it may be? An author and the author's editor(s)/publisher(s) may go through multiple iterations of a manuscript before it's finalized. And sometimes it's never finalized as was the case with the famously unfinished The Last Dangerous Visions. We currently list it as "unpublished", but multiple versions of the text existed at various points in the past. Ahasuerus (talk) 15:15, 30 January 2023 (EST)

Perhaps it would help to view these as types of publication and to designate them for inclusion/exclusion on that basis. That would even allow for some variation of treatment, instead of one-size-fits-all, with a default policy that's in or out and a small set of exceptions to cover the rare cases we agree ought to go the other way. Sort of like we have done for online publications. --MartyD (talk) 15:56, 30 January 2023 (EST)

Patreon Editions

If it's OK, I'd like to break out the Patreon edition discussion into a separate topic from the Advanced Reader Copy topic so that it doesn't get lost again. I think they are significantly different enough that they need to be handled separately.

The last inconclusive discussion in 2021 can be found here. Since then I have been adding the Patreon edition info for Glynn Stewart's books in the title notes. There are about 30 to date with a new one about to be added. In Glynn's case, the covers are the same as the ebook covers but have PATREON EDITION prominently printed on them; the title pages also have PATREON EDITION printed on them. (I'm also keeping a set of cover images for each of them just in case). They are released on Tuesdays anywhere from two to four weeks prior to the public release of the book and may contain some typos that are fixed in the public release. They consistently cost $5.00 regardless of the public release price.

In my mind, they should have separate pub records. That said, I suspect that Patreon editions by other authors may be less distinct but I have no proof one way or another. In addition, I'm not sure if Kickstarter editions should be handled the same way but they would seem likely candidates as well. Phil (talk) 12:37, 28 January 2023 (EST)

I also think that "Patreon editions" are really separate pubs -- as opposed to ARCs -- with a separate publication date, a separate price and (sometimes) a slightly different cover. I think we should be able to create separate publication records for them based on the following statement in ISFDB:Policy#Included:

[Included:] Internet-based publications which are downloadable as electronic files in any number of ebook formats (ePub, Mobi, PDF, etc).

In some ways, they are similar to limited editions, which we create separate publication records for. Ahasuerus (talk) 15:52, 30 January 2023 (EST)

I had the same thought about these being like limited editions. --MartyD (talk) 15:58, 30 January 2023 (EST)