Rules and standards discussions


Jump to: navigation, search

ISFDB Noticeboards
Before posting to this page, consider whether one of the other noticeboards might suit your needs better.
Help desk
Questions about doing a specific task, or how to correct information when the solution is not immediately obvious.
• New post • Archives
Verification requests
Help with bibliographic, image credit, and other questions which require a physical check of the work in question.
• New post • Archives
Rules and standards
Discussions about the rules and standards, as well as questions about interpretation and application of those rules.
• New post • Rules changelog • Archives
Community Portal
General discussion about anything not covered by the more specialized noticeboards to the left.
• New post • Archives
Moderator noticeboard
Get the attention of moderators regarding submission questions.
• New post • Archives • Cancel submission

Archive Quick Links
Archives of old Rules and standards discussions.

1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 · 9 · 10 · 11 · 12 · 13 · 14

Expanded archive listing
Rules and standards changelog

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


Abyss & Apex

I haven't been paying as much attention to recent goings-on as I should have been doing, but has something changed that would allow Abyss & Apex to be included in the database? Nobody has more love and respect for that magazine than I do, but it's a semiprofessional webzine that's very much ineligible for SFWA. I just thought I'd ask if some exception had been made before I went around challenging people's contributions... Dwarzel 17:04, 16 July 2016 (UTC)

I think it's IN under the "It has been short-listed for a major award." exception, if this means "short-listed", that is. --MartyD 22:08, 17 July 2016 (UTC)
Sounds about right. Also, I have been thinking that our current inclusion rules, especially references to SFWA, may be too NorAm-centric given our increased coverage of non-English titles, but that's fodder for another discussion. Ahasuerus 23:16, 17 July 2016 (UTC)
Good point; I'd forgotten that it was indeed on the final Hugo ballot last year, but I see it now. Would that make the entire run of the magazine eligible for inclusion, or just the year's-worth that was considered for the Hugo? Dwarzel 16:55, 19 July 2016 (UTC)
I would say it gets the whole thing in. It would be silly (and confusing) to have it only get the one year in. ···日本穣 · 投稿 · Talk to Nihonjoe 17:02, 19 July 2016 (UTC)

Titles of Barrie excerpts

I would like to link to a discussion on MLB's page about the titles that were used for excerpted novel chapters in Masterpieces of Fantasy and Wonder, about whether they should have (excerpt) after them or not. Anyone have thoughts? --Vasha77 17:17, 4 September 2016 (UTC)

My opinion: We use parenthetical information to disambiguate, not to explain what something is. The reason to add "(excerpt)" is to distinguish between the title of the excerpt and the title of the full work. You should add it if the title given in the publication is the same as the title used for the work from which the excerpt is taken. The same thinking would apply to a sub-part (for example, if a chapter's title is used, include "(excerpt)" if the excerpt is only part of the chapter, but omit it if the except is the full chapter). --MartyD 19:35, 5 September 2016 (UTC)
Thank you, sounds like we should change those titles. --Vasha77 02:43, 6 September 2016 (UTC)
I think that I said most of this, still, it's good it's been codified. While my hard drive got sick and had to be fixed I found my copy of this book. Guess I shouldn't (***sob***) have bothered. Still, if you're interested, the introductory blurb states that these are selected chapters from the novel The Little White Bird. I've never read this book, I've never owned this book, and I know nothing about this novel, still I hope that this bit'o'info helps. MLB 23:58, 8 September 2016 (UTC)

Graphic novels

I've gone through the history of discussions on graphic novels. (Specifically here seems relevant). There doesn't seem to be a standard approach for labeling them, and that is leading to a fair amount of confusion on my part. Many of them are designated chapbooks and others are novels. I recently entered this pub and that highlights a fair amount of confusion. It's quite similar to both this pub which is chapbook and this pub which is a novel. In each case, the stories were initially printed as comics in a monthly format, but only as a limited series. I wanted to add Free Country to the the DC Metaverse series but could not since it's a chapbook. The publication has three distinct 'chapters' with different authors. Should this be a single Chapbook with 7 authors, or should I break it into individual stories each with unique authors. Any clarifications would be appreciated. TAWeiss 21:57, 5 September 2016 (UTC)

Alas this won't help but, AFAIC, there should be strictly NO graphic novels (even by authors "above the threshold") in the db, lest we face such problems and/or an invasion of publications we're not equiped to deal with. Hauck 07:29, 6 September 2016 (UTC)
Amen to that. As for the problem case, I think for any book with different chapters by different people, I'd be inclined to make the publication an ANTHOLOGY and the chapters separate shortfiction works. --MartyD 10:56, 6 September 2016 (UTC)
There are several graphic novels in the db and there have been a number of discussions on how to enter them correctly. I've seen moderators who have entered this content in the database such as this pub. There was specific discussion about authors above the threshold here. I guess my confusion is shared by others.TAWeiss 00:51, 8 September 2016 (UTC)
Regarding adding chapbooks to series, you instead add the chapbook content title record(s) to the series (in this case, add the four short fiction titles to the series). As MartyD mentioned, that chapbook would be better changed to an anthology. ···日本穣 · 投稿 · Talk to Nihonjoe 01:12, 8 September 2016 (UTC)
I agree that the anthology approach would work better in this case. Chapbooks are generally reserved for standalone appearances of short fiction and poems.
As far as the issue of eligibility goes, it has a long and complicated history. In the mid-2000s, some editors supported their inclusion while other editors were opposed. There was some back-and-forth on this issue and the result was a compromise which allowed a limited subset of graphic novels to be included. A couple of years ago the software was upgraded to make it easier to distinguish between prose novels and graphic novel, but the policy issue remains, to a large extent, in limbo.
We'll probably have to revisit the policy issue sooner rather than later because the number of multi-media franchises is growing. A number of popular prose series have had graphic adaptation lately while many manga series have "light novel" offshoots. I watch the numbers climb month after month since I process robotic submissions. Ahasuerus 01:22, 8 September 2016 (UTC)
Upon reflection, it seems more a case of collaborative fiction, where a number of authors have created a single story. None of the chapters stand on their own as a story. If Chapbook is the wrong designation, then I would think it falls in the novel category, which matches most of the other graphic novels that I've seen. It also seems like I should hold off on entering any further graphic novels until there is further clarification, though it leaves some odd holes in the database. There are a large number of books with considerable graphic content, which makes exclusion a difficult judgement callTAWeiss 01:35, 8 September 2016 (UTC)

Variant title dates

These statements are not direct contradictions because the first gives instruction (as befits EditTitle) and the second interprets what is displayed if instructions have been followed (as befits Title). But the second presumes a directly contradictory convention, as I understand it.

Help:Screen:EditTitle > Title: Year

When entering a variant title, enter the first date when the work appeared under any title and any pseudonym; typically, variant titles do not have separate dates.

Help:Screen:Title > Variant titles

If the title/author combination displayed is the canonical version of the title, then any variant titles will be displayed. This will list the variant title, the date of first publication under that variant title, and the author name under which that variant title appeared.

Both conventions are commonly followed in regions of the database familiar to me. I know that I have followed each one sometimes.

Concerning artwork, there may be a related point concerning publication as INTERIORART on the one hand, COVERART on the other hand. In the spirit of the EditTitle instruction, a cover illustration may be dated by its first publication as an interior illustration. --Pwendt|talk 02:14, 9 September 2016 (UTC)

It is my understanding that in practice each variant title has the date that the work first appeared with that title and author and language. This matches the second explanation. Further, the canonical title record should always have the date of first publication under any title. I'm not sure that I see the point in having the dates of the canonical and all variant titles match. --Ron ~ RtraceTalk 02:24, 9 September 2016 (UTC)
Our documentation and our standard practice don't agree on this point. In the past, variants were always dated with the date of the parents. Lately, more and more editors are choosing to date the variant with the first date the variant was used. This has been raised before, but while there were no objections to updating the documentation, there was not much participation in the discussion so nothing was changed. Personally, I think the documentation should be updated to match practice (especially given translations). -- JLaTondre (talk) 02:29, 9 September 2016 (UTC)
I agree with JLaTondre. This makes the database more accurate. ···日本穣 · 投稿 · Talk to Nihonjoe 02:33, 9 September 2016 (UTC)
Reading #Titles of Barrie excerpts, it occurs to me that the same issue concerns titles such as "The Little White Bird (excerpt)" and "Through the Looking-Glass and What Alice Found There [3]". First publication of the work as an excerpt, perhaps enumerated, or its first publication as part of a larger whole. --Pwendt|talk 02:41, 9 September 2016 (UTC)

Some questions about the EditPub titles for magazines

In this help entry for Publications, the Under Titles - Magazine it states "the title should be of the form Magazine Title, Date" and to see the missing and variant dates for more on formatting. That sections has "The date part of a magazine title should be given after the title, following a comma and a space. The month should be given in full and then the year in full. If the issue is a quarterly, or a bimonthly, give the date in the form given on the magazine -- for example, "Fantastic Universe, June-July 1953" or "Interzone, Fall 1979". A hyphen should be used between two months used for a bimonthly issue."

A conflict arises between "give the date in the form given on the magazine" and "A hyphen should be used", if the magazine uses a slash (/). Which is to be given priority? And should be blanks be used to separate the hyphen from the month if the magazine does so?

The second concern is which date to use - in some cases the cover and the contents page differ. The example I have is the "July/August 1992" cover of Science Fiction Chronicle which has "July 1992/Volume 13, Number 10/Issn 0195-5365/Whole #153" leading the table of contents. Which should be given priority?

Combining the two, is issue with cover "December 95/January 1996" and contents of "Issue #187 / December 1995-January 1996 / Volume 17, Number 2 / ISSN 0195-5365".

A final point - a number of magazines have been entered (e.g. Locus) using the Series number as well as the date. Is this acceptable or should it be discouraged? Doug H 16:35, 17 September 2016 (UTC)

A partial answer, with much debate and the apparition of non-US magazines (that are usually sequantially numbered), the series number is used when it's proeminently displayed on the cover (for example Interzone) but not if it's not on cover (for example Asimov's even if each issue has a whole number on toc). Hauck 17:53, 17 September 2016 (UTC)
The competitor magazine for Science Fiction Chronicle was Locus. It includes the series number even though it is not on the cover. There are no images for the first few issues but back in 1980 they had the issue and date on the cover. The existing issues of SFC seem to have been modelled on Locus. The first 60 issues of SFC are not loaded, so I don't know whether they had the issue number or not. I'd prefer to retain the issue number for various reasons (all 60 done this way, with at least 160 unsubmitted issues), gaps are easier to identify - 192 is June 1997, 193 is October 1997), but don't know if (or how, or to whom) I should be arguing the exception. Once these are resolved, maybe we can update the help page. Doug H 19:11, 17 September 2016 (UTC)

Fictional Authors of Fictional Essays

This template states that there is an exception to the rule of recording the author's name as it appears on the title page: "An exception is made for fictional essays, which are written as if by a character in the story, often as an introduction or afterword. Even if these are signed by the fictional character, they should be recorded as by the actual author of the work." I've received a request to change the author of Quidditch Through the Ages from J.K. Rowling to the pseudonym "Kennilworthy Whisp" which is how the book is credited on the title page. The argument is being made that the above exception only applies to true essays, and that essays that we list as short fiction (as is done here) should be credited to the fictional author. My reading of the above exception is that it is there exactly for this kind of work and that the variant titles already created as by Kennilworthy Whisp should be changed to Rowling. Do others have opinions of whether this book is an appropriate application of the exception about fictional authors? Thanks. --Ron ~ RtraceTalk 02:26, 29 September 2016 (UTC)

I do think that this really is a kind of shortfiction as it tells the story of 'quidditch' through the ages. Also, I believe the dissinction between 'fictional essays' and shortfiction a futile one as it can become really tedious to draw a line between the two: where exactly should there be a border established? At the least all this pieces tell a kind of (in-universe) story, however compressed it may be (and this can vary in many degrees). Stonecreek 03:41, 29 September 2016 (UTC)
What author name would people look for the story/book under? If there is a chance that they look for it under Kennilworthy Whisp, that should be the name associated with it. And where would it be filed - under R or under W in a library or in a bookstore (my library put a call number of 796.1 WHISP so if it is just under Rollins, I may decide that the db does not have it). I'd go for the author as written when it is an individual book and not a part of a longer book - even if it is a chapbook. May be wrong but I just do not think that this rule applies here.:) Anniemod 03:58, 29 September 2016 (UTC)
The distinction in my mind is that here, it is a publication credited to the fictional character. The exception might be applicable to the work if it appeared in conjunction with a novel credited to Rowling -- and I believe the spirit of the exception covers this example in such a case -- but here it is a standalone work, so it's not "covered" by any primary work's credit to the real person. --MartyD 10:35, 29 September 2016 (UTC)
I think it would be fine to treat it as short fiction by Whisp, and then make Kennilworthy Whisp a pseudonym of Rowling. That way, people will be able to easily find it. ···日本穣 · 投稿 · Talk to Nihonjoe 23:48, 29 September 2016 (UTC)

"uncredited" publisher

The other day I discovered that about a dozen publications, half of them verified, list their publisher as "uncredited". I don't think it was ever intended, but all that Help says about this issue is:

  • For self-published works, fanzines, bibliographic pamphlets and the like, use the name of the editor/author if no other publisher information is visible.

I think it would be beneficial to add the following sentence to the quoted paragraph:

  • In all other cases where the publisher is not stated, leave the field blank.

Ahasuerus 00:51, 3 October 2016 (UTC)

Looks like a useful addition. But in this case, would the publisher's name appear as "unknown" (as in the cover art field) or as "uncredited" (which would be more appropriate, in my opinion) ? Linguist 08:00, 3 October 2016 (UTC).
I don't know how frequently we run into secondary sources that do not list the publisher, but it does seem the distinction of "uncredited" vs. "unknown" is useful there. I'd rather see an explicit "uncredited" for cases where a primary verifier knows for a fact there is no publisher credit of any kind and leave blank for the cases where we're working with sources from which we cannot tell who the publisher is. --MartyD 10:59, 3 October 2016 (UTC)

ElfQuest graphic novels

I was told some time back NOT to add the ElfQuest graphic novels, but someone has added a whole bunch of them (but not all)

As far as I can figure the prose shorts (the 5 anthologies under the Blood of Ten Chiefs titles) and the novelizations plus the two chapbooks with "new" material (A Gift of Her Own and Searcher and the Sword) and maybe the Gatherums (a, b and the big gatherum) plus the other non-fic stuff (The First 20 Years) should be here but not the other stuff and not the comic books by Wendy for Beauty and the Beast (Night of Beauty and Portrait of Love). How to get rid of:

(I think I got them all) Susan O'Fearna 03:29, 24 October 2016 (UTC)

If I recall correctly, the ElfQuest universe was one of Michael Hutchins's projects and he spent a fair amount of time sorting everything out. I don't recall why he chose to keep the graphic novels, but I assume he had a reason. We probably don't want to delete them until we hear from him. Ahasuerus 15:16, 24 October 2016 (UTC)
Sounds like a plan to me... I just checked my archive (!!) and he's one of the mods who told me original graphics can be added but with rare exceptions (Sandman) collected TPBs (as opposed to gns) are not to be added Susan O'Fearna 17:23, 24 October 2016 (UTC)
I'd say "delete the lot". Hauck 17:27, 24 October 2016 (UTC)
As this is foremost a series of titles with graphic formats I can't see a reason why those should stay. Else we would also have to add all kinds of comics whose universes happen to have an incarnation as one or more prose titles (for example super heroes, Asterix etc.). Stonecreek 11:12, 3 November 2016 (UTC)
Note that this series is now a complete mess. I don't how many changes about it I've moderated (e.g. titles going from NOVEL to CHAPBOOK and back to NOVEL). There was also a lot of work in the cleanup reports (lots of novels with twice the same title included). There is still the matter of the duplicates # in the title series, can someone attend to this before I delete the whole lot? Hauck 09:19, 3 November 2016 (UTC)
If there are no objections forthcoming, I'm going to delete this series in a few days. Hauck 09:30, 6 November 2016 (UTC)
We'll just need to be careful to preserve the (relatively few) prose works that are part of the series. Ahasuerus 12:47, 6 November 2016 (UTC)
No problem. Hauck 13:33, 6 November 2016 (UTC)
Done. Hauck 09:06, 7 November 2016 (UTC)

Pseudonym direction

I need a rules check for a case. I would usually go with Tupper as the main name and Bob Tupper as pseudonym due to number of works involved under the two authors. However - it is a "just family name" against a fuller name so I am kinda inclined to go the other direction and just variant the whole lot... What do the rules say in this case? What is the correct way to pseudonym these two? Anniemod 01:17, 25 October 2016 (UTC)

In our world, "canonical" is not defined to be "most complete", but rather to be "most commonly used". See Help:How to record a pseudonym. --MartyD 11:24, 26 October 2016 (UTC)
Yeah, and I had been following the "most common" rule. That one gave me a pause as the last name only is the common one. Overthinking it (again) I guess. :) I will get it connected the usual way - it can always be reversed of needed one day. Thanks!Anniemod 20:15, 26 October 2016 (UTC)

Tami Hoag

I just noticed there's one novel by Tami Hoag on here... all of her novels, including this one Night Sins are romantic suspense/mystery with no SpecFic... delete it? There are no verifications Susan O'Fearna 05:24, 26 October 2016 (UTC)

OK, done. Hauck 05:39, 26 October 2016 (UTC)

Essay in non-genre book

I apologize for sounding like a total noob. Hauck suggested that I post here after having rejected this submission, which is a book of literary essays including one about Frankenstein -- that was the only regular-title content I included. This database doesn't seem to have a good way of of adding short stories or essays that are in non-genre publications; I hate to clutter up the place with books like The Educational Legacy of Romanticism, but shouldn't the essay be in the DB? What is the opinion of people here? (Would it be possible to just keep the title record for the essay, and have its publication info in a note?) --Vasha77 18:31, 30 October 2016 (UTC)

Not much to add to my rejection message here. If we start to go this way, we will have to accept any essay "about" spec-fic in any publication. I'm personally against such a move that will likely swamp the db. Hauck 18:52, 30 October 2016 (UTC)
I am with Hauck on that - if it is not in a genre media and it is non-fiction, we do not catalog. Otherwise we will end up cataloging articles in TLS or Reader's Digest because they happen to talk about something sf-inal. If it was fiction, what you submitted would be the proper way (we have Nature magazine cataloged like that for example - with only the fiction that is SF) but non-fiction (especially by non-genre authors) just does not belong here. Otherwise we are also opening the doors for sf art in non-SF publications (is non-fiction more important than art?) and all kinds of other borderline cases. At least until there is a different way to separate that kind of content, it clutters the DB. And fiction is what this DB is all about (name kinda says so as well) - we catalog art and non-fiction as supporting elements in my opinion. Just my 2 cents :) Anniemod 22:01, 30 October 2016 (UTC)
It seems to me that there should be room for literary criticism, even if it is in a non-genre publication. Art, no. It's just a matter of defining the line. Unfortunately, most of our rules of acquisition have fuzzy boundaries. For example, debates about the inclusion of graphic novels and letters have been going on for as long as I've been a member. Part of the problem is that the ISFDB is like a monastery, we're mostly closed off to feedback from the outside. I think we need to take a bigger view. The question is how do people use the ISFDB, and what would they like to get out of it? Do we serve only the community of SF readers? Biographers would certainly appreciate references to SF criticism, and authors published outside of the genre. I suggested adding a gender identification field a while back, and that really stirred things up, despite the fact that every published biographical author listing, Encyclopedias,Contemporary Authors, Austlit, etc., identifies the gender of authors. Maybe it's time to update the rules of acquisition with outside feedback and draw as many hard lines as possible?--Rkihara 22:32, 30 October 2016 (UTC)
IIRC this (references to SF criticism, and authors published outside of the genre) approach has been tried once.Hauck 08:09, 31 October 2016 (UTC)
As long as we draw a strict line between criticism and "this article that talks about SF-inal things/books" by someone that had never written anything else about SF... And drawing this line won't be that easy and will depend a lot on what someone is looking for - what I consider criticism may be stricter or less strict than what someone else does. Just thinking aloud. Anniemod 01:11, 31 October 2016 (UTC)
Sharp lines? As if. (To use an Americanism.) Hauck also objected to me adding literary criticism that wasn't in his opinion genre enough. Jorge Luis Borges wrote stories which in his own words "pertencen al género fantástico" and one of which was a finalist for a Retro Hugo last year. It could be argued that the "Ficciones" don't belong to genre SFF as English speakers understand it, since they weren't originally published in that context, but if Hugo voter retrospectively accept them, why disagree? Likewise most of the vast amount of criticism written about those stories isn't written by people interested in genre fantasy, but rather who are interested in epistemology, semiotics, history of literature, etc. Should criticism coming from a non-genre angle be omitted even when it's about genre texts? What about texts which, like "Tlön, Uqbar, Orbis Tertius" have been accepted into the genre but only in hindsight, not when first published? There's no easy answer to that. --Vasha77 02:26, 31 October 2016 (UTC)
About this question "Should criticism coming from a non-genre angle be omitted even when it's about genre texts?", my answer is a clear "No" (see this page where all these books have been entered in the db). I've objected (but sometimes approved them) to some of your proposed additions because a book like this one is IMHO (I never forget that it's in the end a subjective choice) outside our scope as Borges is not above the threshold. For the enough part, as I said I tend to follow a 50% rule (more than half about spec-fic, it's in, less, it's out), it's not perfect but it's a first approach. Hauck 07:32, 31 October 2016 (UTC)

[Unindent] We do our best to be completists when it comes to literary fiction. We do NOT try to do that with non-literary fiction (e.g. movies, comics, fanfic, or graphic novels) nor with SFF art or SFF non-fic, and I think that's a good policy. IMHO, we should only include art books that are entirely speculative fiction or non-fic books that are entirely analysis of, or critique of, speculative fiction. (So Hauck's "50% rule" is much more generous than me.) Partly, we don't need to try to compete with sources like The Science Fiction and Fantasy Research Database, whose goal it is to catalog such non-fiction. And, IMHO, we should try to limit ourselves because we can't really keep up with the fiction by itself!

There is *so* much fiction out there that we still haven't been able to incorporate (especially after going "International") that we seriously need to maintain substantial boundaries on what we try to catalog. I just discovered a publisher we had ignored, who had 60-some books we needed to add. There are thousands of fanzines, many of which have full contents online, that haven't been entered. Half the books in the db have additional OCLC records that we haven't pulled in. WorldCat has thousands of "Title synopses" that we haven't pulled in, and Amazon has thousands more. Most of the big reference guides (Tuck, Schlobin, Reginald, Bleiler, Currey, etc.) have been partly, but not completely entered. There are hundreds, if not thousands, of anthologies for which we have no contents, even though those contents are in WorldCat. There are ebooks, self-published books, non-English works, .... The challenges we have with the range of work we have established are so daunting that it should be clear that we cannot do a respectable job on any attempt to massively expand our mission. IMO, we are better off to keep a narrower focus, and try to do it right, then to establish an "include everything relevant" policy and do a terrible job at it. Chavey 14:54, 31 October 2016 (UTC)

We are volunteers, and if one us decides to enter something "relevant" it doesn't mean everyone else has to. It doesn't become policy because a few people do it. Being above the threshold for inclusion of non-SF books, can be pretty subjective as Hauck points out. In regard to Borges, I've read most of his translated collections and in my opinion he's well over the 50% mark.--Rkihara 15:40, 31 October 2016 (UTC)
Chavey may have a good point, and if other people agree that the DB is for fiction, then I would switch over to just adding contents to anthologies and collections -- I enjoy doing that. It can't be denied that the Cornell University library (which I'm using) is a better resource for criticism than for genre fiction; but it does have enough fiction to keep me busy for a long time. --Vasha77 20:12, 31 October 2016 (UTC)
I agree that we should be extra careful when deciding to expand the scope of the project. As we have learned over the last 20+ years, it's very important to have a good understanding of the additional complexities and the extra volume of work that scope expansion may lead to. Ahasuerus 00:56, 1 November 2016 (UTC)

Acts of Light

What is a collection of illustrated Emily Dickinson poems, Acts of Light(not verified) doing here? I don't see a genre connection at all. Yeah, one or two of Dickinson's poems mention ghosts or witches, but... --Vasha77 20:56, 30 October 2016 (UTC)

Someone cataloged it or it got pulled by the automatic processes based on Amazon tags and somehow slipped in:) I think it should be deleted Anniemod 22:03, 30 October 2016 (UTC)
Done. Hauck 07:21, 31 October 2016 (UTC)


Why must we IGNORE "best practices?"

It's been said that:

"1) a copy (be it made of electrons or of paper) of a magazine is not itself a MAGAZINE (for us), it's transformed into a COLLECTION.

Yet our very own Template for Pub Type informs us quite sensibly:

  • COLLECTION. A publication containing two or more works of SHORTFICTION or POEMs by a single author or authors writing in collaboration should be typed as a COLLECTION. The typing of individual publications which contain works with various combinations of author credit should be discussed on the Community Portal on a case-by-case basis. The title page credit should be the major factor in determining the types of these kinds of publications. Excerpts from other works published after a NOVEL for promotional purposes do not make the publication into a COLLECTION.

It's plain to see that our helpful template makes very clear that the typical MAGAZINE with several authors writing individually, is most definitely NOT a collection!

Perhaps we might seek guidance again from a further Template bulleted item:

  • MAGAZINE. It can be difficult in some cases to determine if something should be regarded as a magazine or a book. Some magazines were published in book format; some books were published as series with letter columns and regular dates of publication. Borderline cases should be discussed on the magazine or book wiki pages, but generally a magazine must have a common title from issue to issue, and an enumeration or dating system of some kind. This still leaves anthology series such as New Worlds Quarterly as judgement calls. In these cases, look for a consensus on the publication bibliographic wiki page. If no discussion exists, use your best judgment and document the decision on the wiki page.

So referring back to the above assertion that a reprinted magazine is no longer a magazine, but a collection-

How is this justified without flagrantly disregarding our own guidelines?
Chrisgherr 11:32, 1 November 2016 (UTC)

There's an error involved in the discussion: most often a reprinted magazine becomes an ANTHOLOGY, not a collection (and that is just what happened in the case that started the upheaval, I think). The title in question is stated to be an ANTHOLOGY, and that's what its respective publications are listed. Stonecreek 11:53, 1 November 2016 (UTC)
Yep, see below (my mistake) all the rest stands.Hauck 12:32, 1 November 2016 (UTC)
Instead of COLLECTION it should be ANTHOLOGY (an artefact of translation), the source of the discussion is in these rejections. Hauck 11:52, 1 November 2016 (UTC)

So after correcting for the faulty translation, we have:

1) a copy (be it made of electrons or of paper) of a magazine is not itself a MAGAZINE (for us), it's transformed into an ANTHOLOGY.

Again referring to our Pub Type template we see:

  • ANTHOLOGY: A publication containing fiction by more than one author, not written in collaboration, should be typed as an ANTHOLOGY. For example, "Late Knight Edition" contains stories by both Damon Knight and Kate Wilhelm, individually; this is an anthology, not a collection. If a book of Conan stories contains stories which are all partly or wholly by Robert E. Howard, it is a collection; if one or more of the stories is by Lin Carter or L. Sprague de Camp, not in collaboration with Howard, then the book is an anthology.

If we were to apply this rule always, we would never have need for the pub type "magazine," since all magazines are indeed "publications" of the type described above!

Perhaps it's useful to put the two pub type definitions one after the other, for easy comparison!

  • MAGAZINE. It can be difficult in some cases to determine if something should be regarded as a magazine or a book. Some magazines were published in book format; some books were published as series with letter columns and regular dates of publication. Borderline cases should be discussed on the magazine or book wiki pages, but generally a magazine must have a common title from issue to issue, and an enumeration or dating system of some kind. This still leaves anthology series such as New Worlds Quarterly as judgement calls. In these cases, look for a consensus on the publication bibliographic wiki page. If no discussion exists, use your best judgment and document the decision on the wiki page.

Looking closely at the two categories it seems that magazines are a particular subset of anthologies; magazines being defined by certain agreed upon characteristics such as letter columns, periodicity, etc.

The question now arises, is a magazine, any magazine, be it "Analog" or "Newsweek," any less a magazine, for having been reprinted?

Conversely, is there any bookish anthology, be it "Dangerous Visions" or any of the "Year's Best Science Fiction," which is no longer an anthology, or is an anthology of another kind, for having been reprinted?

Speaking for myself, I am surprised that there is anything but unanimity on this matter! It seems indisputable that magazines REMAIN magazines; no matter how many times they are reprinted.

I suggest that: (after closely examining both definitions)

We must concede that there is nothing in them which would either suggest or require that we reclassify "reprinted magazines" as something other than "magazines."
Chrisgherr 13:58, 1 November 2016 (UTC)

I suspect that there are two separate issues here:
  • Our software design originally assumed that a magazine would always have one (and only one) printing. Later on we discovered that it was not a safe assumption, but by then it had been embedded in the software implementation.
  • There is also a larger philosophical question here. Is a 2013 facsimile reprint of a 1935 magazine still a magazine? At the moment the rules say that it's an anthology, although some series like Adventure House Facsimile Reprints handle it inconsistently.
Ahasuerus 14:10, 1 November 2016 (UTC)
IMHO the defining characteristic of a MAGAZINE is its distribution system. A MAGAZINE is found at a newsagent's stand (even if it can be found in some specialized bookshops). In France the circuits for books and for magazines are completely separated, and I suppose that this holds true in other countries. The fac-similes or the electronic copies are not to be found at newagents, whence their classification as ANTHOLOGIES that use a more "bookish" (if the term still applies to electronic publications) distribution system. Hauck 15:22, 1 November 2016 (UTC)
But that changed a lot in the last 10 years or so - not just because of e-books. If a newsagent is needed, then none of the e-book and print on-demand only magazines of the last decade are magazines (Is Lightspeed a magazine?). Maybe we just need to fix the help a bit (or look at practices) - because with the e-books now in the picture - is an e-book copy of a magazine created 2 months later than the print version a reprint? If so - should it be an anthology? That will make a really complicated display for that magazine/series. We are not good at handling multi-format magazines as it is anyway. And then there are some magazines that changed to anthologies mid-run (Postscripts for example - and they were never distributed as anything but a book but they were a magazine for years) - thankfully they are not reprinting their early issues or it will be fun. A magazine is a useful form to sort publications that come out more than once a year - outside of that, they are anthologies for all intents and purposes... In my mind there are two different things here: immediate format change reprints (ebooks) - which leaves a magazine a magazine; and later reprints (1935 -> 2016 ones) that we are marking as anthologies due to software restrictions. Just thinking aloud. Anniemod 15:34, 1 November 2016 (UTC)
To be frank and to measure how dinosaurish I am, I personally wouldn't have included any electronic publications in the db without a complete intellectual overhaul (e.g. what does "printing rank" means for an ebook? what is a cover artist for such publication or is there any cover to speak of? what's the publication date of a changing text with multiple iterations? If I change an octet in my ebook, is it a new publication warranting a new record? Etc...). Just for the sake of discussion, we're talking here about a 1930 publication that is not likely to have originally existed in an electronic form. Hauck 15:43, 1 November 2016 (UTC)
That would have made the DB almost irrelevant very fast - as a source of finding information on where a book exists anyway. Which is what I came for - I was looking for options for reading a story I could not get the original media for :) And as much as this will create chaos, we will see more and more of these reprints -- we live in the era of reprints and everything that people want will get reprinted sooner or later. So we might as well address it here. As I said in my last sentence above - old reprints are anthologies under the current rules (software restriction or convention - either way this is the practice). That's the current way and until it is changed, this is the best practice. :) But we do need to start addressing the e-books situation and reprints of one-time media(even when printed) in a way fall under the same hat. As much as I wish we could stop the world, technology seems to be moving faster than expected :) Anniemod 16:42, 1 November 2016 (UTC)
Ultimately a database is an attempt to model (a part of) the world. For better or for worse, the world is near-infinitely complex and encompasses a variety of borderline and corner cases. We, on the other hand, have limited resources, so we can't model it perfectly. Even if we had dozens of developers who could create and maintain a very complex database that would reflect real world complexities more accurately, it would make our data entry rules ever more arcane and difficult to apply -- consider the monster that is the MARC 21 family of standards, particularly their bibliographic standard. And so we compromise.
In addition, the world keeps changing. Of the roughly two dozen binding codes that we recognize, less than 20% existed 200 years ago. Dos-a-dos books with two different covers, which were introduced in the middle of the 20th century, presented unique cataloging challenges. ISBNs were adopted less than 50 years ago. And so on and so forth. E-publications and audio books are just the latest challenges in a very long line of challenges. We try to adapt to them as best we can, but sometimes it requires -- you guessed it -- more compromises. And so it goes :-) Ahasuerus 17:34, 1 November 2016 (UTC)
Except that for some things (e-versions of magazines/fanzines/prozines/sami-prozines for example or the realities of the reprint era) we seem to be closing our eyes instead and wish them gone and leave every magazine/editor to find its own way (some have separate series, some are mixed and have doubles in the same grid (Journey Planet) and some just ignore the e-versions (Analog for example off the top of my head)):) I know I know - I will bail out of this conversation and start another discussion for that :)
Software development is always fun like that - you are right. The world changes and the software always lags behind it. But then we should just find a happy medium - or new editors get very confused (me being a relatively new one can attest to it). Sometimes the moderators have issues explaining very easy concepts (although they all are really trying) because they get boggled into the special cases (somehow almost any early addition ends up being a special case) - and anyone that tries to start with something different from adding a novel is a bit lost in the process. And the fact that you need to understand the intricacies to be able to add anything does not help much. We really need a "easy add" kind of page that allows a lot less choices and can be used by new editors as a start - and then have a more experienced editor hold their hand through the steps to get it more complex. Anniemod 18:23, 1 November 2016 (UTC)
(Coming full circle), I asked at the top of this discussion, "Why must we IGNORE best practices?" It seems (to me) that the consensus reached herein is that IF some "best practice" is being "ignored," it's due to software design limitations. It would also seem that, due to said limitations, there has arisen a "rule" which requires that magazine reprints be typed as anthologies. Also (at the top), I copied the Pub Type Template entry for "magazine." Yet -- I find no such rule mentioned, nor referenced, therein. As a neophyte editor; I rely primarily on such templates for guidance. If the (popup) template provided makes no distinction between a magazine's original printing and any reprintings, on what basis am I to demur? I would also ask, on what grounds should any editor (or moderator) ignore the written rules, presumably arrived at thru a process of discussion and consensus, such as takes place here? Don't the moderators, especially, have a heightened responsibility to see that the rules, AS WRITTEN, are abided by? Furthermore, mustn't the moderators set aside their own subjective feelings; and play the roles of impartial referees, where rule enforcement is concerned?
Chrisgherr 21:50, 1 November 2016 (UTC)
I don't enter reprints, but I do own a couple. One is a hardcover, the other is a soft cover trade edition. I think the practice of entering them as anthologies may have started because calling them magazines would have been confusing.--Rkihara 22:35, 1 November 2016 (UTC)
The ISFDB operates off a set of community standards that have been established via discussion (primarily on this page, but can also occur elsewhere) and/or sometimes by de facto practice. Our documentation has gaps and sometimes even inconsistencies. This can be confusing for new users at times, but the solution to that is to fix the documentation when required. -- JLaTondre (talk) 22:50, 1 November 2016 (UTC)

So time to change the documentation - things sometimes are so well known that noone remembers to edit the documentation to say so - and new editors get confused. What exactly would make that part clearer? What changes in the documentation would have made it clearer now that you know what the whole thing is supposed to mean? Once we find a better way to say things and there is enough agreement, documentation can get changed. :) Anniemod 23:21, 1 November 2016 (UTC)
I mostly agree with Anniemod's conclusion, posted above. It's time to consider changing the documentation! It's apparent to me that as JLaTondre points out above, there has been established a de facto standard, or rule, thru practice and consensus, that magazine reprints are publication typed as anthologies. This de facto rule directly (I believe) contradicts the existing de jure rule, written into the website help, templates, etc. Such a contradiction lends itself at the least to confusion, if nothing more. The solution contemplated above, of changing the documentation, (i.e. the "de jure" rule), would surely be an effective solution, harmonizing these two inconsistencies. However, as a tyro, I must wonder: Why has the written rule lasted so long (presumably), unchanged(?), in it's contrariness vs. the unwritten rule? Is the written rule possibly better than its de facto counterpart? For myself, no defense is needed for the existing written help, templates, etc. (regarding this specific matter). I've not seen a spirited defense of the de facto standard. If we're to rewrite the help, to conform with practice, WON'T SOMEONE HERE ARGUE THE BENEFITS OF DOING SO? (By which I mean: How is what we do now so much better than what the Template advises us to do -- and vice versa.
Chrisgherr 02:20, 2 November 2016 (UTC)
How often do you read documentation for a process that you know? Had you tried editing a document you had read 100 times already and where you know the process in your sleep? Both require fresh eyes :) And a will to help with identifying the gaps and missing "common sense" and "but everyone knows this" parts. And the moderators are very good at explaining things to new editors when things are confusing - may take a few attempts or get a bit complicated but everyone is amazing in explaining and helping. Which makes the help pages a bit... rusty. At least this had been my feeling around here.
I did ask you above - what changes would have made you understand how the magazines are handled better? Which parts confused you? Anniemod 02:52, 2 November 2016 (UTC)
I do hope that someone can very specifically address, (or provide an answer to), my two queries (in bold font) above (just slightly above, from November 2). I don't necessarily expect a thorough or detailed response, it's just that I've noticed that no one has commented specifically on the wonderfulness of the current undocumented practice, vs. the awfulness of the older written guidelines. Remember that I'm a newcomer, for whom things obvious to many here, are complete mysteries to me!
I haven't meant to neglect the two (and now four) queries from Anniemod; I suppose I thought that those answers would reveal themselves more fully in time. As I've now been twice asked, I'll now answer (once). ;)
What exactly would make (the documentation) clearer? I think that I, or any editor, could write one or two lines to add to the Template bulleted item (MAGAZINE) at the very head of this discussion, which would make clear the exception for reprints. I hesitate to write those lines myself, just now, as no one has offered here a compelling argument for the superiority of the de facto practice.
What changes in the documentation would have made it clearer now that you know what the whole thing is supposed to mean? As above.
How often do (I) read documentation for a process that (I) know? As a novice here, there're few that I know- so- not very often! ;)
Have (I) tried editing a document (I) had read 100 times already and where (I) knew the process in (my) sleep? The answer here is no.
I will say that if someone could take a crack at the two bold font queries above, I might well understand things better. :)
Chrisgherr 04:33, 2 November 2016 (UTC)
I was semi-serious with the two of the questions. As part of my day job I need to deal with documentation and two things you will never change - the inability of the human brain to properly read a paragraph they had read 100 times already and the fact that if you know what you are doing, you do not read docs. How does this connect to you and your post? Simple - people that had been here for long enough either never read the template page anymore or if they do, they read what they think is there (in order to copy it). Thus practice changing while documentation does not. :)
So you are basically proposing a sentence/paragraph to be added to the "New Magazine" description that says something like: " If you are cataloging a reprint of a magazine (for example a 2011 reprint of a 1935 issue of a magazine), it needs to be cataloged as an ANTHOLOGY and not a MAGAZINE". That's all that is missing from the documentation, correct?
PS: Editing help pages - please do not that on your own - ISFDB is a collaborative effort; so are the help pages. Discussing changes before making them is the process to use (which is why they do not get updated when someone notices an issue all the time but still...) Anniemod 17:08, 3 November 2016 (UTC)
Your bolded questions: are you asking someone to explain why a de-facto rule is better than the written rule? Because this is what everyone here was explaining above - the help page is outdated, it never got updated when practice led to changes. Making a case on why we should edit an outdated help page instead of fixing the said page is a bit of unneeded. It is not about what is so much better - it is really about what had happened while people had really worked with the records and the rules that had evolved with time. If you believe the written rule is better, show some examples and/or argue it - but without bringing the "the docs say so" argument. If the argument has merit and everyone agrees, rules can be changed. But the de facto policies did not come into being because someone decided to just go off script - ISFDB is a living and breathing organism and sometimes practice trumps even the best laid plans. If you are asking something else - I am really not sure what else remains unsaid :( Anniemod 18:59, 3 November 2016 (UTC)
I very much like "the while people had really worked with the records" bit when you put this in perspective with that which shows that our new contributor has less then 10 edits of the database to its credit . Hauck 14:39, 7 November 2016 (UTC)

What counts as a publication for an e-book?

This question has probably been asked before but I couldn't find it. An indie author creates an ebook and uploads it to Smashwords and Amazon simultaneously (and maybe other platforms)-- do these count as separate publications? If there is an identifiable ISBN at all, it is the same one. --Vasha77 19:02, 1 November 2016 (UTC)

I'm not sure we have a definitive standard on this, but in my observation, we generally create a single record for such an ebook. If, however, the books are published with different ISBNs or as by different publishers (per the ebook title/copyright 'page'), then they would be different publications in my opinion. Fortunately, I haven't seen many cases of those two (other than perhaps people reprinting Project Gutenberg texts), but I'm sure it's bound to happen. -- JLaTondre (talk) 23:02, 1 November 2016 (UTC)
Early on, many e-books were issued with one ISBN per format. I haven't seen much of it lately, at least in the fiction market. Perhaps it's due to the fact that ISBNs are relatively expensive. In most cases Smashwords, Barnes & Noble, Amazon, etc have the same e-books these days.
The main bibliographic problem with the way Amazon lists e-books is that it doesn't display their ISBNs even when they exist. There is a way to look them up programmatically, but human users are out of luck. It makes it harder to confirm that the Amazon version is the same as other places' versions. Ahasuerus 00:37, 2 November 2016 (UTC)
A preview/sample of any kindle book can be obtain for free if you have a kindle (or kindle reader) and it contains the first few pages and almost always the ISBN... Now - publishers that stick the ISBN at the end of the book are a problem but still... And more often than not in newer editions, a printed version of the book will contain the e-book one as vice versa specified on their copyright pages. Anniemod 01:14, 2 November 2016 (UTC)
That makes sense. In that case, though, there is a problem with Amazon Digital Services, CreateSpace, Smashwords, etc. named as publishers, if the same ebook is on several simultaneously. Should such books be treated as self-published with the author's name as publisher? --Vasha77 01:54, 2 November 2016 (UTC)
CreateSpace, which is primarily an Amazon-run printer/distributor, is kind of tricky. As Lawrence Watt-Evans once explained:
  • If you let CreateSpace provide the ISBN, instead of using your own, they'll provide library distribution, but they then show up as the publisher of record.
Some of these CreateSpace-facilitated books have no publisher credit anywhere in the book and end up in our database as published by CreateSpace. Others do have a separate publisher credit, which can be found using Amazon's Look Inside. ReAnimus Press was like that for a number of years, but they seem to be using their own ISBNs now.
I am not sure how Smashwords works these days. Ahasuerus 02:31, 2 November 2016 (UTC)
To take a real-world example, a random book I added to the DB recently was A Dragonlings' Haunted Halloween by S. E. Smith. This exists in the form of a paperback from CreateSpace and an ebook. The copyright page of the Smashwords ebook reads:
IMPRINT: Science Fiction Romance / Science Fiction Romance / A Dragonlings’ Haunted Halloween: Dragonlings of Valdier Book 1.2 / Smashwords Edition / Copyright © 2014 by S. E. Smith / First E-Book Published October 2014
The Smashwords page gives a date of Oct. 20, 2014, and an ISBN of 9781311508980. On Amazon, the ebook is listed with a date of Oct. 19, 2014, and no ISBN. There is no space labeled "Publisher", but instead "Sold by: Amazon Digital Services LLC". The Kindle preview gives the following copyright page:
Science Fiction Romance / A Dragonlings’ Haunted Halloween: Dragonlings of Valdier Book 1.2 / Copyright © 2014 by S. E. Smith / First E-Book Published October 2014
So... I do not think that Amazon Digital Services (in this instance at least) should be named as a publisher at all; if I was only entering the book from Amazon, I would want to say that the publisher was "S. E. Smith". But what to do with it being on Smashwords? Does the notation "Smashwords Edition" on the copyright page make that a separate publication, and would Smashwords be the publisher? --Vasha77 14:01, 2 November 2016 (UTC)
Addendum: More complexity. This book is on Kobo with the publication info "S. E. Smith LLC, February 2015", with the same copyright page as Amazon; and an "ISBN" of 1230000305247. (Smith incorporated herself as an LLC?! Surely not all authors would do so. That shouldn't be necessary in order for the author to be listed as the "publisher".)
And on Barnes and Noble, we have "ISBN-13: 2940151275286 / Publisher: S.E. Smith LLC / Publication date: 06/04/2015"; and again the same copyright page.
These Kobo and B&N "ISBNs" aren't that, but rather ID codes. I want to know what the Amazon edition's ISBN is -- if it has one. --Vasha77 14:43, 2 November 2016 (UTC)
An interesting example. Using the previously mentioned programmatic method of querying Amazon, I have checked ASIN B00OOWTW1I and ISBN 9781311508980. Neither lookup returned any matches. Of course, 1230000305247 and 2940151275286 are not, as you noted, real ISBNs since 13-digit ISBNs must start with 978 or 979.
As far as "Amazon Digital Services" goes, I agree that it's not a publisher. Here is what Helps says about these types of situations:
  • For self-published works, fanzines, bibliographic pamphlets and the like, use the name of the editor/author if no other publisher information is visible.
I am not really familiar with Smashwords' self-publishing model, but I note that their list of "Global retail distribution" does not include Amazon. Taken together with the fact that Amazon doesn't recognize the Smashwords ISBN, it suggests that there are two different editions involved. Ahasuerus 14:59, 2 November 2016 (UTC)
P.S. On the other hand, another self-published e-book available on Smahswords, Rise the Renegade: A Rork Sollix Space Opera Adventure, is also available on Amazon with the same ISBN, 9781941939062. Curiously, Amazon's publication date is 2016-02-23 while the one on Smashword is 2016-10-24. Ahasuerus 15:15, 2 November 2016 (UTC)
It seems that even if an ebook has an ISBN, most retailers will display their own ID code in place of it. That does make it hard for a casual user to tell how many editions are out there. Cutting through the Gordian knot would be nice, (all ebooks are the same?) but that might miss some actual differences in publication... --Vasha77 17:00, 2 November 2016 (UTC)
This book is on Smashwords and also has a publisher ( Gone Writing Publishing) listed, whereas the other two examples have no mention of a publisher. Do you think the other two should have Smashwords listed as their publisher or the author's name? --Vasha77 01:32, 3 November 2016 (UTC)
My gut feeling is that the author's name would be safer, but I don't know enough about the way Smashwords facilitates self-publishing to be sure. As our experience with CreateSpace and some other online places shows, a self-publishing company can support multiple models at the same time. Ahasuerus 17:07, 3 November 2016 (UTC)
This is one of the reasons I always thought it was a poor design to separate author authorities and publisher authorities as different record types (but it seems we are stuck with the design now; unless we want to add new authority types like translator but we also include artists in our author records so things are confusing). I do not see the difference when a company is credited with writing a work or a single person is credited with publishing a publication. Perhaps we want to separate authorities by whether they contribute to a work, or a publication of such a work. Regardless this seems like it introduces redundancy (and that always spells maintenance issues). Uzume 21:28, 12 April 2017 (UTC)

Discussion continued at Community Portal

Improve "How to enter foreign language editions" page regarding the publication date of translations

Proposal for a small improvement based on existing rules. This is what bullet point three in Template:TitleFields:Date says regarding the publication date of variants/translations: "When entering a variant title, enter the first date when the work appeared under any title and any pseudonym; typically, variant titles do not have separate dates. However, if the text was heavily revised (or translated) when it appeared as a variant, the date should be that of the first appearance of the revised version of the text."

Based on this I think it would be good idea to add a paragraph like this as third bullet point to the Help:How_to_enter_foreign_language_editions page:

  • The publication date of the Variant Title should be set to the first appearance of the translation, not to the first appearance of the title of the translation. This means that if a translation of the same translator has been released with different titles, each title would receive it's own Variant Title record, but all of these Variant Titles would have the same publication date. See Template:TitleFields:Date for more details.

Jens Hitspacebar 21:30, 7 November 2016 (UTC)

I think that had been done exactly this way (so we just need to get the help page to agree with practice :) Which is to say "Agree, that page needs an update" :) Anniemod 22:21, 7 November 2016 (UTC)
Yes, but wasn't there just a discussion where most people agreed that under current practice, the date of a variant title IS the first appearance of that title? --Vasha77 22:59, 7 November 2016 (UTC)
If there was a discussion about this and it produced undocumented results the documentation should be updated. Links, please! :) Actually, the idea for my proposal also comes from a (though very brief) discussion on my talk page. Quoting myself from there:
"I always thought that's the way we enter translated titles: use the first appearance of the same translation with the same title, not of the appearance of the translation of the same translator (regardless of the title). No idea where I got that idea from - I probably deduced that from other titles which already existed in the database. A quick search brings up several examples which have been entered the same way, see this and this, or this and this, or this and this. I'm sure there are many others.
However, the TitleFields:Date help says: When entering a variant title, enter the first date when the work appeared under any title and any pseudonym; typically, variant titles do not have separate dates. However, if the text was heavily revised (or translated) when it appeared as a variant, the date should be that of the first appearance of the revised version of the text.
Therefore it looks like you're right. And that we have lots of records which have not been entered the correct way."
Jens Hitspacebar 23:44, 7 November 2016 (UTC)
Can you post a link (or links) to an example so we have a real world reference for what you are suggesting? I am not completely understanding what you want to do and I think having an example would clear it up. ···日本穣 · 投稿 · Talk to Nihonjoe 23:10, 7 November 2016 (UTC)
Sure. Look at the German variant titles of Perdido Street Station: there are three different German titles, but all of them are the same translation by the same translator. The latest German edition is from 2014, using the same translation as the others before, but the title Perdido Street Station for the first time. Nevertheless, its title record has a date of 2002, not 2014. Jens Hitspacebar 23:44, 7 November 2016 (UTC)
Variant dating is another case where our documentation and practice do not agree. The trend has become to date the variant at the first appearance of the variant. This has been raised before (see one and two), but with no resolution. A quick check of the last database dump shows that, where both the parent and the variant are English, only 48% of variants have the same date as the parent (I quickly scripted that so I won't swear it's absolutely correct, but am pretty confident). While each conversation has only minimal participation, there have been no objections and people are not adhering to the current documentation. I think it is time to make the change. -- JLaTondre (talk) 23:45, 7 November 2016 (UTC)
My improvement proposal was especially to clarify how to enter translations, not about variants and parents of the same language. The current documentation (I quote above) treats these two types of variants the same. I'm not sure, but: is all this probably a hint to treat translations differently (i.e. like documented) than same-language variants? I'm not advocating this, because it may overcomplicate things. Just wanted to think aloud. I'm actually fine with either of the solution, as long as we get the current standard documented correctly and up-to-date. Jens Hitspacebar 00:20, 8 November 2016 (UTC)
I entered an example today of same translation, different title. La región más transparente by Carlos Fuentes, translated into Italian by Germán Quintero and Luigi Dapelo, appeared under the title L'ombelico della luna in 2000, La regione più trasparente in 2011. I was planning on giving those variants different dates. But then that would contradict the Perdido Street Station example. Clarification is much needed. --Vasha77 00:33, 8 November 2016 (UTC)
A translation is a different version from the original. Which is why we are varianting. If we keep the same date as the original, we are losing the information about the translation (and even if you may be able to find it by looking at the publication, we have that field that should be usable for something (otherwise we should just remove dates for variants altogether... A translation is a revision in a way - so I think the current rule is covering it but adding it explicitly won't hurt. Now, I think that any variant should carry its earliest date - so it is visible in a glance how the dates and variants progressed (and that seems to be the de facto policy anyway) but I am more concerned about the translations than the straight variants. Anniemod 02:32, 8 November 2016 (UTC)
Translations are dated separately already per our documentation. As I understand it, Hitspacebar is talking about occurrences of the same translation (i.e. by the same translator) under different titles and/or author credits. -- JLaTondre (talk) 11:45, 8 November 2016 (UTC)
Exactly. Jens Hitspacebar 13:01, 8 November 2016 (UTC)
@Hitspacebar: I recognize that your proposal was about translations. However, it is the same principle. Either we believe there is merit in recording the first time that a particular title & author credit was used or we don't. If we do, than it should apply to translations as well. Given most users seem to think there is, the documentation (which would apply to multiple pages) should be updated to state that variants (including translations) should be dated to the first appearance of that specific title and author credit. -- JLaTondre (talk) 11:45, 8 November 2016 (UTC)
I agree with both of those points. It is the same principle, and I think there's merit in recording the date of the first appearance. That said, I would be interested in seeing an explanation of the reasoning behind the always-use-the-canonical-date rule. Is that doing something we would lose if all variants used their origination dates instead? --MartyD 12:05, 8 November 2016 (UTC)
I currently don't have an opinion about what's best for same-language variants, but as for translations I think it feels more intuitive (especially for the occasional user of the site) to record the first appearance of a variant title under its title and not (as it's currently documented) the first appearance under any variant title. This already seems to be the current practice of several editors if I understood the discussion correctly so far. This would mean that the date of the title record of the Perdido Street Station example I posted above should be changed from 2002 to 2014. The information lost with this practice would be the immediate answer to the question "when was a translation by certain translator first released". However, this information can still be retrieved from the NOTE field of the title or publication records as long as the translator is stated there. Jens Hitspacebar 13:01, 8 November 2016 (UTC)

(unindent) I think it would be best to consider all variant-dating rules, including the rules that apply to same-language and translation VTs, at the same time. If we handle them piecemeal, there is a good chance that we will forget some permutations.

Re: the original reason why Help says "typically, variant titles do not have separate dates", as I recall, it had to do with the way variant dates were displayed in the Contents section of the Publication page. The software has changed in the last 10 years and I think that it handles different variant dates correctly now. We'll need to double-check, though.

My current take on the larger issue is that in most cases (including this one) it is best to capture as much granular information as possible. In this case it would mean updating Help to something like "A new date is assigned to the first appearance of a new version of a text [including new translations] and/or a new title". Ahasuerus 14:15, 8 November 2016 (UTC)

I think that the last paragraph is the way to go - update the help page, have the date the first time the variant started to exist as a date -- this way you can see at a glance when a new title started to be used for example... Anniemod 23:12, 8 November 2016 (UTC)
I looks like this seems to be the consensus in all answers if I didn't overlook something. Is more feedback by other moderators needed or should we go ahead and update the documentation? Jens Hitspacebar 15:23, 11 November 2016 (UTC)
Anyone...? Hitspacebar 18:23, 12 November 2016 (UTC)
If there are no objections, I will change Help in 24 hours. Ahasuerus 19:03, 12 November 2016 (UTC)
Template:TitleFields:Date has been updated. Hopefully the wording is clear. Ahasuerus 20:44, 13 November 2016 (UTC)

Proposal to create a rule changelog

With so many rules split across lots of pages, and with so many changes going on in the Wiki every day it can be pretty tough to find out if important things have changed - especially regarding rules and if you're not here on a daily basis. I think it would help a lot if we'd do two things:

  • Create a changelog for the rules (similar to Development/Recent_Patches) which should contain the following information for each rule change:
    • Date of the change.
    • Very short description (e.g. "Date of variants").
    • A link to the wiki page which conatins the rule.
    • What was the rule before?
    • How has it changed, what has changed?
  • Ensure that every rule change is logged there. We could create a small wiki template which would be added at the beginning of every thread started on this discussion page in order to verify its outcome:
    • Is the thread about a rule change at all? [yes, no]
    • Is there a result? [yes, no]
    • Has the documention been updated to the changed rules? [yes, no, not applicable]
    • Has the rule change been logged in the rule changelog? [yes, no, not applicable]

Jens Hitspacebar 16:03, 11 November 2016 (UTC)

I like the idea. When an editor comes back after a hiatus, it can be hard to determine what, if anything, has changed. Ahasuerus 16:44, 11 November 2016 (UTC)
P.S. I should add the we already have What's New, which is mostly dormant. We'll have to think of ways to integrate it with the proposed changelog page. Perhaps we could make the changelog a sub-page or a section of the "What's New" page. Ahasuerus 16:47, 11 November 2016 (UTC)
I think it shouldn't be section of the What's New page but a new separate page because rule changes are (mostly) a different thing than new or changed features (though they can overlap of course, e.g. if a new feature needs a new rule). If you want to revive the What's New page and put all on one page there might again be too much clutter to detect which rules have changed. Jens Hitspacebar 18:21, 12 November 2016 (UTC)
Well, we could create a new Wiki page for the proposed changelog and then link it from "What's New" as well as from the main Help page. Ahasuerus 20:45, 13 November 2016 (UTC)
Yes, that's better. We might want to wait a few days for more editor feedback, and then I could create the page (and the templates I mentioned above) over the forthcoming weekend. Jens Hitspacebar 17:16, 14 November 2016 (UTC)
I'm not sure what y'all mean by a 'rule', but it seems to me that if it's a rule, it's documented. If it's not documented it's a habit. So if there's a discussion of a 'rule' - be it an interpretation or request to change / create / delete - you should be able to reference that rule. So to automate your log (generally the only hope of keeping it accurate and complete), you need to track changes to the places where rules are kept and track references. I presume the wiki pages that contain the rules have tracking, so what would be useful is a way to reference a rule in wiki discussions that will allow for automatic tracking. Doug H 03:39, 14 November 2016 (UTC)
It's true that the Wiki software keeps a history of every Wiki page, including Help pages. However, although you can use Wiki history to find policy changes, it can be time-consuming once you get past the last few changes. Also, some changes are minor and don't affect policy, so you may have to wade through a lot of editorial changes to find what you are looking for. Finally, some policy changes, e.g. translation changes, applied to multiple Help pages, which would make reverse-engineering the policy change from Wiki history very hard.
In addition, we don't have unlimited disk space, so we have to purge Wiki history (all but the last 50 versions of each page) periodically. Ahasuerus 14:52, 14 November 2016 (UTC)

"Omnibus" whose only novel is non-genre

I would like some other opinions on the subject of this discussion -- the situation just seems weird. There is an omnibus that contains only a non-genre novel and a genre novella, so what title type is it? (BTW I don't want to add La muerte de Artemio Cruz to the DB-- Fuentes has too many non-genre works and too many editions and I don't want to even start letting any of them in.) --Vasha 16:51, 13 November 2016 (UTC)

Our publication type definitions make no distinction regarding contents being genre or non-genre. And that to me seems to be proper. From a bibliographic perspective, the publication remains the same type of publication even if we choose to not list certain contents. Our cleanup scripts are designed to improve the quality of the database. In a case like this, we shouldn't degrade the quality just to satisfy the script. Instead, a better solution would have been to ask Ahasuerus if he could have the script ignore this record (this is uncommon enough case that adding a moderator ignore capability is probably not worth the effort). -- JLaTondre (talk) 13:01, 14 November 2016 (UTC)
We have lots of Collections that only contain one genre short story. We do not want them listed as "chapbooks". The same philosophy would apply here. Chavey 20:38, 14 November 2016 (UTC)
I agree that collections, anthologies and omnibuses that contain only one SF work should be entered as collections, anthologies and omnibuses respectively. Ahasuerus 20:56, 14 November 2016 (UTC)
Thanks for that. Shall I make the change to OMNIBUS or do you want to do it? --Vasha 21:31, 14 November 2016 (UTC)
Well, there is no hurry, we can wait and see what other editors think and then adjust Help accordingly.
I should also add that I am not the ultimate authority as far as policy and data entry rules go. I just tweak the software to support the community consensus. There have been times when I proposed new features only to drop them for lack of support. Ahasuerus 21:48, 14 November 2016 (UTC)
Unless there are objections, I will adjust the cleanup report to let moderators ignore "OMNIBUS pubs with no NOVEL titles" tomorrow. Ahasuerus
The software has been updated. Moderators are now able to "ignore" OMNIBUS publications without NOVEL/COLLECTION/ANTHOLOGY/NONFICTION titles. Ahasuerus 23:38, 17 November 2016 (UTC)
(outdent) I've update the record to be an OMNIBUS. It can be ignored in the next cleanup report. -- JLaTondre (talk) 00:17, 19 November 2016 (UTC)

Print and Ebook Editions of Clarkesworld Magazine

Hi all,

I have been adding the 2016 issues of Clarkesworld magazine and just noticed today that there is also a magazine Clarkesworld (print). These issues mostly do not have the table of contents filled in. From the issues that are filled in, it seems that the ebook and print editions are nearly identical. The only differences are 1. price 2. the inclusion of page numbers in the print edition. Why do we need both of these series? Is the print series being automatically generated from Amazon and that is why the table of contents is blank on non-verified issues?

Sorry if these questions are easy to answer; I am new here. m (I also don't know why this isn't showing up as a new section...sorry person whose section I have hijacked) Amoeba of horror 05:06, 14 November 2016 (UTC)

Yes, the print issues are straight from amazon via Fixer (our robot) and massaged into shape by a moderator (usually me) that, alas, has no time to enter the contents without a physical copy. Don't hesitate to populate the print issues by using the "Import Content" link on the left and entering the titleID# of the electronic issue in the top part of the screen (in this case do not use the bottom part). Hauck 07:15, 14 November 2016 (UTC)
I think you meant "publication ID ... in the top part of the screen" :-) Ahasuerus 14:42, 14 November 2016 (UTC)
To add to what the guys already said (and if you know that already, ignore it) - a series grid contains publications. So if we add both the publications in the same series, you end up with something like this grid: Journey Planet. Which works for smaller series but gets too crowded for ones that have issues every month (or more often) and had had a lot of issues. So a second series is sometimes created for the second format - we kinda do not have a specific rule on that but that is the only way to have them and not mess up the formatting of the grid. :)
PS: For making a new section, when you are adding, use the "+" on top of the screen and not "Edit". That will ask for a title and make a new section for you. Anniemod 05:28, 15 November 2016 (UTC)

Collected Works as a series?

Do you think it's a good idea to put the volumes of a Collected Works edition in a series or not? There aren't very many examples of collected works in the DB to compare. Ambrose Bierce has a series made, G. K. Chesterton does not. --Vasha 03:35, 16 November 2016 (UTC)

If they are published as a series and they are unique selections and are connected in some other way besides being written by one person (such as this one) - yes. But if one of the volumes is a single novel for example (that can be published outside of the series), then I would make a publication series instead.. Looking at your examples - both can be either. Just my 2 cents :) Anniemod 03:47, 16 November 2016 (UTC)
Usually it would be only a pub. series. There really are only a few examples of collected works as a title series; and it's only meaningful if they are published for the first time, such as Philip K. Dick's collected shortfiction (but as Anniemod wrote: if they had enclosed his novels also, it likely wouldn't have been catalogued as a title series). Stonecreek 05:21, 16 November 2016 (UTC)
At the moment we have a few dozen regular "Collected" series and six "Collected" publication series. Translated "collected works" like this one are usually entered as publication series.
Another thing to keep in mind is that support for publication series is relatively recent. Early on editors had no choice but to use regular series. There is a good chance that many of them would have entered regular series like The Collected Stories of Robert Silverberg (Bantam) and The Collected Stories of Robert Silverberg (Subterranean Press) as publication series if the functionality had been available in the 2000s. Ahasuerus 18:52, 16 November 2016 (UTC)
Additional thoughts after sleeping on it:
Suppose we converted The Collected Stories of Robert Silverberg (Subterranean Press) to a publication series. At this time the series is up to 9 volumes, each one with a different title, e.g. "Trips: 1972-73" or "Hot Times in Magma City: 1990-95". If the title series were eliminated, these 9 titles would be interspersed with other Silverberg collections on the Summary page. There would be no easy way of telling that they are related, which would seem like a step back. Ahasuerus 18:29, 17 November 2016 (UTC)
That's what I had been thinking about with my "unique selection" comment above -- there is a difference between collected stories and collected works in my mind - collected works should almost never be proper series - because of the inclusion of novels and/or preexisting collections -- while collected stories will almost always be unique and grouping them makes sense on the author page - which leads to a proper series... Anniemod 19:01, 17 November 2016 (UTC)

Proposal to group pubs without a publication series

Hi,there are some publishers (such as this one) with hundred of titles grouped into several publication series. Only a very few titles (see this) are not assigned to any pub series (for whatever reason). That's why these titles are hard to find in the publisher's book list: it is necessary to click through every single publication year to view the column 'publication series'. My proposal is: If a publisher has one or more publication series then pubs not assigned to a pub series are grouped into a special 'pub series', e.g. [no publication series]. This will be very helpful. Boskar 10:00, 18 November 2016 (UTC)

Having thought a bit about that problem, I'd think that such an installation may lead to some problems: 1) It would contradict the help pages: A Publication series is a group of publications marked out by the publisher in some way. (that's why I removed such a 'pub. series' in the first place). 2) Even more grave is that when we state [no publication series] we'd assume something that we don't know about, at least for the loads of unverified publications: they may have a pub. series stated, but as no one of us has looked into the publication, we can't just state that there isn't any. Or would there be a way to circumnavigate this? Stonecreek 15:54, 18 November 2016 (UTC)
If the problem is finding publications that are not assigned to a publication series, then we can add a new link, "View publications not in a publication series", to the Publisher page. Ahasuerus 16:57, 18 November 2016 (UTC)
If that'd be not much trouble, it'd be wonderful! Stonecreek 17:03, 18 November 2016 (UTC)
OK, FR 948 has been created. I hope to finish the monthly upload of Fixer-identified ISBNs tonight, at which point I should be able to go back to development. Ahasuerus 17:29, 18 November 2016 (UTC)
The software has been changed to support this functionality -- please see the Festa page for an example. Ahasuerus 23:46, 19 November 2016 (UTC)
Hey! That's not just great, that's perfect! Thank you! Boskar 13:33, 20 November 2016 (UTC)
Excellent! :) Ahasuerus 18:01, 20 November 2016 (UTC)

Reviews of Magazines

Under the Rules of Acquisition Policy it is stated:

Reviews of media products, stage productions, magazines and fanzines (regardless of their genre), and books that are not eligible for inclusion in the database, should not be entered into the "Reviews" section of the data entry form. A record should be created in the "Regular Titles" section typed as ESSAY.

However, this appears to be contradicted by the Help Screen for New Pubs, which states:

Note also that only books, magazines, and short fiction are entered; if the column reviews fanzines, you don't need to enter the review records for these, only the ESSAY record.

Albinoflea 06:03, 19 November 2016 (UTC)

Minor quibble - call them Regular Title records, not ESSAY records to be consistent. You're mixing interface and database terms. Doug H 16:11, 19 November 2016 (UTC)
Minor quibble(2) - why call them EDITOR records? The screen I use to get the record number to link to calls it "ISFDB Title Record #". The word EDITOR shows up 4 lines down as the Type: Doug H 19:06, 21 November 2016 (UTC)
Title records have different "title types". Some pages call them "Entry types", which may be confusing. The full list of supported title types is available as a drop-down list in Edit Title. It is also available in the Regular Titles sub-section of the Content section of the Edit Publication page as well as on the Advanced Search page. Similarly, publication records have "publication types".
Each publication record must contain a "reference title". In most cases the type of the reference title matches the publication type, e.g. "NOVEL"/"NOVEL" or "COLLECTION"/"COLLECTION". The only exception is magazine and fanzine publications. Their publication types are MAGAZINE and FANZINE respectively, but the title type of their reference title is "EDITOR". Ahasuerus 04:10, 22 November 2016 (UTC)

I was taught/advised long ago that even though entering reviews of magazines was not in itself bad form, since they can't be linked to their TITLE records (but rather their EDITOR records) that it was better to leave these as ESSAY entries rather than REVIEWs.

I'm wondering what the current consensus is on this as there are a group of LinkReview requests in the moderation queue now attempting to link REVIEWs to EDITOR records, and I'm not sure how best to proceed.

It would appear the three basic options are:

  • a) Enter reviews of magazines as ESSAYs
  • b) Enter reviews of magazines as REVIEWs, but leave them unlinked
  • c) Enter reviews of magazines as REVIEWs, and link them to the corresponding EDITOR records

What's the thinking around each of these options? Thanks, Albinoflea 06:03, 19 November 2016 (UTC)

Certainly a) for the reasons you state, plus it may be that the magazine/fanzine is not in the db, in which case you would not have a record to link to anyway. PeteYoung 09:12, 19 November 2016 (UTC)
If a book reviewed is not present, you have nothing to link to. Is it normal to create a stub in such cases? Doug H 16:11, 19 November 2016 (UTC)
If the book falls within the Rules of Acquisition, yes. I can't tell you how many hundreds of new titles I've added just from reviews in SF Commentary alone. PeteYoung 22:09, 20 November 2016 (UTC)
(After edit conflict with Pete) When an editor comes across a review of a book that is not in the database, it is usually the starting point of an investigation. For the most part it turns out that the book ought to be in the database. If that is the case, we use the review as well as other secondary sources to create a record. Other times it turns out that the book is not eligible for inclusion. Occasionally we run into more complex cases, e.g. a book reviewed under its pre-publication title, a cancelled book or even a joke review of a non-existent book, so we have to be careful. Ahasuerus 22:16, 20 November 2016 (UTC)
Regarding b), I would consider this bad form. To my eye it looks like an unfinished publication record to a database user, who may not understand why no link is provided.
Regarding c), If a magazine or fanzine is grouped by year (as is thankfully more common now), the review as it appears on the Editor record will not be seen to refer to a specific issue. I've added a fanzine review to the bottom of this pub (which I will remove later) so you can see how it appears here: the specific issue reviewed does not show up on the Editor record, it just looks like a review of the whole year's publications. Also, sometimes a magazine or fanzine review can refer to more than one issue, eg. Review of the fanzines Broken Toys #17 & #18 by Taral Wayne so, if we're lucky enough that they are in the database, which would you link to? PeteYoung 09:12, 19 November 2016 (UTC)
We've arbitrarily decided to group magazines by year and should expect some fallout. This one is that reviews of all issues are combined in the list of links. The review's title could/should specify the particular issue(s). If a review covers two book titles, wouldn't you separate them into two reviews? Couldn't you do the same for issues that span years? Doug H 16:11, 19 November 2016 (UTC)
I agree that the word "magazines" in the Help Screen for New Pubs is misleading and should be removed. My two cents. PeteYoung 09:12, 19 November 2016 (UTC)
It looks like the referenced section of Help is out of date. It says:
  • Non-sf works should be entered but if an onerous number of non-sf-related works are reviewed in a column you are entering, discuss the situation on the Bibliographic Rules page to decide what can be eliminated.
even though the relevant policy was changed a while back. The new policy is explained on the same Help page:
  • ... books that are not eligible for inclusion in the database (graphic novels, nongenre novels by authors that are below the threshold, nonassociational nonfiction works), should not be entered into the "Reviews" section of the data entry form. A record should be created in the "Regular Titles" section typed as ESSAY.
Ahasuerus 22:23, 20 November 2016 (UTC)
I think the question is - Does the new policy include magazines as 'books' to be put in the "Reviews" vs "Regular Titles" section? IF they go under "Reviews", they must be linked to the title record # (that's what the screen says) for the issue's publication entry. This loses the issue information, which could be included in the review title. Multiple issues would mean multiple review entries. Missing issues would not be a problem if there are issues for the year. If the year has no issues - a stub would be required. If the magazine is not present, it's a whole new layer of stubs. Doug H 19:06, 21 November 2016 (UTC)
The Help excerpts that I quoted immediately above are not related to the magazine issue. I posted them as an example of this Help paragraph being out of date in other ways as well. Ahasuerus 03:28, 22 November 2016 (UTC)
The idea that "The review's title could specify the particular issue" has its charm. Up to now I've only added reviews for magazines that have single title entries (for example because of the schedule of publishing: only once or twice a year). Stonecreek 17:04, 21 November 2016 (UTC)
One problem with this approach is that the way the software currently works any title assigned in the Review is stripped out when it is posted on the TITLE/EDITOR record; see Dead of Night - 1995 for an example; even though the title of the review is "Terminal Fright #10, Winter 1995" it only displays as "Review"; Pete alludes to this in his comment above.
If the desire is to discourage the linking of reviews to records with type EDITOR, then perhaps we should flag those on the moderation screen? I assume they would never be matched by the system and are always going to show up as manual LinkReview entries in the queue.
Or is it a situation where if someone wants to make the added effort to create these links, even if they're not perfect, moderators should pass them through? Albinoflea 01:30, 22 November 2016 (UTC)

(unindent) I have been thinking about this for the last couple of days. The underlying issue appears to be that we merge EDITOR records for magazine publications. If we didn't, then we could link REVIEWs to individual issues and the problem would go away.

Now, the reason that we decided to merge EDITOR records back in the late 2000s was that some authors/editors like John W. Campbell Jr. and Stanley Schmidt are responsible for hundreds of issues. Our concern was that displaying one line per issue would result in very long Summary and Series pages that would be difficult to navigate. Unfortunately, the implemented solution has proved problematic in other ways as seen here.

Perhaps what we need to do is to take a step back and examine whether it would be practical to:

  • convert to a "one magazine issue = one EDITOR record" system, and
  • handle any display issues with extra-long Summary/Series pages in the software.

The "unmerges" that will be required to change to the new system can be done automatically. Ahasuerus 03:22, 22 November 2016 (UTC)

I would definitely be in favor of that; the collision of presentation and data structure that was created when this decision was made seems to consistently trip up editors working on more complex magazines. Which is not to say the decision didn't serve some purpose when it first came about.
It's good to hear that the unmerges will be able to be automated; a few other things that come to mind:
  • We'll likely need a cleanup report to identify reviews that are currently linked to EDITOR type records with multiple pub records associated with them, so they can be re-linked to their proper issue.
  • There will be numerous reviews of magazines that have been entered as ESSAY type titles under the current guidelines that will be eligible for conversion; these should be somewhat easy to programmatically identify if they're titled with the current naming convention.
I suspect the display issue will be the trickiest bit of the process. Albinoflea 07:36, 22 November 2016 (UTC)
The one area makes use of a single (or as few as possible) EDITOR records per year is that of awards. We should consider the best way to handle those, not to mention how to handle the awards that are currently linked to annual EDITOR records. We could convert them to untitled awards, or perhaps change the software to enable us to link award records to magazine series. --Ron ~ RtraceTalk 11:14, 22 November 2016 (UTC)
Ahasuerus says the unmerge can be automatic. Does that include the linking of existing reviews to a particular issue? This isn't always obvious since the title may say June 2010, but the review says Issue #123. It's also possible that the particular issue doesn't exist - no stub was created because it would be the same EDITOR record. I don't know how big an problem (almost said issue) these present. The mis-matching of details in the titles also presents a problem when creating stubs for a review - presuming we do the same thing for an issue as we do for a book stub which is to fill in as much as you can using as many sources as you can reasonably find. This is easier for new issues of an existing magazine series, but harder if no such series yet exists. BTW - will unmerging affect how the magazine issue list / grid are assembled and/or presented? Doug H 14:16, 22 November 2016 (UTC)
I don't know how good of an idea this is, but that's never stopped me before: It seems that the variant title mechanism -- plus some display changes -- might give us a lot of the mechanics we want. Really, we'd like one title (EDITOR) record per issue (which would also then let us handle reprints/reissues and multi-format, BTW). Such a record could be linked for a review of a specific issue. We could then have a parent record that represents some logical grouping (might be calendar year, might be volume, might be something else) to a set of these could be varianted. Not unlike our handling of serializations of novels. That parent record could be used for issue-less situations, such as awards. --MartyD 14:25, 22 November 2016 (UTC)
I had been typing something in that direction as well and walked away to find coffee and MartyD posted :) We may have a "best of the two worlds" solution -- leave the existing yearly records in place but create variants for each of the separate issues. That allows the awards to stay on the year level, any review of a whole year can stay attached as well but gives us the individual entries. It will also open the door for a new user preference "Show individual issues" which will allow the user to tweak if they want the current view or a view that shows the individual issues as well. Although I am not sure if you can enter a review to a variant? If we cannot, we need this tweaked anyway - to allow a review of a translated work to be linked to the proper variant (and then allow the display to show it on the parent record)... Anniemod 14:36, 22 November 2016 (UTC)
Yes, you can link reviews to variant titles. For example, this review is currently linked to a VT. Reviews of variants are listed on the variant page as well as on the parent page.
The more I think about this approach, the more I like it. It should be easy to suppress the display of EDITOR variants on Summary and Series pages. You can have your garash cake and eat it too! :) Now we have to think through the proposed solution's implications and consider any unintended side effects. Ahasuerus 20:29, 22 November 2016 (UTC)
And here is the fist unintended side effect. Consider Charles Coleman Finlay's Summary page. He has been editing The Magazine of Fantasy and Science Fiction as "C. C. Finlay" since mid-2014. Under the current system, his EDITOR titles are attributed to "C. C. Finlay", then varianted to "Charles Coleman Finlay". His Summary pages displays them as:
  • The Magazine of Fantasy & Science Fiction, July-August 2014 (2014) with Gordon Van Gelder [only as by Gordon Van Gelder and C. C. Finlay ]
  • The Magazine of Fantasy & Science Fiction - 2015 (2015) [only as by C. C. Finlay ]
  • The Magazine of Fantasy & Science Fiction - 2016 (2016) [only as by C. C. Finlay ]
Under the proposed system the variant EDITOR titles will still be attributed to "C. C. Finlay", so there will be no loss of information. However, if variant EDITOR titles are hidden on Summary and Series pages, you will have to drill down to the parent title level to see them.
I guess we could tweak the display logic some more and replicate the current "only by" messages based on authorship. Ahasuerus 21:19, 22 November 2016 (UTC)
Or change the rules a bit - and have the title records for EDITOR to always contain the canonical name as we do for interviews (I think?). We won't lose information because the publications will have the correct names. If you want to limit to ones that have publications so the yearly records can be varianted that is fine as well I guess. But that will solve the immediate problem. Anniemod 22:26, 22 November 2016 (UTC)
Reviews and interviews are a special case. Regular titles capture one sets of authors/editors per title record. Reviews and interviews capture two sets per title record: one set for reviewers/interviewers and another one for "reviewees"/interviewees. The data entry rules for reviewers and interviewers are the same as for regular titles. The rules for "reviewees" and interviewees are different because they not really "title authors".
Also, AFAICT changing the data entry rules to use canonical names for EDITOR records would have the same impact as simply hiding variant EDITOR titles on Summary and Series pages. Ahasuerus 23:02, 22 November 2016 (UTC)
Yes, I understand why the special cases are as they are. I am not sure I follow why it is the same impact (or I am misunderstanding what is proposed to be hidden). If the editor records carry the canonical name, they will show up on the page authomatically (which won't happen if they are on the pseudonym as we cannot variant as they are already variants). If the change proposal is to do that behind the scenes, then yes, that would be enough. Anniemod 02:28, 23 November 2016 (UTC)

(unindent) For now I have removed the offending Help paragraph. The information that it contained was either duplicative or obsolete. Ahasuerus 21:24, 22 November 2016 (UTC)

This is interesting but I have two question: How do I enter magazine reviews now? Will the system and moderators accept requests that follow this process? Doug H 20:07, 28 November 2016 (UTC)
In re-reading this in search of answers for the same question Doug H asks, I notice it never was answered. Are we still in limbo on this or was it answered some place I don't know about? Thanks, Doug / Vornoff 06:35, 3 March 2017 (UTC)
Many R&S discussions never come to a resolution.... IMO, you are best off thinking "EDITOR" == "TITLE" and following the general review-entry rules. If the magazine being reviewed is eligible for inclusion or is already in the database, link to the closest available EDITOR record. If it's been merged and is for a full year, so be it. The reason I suggest that is because if any of the above suggestions are ever implemented, we'd easily be able to find such a review and do something with it. If it's entered as ESSAY, we'd have a much more difficult time finding it. --MartyD 12:30, 4 March 2017 (UTC)

Weird Heroes: publication or title series

We have this as a title series, probably entered long before we had publication series. This prevents Ron Goulart's Quest of the Gypsy (two novels and one story) to be in a title series of their own. I would like to convert Weird Heroes to a publication series, but that wolud leave the 2004 i-books reprint of volume 1 out of the series. any ideas (or other problems)? --Willem 20:33, 20 November 2016 (UTC)

Given the fact that these books contain titles which belong to various unrelated title series, I think it makes sense to convert it to a publication series. Ahasuerus 21:56, 20 November 2016 (UTC)
Looks like there were no objections. Conversion is done. --Willem 21:14, 26 November 2016 (UTC)

Are Baudelaire's prose poems POEM or SHORTFICTION?

Currently there are seven of Baudelaire's "Petits poèmes en prose" in the DB; "La chambre double" and "Le joueur généreux" have the title type POEM, the rest SHORTFICTION. --Vasha 21:50, 20 November 2016 (UTC)

As I recall, at least one editor worked on sorting out Baudelaire's works a while back. Unfortunately, I don't remember who it was. Ahasuerus 21:58, 20 November 2016 (UTC)
Various people verified publications with one of these prose-poems in them: Rtrace, Don Erikson and Taweiss verified Le joueur généreux as POEM, and according to a notation by ChrisJ, it may be listed as a poem in Contento too. Only Linguist verified one of them (Mademoiselle Bistouri) as SHORTFICTION. Personally, I think POEM makes more sense. Everyone seems to agree that a "prose poem" is a thing that exists. --Vasha 22:10, 20 November 2016 (UTC)
"Prose poems" are on my list of things to add to the set of recognized title types, but we'll need to have a comprehensive discussion of various permutations first. Ahasuerus 22:19, 20 November 2016 (UTC)
OK, another data point then: many critics describe the little snippets of prose that Jorge Luis Borges published in El hacedor (a collection of half verse, half prose) as "prose poems"; currently they, and a few similar pieces he included in other poetry collections, are listed as SHORTFICTION. (To complicate matters, they were included in Andrew Hurley's translation Collected Fictions but also some of them found their way into the companion volume Selected Poems as well.) They certainly are rather more lapidary in style than Borges's earlier and later "fictions". --Vasha 22:37, 20 November 2016 (UTC)
I don't see why there needs to be a separate title type for prose poems, though. They are already unproblematically recognized as poems when they occur in anthologies of poetry (like "Paint" and "I Last Saw My Mother" in this anthology). Sure, there is plenty of confusion in individual instances as to what exactly an item is; but that would only carry over to a new distinction, with people not sure whether to call something SHORTFICTION or PROSEPOEM. I think what is needed, instead, is a note added to the page on title types confirming that prose poems are poems, and leaving the problem of recognizing them (which will only occasionally be difficult) up to individual editors. --Vasha 21:37, 21 November 2016 (UTC)
Unless anyone strongly objects, I'm going to change Baudelaire's "shortfictions" to POEM to make them all consistent, but leave the other questionable cases be for now. --Vasha 02:01, 27 November 2016 (UTC)

Multi-volume "Omnibus" works with only one novel

On an issue similar to the one recently discussed, we have at least a few examples of "The Collected Works of XXX", published in several volumes, but with only a few volumes containing genre works. Should the entire "Works" be entered, or only the volumes containing genre works? This certainly applies to many editions of the Works of Shakespeare, for example, but the one that jumps out at me is Jonathan Swift. All 5 of the English editions of Gulliver's Travels that we have listed between 1735 and 1765 are 1-2 volumes of "Collected Works" editions. Four of them were entered (by me) as novels with just the volumes for Gulliver's Travels; one was entered as an omnibus with all of the volumes mentioned. (In addition to Gulliver's Travels, the "Works" also includes the "Battel of the Books", of genre interest.) Which is the correct way to enter these works? Chavey 21:56, 21 November 2016 (UTC)

I would enter just the volumes containing genre works. I would then create separate publication series for each "Collected Works" version. Ahasuerus 22:15, 21 November 2016 (UTC)
Thanks. Chavey 23:19, 21 November 2016 (UTC)
It would also be good to include a clear note on the series indicating why some volumes weren't included. ···日本穣 · 投稿 · Talk to Nihonjoe 20:35, 22 November 2016 (UTC)

Composite "story length" values for OMNIBUSes

Re-checking my recent changes to Template:TitleFields:Length, I see that the previous version of this template read:

  • Other possible uses include "/1N/2C" to indicate the omnibus contains a novel and two collections

The current version reads:

  • For other types of omnibuses, enter the count of included titles, e.g. "/3N" or "/2N+2C" where "N" stands for "Novel" and "C" stands for "Collection"

Note the difference between "/1N/2C" and "/2N+2C".

This was not an intentional change. I am so used to plus signs being used as separators that it didn't register that the 2008 version suggested using slashes.

Checking the database, I see that we have 12 titles that use slashes and 94 titles that use the plus sign. Given the current usage, I suggest that we make the plus sign our official choice and convert the 12 titles that use slashes. Ahasuerus 17:40, 29 November 2016 (UTC)

I think that plus is more logical. Anniemod 18:07, 29 November 2016 (UTC)
There isn't the slightest bit of consistency in how people describe the contents using the plus notation. (Go through half a dozen items in that search list and you'll get an eyeful.) But in all the ones I looked at, I could figure out what they meant, so I don't think it's much of a concern. Do you want to give more detailed instructions, though?
One thing that might be plain wrong is that some people wrote something like "/1-4+nv" meaning volumes 1-4 of a certain series plus a novella. --Vasha 18:12, 29 November 2016 (UTC)
Well, "/1-4" is standard usage for omnibuses containing volumes one through four. How would you propose entering an omnibus which collects volumes 1-4 plus a novella? Ahasuerus 18:28, 29 November 2016 (UTC)
Give the novella a proper number in the series and then just use /1-4+3.5? More seriously: we also may have options like /1-3+4ss for example. I would say something like - list the numbers of the known elements and then after the + add all other elements in the order N, C, nv, nvt, ss, sf (So /1,2,3.6+1N+2nv+2ss) - which means series members 1,2 and 3.6 and then one novel, 2 novellas and 2 stories that are not numbered? Anniemod 18:40, 29 November 2016 (UTC)
I think it may be an overkill considering that only 106 titles are affected by this rule. As Vasha pointed out, the codes are usually self-explanatory. Ahasuerus 22:18, 29 November 2016 (UTC)
Probably a valid argument :) Always can be revisited when the number of strange editions increase. Anniemod 22:22, 29 November 2016 (UTC)
What do you do if something is both an omnibus and juvenile? Two kinds of data to record in the same field. --Vasha 22:00, 29 November 2016 (UTC)
Omnibus data takes precedence in my opinion, jvn gets re-delegated to the notes (but this is exactly why I think that the non-length identifier should be sent to their own checkboxes. Plus jvn as length is a bit... funny. :) Anniemod 22:02, 29 November 2016 (UTC)
It's also possible for a title to be "novella", "juvenile" and "novelization" at the same time. Which is why FR 163 was created a while back. I hope to implement the FR soon (tm) and convert the "storylen" field to a drop-down list for SHORTFICTION titles. Ahasuerus 22:17, 29 November 2016 (UTC)

(unindent) OK, so the plus sign is the new standard just like the new Help text states. I have converted all eligible slashes to plus signs. Ahasuerus 16:34, 2 December 2016 (UTC)

Publication format clarification, please -- moved from CP

I've been using the format "pb" to describe the European "pocket" size of paperback, which is 19-20 cm tall (and contrasts with non-pocket paperbacks which are generally 23 cm or more). But I now notice that the template description for PubFormat says "For books as tall as 7.25" (19 cm)... use "tp"." Is that really the rule? It doesn't even fit the new practice of making American mass market paperbacks 7.5 inches tall. --Vasha 16:48, 10 December 2016 (UTC)

I've come across this as well several times regarding German paperbacks. The current size rules are already tight for old German paperbacks (which are often exactly 11.5 cm wide), but newer German books published as paperbacks are often exactly or a bit bigger than 19cm and very often almost (or more than) 12cm wide. Based on the current rules they must all be entered as "tp", though they are "just" paperbacks actually (at least if we think of a trade paperback as a "higher-quality paperback book"). Jens Hitspacebar 17:18, 10 December 2016 (UTC)
In publishing the term "mass market paperback" describes paperbacks which are not returned to their publisher if the retailer fails to sell them. Instead the cover is removed and returned to the publisher while the rest of the book is pulped. In most cases what publishers call "mass market paperbacks" are smaller than trade paperbacks, but there are exceptions.
However, the ISFDB uses the term differently. We use the book's dimensions as the only criterion when deciding whether it's a "pb" or a "tp". Thus all previously mentioned "tall mass market paperbacks" are considered "tp" rather than "pb".
We have discussed this issue a few times and there are concerns about the viability of the traditional "pb" format going forward. However, no changes have been made so far.
My personal take on it is that we shouldn't change the current definition of "pb". Instead we may want to consider adding one or more new codes for different types of "tp"s. Ahasuerus 19:27, 10 December 2016 (UTC)
If you don‘t want to change what pb means (and I agree that that would be problematic because it would alter the meaning of so much data that‘s already been entered), then there's definitely a need for an intermediate category for paperbacks that have a distinct small size that distinguishes them from trade. Let's say, between 7.25 and 8 inches/18-20.5 cm, or whatever exactly seems best. I do think it would be worthwhile distinguishing that size category which corresponds to a marketing category. How hard would it be to implement this? --Vasha 20:14, 10 December 2016 (UTC)
(afetr resolving an editing conflict) Adding new codes will be tricky though - no way to know where the current tps will fall. One option may be to leave the current "tp" as is and as a catch all and allow other "tp" ones -- so the old ones remain as is, one can still use "tp" if they do not know where it falls and allow better separation... I would love to be able to differentiate between small tps and the real big ones. Maybe instead of playing with the type, add a new field to allow dimensions to be added? This way the tp will basically mean "softcover and not too small" and the sizing will allow different international formats to start getting grouped together... Anniemod 20:20, 10 December 2016 (UTC)
Yes, the real objection to any sort of change is how to convert old data -- much too difficult, aside from doing things like finding German books that are published by a publisher that specializes in Taschenbücher. Maybe just an additional data point for books that are tp? "Is it under 21 cm" and for existing data this would be unspecified? Does that make sense at all? --Vasha 20:32, 10 December 2016 (UTC)
Let me first mention that it would be very easy to add another supported "binding code". It would be significantly more time-consuming to add a new field.
Having said that, one thing that we may want to do is compile a list of more granular paperback formats that we have to deal with. Off the top of my head, there are:
  • UK: "B Format" -- 130mm x 198mm (5.12″ x 7.80″)
  • UK: "C Format" -- 135mm x 216mm (5.32″ x 8.51″)
  • US: "Tall but thin mass market paperbacks" -- ?? dimensions
Are there other standard sizes? Ahasuerus 22:21, 10 December 2016 (UTC)
Here is a summary of the sizes that German publishers use. The standard Taschenbuch is 130 mm x 190 mm (5.12″ x 7.48″).
Here is a French printer listing formats they offer; there apparently are a variety of small formats, and googling suggests that there may be even more than the three listed on that page (11 x 17 cm, 12 x 18 cm, 11 x 20 cm). I actually found a journalist, asked to research the standard size of French paperbacks, writing: "Je n’ai pas trouvé non plus de normes sur le format du livre dit de poche" (I didn't find any standards for the format of so-called "pocket" books). --Vasha 22:54, 10 December 2016 (UTC)
Most Turkish publishers use a height of 195 mmm (7.68″), while the width varies from 125 to 135 mm (4.92″ to 5.32″). --Vasha 22:41, 10 December 2016 (UTC)
Bulgarian genre books come in 2 standard formats (except for special cases like a few series in pocket formats or a publisher trying to be interesting which can be handled separately) - both end up as tp here but they are quite distinctive (the smaller one is ~130 mm/200 mm so it will fit with the UK "B" format if we give it some leeway; the bigger one is ~160x240, which is bigger than the UK formats. I am getting exact dimensions of the two formats but these are close to it. Anniemod 22:50, 10 December 2016 (UTC)
As for the German paperback sizes linked by Vasha above: these are the current ones. But they have been smaller in earlier years (the current one for "Taschenbuch" is a lot bigger than the one from the 70s or 80s). It looks like the "tp" and "pb" sizes change over time and also vary between countries (if we only want take the size of the books into account for the "tp/pb" distinction and not the print/paper quality). The current "tp" and "pb" size rules seem to have been set for US/UK formats, probably at a time when the database contained mostly English publications. Is it possible that the distinction between "tp" and "pb" should be made language-dependent (or country-dependent)? Though this might be a bit over the top regarding rule or software changes... If I may ask a heretical question: is it necessary to record the "tp" and "pb" distinction at all? Why not just "hc" versus "sc" (softcover)? Just asking... :) Jens Hitspacebar 23:12, 10 December 2016 (UTC)
We first made the distinction between "pb" and "tp" in the mid-1990s during the initial design phase. At that time mass market paperbacks still ruled the publishing world, although their reign was coming to an end.
I think the "pb" format is worth preserving in order to accurately represent the way things were done in the 20th century. Based on current trends, mass market paperbacks will soon become a shrinking ghetto and "tp" will become the de facto equivalent of what Jens has described as "sc".
Given the number of possible permutations that we were able to identify in under an hour, I am not sure that coming up with additional "binding codes" for different dimensions would be viable. If we were to add a dozen+ new codes, it would make life more difficult for editors as well as for users. Ahasuerus 23:41, 10 December 2016 (UTC)
Actually, if you allow a slight amount of leeway, a width of 13 cm (plus or minus .5) and a height of between 19 and 20 cm covers a vast number of books being currently printed: tall American mass market, B Format, modern Taschenbuch, Turkish, the smaller Bulgarian size... --Vasha 23:58, 10 December 2016 (UTC)
That might work, but too much leeway could blur the lines between pb and tp. I just did some samples with Germam tp, and they can be just 20.4 x 13.0 cm. Jens Hitspacebar 00:23, 11 December 2016 (UTC)
How about, larger than pb, but no taller than 20 cm and no wider than 13 cm? That would probably leave out a few books that call themselves pocket size, but not very many. --Vasha 00:35, 11 December 2016 (UTC)

(unindent) It looks like the latest version of the proposal is to redefine the paperback formats as follows:

  • "pb" stays the same and covers paperbacks up to 185mm tall and 110mm wide
  • "tp" covers paperbacks that are larger than "pb", but we don't know the details
  • "new format X" covers paperbacks that are between 186 and 200 mm tall and between 111mm to 130mm wide
  • "new format Y" covers paperbacks that are >200mm tall and/or >130mm wide

Is this about right? Ahasuerus 05:02, 12 December 2016 (UTC)

That‘s so, yes--Vasha 15:35, 12 December 2016 (UTC)
In this definition the "pb" format doesn't stay the same. Its current maximum width is not 110 but 115 mm (Template:PublicationFields:PubFormat). Jens Hitspacebar 20:36, 12 December 2016 (UTC)
Oops. Sorry about that! Ahasuerus 21:31, 12 December 2016 (UTC)
The "pb" size should probably keep this maximum width (or leeway), otherwise at least a lot of the older German paperback pub records need to be changed to the "new format X". They are often exactly 115mm wide (e.g. the pubs from the big Heyne Science Fiction & Fantasy pub series). Jens Hitspacebar 20:36, 12 December 2016 (UTC)
With the non-change in pb, that sounds about right (meanwhile got my confirmation from back home - Bulgarian big format has two varieties (145×210 (the new big) and 165×240 (the old big)). Both will comfortably fit in the biggest format. The "normal books" are confirmed at 13×20 (which is where the smaller new format go) so these will be perfect - biggest book that can go in that category) :)
We may also want to start a new page "per country of origin" or something to help new and not so new editors find what format to use easier... with examples of books from that country. Anniemod 20:46, 12 December 2016 (UTC)
@Jens: that extra quarter inch was to allow for inexact trimming, and I think we need to keep allowing that; we should be able to tell (I hope) whether a book is larger because of the format or just because of trimming. If those books were always 115, then yes, they belong in the intermediate format... --Vasha 21:36, 12 December 2016 (UTC)

(By the way, whether or not this new format is added, I need to go back and find all the 19-20 cm books I mistakenly put in as pb -- would there be some efficient way of searching my edit history to pull up all the NewPub, AddPub, and ClonePub?) --Vasha 21:19, 12 December 2016 (UTC)

There is a Feature Request to "Add the ability to search submissions", but it hasn't been implemented yet. For now the most direct way to get this information is by paging through "My Recent Edits". Alternatively, if the problematic submissions followed a pattern, e.g. they were by associated with certain publishers, you can try Advanced Publication Search and limit it by publisher name and binding code. Ahasuerus 01:18, 13 December 2016 (UTC)

(unindent) I guess we'll need to have names for "new format X" and "new format Y" mentioned above before the proposal can be put up for consideration in a new section. I am not sure what they should be since they cover a number of different scenarios. Ahasuerus 01:32, 13 December 2016 (UTC)

Small tp and large tp? Anniemod 01:36, 13 December 2016 (UTC)
"tp" stands for "trade paperback", which the smaller proposed format is not. --Vasha 01:51, 13 December 2016 (UTC)
Based on what definition and in which country? The B format of UK is technically a trade paperaback. Both of those are now called "tp" around here already, we are just splitting them in two categories.:) Anniemod 01:58, 13 December 2016 (UTC)
True... the term "trade paperback" doesn't make sense in an international context. Will this be like the AARP which, when they wanted to emphasize that not all their members were American (or even retired), stated that the letters didn't stand for "American Association of Retired Persons" but were just letters? --Vasha 02:07, 13 December 2016 (UTC)

(unindent) So - are we going to be adding new formats and if so, when? :) (I am sitting on a batch of books that will fall under one of those new ones - so want to know if I should add them (and add a note so I can find them for changing later) or if I should just sit on them a bit longer. Anniemod 21:22, 15 December 2016 (UTC)

I think the next step is to do a short write-up and post it as a separate section. That way editors will be able to decide whether to vote "yes" or "no" without having to follow the intricacies of the debate.
Before we can do that, we'll need to clarify a couple of points. First, just how tall are the American "tall" mass market paperbacks? Second, what are the proposed names for "new format X" and "new format Y"? Ahasuerus 22:21, 15 December 2016 (UTC)
7 1/2 x 4 1/4 inches according to this. I have a tall mass market paperback on my night stand at home at the moment so can confirm the exact when I get there in a few hours (unless if someone has one handier) :) There might be a third format as well (wider (5 inches or a bit more than that) but not as tall as the tall mmpb - but I am not sure how many of these are around - ha, found one of my sources here). For the names - see my proposal above? Anniemod 22:32, 15 December 2016 (UTC)
That thread you linked says that those wider books are digest size, and also that they may be produced only for Wal-Mart? anyhow I can't find what the exact size is -- couldn't find them on the current Wal-Mart site. --Vasha 22:58, 15 December 2016 (UTC)
Yeah, looks like that. And just to be annoying, the tall mmpb I can get easy access to is 7 1/4 x 4 1/8. But I know that I have a few that are 7 1/2. Anniemod 01:48, 16 December 2016 (UTC)
OK, so they are in the 7.25-7.5" range.
Next, we need to clean up our math. PubFormat says "For books as tall as 7.25" (19 cm)... use "tp"." However, 7.25" is 18.145 cm rather than 19 cm; 7.5" is 19.05 cm.
We need to decide whether the cut-off point is 7.25" (and change 19 cm to 18.15 cm) or 7.5". Since we have been entering 7.25" paperbacks as tp, I vote for the first option. Ahasuerus 16:56, 19 December 2016 (UTC)
It almost starts sounding as we should rename the pb to pocket book (no rename - the name works like that) and the tp to sc (for softcover) and be done with - this way the current data will be valid, we will be closer to the hardcovers (because we do not split them there) and cause less confusions. The only issue will be books in the same series being sc and tp depending on who adds them but it will always be happening with people needing to measure their books (and we can give some guidelines per language for the big series I guess)...
On the math, I vote for going with the current inches values. Annie 17:14, 19 December 2016 (UTC)
Except see here, where Christian tells me that he has been adding German Taschenbücher up to 19 cm as "pb" all along -- lots of them, I think. There's at least one vote for redefining pb to be 7.5 inches/19.05 cm. I would be OK with that -- in that case, the B format would just be tp (or sc per Annie).
I think that defining pb as less than 19.1 cm would resolve some unclarity in existing data; not only the Taschbücher issue, but also some people, misled by the occurrence of "19 cm" in the help, may have added some books over 7.25 already -- the space between 7″ and 7.5″ is a grey area. Such a redefinition would bring "pb" a little closer to a general "pocket book".
I am not sure that having three categories would be a good idea, but if clarifying the definition of pb was to make it 19cm instead of 7.25″ I think that would be an improvement. --Vasha 05:27, 20 December 2016 (UTC)

Non-Genre Magazine Cover Art

Where do we stand on non-genre magazine cover art? The help says they are not to be entered unless they illustrate a genre story or are by a genre artist. However, we seem to have quite a few of them entered. I have an edit pending on hold related to this. The editor's note to moderator states "Not sure if I should download a cover image, but all the listed issues of this magazine have cover images." I think their confusion is justified so am seeking broader opinions. Thanks. -- JLaTondre (talk) 20:39, 17 December 2016 (UTC)

The last time the question came up, the majority was against including non-genre cover art. As I recall, one of the arguments was the sheer number of non-genre artists that we would be adding. Ahasuerus 20:48, 17 December 2016 (UTC)
The only reason we have the non-genre magazines at all is so that we do not need to invent another way to add the stories they contain (please correct me if I am wrong). So unless the cover is the story's art, it really does not have much of a space here -- otherwise how is that different from a non-genre article in the same magazine? Or a non-genre story? Anniemod 21:04, 17 December 2016 (UTC)
As somebody who has been called into the principle's office a number of times for this, I'm ambivalent about non-genre covers, unless they are directly related to the listed speculative content(s). I have posted cover images for some non-genre items because either the covers have something speculative about them, or they illustrate something speculative inside. On the other hand, I have listed something speculative in any number of Alfred Hitchcock's or Ellery Queen's and except for the covers of about two, the rest have cover images that I didn't post, wasn't aware of until recently, but which are still there. Listing the cover artist, I think, is fine, but if we don't accept cover images, somebody is going to have to go back and do a lot of deletions. Maybe a cut-off point? Like nothing past the fifties. Don't know, just asking. I suspect as more editors get involved with this site, and more non-genre sources, like the slicks and pulps that are constantly being looked at, this question is going to come up again and again in the near future. MLB 22:40, 17 December 2016 (UTC)
I'm also opposed to including cover photos or cover artists for non-genre magazines, except for those rare circumstances where the cover art is for the genre story. I would be rather uncomfortable with including a bunch of Playboy covers just because the magazine included an SF story. Chavey 06:59, 18 December 2016 (UTC)
It's a likewise standpoint with me. It'd still be possible to link to amazon, if the cover's there available. Stonecreek 19:42, 18 December 2016 (UTC)
Based on the discussion, the submission was rejected, the image was deleted, and additional images from other issues removed. Thanks. -- JLaTondre (talk) 00:10, 21 December 2016 (UTC)

Web or online exclusive supplements to a printed book

This is pretty common lately with the journals and reviews published from the various colleges. Most of them are not genre so I am not sure if it ever came up (cannot find anything - if it was already discussed, please let me know). The latest issue of Conjunctions is Speculative Fiction one (most of it for sure, they claim the whole of it - had not read it yet so cannot say). And they do have a few stories in a "Web-exclusive supplement" as usual for them lately and at first glance at least one of the stories is eligible for inclusion (if not all of them actually). So what is the policy for this kind of cases? Annie 01:59, 20 December 2016 (UTC)

As to whether to include all the contents, the policy page says to include "non-genre speculative fiction" so I guess if Conjunctions says they think these stories are speculative, they're in. --Vasha 03:15, 23 December 2016 (UTC)
I am more concerned about what to do with the online stories (do we include or not and if yes, how), not the content itself - the printed ones will go in in full. :) Annie 03:21, 23 December 2016 (UTC)

Need help identifying book as spec fic

I'm unclear as to if the following book is spec-fic or not.:

Phantom Sun/White Needle by Carl Bowen

The book is about a special forces group called "Shadow Squadron" tasked to recover a crashed unidentified flying object from Antarctica. The object turns out to be a hunter-killer drone satellite, well in the zone of spec-fic. It was published by Scholastic, dos-a-dos with another book in the series. White Needle, not obviously spec-fic (Syrians and chemical weapons). The book is glossy and illustrated, first published by Stone Arch Books.

If one is spec-fic and the other isn't, can it still be included?--Auric 22:35, 7 February 2017 (UTC)

If one of the two stories is genre, then the publication is in. I think that given it's dos-a-dos, and the author has a bunch of spec-fic works, it would be ok to index the other story and mark it non-genre. Even if we don't consider Bowen "above the threshold", it's a pretty clear case that wouldn't necessarily open any floodgates. But I do have inclusionist tendencies.... --MartyD 02:45, 8 February 2017 (UTC)
Thanks. I've just submitted it.--Auric 13:01, 8 February 2017 (UTC)

Capitalization of "a.m." and "p.m."

I realize the English capitalization rules leave a lot to be desired, but I'll try to ask a simple, focused question and hope we can keep the discussion to just that: Should "a.m." and "p.m." be among the things that are always lower case? We are rather inconsistent, and to me "A.M." and "P.M." in mixed-case titles looks weird. I don't think that is my U.S. English handicap speaking.... --MartyD 11:48, 27 February 2017 (UTC)

As they are in practice abbreviations of words that would have been with capital letters, it is more logical to have them capitalized. At least for me anyway. Not that we follow the most logical rules all the time though :) Annie 18:28, 27 February 2017 (UTC)
A little googling finds the following quote (note the bad HTTPS connection) from "Common Errors in English Usage", third edition, 2013:
  • formal writing it is still preferable to capitalize them, though the lower-case “am” and “pm” are now so popular they are not likely to get you into trouble.
Ahasuerus 21:16, 27 February 2017 (UTC)
Well, Paul Brians isn't one of the most commonly cited authorities. The Chicgo Manual of Style, the AP Stylebook, and Garner's Modern English Usage go with a.m., all of them more respected sources than Brians. So I guess Marty has the majority behind him (to my surprise -- it's not what I expected to find when I started looking at stylebooks). --Vasha 22:50, 27 February 2017 (UTC)
For Title Case? I know that their recommendation for regular writing is to leave them small letters but last time I looked (I'd admit it was awhile), they were not on the exception list for Titles Case capitalization Annie 22:59, 27 February 2017 (UTC)
Ah, good point, I wasn't thinking! I really can't find anything about that. However, two letter abbreviations like vs. and e.g. don't get capitalized in titles (I think!), so I guess maybe not?? --Vasha 23:13, 27 February 2017 (UTC)
Yeah, I know. But something is bugging me for this case specifically - it had stuck in my mind as always getting capitalized because of their function (unlike e.g. for example). Bulgarian does not use title case (Titles use the same capitalization as standard writing) so the rules were a pain to remember and some of them had been stuck for awhile :) Annie 23:28, 27 February 2017 (UTC)
There are three widely used "style manuals" in the U.S. -- MLA, APA, and Chicago. MLA says to use "a.m." (lower-case, periods, no extra spaces). APA says to use "a.m.". The Chicago Manual of Style says to use either "a.m." or "AM" (small caps, no periods, no spaces). It's not clear that these rules follow over into titles, but the rules are surprisingly consistent for other uses. Chavey 06:18, 28 February 2017 (UTC)
Yes, and Garner (Oxford University Press) is one of the usual sources for the UK -- same again. The problem is I don't have an actual copy of any of them to search if they go into nitpicky details about titles. --Vasha 09:32, 28 February 2017 (UTC)
The Random House Handbook claims there is no consensus on capitalizing A.M./P.M. In any case, we should be consistent. I would vote for capitalization in titles, since it looks wrong to my eye not to do so.--Rkihara 18:19, 28 February 2017 (UTC)

(unindent)I have copied out some sections from the Chicago Manual of Style here. The book does not address this particular issue at all. However, the sections I copied address some issues which (I believe) have not been standardized here. If they're ever debated, the reference material is now available. I'm now off to check other reference books. --Vasha 18:48, 28 February 2017 (UTC)

Unsurprisingly, the other works consulted didn't answer the question of abbreviations either.
  • The APA Publication Manual (5th edition) goes into very little detail, saying only "Capitalize major words in titles of books and articles.... Conjunctions, articles, and short prepositions are not considered major words; however, capitalize all words of four letters or more. Capitalize all verbs (including linking verbs), nouns, adjectives, adverbs, and pronouns. When a capitalized word is a hyphenated compound, capitalize both words. Also, capitalize the first word after a colon or a dash in a title." This is much too vague, and leaves open the possibility that there may be words other than "conjunctions, articles, and short prepositions" that are "not considered major words". One difference from the Chicago Manual of Style to note is that, where CMoS recommends not regarding words after a dash as a subtitle for capitalization purposes (their example is Chicago—a Metropolitan Smorgasbord) APA says "capitalize the first word after a colon or a dash in a title". In this case, at least, the APA makes more sense to me.
  • I have copied some sections from the MLA Handbook (7th edition) here, for general reference. On the question of capitalization after a dash, they side with CMoS.
--Vasha 21:01, 28 February 2017 (UTC)

(later addition) Being as the books give no answer, we should just choose a practice and stick with it. I suggest "Always capitalize abbreviations unless they are a preposition like vs.; capitalize every initial before a period (A.M., N.B.)" --Vasha 00:33, 9 March 2017 (UTC)

Disambiguating "untitled" Interior art

We have a policy to disambiguate Introductions and similar items but it seems like we do not have one for Interior Art titles called "untitled". As a result, for example has multiple titles called like that (and in the same year look at 1953 for example). Which makes it meaningless in her bibliography. Shouldn't we follow a similar rule for these as we do with Introductions for example and mark where this generic title belong to? There are 208 entries like that in the DB now (list) so it should be easy enough to convert them. But we need a rule for that. Annie 20:31, 27 February 2017 (UTC)

It seems natural to extend the scope of "generic titles" treatment to cover this case. --MartyD 03:38, 28 February 2017 (UTC)
I agree. I've generally entered "Title of the Book" as the title and set the title type to "INTERIORART". ···日本穣 · 投稿 · Talk to Nihonjoe 19:22, 28 February 2017 (UTC)
That's what I would do for an untitled art - name it after the work it belongs to or if it does not belong to a specific one, use the book name. But there are a lot of Untitled ones out there. Thus the question :) Annie 20:44, 28 February 2017 (UTC)

Handling of suffixes on author names

It seems actual practice for the Legal Name field on authors deviates substantially from the help's: .... No suffixes or prefixes should be used; "Jr.", "Capt." and so on are not recorded in this field.. Some examples:

One problem I noticed from an editor's submission I processed is that if the author's actual name contains such a suffix, but the author does not use it (so no canonical name having it), there is nowhere in the author record to capture it.

Should we change the policy for this field to reflect practice, and, if so, what should the standard for position of the suffix be (at the end, or after the surname)? I included the III example because the first page of results illustrates the present inconsistent placement. --MartyD 03:45, 1 March 2017 (UTC)

As we are already forcing Last, First (which is not the usual way for a lot of countries), I am in favor of Last, First, Suffixes. The only issue will be if First contains commas which I am not sure happens often enough to worry about it. Annie 22:32, 1 March 2017 (UTC)

(unindent) There is definitely an issue where a suffix is part of the author's legal name and should be recorded. I propose we reword this portion of the existing Template:AuthorFields:LegalName text:

No suffixes or prefixes should be used; "Jr.", "Capt." and so on are not recorded in this field.

as follows:

No prefixes should be used: "Capt.", "Mrs.", "Sir", and so on, are not recorded in this field. Suffixes that are not part of the legal name -- for example, "Sr.", educational and professional designations such as "Ph.D." or "M.D.", ranks and orders such as "Col.", "K.G." -- also should not be recorded. Certain suffixes such as "Jr.", "II", "III", and so on, may or may not be part of the legal name. We assume they are not, unless we have reliable evidence to the contrary. If a suffix is part of the legal name, use the format "Lastname, Firstname Middlenames, Suffix".

I also think the existing text's "The format of the legal name should be" can be tightened up to "Use the format". The full result (with changed text highlighted here, but not in the final presentation):

Legal Name - This field should contain the most recent legal name for the author. For example, James Tiptree, Jr. was a pseudonym of Alice Sheldon, who was born Alice Bradley. Alice Sheldon is the most recent legal name for this author. Use the format "Lastname, Firstname Middlenames", with all names being given in full. The reason for this format are names like “Patrick Nielsen Hayden” where you can't readily tell whether the last name is “Hayden” or “Nielsen Hayden.” No prefixes should be used: "Capt.", "Mrs.", "Sir", and so on, are not recorded in this field. Suffixes that are not part of the legal name -- for example, "Sr.", educational and professional designations such as "Ph.D." or "M.D.", ranks and orders such as "Col.", "K.G." -- also should not be recorded. Certain suffixes such as "Jr.", "II", "III", and so on, may or may not be part of the legal name. We assume they are not, unless we have reliable evidence to the contrary. If a suffix is part of the legal name, use the format "Lastname, Firstname Middlenames, Suffix".

I suspect it might be useful to bold the two format strings to make them easier to pick out. Anyway, what do people think? Is it clear? Concise enough yet comprehensive enough? Acceptable? Oh, and an aside: Should "Lastname" be changed to "Familyname" to be consistent with the change we made (back in 2014) to the label for that separate field? --MartyD 13:25, 4 March 2017 (UTC)

The motion carries due to lack of dissent. --MartyD 13:12, 12 March 2017 (UTC)

Regularization of Publishers

As Bill and now Dirk are now discreetly (without notifying PVs) de-regularizing publishers (for example yesterday changing back "Coronet" to "Coronet Books", or "Magnum" to "Magnum Books") that were specifically regularized, my question is simple: "Do we still regularize publishers?". If this policy is officialy abandoned, it would be nice for the people that are coaching new contributors and try to teach them our idosyncrasies (regularizing publishers is one of them) to be informed of this fact (and so the corresponding cleanup report that I exploit every day be deleted). If it's not the case (we keep on regularizing), it would be similarly nice for every contributor to follow the agreed-upon rules.Hauck 07:55, 8 March 2017 (UTC)

And just in case my question is for NOW, so please do not deviate on a hypothetical future system of publisher's variants. Hauck 10:12, 8 March 2017 (UTC)
I prefer "as in the publication", rather than regularized, as I think it's easiest to follow and to explain (although it's also the most difficult for the moderators to notice errors). But it seems we'd need some sort of pseudonym-like, or other linking, functionality to allow one to find all publications by a publisher, independent of vagaries of name. Without that, it seems some amount of regularizing is the most practical approach. My understanding (and what I've been enforcing/allowing) is that we regularize. --MartyD 11:40, 8 March 2017 (UTC)
I like the idea above of creating variants. We'd have to carefully think out how to apply it in order to deal with publishers merging and splitting over the years, but we could likely come up with a good way to handle it. ···日本穣 · 投稿 · Talk to Nihonjoe 07:20, 20 June 2017 (UTC)
When I am adding, I always check how we have the publisher in the DB already and use that name. Going back to "as is in the book" will make all of the old regularized records incorrect... We may actually be better served with two separate fields - regularized and "as is" Publisher - one used for showing the lists, one used for completeness. But that is going off script so I will shut up. Annie 16:12, 8 March 2017 (UTC)
My preference would be to use "as in the publication", with a limited number of exceptions. For example, I don't see much value in including ", Inc." or ", LLC", but I do think there's a value in distinguishing between "so-and-so", "so-and-so and Company", and "so-and-so & Co.", since these variants sometimes tell us about particular year ranges for the publication, which then is useful for pubs without years. Chavey 03:47, 9 March 2017 (UTC)
For good reasons, I think, we regularized publishers: Like ", Inc." or ", LLC" the addition of Publisher, Books, Editorial, Editore, Verlag etc. not only doesn't tell us much, it also would devaluate the summary page for the publishers, as it would disintegrate with every slight change of name. In effect, the reasons are the same as for regularizing titles and author's names. And as Marty hinted at, moderating the publishers side, will be extremely more difficult! Christian Stonecreek 04:23, 9 March 2017 (UTC)
Notes on the book level (or even on a publisher level) can add all those details on the "used name at a certain time". Unless if we have a good way to connect the different ones (2 fields or the current over-regularized ways), we will start losing the connections inside of a publisher. I am all for complete "as is" record but if that makes it impossible to find all books published by a publisher in a year, the DB simply does not work :) Just thinking aloud... Annie 04:42, 9 March 2017 (UTC)
An interesting issue cropped up in a submission today. Tolkien: min vän Ronald och hans världar had a publisher credit of Atlantis, which is the German "Atlantis Verlag", while the book is in fact published by the Swedish "Bokförlaget Atlantis". "Bokförlaget" is the equivalent of "Verlag" (notice the förlag). So, which is better/worse? "Atlantis Verlag" and "Bokförlaget Atlantis" or "Atlantis (Germany)" and "Atlantis (Sweden)"? For now, I let the "Bokförlaget Atlantis" change through and did not touch the existing Atlantis, since some form of disambiguation is required, but that's clearly inconsistent, so one of the two should be changed further. --MartyD 10:39, 18 June 2017 (UTC)
In this case, I would personally prefer Atlantis (Germany)" and "Atlantis (Sweden)", since I think those are both more readable to those without extra language skills, and more likely to be usable to someone entering data from, say, WorldCat. In a random sampling of 30 books in WorldCat from "Atlantis" (Germany), 9 of them included the word "Verlag", but all 30 included the city from which it was published. Chavey 01:59, 20 June 2017 (UTC)

Stories initially published in non-genre media

We seem to have two separate ways to handle these cases:

  • Note in the title record mentioning the first publication
  • Adding the magazine/newspaper/book as non-genre and with limited content (as per the rules).

Is there any reason to have both ways and if so, what is the preferred one in which case? For the record - I prefer the second way - it allows the first publication to be visible in the publications table PLUS it allows to spot magazines/newspapers that print our stories often. So is this one of those "editor decision" things or am I missing a distinction and/or a rule here? Annie 00:27, 16 March 2017 (UTC)

I prefer the second way, and generally use it for the reasons you give. ···日本穣 · 投稿 · Talk to Nihonjoe 00:37, 16 March 2017 (UTC)
I prefer the second way, but additional factors come into play: There's no requirement that non-genre publications be entered, so some editors may choose not to do so. Sometimes the information comes from attribution/credits given in another publication that's being entered, so all the editor has available is that information. In both of these cases, we certainly prefer the note to nothing. And once the note is there, do we care if it remains, should the publication cited be entered (if we even noticed)? --MartyD 01:29, 16 March 2017 (UTC)
Yeah, for the stories where we do not know the exact issue/publication, the note is the only way anyway. I was really thinking about the ones where we do know all the details - they were just never used to create the parent record. Editors have a choice not to add them of course - but if it is noted already and someone stumbles on it, is there a valid reason not to add it? Annie 01:59, 16 March 2017 (UTC)
As a general rule I like having publication records because they can be verified. In addition, even if they can't be verified, they let us capture whatever bibliographic data we have about the title's appearance.
If I recall correctly, the main argument against creating pub records was that it could greatly increase the number of non-genre records in the database. I am not sure it's a significant concern at this point given the limited number of pub-less titles that we have on file. I may be forgetting some other arguments, though. Ahasuerus 02:13, 16 March 2017 (UTC)
Well, a lot of the Bulgarian first translations and stories were published in non-genre magazines (genre magazines did not develop until very late) - some of them similar to Nature - one story per issue; some of them doing one in a blue moon. It still is important information and as we are a DB, cross-indexing is important. Which made me wonder why the practice is as is. Thanks for the background. Annie 02:24, 16 March 2017 (UTC)

(discussion of the proposed cleanup project moved to this section of the Community Portal)

Author's wishes vs ISFDB practices

According to any rule we have (written or spoken), the canonical name for an author will be the most used one (or the most recently used one anyway - as it tends to become the most used sooner or later). Apparently this does not apply if an author prefers a different version to be their canonical ones. Can we have our help page edited to reflect this?

The case that brought that is W. C. Roberts - ~65 titles, including all current ones under that name and yet, the canonical is set to Willis Couvillier. My attempt to reverse the pseudonym was rejected due to author wishes (no note anywhere on the account about it). If this is to remain this way, we need a few things to happen:

  • A note need to be added to the canonical name in the effect that: "Despite ISFDB accepted practice, the canonical name is set to a name that is not the common name due to the wishes of the author". That will stop someone (me including in a few months when I forget) to try to "fix" it again.


  • all those 64 remaining titles need to be varianted to go under the canonical.

Can someone confirm that this is the correct approach? :) Annie 22:19, 21 March 2017 (UTC)

If there is agreement on the author's wishes trumping standards, than that would be the best way to handle it. Personally, I don't care too much, but I think that is starting a slippery slope. However, whichever moderator accepted the original pseudonym and/or rejected your change should be the one that does the varianting. They should never have left it in that state. If they want to trump standards, they should do the cleanup and not leave it for others. -- JLaTondre (talk) 00:02, 22 March 2017 (UTC)
It's exactly why I stopped cleaning-up (except my own and the approved-by-me mistakes). Hauck 07:04, 22 March 2017 (UTC)
I don't recall a decision to make the author's wishes the decisive consideration when determining the canonical name. Granted, sometimes an author can provide additional information about the way he or she has been credited, which may help us decide which name to use as the canonical name. At the same time an author who is not an active editor may not even fully understand how we use "canonical names", so the ultimate decision needs to be made by editors and moderators. Ahasuerus 00:57, 22 March 2017 (UTC)
It was at some time a common phrase that we are an author-friendly site: that is, we do recognize and try to fulfill author's wishes, if possible. If that attitude should have changed, please someone should do a general revocation. (This may have in fact have changed with some of the moderators who were responsible in those far-away times are now absent). Stonecreek 21:20, 22 March 2017 (UTC)
I don't think we have ever had a general policy of trying to comply with author wishes. For example, when authors ask that their works be removed from the database, their requests are uniformly denied. There have been enough requests to warrant updating our Policy to say "the ISFDB doesn't delete publications/titles that are known to have been published".
At the same time, we do consider author requests and sometimes adjust our policy to accommodate them. For example, early on we received an author request not to add her pseudonyms which had not been made public at that point. That's when we amended our Policy to say that "only publicly available sources are used". Just this month we updated Policy to formalize the process of removing biographical data based on author requests.
So I wouldn't say that the general policy has been to comply with authors' wishes/requests, but we try to take them into account when formulating our policy and/or our data entry rules. In this case the rules say that "For authors who publish under multiple names, the canonical name is the most recognized name for that author." Of course, if there is a proposal to change this rule, we can start a discussion on this page. Ahasuerus 21:47, 22 March 2017 (UTC)
So the agreement is that there is no agreement? I do not like accepting author's wishes for something like that - it is different for personal information but if they are using multiple names, they cannot tell online bibliographies how to present their data if the two names are already connected. If we decide to do that, I will follow suit but I really do not like it. Just saying... Annie 18:24, 22 March 2017 (UTC)
It seems to me that everybody that gave an opinion think that WE (our select club of supreme bibliographers) should decide (so do I). Perhaps should the person that created the pseudonymistic link and rejected your changes express him/herself. IMHO the only situation where the author can have his/her say on the matter is when there are sex/gender changes (IIRC we've already had such cases). Hauck 18:52, 22 March 2017 (UTC)
I was the one who did the rejection, but I think (if I remember correctly) that I was only seeing the acceptance of the pseudonymous relationship now given as submitted by the author (if that was not the case, and I was in fact the acceptor, my apologies). The fact remains that the author himself did the organisation, and we should be indebted to the author's wishes, as this was what we set out to fulfil at some time. Stonecreek 20:14, 22 March 2017 (UTC)
So you think we should leave the works split between the two pages? That makes no sense in the context of how we organize data. I am sorry but I disagree with you here - we have a practice to put all works of the same person on the same page regardless of how they had been credited. If they need to be on split pages, then the pseudonym needs to be broken altogether, pages left as two separate entities (with no cross references at all in terms of variants and the like) and notes added on the reasons for this so they never get reconnected. Otherwise this whole thing will open a can of worms that will render the DB unmanageable. Annie 20:22, 22 March 2017 (UTC)
I don't know where you did get that idea from! I have varianted twenty or so items to the parent author, but will leave the others to variant to other editors. After all, it's a community effort. Stonecreek 20:33, 22 March 2017 (UTC)
No change in the accounts for a week after we initially talked about that kinda made me think that you are ok with the split pages. That's all. Sounds like we are on the same page regarding this. :) Annie 21:51, 22 March 2017 (UTC)

(unindent) All variants and the note for the account explaining why the pseudonyms are set that way are now submitted so that someone does not try to improve it later on. All it needs is a moderator to approve them all. Annie 18:40, 23 March 2017 (UTC)

Done. Hauck 20:08, 23 March 2017 (UTC)
Thanks! Annie 21:33, 23 March 2017 (UTC)

LibraryThing as an image source

I recently tried to replace a dead Author image with an updated version, and got a notice that LibraryThing is not on the approved link list. lists three authors with LT images, two of whom, Frank R. Stockton and Lord Alfred Tennyson also have non-functioning images. If LT is not approved, how were they linked in the first place?--Auric 09:39, 29 March 2017 (UTC)

I'm not sure that finding by who and/or why and/or if an eventual mistake was made is very useful. The reason is very likely that, at the beginning of times, there was less caution about what was included in the db. I've deleted the links. Hauck 10:42, 29 March 2017 (UTC)
Thanks for fixing that.--Auric 12:35, 29 March 2017 (UTC)

Nongenre works (Part I): editors

In accordance with this discussion, I've been adding a few issues of nongenre magazines when they contained the first publication of a genre story. Likewise, I've added nongenre anthologies. My question is, the same way that the magazine's editors are stated as "Editors of [Magazine Title]", is it the usual practice to list the anthology as by "Editors of [Anthology Title]"? The help page says "Any issues of any periodical containing works of speculative fiction can also be entered in the same way as a non-genre magazine. So can anthologies that have some SF content, but contain mostly general fiction." Does "in the same way" extend to the matter of editors? --Vasha 14:28, 4 April 2017 (UTC)

OK, since no one has an opinion, I am going to do it that way, with anthologies credited to "Editors of". --Vasha 15:30, 5 April 2017 (UTC)
Non-genre anthologies are credited to their actual editors. "Editors of" is only used for non-genre magazines. That seems reasonable to me as creating a single author record per anthology is overkill and many non-genre anthologies are by genre authors. -- JLaTondre (talk) 18:22, 5 April 2017 (UTC)
On the other hand, this would also prevent there from being author records for people with no genre work. Maybe only editor credit for "above-the-line" authors? Admittedly that would not be easy to implement.
As for the concern about having too many editors-of records, those are not intended to be used for anything anyway, and the nongenre anthology editors often only have one credit themselves. --Vasha 19:11, 5 April 2017 (UTC)
If an editor published a speculative fiction story once, they are more likely to do it again when they do a new anthology. So even if we will have a lot of individuals, we will see repetitions. Magazines are different because there are multiple editors or they change... Annie 19:18, 5 April 2017 (UTC)
I don't think that's a major concern... but I realize I may be the only person who feels this way! --Vasha 19:34, 5 April 2017 (UTC)
I think you're right, Vasha. If the anthology is edited by authors with at least one genre title, we should credit them by their name, but many don't show any tendency to publish in our area. If that's the case we shouldn't credit them by name. Stonecreek 19:37, 5 April 2017 (UTC)
That seems to create more complications. Take The Best American Short Stories as an example. Some editors are genre and some aren't, including within the joint editors of the same pub. Plus some editors are only in the database otherwise because of an essay (some not even really genre, like Martha Foley) and not a story. Crediting a pub as "Stephen King and Editor of Best American Short Stories" seems weird to me. Plus the complications if they do then publish a genre story having to know to go back and change the prior credits. Magazines make sense to me because of the volume of records. -- JLaTondre (talk) 19:48, 5 April 2017 (UTC)
I know it looks weird, but IMHO it's still preferable to amassing surplus entries by assumed genre editors; and there's really a vast bulk of non-genre anthologies out there that contain one or a few genre titles: they may even easily outnumber magazine titles, we just didn't have time to call them in.
Those special editors with effectively no genre titles then should be left out: it's a moderator's judgment in each case. Stonecreek 20:35, 5 April 2017 (UTC)
Full disclosure: I'd rather we recorded actual editors whenever we know them, even for non-genre magazines. So this thought is probably biased, but... I think a difference between anthologies and magazines is that someone doing research is more likely to look for an anthology by editor (vs. title), yet would be more likely to look for a magazine by title (vs. editor). So there may be some benefit to recording editors for non-genre anthologies, even if the editor is not "genre". --MartyD 01:05, 6 April 2017 (UTC)
Yes, but to many users it certainly looks weird that we don't list all titles by all featured authors (because some of them fall below the threshold): it's one of the idiosyncrasies of our little web site. But if the consensus is to list the editors for non-genre anthologies this should be made part of the guidelines of rules. Stonecreek 04:31, 6 April 2017 (UTC)
Another thought: couldn't anthologies (and especially series like The Best American Short Stories) illustrate our politics quite easily if we'd choose to credit the regular (non-genre) editors as 'Editors of'(for example Editors of The Best American Short Stories)? Stonecreek 14:57, 6 April 2017 (UTC)
Now that we have the notes for the authors actually showing on the page itself and not in the wiki, we can use the space to explain there why not all works are included for an author especially for the ones that are major authors but just not SF major ones... Annie 04:35, 6 April 2017 (UTC)

Nongenre works (Part II): Cover art

As I understand it, nongenre books are not supposed to have a cover artist specified because the database shouldn't contain records for nongenre art. But I know there are some. They should be found and removed. --Vasha 18:05, 5 April 2017 (UTC)

Only non-genre magazines are explicitly restricted from having cover art or cover artist credit. Not sure why it was limited to those, but a wider discussion should happen before implementing this. -- JLaTondre (talk) 18:24, 5 April 2017 (UTC)
OK, I don't understand that either. Isn't the idea that there shouldn't be art records for any art not associated with a genre work? If so, the help files should be updated. --Vasha 18:58, 5 April 2017 (UTC)
Which help files? The only help files that mention it (that I'm aware of) is the non-genre magazine one. -- JLaTondre (talk) 19:03, 5 April 2017 (UTC)
Right; one thing that's been bugging me for the last few days is that there isn't much said in the help about nongenre anthologies. It is left vague whether they should be treated exactly like nongenre magazines. Do we need a R&S discussion about this? --Vasha 19:26, 5 April 2017 (UTC)
That is just it, they aren't treated the same. If you want to change the status quo, yes, a R&S discussion should be had. -- JLaTondre (talk) 19:32, 5 April 2017 (UTC)

Cover art borrowed from previous artistic works

I don't know whether this topic has been discussed before : it might very well have, but in this case, this is just to request some enlightening. In the case of a book cover illustrated with the pre-existing work (or a detail thereof) of an artist, would it be possible (or could it be made possible) to variant all the concerned book covers to the original title of the work (+ date, if known), illustrated with the original painting ? I was particularly thinking of Hieronymus Bosch, whose paintings have been the source of so many cropped details that could all be easily varianted to the complete original. But I think that applying this system to all other cases might also be a useful way of clarifying sources. Any opinions about the matter ? Thanks, Linguist 10:06, 24 April 2017 (UTC).

I'm against this move, lest we start the slippery slope to being a kind of gallery of art, listing not only a bunch of paintings by Bosch or Magritte, but logically facing the obligation of extending this to all paintings of Foss, Hardy or Siudmak (this will nearly double the number of such records). Our "art" cannot be separated from the book it illustrates and doesn't "exist" outside of it. Hauck 10:36, 24 April 2017 (UTC)
Well, it would be only logical to induct such attempts: in fact, it has been done in some cases where re-used covers only display half (or a detail) of the original work. It may be looked upon as a sad fact, but at least with allowing interior art (if not already by allowing cover art) we also have become a kind of art gallery, and I feel we should handle it carefully and in a responsible way. This would mean that we should variant art details to their respective parent. Stonecreek 18:47, 24 April 2017 (UTC)
Where would we put the parent? We aren't actually an art bibliography, so I would be opposed to having a separate record just for The Garden of Earthly Delights. (Or we would risk the precedent to just start including every piece of art by a fantasy artist.) So do we try to find the book that has the "largest" chunk of that painting and variant to it? Or do we variant to the oldest book that used that? Almost any solution would seem to run the risk of having cover A be a variant of cover B, when those two books have NO actual overlap in the portion of art they're using! IMHO, the inherent conflicts that would arise speak against attempting to do this. Chavey 05:42, 25 April 2017 (UTC)
In the case of Bosch I think there's no reason to not include art books on or by him, so there is no reason to avoid a variant to The Garden of Earthly Delights, because this work will eventually pop up anyway. I think there's also no reason to avoid varianting full reproductions of other art.
As I understand it there wouldn't or shouldn't be varianting to covers but to the original art title. (As we list the occurences of art pieces, I'm afraid that it seems we actually have become an art bibliography (it was a can of worms when opened, we just didn't realize it, I think), and I don't see an actual reason to avoid applying our quality standards. Stonecreek 07:41, 25 April 2017 (UTC)
IMHO it's not a problem of quality standards (even if they would be better enforced if the cleanup reports were exploited by all non-self moderators), just of logic. We're a database of books (in a general sense, of course) not a database of paintings. In the case of "extracts" of a larger painting, if there is no book in the db that is adorned with the "full" painting, there is just nothing to variant to, except if we start creating publess records for such an usage which is not a good idea. Hauck 08:08, 25 April 2017 (UTC)
As said above: caused by the works of several editors we have become a database for art as well, at least that's the way users will perceive us, if we give so much emphasis on art: listing artist's biographical data, detailing the reproductions of art works etc. Stonecreek 09:07, 25 April 2017 (UTC)

(after edit conflict)

Just to make things clear :

  1. I was mainly thinking of works that were not meant to be spec-fic illustrations, but were used that way — which limits the scope a bit.
  2. In my mind, I was just considering this as an option, not something systematic, that could be occasionally chosen when particularly useful.
  3. I don't see why, in the case of works that were neither book covers or inside art in the first place, we couldn't have the original piece as the parent, possibly with some indication like "(original artwork)" to make it clear it wasn't included in any publication in the beginning. I must say I don't understand the reason of Chavey's list of questions and remarks : we are already doing all that when dealing with such cases, and this is precisely to avoid these problems that I was considering the possibility of entering original works. What we are actually doing at the moment is "inventing" a parent which does not exist, because there is no original book cover in these cases.
  4. It is agreed that we are a bibliographical database, but doesn't this mean we ought to be as thorough as possible as far as the origin of the illustrations is concerned ? And having the possibility to view at a glance the original and all the derived book covers seems to me to go that way.

But this being said, it was just an idea that had crossed my mind… Linguist 09:36, 25 April 2017 (UTC).

Right on to the point, I'd say. As much of our work it will have to be grounded in the will of editors to pursue such a task for their chosen works, so a systematic approach seems not to be achievable. But still we should try, I think. Stonecreek 17:29, 25 April 2017 (UTC)

Precedence of ISBN vs. SBN

I found a copy of The Early Asimov: Book Two. At present the pub record uses catalog # P2323. Prior verifiers, both retired from ISFDB, either missed or did not document that the publication also has an SBN on the spine. There's no ISBN/SBN on the copyright page. I've added a pub note about the SBN. It brought up an issue that's not covered on Template:PublicationFields:ISBN. Is there a consensus what to enter the catalog # field when a publication has both a catalog # and an SBN? Do we enter the catalog #, SBN, or translate the SBN to an ISBN and use that? I've been in the habit of translating the SBN to an ISBN and using that in the Catalog # (along with a pub note). That time though the pub record is already there and PVed which made me wonder if others believe the catalog # gets entered. --Marc Kupper|talk 08:33, 13 May 2017 (UTC)

As per Help:
  • If a publication has more than one number, then enter the one that seems most reasonable in the ISBN field but then also add a comment to the notes field about the various numbers and where they are located in the publication. For example, a book may have both an ISBN and a catalog number. In this case the ISBN would go in the ISBN field and you would make a note, for example, that the ISBN is on the spine and catalog number is on the front cover.
Please note that, as per Roadmap 2017, the ISBN field will be revamped in the near future:
  • ISBNs (mostly to support internationalization, but will also help in other cases):
    • Move catalog IDs to a separate field, which will help with pubs that have both an ISBN and a catalog ID, e.g. book club publications.
    • Support multiple ISBNs per publication
    • Add two checkboxes next to each ISBN: “derived” and “corrected”. The latter will let us capture two versions of invalid ISBNs, the stated one and the corrected one.
Ahasuerus 13:28, 13 May 2017 (UTC)
Thank you. I suspect I'd read that help before which is why I put ISBNs before catalog #s. In this case it's an SBN and catalog # which is not addressed in the help. Once that roadmap 2017 item is done then we'll be able to have all three values (Catalog #, SBN, and derived ISBN) meaning searches for any of them will work. For now I'll stick with the catalog # for that publication record on the theory that two previous verifiers saw it but did not seem to notice the SBN. Someone else with a copy of the publication will likely also look for the catalog #. I've already added a pub note about that it has a catalog #, SBN, and provided the derived ISBN. BTW, in my book list, which is a spreadsheet, I support entering "SBN 449-02323-125" in the ISBN field. There's logic that looks for the SBN prefix and then derives the ISBN from a variety of SBN formats. --Marc Kupper|talk 08:58, 14 May 2017 (UTC)

Titling conventions for magazine titles grouped by year

I have a question about naming conventions for series grouped by year. I know the grouped name changes if the editor(s) change, as can readily be seen by the Astounding/Analog series, but what about the EDITOR record title changes? For example, all of the titles in the EDITOR record for "Astounding Stories of Super-Science - 1930" match the grouping title. But then, in "Astounding Stories - 1931" the first title in the EDITOR records doesn't match (it retains the title of the previous year) but all the others match. Should it not be in this grouping because of the title change, or is it ok to have differing titles in the EDITOR record? The same thing happens again in 1938. In 1960 the same thing happens for the change from "Astounding Science Fiction" to "Astounding Science Fact -> Fiction", but then, also 1960, it changes to "Analog Science Fact -> Fiction" and it gets a new grouping. Is that because it's a radical title change? So if it's okay to have different titles for minor changes in the EDITOR record for a group, how is the name chosen? By which title has the most entries? Thanks for any input on this. Doug / Vornoff 20:20, 8 June 2017 (UTC)

It's usually decided on the spot. IIRC the grouping is not mandatory (it's just a way to avoid too much cluttering of the pages) so there are as many ways to proceed (in such complicated cases like Astounding/Analog) as there are editors and moderators. I personally find that Astounding/Analog's page was split too much (e.g. at the 1960 juncture) and so nearly unreadable for someone not aware of the specificities of the magazines and its title changes. If pressed, I would say that the EDITOR record at publication level should be the more precise and I'm ready to accept this kind of grouping. Hauck 06:14, 9 June 2017 (UTC)
I agree. I think the only situation where a split is "mandatory" is if the actual editorship credit changes mid-year/volume. Having the grouped name be sort of canonical, with variations in wording of the individual EDITOR records, seems acceptable. I do think a major name change, such as "Astounding" -> "Analog" can be worth splitting into two groups to make it easier to see when the change occurred. --MartyD 12:07, 10 June 2017 (UTC)

Definition of "juvenile"

It seems like we are using juvenile for both children's books and for young adult ones - effectively making it mean "not adult". Is that the desired effect? And if so, maybe we should change the name of it so it does not sound as if we call those books not serious or even too childish to be of any considerations(as much as I am fond of the word juvenile, it is not exactly used anymore in the context it was in Heinlein's days for example). So what is the intent behind the category? Annie 23:05, 9 June 2017 (UTC)

I'm not sure about it. Nowadays many of them are also denominated 'all-age' literature, which contradicts to some degree 'juvenile'. Stonecreek 04:13, 10 June 2017 (UTC)
I've never used the "juvenile" label for the reasons given by Annie (it's IMHO quite meaningless), it's another can of worms that I don't want to open (tens of posts to arrive at the statu-quo ante). In fact, I find nearly all labels either useless (if it's "non-genre" it should not be in the db, if it's "graphic" it should not be in the db) or so arguable (What does "juvenile" mean? What constitues a "novelization"?) that we may be better without them. Hauck 06:32, 10 June 2017 (UTC)
Yeah, even Heinlein's original juveniles are read by and published as for adults; so, they aren't exactly juvenile anymore. Stonecreek 09:10, 10 June 2017 (UTC)
I would be happy to see it go, but I don't use it, so that's easy for me to say. It would be helpful to hear the thoughts of someone who does use it and finds it valuable. --MartyD 12:09, 10 June 2017 (UTC)
Actually, I do. For one thing, novellas and picture books both are in the database as "CHAPBOOK containing SHORTFICTION" and I do like having a way to tell them apart. As you probably know, I've made it my project this year to try to get correct information into the database about short fiction published in 2017; it'll come in handy next year when people are making up their awards lists, for one thing. It is harder to get info about children's books, which often don't have Amazon previews. The same goes for juvenile ANTHOLOGYs and COLLECTIONs as CHAPBOOKs. I decided not to concern myself with children's literature; people who want info about that probably have other sources, and if they are using this DB, well, I'm not the one helping them. So it is very useful to me to mark things as juvenile.
I find that cataloguing young adult books doesn't suffer from the same difficulties as children's. I think my dividing line for what is juvenile or not would be the stage at which they stop having special formats: middle grade books often are a different size, more often illustrated, may have larger type, etc. Whereas young adult books generally resemble adult books.
Just my two cents... --Vasha 17:52, 10 June 2017 (UTC)
That is not how juvenile is usually defined. Juvenile is typically defined as marketed for those under 18. These days, children's literature is a more widely used term that has replaced juvenile (as well as a tendency to divide such a broad category into more age targeted marketing terms, like young adult). I don't care if we keep the flag or not, but let's not create a site specific definition as that will just cause more confusion. -- JLaTondre (talk) 18:25, 10 June 2017 (UTC)
And in my non-English native ear (or it may be a combination of language and culture references), juvenile sounds like sub-standard. Annie 19:27, 10 June 2017 (UTC)
"Juvenile" can be used to mean "immature", "childish", etc, but that's not how it's used in publishing. Ahasuerus 19:37, 10 June 2017 (UTC)
Yeah, I know :) Did my digging when I encountered it here. Retraining my mind will take longer Annie 19:56, 10 June 2017 (UTC)
Which is what made me ask... Maybe we need a YA flag (for what Vasha is trying to do)... and a different one for pure children ones. Although I would not be too broken up to lose the flag altogether. I used it when adding some pure children books a few months ago - and it kinda made sense but had been thinking on it since. Annie 19:27, 10 June 2017 (UTC)
Sorry, but no, this is a slippery slope with potential multiple age categories to be defined and all the fun coming from debating to decide if a specific text is YA or juvenile or adult (who say Heinlein?). That's why I'm against all those unclear categories (or those nebulous definitions) that we're forced, because of history, to pull along. We're not marketers, we're just bibliographers (IIRC Dewey doesn't, for fiction, makes a difference between adult, young-adult, not-so adult, nearly-adult, not-at-all adult or really-not-adult texts). Hauck 19:46, 10 June 2017 (UTC)
I do not disagree with you - especially with publishers marketing books any way that will get them more sales... I was merely thinking aloud Annie 19:56, 10 June 2017 (UTC)

(unindent) I can think of a few reasons to keep the "Juvenile" flag:

  • It facilitates searching the database, which is very helpful when performing tasks similar to what Vasha has been working on.
  • It matches the data that other catalogs provide under "Material Type", e.g. see this OCLC record.
Anyone know where OCLC policy for defining this is? If there is one. We couldn't rely on them entirely, but at least using their classifications for new materials might speed things up. Vasha 20:17, 10 June 2017 (UTC) --Vasha 20:17, 10 June 2017 (UTC)
  • Deleting it would discourage the editors who have turned this flag on for approximately 21,000 of our titles.

That said, these days I find that tagging is a better match for what we are doing here. For this reason I usually do both: flip the "Juvenile" flag and assign a tag like "juvenile sf", "juvenile fantasy", "young-adult horror", etc. Ahasuerus 19:48, 10 June 2017 (UTC)

So maybe we need a better definition of the flag so it gets a bit more consistently applies instead of every editor coming up with their own rules? And I need to remember to tag... Annie 19:56, 10 June 2017 (UTC)
The problem with using the tag instead of the flag is if you search for "tag does not contain juvenile", it only gives you other tagged records, not untagged ones. Vasha 19:58, 10 June 2017 (UTC)
In addition to, not as a replacement :) If we keep the flag, we just need consistent usage (and help update for that) Annie 20:03, 10 June 2017 (UTC)
I certainly can't disagree that the term should be defined, and I'll go along with any definition. --Vasha 20:09, 10 June 2017 (UTC)
The real fun will start here (although without me). Hauck 20:11, 10 June 2017 (UTC)
Also, if you flag a chapbook, please flag the contents too, and vice versa. Vasha 19:58, 10 June 2017 (UTC)

(unindent) The Library of Congress generally has MARC codes that they view as specifying the level of the target audience. That becomes a natural way to define what "juvenile" means: If the Library of Congress says it's juvenile, then we say it's juvenile. The MARC codes of interest to us are: "a" pre-school; "b" primary (K-3); "c" elementary and junior high (grades 4 - 8); "j" juvenile (through age 15 or grade 9); "d" secondary (grades 9 - 12). IMHO, "a" shouldn't be included here; "b" might be included, or might be rejected; "c" and "j" would be juvenile (to us); "d" might be "YA" to us. To find the MARC codes, go to the LCCN record, at the very top are two tabs: "Full Record" and "MARC Tags". Look at the tags for the "008" line. For example, with A Coalition of Lions, this line is:

008   020930s2003 nyu j 000 1 eng

The only thing important to us here is the "j" in the middle. That says we should tag that book as "juvenile". With A New Alice in the Old Wonderland that line gives us:

008   140520s2009 ie a j b 000 1 eng

That tells us that it's juvenile, and furthermore the "a" and "b" surrounding the "j" tells us that its target audience is pre-school to grade 3. This would qualify as "Children's", although for now we have no formal way to indicate anything about this other than that it's "juvenile", which we should have tagged (we don't). The advantage of this approach is that we are relying on Library of Congress judgement, and avoiding getting into our own definitions or debates. The disadvantage is that if there isn't an LCCN record, or if that record is old enough to not have MARC codes, then we still need to decide ourselves. But at least we have the Library of Congress definition of target audience. Chavey 23:06, 10 June 2017 (UTC)

That sounds really good to me! And if their dividing line for j is approximately age 15, that helps convert to other systems of classification. --Vasha 23:44, 10 June 2017 (UTC)
A couple of things about the MARC21 standard or, technically, family of standards. Although the Library of Congress is one of its maintainers, it's an international standard and used in a number of countries. The "target audience" codes that Darrah mentioned earlier are as follows:
  • # - Unknown or not specified
  • a - Preschool: "Intended for children, approximate ages 0-5 years"
  • b - Primary: "Intended for children, approximate ages 6-8 years"
  • c - Pre-adolescent: "Intended for young people, approximate ages 9-13"
  • d - Adolescent: "Intended for young people, approximate ages 14-17"
  • e - Adult
  • f - Specialized
  • g - General
  • j - Juvenile: "Intended for use by children and young people, approximate ages 0-15. The code is used when a more specific code for the juvenile target audience is not desired."
  • | - No attempt to code
In theory, a, b, c, d and j match the "juvenile" flag in our system. In practice, however, things can be rather messy. Back when I was coding the MARC21 module of Fixer, I examined hundreds of MARC records from different publicly available data sets. I found that the use of this field was inconsistent. Some catalog records had very good data while certain other ones didn't. When set, this field can be useful, but I wouldn't treat it as gospel. Ahasuerus 00:43, 11 June 2017 (UTC)
I think the MARC codes provide us with a good definition, but as you say, the implementation can be weak. In cases where there are several LCCN records (generally, when the title has had multiple US publishers), there are multiple records to consider for this classification, which would generally lead to a more reliable decision. There was an effort on the part of OCLC to try to make a major analysis of the target audience level where they combined evidence from the MARC records together with the libraries that held that book: E.g., if elementary school libraries held the book, it was likely targeted at a children's audience (or at least a juvenile audience). Their report on this effort is available here. But unfortunately, I didn't find any concluding data from this effort. They discussed having a service where you could enter the OCLC number and have it report the target audience based on this criteria, but I don't see that they did this. Chavey 01:17, 11 June 2017 (UTC)

Using spelled-out initials to disambiguate abbreviated author names

In the case of S. Dorman (Sonya Dorman) and S. Dorman (I) (Susan C. Dorman), where the ambiguity is due to use of initials, what do we think about using spelled-out initial(s) as disambiguator? E.g., "S. Dorman (Sonya)" and "S. Dorman (Susan)"? --MartyD 12:07, 17 June 2017 (UTC)

It will be clearer than the current (I) way. So I like the idea :) Annie 17:12, 17 June 2017 (UTC)
I think that's an excellent idea. Chavey 04:13, 18 June 2017 (UTC)

The Capitalization Post!

Help:Screen:EditPub contains guidance on standardizing the capitalization of titles. The current guidelines are a good attempt at providing simple information for editors, not all of whom are native English speakers. But they could be better. The biggest problem is the attempt to avoid having rules for capitalization by simply listing words that should not be capitalized. This list is too short. The results it produces don't correspond to any standard style. Also, it neglects certain exceptions that are found in all standard styles.

The basic principles of all standard English capitalization styles are that articles, coordinating conjunctions, and prepositions are exceptions to capitalization. There are three competing approaches to how to treat prepositions: Capitalize all those of four letters or more; all those of five letters or more; or none at all. There is a comprehensive list of prepositions on Wikipedia; people who want a definition of what they are can consult A Student's Introduction to English Grammar -- warning: like all grammatical questions, it isn't simple.

At this point I am going to quote MartyD: "One suggestion: Try to find an authoritative, semi-public standard for us to follow. Having our own only makes it certain that no one will know what to do without 'training'." I wholeheartedly agree with Marty. For this reason, I've been making a survey of various editorial style guides: For the US, the APA, MLA, and Chicago styles; for the UK, the University of Oxford Style Guide. By far, the most comprehensive, clear, and detailed one is the Chicago Manual of Style; it also contains copious illustrative examples. Choosing it as an external standard for this database would be natural, if not for the problem that its latest edition is available only in libraries: in hardcopy or online by subscription. However, I did copy out relevant passages; I hope my excerpts may be brief enough to pass for fair use. I also copied corresponding passages from the MLA handbook, which is the second-most-detailed guide. You will see that the two guides largely agree.

[These excerpts also contain considerable guidance on punctuation and such matters, which I think it would likewise be good to submit to an external guide; but let's stick to capitalization in the current discussion.]

Here follows a comparison of the current ISFDB guidance with summarized Chicago and MLA suggestions; with problems and options.

A. From Help:Screen:EditPub. Regularized case means that the first word is capitalized, and all later words are also capitalized except for "and", "or", "the", "a", "an", "for", "of", "in", "on", "by", "at", "from", "with", and "to".

Chicago: First and last words should be capitalized; all other words should be capitalized except for:
  • articles (a, an, the);
  • certain conjunctions (and, but, for, or, and nor);
  • all prepositions (see Wikipedia list) except when they are not used as a preposition but rather as part of a verb (Turn Up, Help Out), as an adjective (The On Button), or as a subordinating conjunction ("Equal before the Law" BUT "We'll Hide Before He Returns");
  • to when it is part of an infinitive; as in any grammatical function.
MLA: First and last words should be capitalized; all other words should be capitalized except for:
  • articles (as above);
  • certain conjunctions (and, but, for, or, nor, so, yet);
  • all prepositions except when they are not used as a preposition but rather as part of a verb, an adjective, or a subordinating conjunction (see above);
  • to when it is part of an infinitive
PROBLEM: Both Chicago and MLA expect people to be able to distinguish the grammatical function of the words on the list of prepositions and capitalize them or not accordingly. Even for me, a highly literate native English speaker, this is occasionally frustrating. For some users, it will be completely baffling. I see three possible solutions (anyone have others?)
1. NEVER capitalize the words on the preposition list unless they are the last word. However, this would depart from standard practice.
2. Adopt the style guide policy and have a capitalization help desk for people to ask about titles they aren't sure what to do with (I, for one, would volunteer to monitor it)
3. Adopt the style guide policy and don't worry about standardization. If people get it wrong, so be it.

B. From Help:Screen:EditPub. Hyphenated words have the first letter after the hyphen capitalized.

Chicago: Capitalize the first letter after the hyphen except if it is a word that would be lowercase anyway (article, conjunction, preposition), or if the word before the hyphen is a prefix that isn't a word on its own (Re-entry, Anti-establishment).
MLA: Capitalize the first letter after the hyphen except if is a word that would be lowercase anyway. [MLA does not mention anything about hyphenated prefixes.]

Have I gone on long enough yet? :-) --Vasha 19:45, 26 June 2017 (EDT)

ADDENDUM: A couple of more-extreme solutions to the capitalization problen:

1. Capitalize every single word. This is legible enough, but amateurish-looking.
2. Follow the style of Worldcat and most academic bibliographies: Sentence-case.

--Vasha 20:40, 26 June 2017 (EDT)

Re: "except when they are not used as a preposition but rather as part of a verb (Turn Up, Help Out)", I think our current de facto standard is to capitalize postpositions, i.e. prepositions which appear after the verbs that they are related to. Ahasuerus 23:09, 26 June 2017 (EDT)
Actually, those are what Chicago/MLA call "prepositions used adverbially," or some people call verb particles. Postpositions, on the other hand, are simply prepositions that occur after their noun: two years ago, two miles away. Be that as it may, both the verb-following ones and the postpositions are standardly capitalized. And editors who've internalized standard practices have been doing it even though the Help doesn't say to. --Vasha 05:39, 27 June 2017 (EDT)
It sounds like all we have to do in this particular case is change Help from:
  • except for "and", "or", "the", "a", "an", "for", "of", "in", "on", "by", "at", "from", "with", and "to".
to something like:
except for "and", "or", "the", "a" and "an". "For", "of", "in", "on", "by", "at", "from", "with", and "to" should not be capitalized unless they follow the verb that they modify, e.g. "Get Up".
Ahasuerus 11:44, 27 June 2017 (EDT)
The easiest way to tell if the preposition is by itself as part of one of these verbs, and not part of a preposition phrase, is to see if it can be moved to the end: "Bring in the newspaper" Yes: "Bring the newspaper in". "Swim in the pool" No: "*Swim the pool in." (asterisk being a linguist's way of marking a wrong sentence). I think it would actually be easier to specify that prepositions be capitalized everywhere EXCEPT if they form a phrase with a following noun. The cases where they ARE capitalized are so numerous: "The On Button", inverted constructions ("Where Did We Come From?") and so on and so on. So, only phrases (which are usually used to indicate things like relationships of location, time, agent, etc.): "sit on the chair," "a calm after a storm," "written by me," "a gift for him," etc. The only exceptions then are to in infinitives ("The Way to Go") and (maybe) as in comparisons ("As Hot as You Like", if we decide to use that rule).
The current list is too short, though. At the very least it should include the commonest prepositions, which is the current list plus as; wouldn't it be better to simply not capitalize all prepositions of four letters or less (a very common practice) or all prepositions whatsoever (the practice recommended by Chicago and MLA)? --Vasha 12:54, 27 June 2017 (EDT)
Not so easy to see if you can move the preposition somewhere if English is not your native language though (especially if in your language prepositions work a bit differently). We need a rule that does not require a degree in English to be understood. It is great to follow a standard guide and all that but if the rule becomes too complicated, we will end up with people ignoring it and moderators needing to correct a lot of entries. :) Annie 13:10, 27 June 2017 (EDT)
If your proposal is to "simply not capitalize all prepositions of four letters or less", I am not sure why you posted all the long winded explanation above :) So what IS the proposal - follow the manuals or go for anything under 4 letters? Annie 13:10, 27 June 2017 (EDT)
I agree with you Annie, standard English capitalization rules are not very good because they DO require strong knowledge of English to follow correctly. That's what I said in the section marked PROBLEM above, and that's why (not very seriously) I proposed the two extreme alternatives in the addendum.
The rule would be "Do not capitalize any preposition of four letters or less ONLY WHEN it forms a phrase with a noun or pronoun following it; do not capitalize to when it is used in an infinitive." That is short to say, but not necessarily easy to put into practice. In the examples I gave above, would it have been clear to you that the phrase structure was "[bring in] the newspaper" and "swim [in the pool]", so that only the second one would have in not capitalized? --Vasha 13:31, 27 June 2017 (EDT)
English grammar used to be a hobby of mine for awhile so recognizing a phrasal verb is not a problem. And I've done some writing under the Chicago rules - overeager teacher and all that. So I can follow just fine :) However, trying to explain the intricacies of the phrasal verbs to someone without a firm grasp of the grammar is a tiny bit hard (the good old example of "go on" vs "go to") :)
I know we need an update of the current rule but having a definitive list makes it easy to understand and less prone to each editor having their own interpretation. I just wish that English was like Bulgarian or Russian in that regard (sentence case in the titles in both):) Annie 14:17, 27 June 2017 (EDT)
The list of 2/3/4-letter prepositions, omitting some particularly rare ones, is: amid, as, at, but, by, ca. (abbreviation of circa), down, for, fore (or 'fore), from, in, into, less, like, mid (or 'mid), near, of, off, on, onto, out, over, past, per, post, re, save, than, 'til, till, to, up, upon, via, vs., with. That's 36. I bet I can find an example in the database using almost any one of them. --Vasha 14:39, 27 June 2017 (EDT)
Probably you can. And 36 is fine I think :) It is a definitive list. Add "a", "an", "the", "or", "nor" and "and" and we have a list. Although "save", "past" and "post" are used more often in their non-preposition roles and having them on the list is going to confuse people. I would leave those to get capitalized in all cases if I was making the rule. Annie 14:56, 27 June 2017 (EDT)
Well, past is a very common preposition and should be included. It is not hard to distinguish between save used as a preposition (All Is Lost save Honor) and as a verb. But I would omit fore/'fore, less, mid/'mid, pre, and post as rare and confusing. Leaving us with 31. And it really would help to divide the list into sections: articles, conjunctions (and, but, for, or, nor) and prepositions. --Vasha 15:08, 27 June 2017 (EDT)
I do not have issues with splitting into categories but you added "but" in your initial list of 36 so sounded like you do not want to split them. :) Agree on "less" - not sure why I did not add to the 3 above. But I still disagree on "past" and "save". They can be normal prepositions but their other grammatical roles are a lot more prevalent adn quite honestly, leaving them capitalized won't hurt anything (and will eliminate the confusion). We are looking for a rule that is easy to explain, right? Annie 15:22, 27 June 2017 (EDT)

But is a preposition as well as a conjunction -- "I worship no god but Yog-Sothoth!" I agree that save could easily be omitted; but I don't see what is controversial about "The unicorn ran past the hunters." How could that be confused with "Unicorns were more common in the past"? —The preceding unsigned comment was added by Vasha (talkcontribs) .

(unindent) Here's proposed language for the help, if we decide not to go with an external source:

Regularized case means that the first word is capitalized, and all later words are capitalized except for the following: articles (a, an, the); coordinating conjunctions (and, but, for, nor, or); the to of infinitives; prepositions of four letters or less IF they form a prepositional phrase with a following noun or pronoun (amid, as, at, but, by, ca., down, for, from, in, into, like, near, of, off, on, onto, out, over, past, re, than, 'til, till, to, up, upon, via, vs., with) prepositions of two or three letters (as, at, but, by, ca., for, in, of, off, on, out, re, 'til, to, up, via, vs.) plus from and with IF they form a prepositional phrase with a following noun or pronoun [Amended 06-28]. The same rules apply to the word that follows a hyphen.

I just don't see how to be simpler than that and still be correct and complete. --Vasha 15:46, 27 June 2017 (EDT)

I guess I would like an up-or-down vote on changing to sentence case. It would be unusual, and surprising to visitors, but SO much easier to do. --Vasha 17:41, 27 June 2017 (EDT)

"Ca.", "re", "via" (when used as a preposition) and "vs." seem like natural additions to the list. The rest feel kind of awkward. Ahasuerus 18:06, 27 June 2017 (EDT)
I don't know about you, but I think it is far easier to remember "all preposit'ons of four letters or less" than a specific list of 15 or 19 allowed prepositions. I would tear my hair out if I had to be always checking "is 'into' on the allowed list?" or having a moderator say "I can't approve that submission because 'into' isn't on the allowed list." --Vasha 18:20, 27 June 2017 (EDT)
Hm... Come to think of it, if we were to add the four words that I listed above, then our exclusion list would be effectively "all prepositions of three letters or less plus 'from' and 'with'", right? That should be fairly easy to memorize. Ahasuerus 18:30, 27 June 2017 (EDT)
Yes it would be, and semi-justifiable since "from" and "with" are vastly more common than other four-letter prepositions. Any sort of rule is better than memorizing a list. --Vasha 18:39, 27 June 2017 (EDT)
I like the "all prepositions of three letters or less plus 'from' and 'with'". I think we should still list them just so people do not need to check what a preposition is at 2 am in the morning but for memorization, this will work beautifully. Annie 19:03, 27 June 2017 (EDT)
Are we agreed then, to adopt the above amended description of English capitalization in the Help? --Vasha 20:03, 28 June 2017 (EDT)
Such a big change will require a lot more editors to at least chime in. And not everyone is around every day or has the time or the will to read the long explanations. Annie 22:03, 28 June 2017 (EDT)
True. Let's wait until after the weekend. Is there anyone in particular who should be pinged to ask for an opinion?
Also, I thought of another addition. We should probably add "the as of comparisons" so as not to produce titles like "An Ogre As Big as a Mountain" -- that would look better with both as's lowercase. --Vasha 22:18, 28 June 2017 (EDT)
I'd say take the weekend to get the proposal cleaned up with all the changed. Because it is English, we are looking at about 1.6 mln records influenced between titles and publications. Then post it - short and to the point and give it a week or two for people to comment. Dunno. It is the curse of the language that everyone edits in. Up to you - I am just posting an opinion. Annie 23:07, 28 June 2017 (EDT)
We should keep the rule simple. I don't care that much what we capitalize or not, but it should be a simple consistent list. It should not be a complex thing that makes it difficult on causal users (and users for who English is not a first language). Also, maybe I missed it, but I don't see anything above about how punctuation effects case. Our current help doesn't address that either, but it should (i.e. "Title: A Subtitle" should remain as such). -- JLaTondre (talk) 08:56, 2 July 2017 (EDT)
I wish with all my heart that the rule could be kept simple... but unfortunately English "title case" is not simple (it was devised by printers who didn't care whether mere laypeople could use it), and it absolutely cannot be described in a single line... unless that line is "Words belonging to certain parts of speech are lowercase at times, depending on where they fit into the structure of the title." It's a case of "Brief, precise, or correct-- pick two, because you can't have all three." My proposal 1A below is my very best attempt at being precise while also fairly brief and nearly completely correct. --Vasha 10:55, 2 July 2017 (EDT)
There is no such thing as 'correct' title case. There are multiple, conflicting standards. -- JLaTondre (talk) 11:47, 2 July 2017 (EDT)
Absolutely... and NONE of them can be summarized in one line. The suggested new language follows the standard from the Chicago Manual of Style because the CMoS is widely used and well-explained (in several pages). --Vasha 11:54, 2 July 2017 (EDT)
Our current one can be. Adding complexity for the sake of compliance to something that is not truly a standard is pointless. -- JLaTondre (talk) 12:01, 2 July 2017 (EDT)
Now *There* is a sentence I can completely agree with! ^^^ Chavey 02:49, 3 July 2017 (EDT)


As we know (to the frustration of many of us) the rules of standard "title case" capitalization in English are quite complicated. And as many of us have noticed, although the paragraph in Help:Screen:EditPub dealing with "regularized case" is a brave attempt at describing the principles of title case in a short and simple manner, its advice isn't actually sufficient to produce correct results. It states:

Regularized case means that the first word is capitalized, and all later words are also capitalized except for "and", "or", "the", "a", "an", "for", "of", "in", "on", "by", "at", "from", "with", and "to". Hyphenated words have the first letter after the hyphen capitalized.

Currently, users achieve nearly-correct results by filling in the gaps in these instructions from their own knowledge and ignoring them when they conflict with their knowledge of standard practices. Therefore, there'a a lot of variation in the database. Consider the following existing titles: Where to, Please?; Letter (Planet Stories, Fall 1940): They Always Seem to, Don't They?; Who You Talking To, Zone?; I Just Have To, Baby. The first two accord with the instructions given by Help; the last two are actually standardly correct.

Therefore, I'd like to propose modifying Help. There are two main options for doing so, I believe:

1A. Describe title case capitalization more completely. This has been the subject of the discussion above; the result was wording which essentially compresses the stylistic recommendations of The Chicago Manual of Style into a single paragraph, with the important difference that instead of leaving all prepositions in lowercase as per the CMoS, it mandates that only a subset of them should be lowercase (nearly the same subset as those specified by the existing wording):
Regularized case means that the first word is capitalized, and all later words are capitalized except for the following: articles (a, an, the); coordinating conjunctions (and, but, for, nor, or); the to of infinitives; the as of comparisons; prepositions of two or three letters (as, at, but, by, ca., for, in, of, off, on, out, re, 'til, to, up, via, vs.) plus from and with ONLY IF they form a prepositional phrase with a following noun or pronoun. The same rules apply to the word that follows a hyphen.
1B. On the help page, simply say that titles should be in title case, and link to a page that explains title case in detail, with examples.
2. Switch to using sentence case capitalization for titles. This is what Worldcat and most academic bibliographies do (and what nearly all languages other than English do). It would be a radical change, and initially surprising to casual users of the site; but it would be an easier standard of regularization for editors to use.

What are your thoughts on which proposal, if either, to adopt? --Vasha 22:48, 1 July 2017 (EDT)

Neither. Remain with the status quo. -- JLaTondre (talk) 11:47, 2 July 2017 (EDT)
I see no advantage to any of these three proposals over the current rule. I vote for the status quo. Chavey 02:52, 3 July 2017 (EDT)
I suspect that one of the issues here is the scope of the proposal. It covers a number of different areas and would necessitate reviewing and potentially changing hundreds of thousands of records. It's a daunting proposition.
I think it may be safer and easier to handle things piecemeal. For example, the proposal would add "ca.", "vs." and "via" to the list of prepositions that should not be capitalized. Checking the database, I see that we have 327 titles with " ca. " and only one title with " Ca. ". Similarly, we have 461 titles which use " vs. ", 48 titles which use " Vs. " and 1 title which uses " VS. ". If we were to change the data entry standard not to capitalize them, it would be in line with the current usage and require minimal changes to the existing data. On the other hand, " Via " and " via " are more evenly divided: 59 vs. 41 with some of the former having a different meaning as in "La casa in Via del Cimitero". Ahasuerus 11:56, 3 July 2017 (EDT)
Actually, I think that the fact that title case does NOT have those words lowercase in all circumstances is a bigger problem than exactly which words are on the list (as needs to be added too BTW). The list-based approach only gives a very weak approximation of title case. The reason that the data currently looks as good as it does is that people who know the full rules have been going beyond the list-based approach when they work. --Vasha 12:36, 3 July 2017 (EDT)
For reference purposes, we currently have 2,118 titles with " as " and 553 titles with " as ". Ahasuerus 13:31, 3 July 2017 (EDT)
The advice from one web site on choosing case was: "So, should you use sentence case or title case? If your school, college, or business has a house style guide, that decision has been made for you. If not, simply pick one or the other (flip a coin if you have to), and then try to be consistent.". That would mean sticking with title case. Given the number of forms for title case, I see nothing wrong with being more explicit in the Help. What I wonder is: if we achieve "nearly-correct results", how big is the problem really? If the rules are clear (albeit nearly impossible to follow if they depend on precise parsing of the title) then let people get it nearly correct (as they do pretty much everything they enter) and allow a few 'experts' clean up periodically. Doug H 14:41, 3 July 2017 (EDT)
You would get somewhat closer to right (though still not covering all cases) if you said that the words on the list are only capitalized when they are the first or last word or before a punctuation mark.
I diffidently suggest a variation on Doug's proposal: What about saying, "Titles should be in title case. Approximately speaking, this means that all words in the title should be capitalized, except that the words on the following list should only be capitalized when they occur first, last, or before a punctuation mark. For a fuller description of title case, see [page]." And allow people to use either the approximate or the expert version. In case of disagreement, the expert version takes precedence. Not good to have two conflicting standards, but I think this could be worded to make it clear that one is the real standard, the other is an acceptable way of getting close to it if you don't understand it... --Vasha
And what is "right"? That is the whole problem here, isn't it? There isn't one "right" thing. Annie 17:09, 3 July 2017 (EDT)
That is a good reason to cite an external authority such as the CMoS (which particular authority? That is essentially an arbitrary choice.) --Vasha 17:39, 3 July 2017 (EDT)
And how is that better than our home-grown system besides the "it is not standard" and "I do not like it"? The standards were never created with non-native speakers in mind - I would make a case that they are there partially to help with the grammar and mainly to cut on weird clauses and make things uniform in publications. Which our standard already does. There is a reason why there are multiple standards and a lot of in-house ones - there is no real right and wrong. Annie 22:42, 3 July 2017 (EDT)
I've been thinking on that whole thing in the last days. I will voice my problem with this again: "explain "prepositional phrase" in 2 sentences that do not rely on "but it is obvious that "go to the hospital" and "go on dancing" are different"" - if you can do that, every ESL teacher in the world will love you. That will end up with the moderators needing to explain grammar to contributors (or contributors just ignoring the whole thing because they cannot even understand when to capitalize). We can add a few more words to the list but having a list is easy to understand and to follow. Plus a list allows computers (and bots) to capitalize properly as well - allowing better consistency. I believe that consistent usage is more important than being correct according to one standard or another... Annie 17:09, 3 July 2017 (EDT)
Computers would be able to follow the first/last/punctuation rule too. Are you opposed to that amendment? --Vasha 17:39, 3 July 2017 (EDT)
Punctuation is tricky in titles (because it does not follow the standard rules). Commas (or their lack) should not really have any bearing on the capitalization. Annie 17:40, 3 July 2017 (EDT)
The punctuation part of that sentence is meant to deal with examples like "Where To, Please?" and "I Just Have To, Baby", which are capitalized the same as "Where To?" and "I Just Have To", and with all of the titles in this search; and to avoid having to explain that "last" also means the last word of the main title before the colon that marks the subtitle. --Vasha 22:11, 3 July 2017 (EDT)
You know that. Someone reading the sentence does not. It is again complicating a rule just because some authorities say so. Keeping it simple makes it easy to understand and implement. Annie 22:42, 3 July 2017 (EDT)
All but two of the titles in that search already do follow the "capitalize before punctuation" rule. Do you want to change thousands of existing titles just to leave five words out of a definition? It is really not a long or difficult rule. --Vasha 22:54, 3 July 2017 (EDT)
Plus allowing two different versions of the rules based on editor's choice is probably the worst we can do - it will lead to editors wars - I want it exact, the other PV on the title wants on the relaxed rules and both of us are adamant that if we PV, the titles follow the rules - now what? Saying that the expert one takes precedence brands the other as what? The lazy one? The one for the people that do not care? Let's not open that door. Or a contributor that prefer the relaxed way but makes a mistake and a moderator converts all to the complete because there were 2 fixes needed to get there and that is what they prefer. And we will end up with stories in the same anthology with different capitalization because they were entered at different times by different times and just imported... So yes, I am against any change that will officially allow two different standards. We either go with the punctuation/first/last rule at all times or never at all. Annie 17:40, 3 July 2017 (EDT)
I feel an instinctive aversion to adopting a rule that would say that the capitalization of "Go On Dancing" is wrong... Maybe I am just too hung up on following professional editing standards because that's my profession :-) If having the titles in this database capitalized differently than the way other people do it is the price of avoiding edit wars and wrangling, then it's a price worth paying. So yes, I think the list plus first/last/punctuation would be close enough. --Vasha 17:53, 3 July 2017 (EDT)
And if you need to edit for a paper/school that had adopted a different standard compared to what you are used to, would you turn down the job and/or try to tell them that their standard is not good? :) If anything, the current status quo (with a very small list of what we do not capitalize) is easy to implement and use. Adding a few more to the list to cut the confusion is workable. Something bigger? Changing to one of the standards won't improve the data here, will not help with simplifications - it will just confuse people and cause a lot of work that will need to be done to bring the current DB to that standard. Can you give one argument about how changing the rules will make the data better? Search is case insensitive for Latin-1, the titles now are consistent (kinda), the reading is not impeded from the current system. And it is very easy to explain and implement it.
And... all those standards are US-based. UK has their own. So do Australia or any other English speaking country. Why use a US-based standard as a base? :) Ours do not care about the country; it only cares that the title is in English. Annie 22:42, 3 July 2017 (EDT)

Prices and weird currencies - reading verification needed...

Our current help is very specific: "Enter a single price, preceded with a currency symbol... For books priced in other currencies, use the appropriate symbol or the accepted alphabetical alternate.". In my reading it means that a price field can never start with a number. Had that changed for non-standard currencies? I had seen some of the Bulgarian and all of the Croatian books (I let a few in as well because all the previous ones were like that...) go with numbers first (15 leva, 10 din and so on). There are probably more than these. Before I go and fix that, I want to make sure I am not overreacting. Thanks! Annie 18:23, 30 June 2017 (EDT)

I wonder if it might be good to create a drop-down with currencies next to the field for the actual price? It would require some cleanup beforehand, perhaps (a lot of it), but it might be good in the long run as the database becomes more international. ···日本穣 · 投稿 · Talk to Nihonjoe 18:29, 30 June 2017 (EDT)
That actually is a great idea. We will need a "Other/Not listed" that will require a note in the Notes but having a drop down may be very very useful. Allow a user to have a default (as with the language) and that should eliminate a lot of funny errors. My only worry will be that this list can get very long when we add international books... so we will need some filtering on what an editor see in a specific book - if the list has 200 entries, it is worse than not having a list):) Annie 18:40, 30 June 2017 (EDT)
Maybe make it default to the 5-10 most common currencies in the database, and allow the editors to got to a settings page where they can check boxes to add (or remove) additional options to the list if they wish. That should cover 90%-ish (or more) of those entering items. ···日本穣 · 投稿 · Talk to Nihonjoe 02:56, 1 July 2017 (EDT)
Additionally, this might add the opportunity to add multiple currencies to publications (so those can be moved out of the notes, too). ···日本穣 · 投稿 · Talk to Nihonjoe 02:57, 1 July 2017 (EDT)
The other day I was thinking about the changes that will need to be made to add support for multiple prices as per Roadmap 2017. Standardizing currencies was on my radar screen. However, we have almost 400,000 pubs with prices, so we'll have to do quite a bit of analysis first. For example, we have a few thousand "old UK" prices which look like "4/6". The good news is that the number of purely numeric prices is low, a bit over 100, and most of them are "0.00". Ahasuerus 11:32, 1 July 2017 (EDT)
And they are probably US dollars. Can you run a query on how many we have that start with a number - back to my original question? :) And am I misreading or do they need fixing? 14:06, 1 July 2017 (EDT)
Spot checking the 100+ "purely numeric" prices suggests that all of them are likely in error. The superset of prices which "start with a number" are a different story, though. There are 12,445 of them and they belong to different subcategories. 7,584 are "old UK prices" like "2/6" and "9d". The rest mostly have the currency abbreviation appended rather than prepended: "24.00 FF", "8,000 Lire", etc. We also have a number of oddball cases like "1 ruble 60 kopec" and "15$00".
We already have a cleanup report that checks for invalid prices, but it's currently limited to a few common errors like "$CDN" instead of "$C". We should probably expand it to look for other oddities as well. Ahasuerus 14:47, 1 July 2017 (EDT)
I've seen something in that report probably 3 times so I did not even remember we have it. And I forgot about the old shilling ones not having a symbol at the start (I think I am blanking their existence from my mind at all...). I can fix all the Bulgarian ones before I start adding the rest to be "Lev XX" and consistent and then we will work on the other languages. And the oddballs will need cleanup (R 1.60 looks so much better than 1 ruble 60 kopec :). Annie 17:05, 1 July 2017 (EDT)
It looks nicer, but I am not sure how many users will be able to tell what the "R" in "R 1.60" stands for. Spot checking the database, I see "100 Pts", "10.50 FB", "299 Kč", "48.00 元", "149 UAH", "9.90 RON" and any number of other abbreviations and symbols. Ideally, we will want to have mouse-over Help bubbles available for all currencies just like we do it for transliterated names. We'll need to make all currency symbols table-driven first, though. Ahasuerus 17:28, 1 July 2017 (EDT)
Oh, I did not mean just R 4.50. It falls under the less-known currencies so it will get also a "Price is in Russian Rubles" note in the Notes. :) I was just talking about the field itself. On the other hand anyone that cannot decipher R in a Russian language book price will be even more confused by "kopec" :) Annie 17:49, 1 July 2017 (EDT)
Well, a Google search on "kopec" finds this Wikipedia disambiguation page in two clicks. OTOH, a Google search on "R currency" suggests that "The currency code for Rand is ZAR, and the currency symbol is R."
The more I am thinking about it, the more I like the idea of making currency symbols table-driven... Ahasuerus 18:03, 1 July 2017 (EDT)
But if you are looking at a Russian book and you are interested in the price, the R will be obvious. If you need to google it, chances are that you need to google rubles as well anyway. I do see your point but I would prefer consistency when possible... And having a table will be awesome - but in the meantime the notes in the Pub notes will need to do. Can we get a template "Price is in "currency""? So we can find those and clean them up when we switch to another system? Not too many editors are adding a lot of non-major for the DB languages books as it is and getting them to use a template should not be that hard. Annie 18:09, 1 July 2017 (EDT)
IIRC, we already had a heated discussion on the subject years ago. My suggestion was quite simple: there is an ISO table of currencies (I used it when I was in the banking field, you can skip the bit about the position) that we can use instead of creating another custom one. The major drawback is that a "$6.99" price would in this case be transformed to "USD 6.99" which seemed at the time to annoy some anglocentric major contributors. Hauck 01:55, 2 July 2017 (EDT)
The ISO table will also solve the whole issue of everyone picking up the name for their currency any way they want and needing to put a note explaining what that might be. And if people's biggest issue is USD 6.99, it won't be that hard to code a few currencies in a special way I suspect (the ones that are now character based maybe?) and leave the exotics to follow the ISO. Actually the more I think about that, the more I like the idea - fully ISO or ISO for all except a handful of currencies. And it will be a lot easier for international editors. Annie 03:25, 2 July 2017 (EDT)
That's such "solutions" that I profoundly dislike. Why should we be making an exception for "a handful of currencies"? It's ISO or not, it's as simple as that. As a moderator, you'll find that it's exactly such idiosyncrasies (rules that are "generally" followed, but not always -who say regilarization of suffixes?-) that are the less easily understood by new contributors. Hauck 04:29, 2 July 2017 (EDT)
Most new contributors work in one language (or a handful anyway). My point is that if that is needed to convince the anglo-centric world to use the ISO table, then so be it. It gets us one step closer to a generic ISO based rule - and will be a vast improvement over where we are now. Small steps sometimes work where the big ones don't - and then the gap is closed a lot more easily :) I would rather leave the USD to be an exception and get all the rest fixed than have everything in a mess because we cannot convince everyone to adopt USD. Annie 04:46, 2 July 2017 (EDT)
The issue of prices can get complicated, e.g. see this 2011 discussion and this 2015 discussion. Creating a drop-down list of currency codes and/or symbols (a big can of worms in itself) may help address some of these issue, but further discussions will be needed. Ahasuerus 10:57, 2 July 2017 (EDT)
As usual it will go nowhere (space vs. no space, comma vs. period, before vs. after, USD vs. $, and I spare you the "space if it's DM, no space if it's €" juicier bits). Either we propose something and we vote or we wait for another two years to have another soul-searching round of discussion. As we're an english speaking outfit, the ISO norm is quite clear (even if didn't please Michael), and boils down to "XXX<space>888.88" where XXX is the ISO code.Hauck 11:14, 2 July 2017 (EDT)
Adding a drop-down list would address many of these issues if we were to add a table defining the formatting of each currency: spacing, "before vs. after", etc. It would eliminate the need for complex data entry rules. Or at least most of them anyway. Ahasuerus 11:39, 2 July 2017 (EDT)
IMHO it's here that we err. The entry standard (before/after, space/no space) should be the same for all currencies and set once for all. We're a bibliographical outfit and strict adherence to national typographical norms of price writing is not our problem. Hauck 11:50, 2 July 2017 (EDT)
Well, once we make currencies software- and table-driven, it will be easy to change the display component any way we want, including standardizing the formatting if that's what the majority prefers at some point.
My main concern is the data entry side: with 400,000 prices in the price field and another (estimated) 100,000+ prices in the Note field, manual changes become prohibitively labor-intensive.
My current thinking is that the project that will add support for multiple prices per publication will need to add support for a drop-down list of currencies at the same time. If we do it as a two-step process, we will have to re-do 100,000 prices, which will be unnecessarily painful. Ahasuerus 12:03, 2 July 2017 (EDT)

(unindent) (because my phone started making the letters above very small so it can show me more than 2 letters per line) Not all 400 000 prices will need manual touching - most of them are standard and will just get a conversion. We are looking ta the 12K or so that are non-standard. And even in them there are some patterns (the Croatian "XX din.", the Bulgarian "xx leva", most of the shilling prices and so on. The Notes ones will need a manual moving no matter what we do (some may be better formatted than others but from I had seen, it is editor preferred note) but as long as the first move is done to a standard, the second, if you go for 2 steps, will be automatic. I am with Herve though(albeit in a bit softer position - I will allow for a few special cases) - we do need to have the options for every single currency - go ISO (currency at the start) and we are all set.

And if we do not change it now, we will have even more entries in 2 years. Problems in consistency of data do not disappear, they just get worse as we all know. Annie 18:43, 2 July 2017 (EDT)

I am sorry, my formatting may not have been clear. When I wrote "manual changes become prohibitively labor-intensive", I didn't mean to imply that we didn't want to make currencies table-driven. It was an argument in favor of a single-step conversion rather than a two-step conversion. I should have kept it as a single paragraph, but the extreme indentation threw me off.
As far as the implementation details go (ISO 4217, currency symbols vs. currency codes, etc), there will be various issues that we will need to discuss before we finalize the design. For example, ISO 4217 only covers current and "recent historical" currencies while we need to support older currencies as well: the ducat, the livre, the real, etc. Some editors prefer currency symbols like £, $, €, and ¥ to currency codes, but it may be less of an issue if we add mouse-over help. Ahasuerus 21:09, 2 July 2017 (EDT)
So let's start discussing? :) I understand that there are some complications because of older currencies but then we implement ISO+ (ISO for where it exists; custom table otherwise). And I am not sure that $3.12 is any better than USD 3.12 but I won't argue that one... Annie 21:13, 2 July 2017 (EDT)
I find that it usually works best when we start discussing the details of Change B while Change A is being coded. If we have too many balls in the air at the same time, I start running out of bandwidth, especially when there are Fixer-related issues that require my attention. I hope to wrap up the current Amazon emergency (one way or another) in the next few days, then I will need to implement the security fixes that I have been working on, then we can separate catalog IDs from ISBNs, then we can add support for multiple ISBNs and then we can handle prices :-) Ahasuerus 22:10, 2 July 2017 (EDT)
Well - this started with me asking "do I misread the rules or should we always start with the currency on non-standard currencies?" and sounds like I do not misread the thing so I will proceed accordingly for now. :) Annie 22:59, 2 July 2017 (EDT)

(unindent) Why do I keep finding things to fix instead of working on adding my books? :) Annie 17:05, 1 July 2017 (EDT)

It's part of the standard progression:
  • Read SF
  • Collect SF
  • Catalog your SF collection
  • Catalog SF in general
  • Help design software to catalog SF
  • Write software to catalog SF
  • Create and maintains robots which will catalog SF on your behalf
  • Wonder how you ever had the time to actually read the stuff!
Ahasuerus 17:36, 1 July 2017 (EDT)
You are forgetting "contemplate learning a new language so you can read the books noone wants to translate in one of the languages you know". I almost learned French so I can read Brussolo (and a few more guys whose names I do not even remember anymore) once upon a time. Russian came to the rescue though so I was spared that fate. :) Plus if that is the progression, I am working on it in a weird order (which is not very unusual for me). :)Annie 17:49, 1 July 2017 (EDT)
Personally, I like seeing symbols in the currencies with which I am familiar. They're more quickly recognizable than ISO codes. But it seems to me that the natural thing is to have everything stored internally as ISO codes and allow the user to set preferences for symbols. Chavey 03:03, 3 July 2017 (EDT)
It's been my experience that any coding system designed by humans is likely to change over time. In this case, we have the Russian ruble, whose ISO code was changed from "RUR" to "RUB" in 1997. For this reason I try to use integers as unique identifiers in ISFDB tables and then I put human-assigned codes in separate fields.
There are other permutations to consider. For example, the ISO standard assigns a new code when a currency is revalued, so the code for the Mexican peso changed from "MXP" to "MXN" when the peso was replaced with the "new peso" ("nuevo peso") in 1993. In 1997 the word "nuevo" was dropped, so it's now back to just "peso". However, the ISO code has remained "MXN". If we were to use ISO codes, what should an editor do when entering an undated Mexican books whose price is listed as "100 peso"? Depending on whether it was published prior to 1993 or after 1996, the correct ISO code should be either MXP or MXN, something that most of us couldn't determine without a fair amount of digging.
To go back to the Russian example, the ISO code for the Soviet ruble was "SUR". When the USSR was dissolved at the end of 1991, the code was retired. It was replaced with "RUR" (later "RUB" as per the discussion above) for the Russian ruble and "BYB" for the Belarusian ruble. The latter was replaced with "BYR" in 2000 and then with "BYN" in 2016.
I am sure we can sort it out and come up with a design that will work for our purposes, but it will require a fair amount of bandwidth. Ahasuerus 11:23, 3 July 2017 (EDT)
And how important it is to know which peso it was? I know it is a piece of information but unless if we get a Mexican contributor that really knows how to determine that without spending hours digging, it will be peso anyway. We can use a blended MXP/MXN and when one of them is verified, it will be fixed. As for the Belarusian, using the current BYN will not lose any information (as using RUB for the SUR and RUR won't). If we know which one it is, then fine - safe it. But in the current rules, all the rubles will be just R. So nothing lost. And a system can be amended and improved.
There is a thin line between finding a good design and getting stuck because we are looking for a perfect design. I am not saying not to try to find something that catches everything but sometimes good enough is better than trying for perfection. Even the ability to add free text ones that land on a report for manual inspection will probably work good enough - do we really expect a surge of thousands of non-standard currencies in a week? Annie 17:28, 3 July 2017 (EDT)
I am not proposing a design yet, I am just contemplating requirements :-) We'll need to have a good working understanding of what the challenges are before we come up with a design. For example, we'll need to keep in mind that currency signs/symbols are sometimes duplicative, e.g. the "naked" dollar sign ($) is used both for US dollars and for Mexican pesos. Or, as we discussed earlier, the abbreviation "R", which we have been using for the Soviet/Russian ruble, is also the official sign for the South African rand. Ahasuerus 19:14, 3 July 2017 (EDT)
Still - trying to think of all special cases will never work - there will always be that one country that has messed up their monetary system so badly that no rule can catch it. Anyway - I guess will pick up that again when the current projects are done. :) Annie 21:41, 3 July 2017 (EDT)
Yup! 👍 Ahasuerus 22:30, 3 July 2017 (EDT)
Obvious settings might be "Always use ISO; Use symbols for the most common currencies; Use symbols for every currency we know symbols for; Use symbols for currencies I have checked in my preferences." Pushing this a bit further (Currency v. 2.0), one can imagine a preferences setting that allowed US $ and Australian $ to be seen as ($, AU $) by an American and (US $, $) by an Australian. Or allowed a Nigerian visitor to specify that their currency symbol be "₦", at least for them, even if we didn't have that entered in our table yet. (Ok, so far that would only apply to one book. But our Nigerian visitor might add some more!) Chavey 03:03, 3 July 2017 (EDT)
I too considered this approach at one point. It would take some work, but it's certainly doable. However, my concern is that it would result in our bibliographic pages appearing differently to different users. We already have User Preferences to suppress certain data elements, but displaying the data differently for different users would mean opening a different can of worms. Ahasuerus 10:31, 3 July 2017 (EDT)
Aside: A fundamental principle of software systems, which I try to teach to my students, is that the system for storing data for manipulation by the computer is generally completely separate from the system that is most convenient for the user to see and read. You should not try to force the computer system to deal with what looks nicest to the human; and you should not try to force the human to read what is best for the computer. Assume early on that you will have a conversion layer that turns one viewpoint into the other. That also makes it easier to change how things are presented to the user, without having to mess around with internal details. (In fact, I just gave that lecture to my research students last week.) Chavey 03:13, 3 July 2017 (EDT)
In a way, we have had this separation for as long as we have had machines. Consider a 50-year-old car. It had no computers in the modern sense of the word. However, its dashboard, which was designed to "look nice to humans", was only indirectly related to the way the car operated. It also required conversion layers between what the human operator saw and did and what the actual machinery did.
Of course, we have a lot more flexibility when implementing conversion layers in the software. The "conversion layer" of a WWII-era battleship consisted of things like 4-foot-long levers :-) Ahasuerus 10:46, 3 July 2017 (EDT)
I am horribly late to this conversation, and I have no strong opinion about the formatting (symbol before or after, space or no space, etc.)... I'm happy to let the dust settle and have those with stronger opinions tell me what I should be doing and enforcing.
That being said, one thing that has always troubled me about the price field is that in many cases it doesn't reflect what's printed on the item but attempts to translate it into a standard, but the standard is rather loose and has perhaps more exceptions than actual rules.
In my dream database there's a field where you can record each price directly as it's listed on the pub, then a separate field (or fields) where you can enter a proper amount and identify the currency for sorting and comparison purposes; the conversion layer could work on the latter two standardized fields.
Just my USD 0.02 Albinoflea 23:17, 3 July 2017 (EDT)
I agree with this idea. I work in a world of compiling information from disparate sources where there is no uniform reporting standard, and we have found it best to capture and preserve as-reported and to maintain normalized/computed separately. --MartyD 09:26, 4 July 2017 (EDT)
A semi-aside about the "15$00" mentioned above. It means 15 escudos, and the "$" should really be the cifrão symbol (an S with two vertical lines through it), but there's no Unicode character for that. --MartyD 09:26, 4 July 2017 (EDT)
We could use a PNG of the symbol. I grabbed one from Wikimedia Commons. We'd just need to specify an appropriate size for it: Image:Cifrão symbol.png. ···日本穣 · 投稿 · Talk to Nihonjoe 14:45, 4 July 2017 (EDT)
PNGs in the middle of a text field make a real mess of the view on smaller screens (I work from my phone a lot) or when you need to make the letters a lot bigger because of failing eyesight. Annie 15:24, 4 July 2017 (EDT)
Okay, what about this: $? It's messier, but produces the correct symbol. ···日本穣 · 投稿 · Talk to Nihonjoe 15:55, 4 July 2017 (EDT)
An interesting point. Suppose we have two books and each one says "$100" on the cover. In one case it may mean "100 pesos" while in the other case it may mean "100 dollars". Of course, an English language book would be more likely to be priced in dollars and a Spanish language book would be more likely to be priced in pesos, but it's not dispositive since there are many Spanish language books being published in the US.
If you think about it, we already make this distinction ("as-reported" vs. "normalized") for authors and titles. We capture authors/titles as they appear in publications and then create variants if needed. In addition, we plan to start capturing "stated" vs. "corrected" vs. "derived" ISBNs as per RoadMap 2017. There have also been preliminary discussions re: this issue as it applies to publishers and imprints.
That said, we do some normalization of the captured data even when we aim to enter it "as is". For example, we normalize capitalization and some (specifically Latin-1) accented characters as in the case of Philip José Farmer. We'll have to think about these issues when tackle price changes. Ahasuerus 13:23, 4 July 2017 (EDT)
Not only do we "normalize", but we also practice some outright changes to the data. I'll always remember my disappointment when I was forced to change the "Kurt Vonnegut Jr" that I was seeing on the title page of the book just before my eyes into "Kurt Vonnegut, Jr.". It's from this day that the promised magic of "as is" disappeared for me. Hauck 13:44, 4 July 2017 (EDT)
Out of curiosity but is the author name normalization from the days before pseudonyming? It always made the Latin alphabet languages a bit weird in cataloging here (you cannot apply the rule in the Cyrillic alphabet languages because that format does not exist and there are no shortened versions of words like Junior (and adding it in Latin will be... weird) - so we are much closer to As Is there). I am not challenging the practice, I am just curious. Annie 15:24, 4 July 2017 (EDT)
VTs and pseudonyms were part of the original ISFDB 2.0 implementation which was beta'd in 2006. (Pseudonyms couldn't be deleted until late 2009, but it was an unrelated headache.)
The reason why the issue of "name regularization" originally came up was that some publications, especially magazines, were using unusual forms of author names which many editors felt didn't deserve a VT/pseudonym. For example, if a magazine issue credits Heinlein as "robert a heinlein" on the title page, should we create a new author record? Once we decided that the answer was "no", we kept finding more and more oddball cases which had to be addressed. Hence the current rules. Ahasuerus 12:38, 5 July 2017 (EDT)
Thanks for the background. It just works a bit weirdly now that we allow non-Latin author names - because there is effectively no standardization there. I am coming from the LT system where we just combine the authors and it does not matter how many different versions are there (although no varianting so translations are not caught) - whoever is the main author ends up with the list of all combined ones and that is it. Different systems, I know :) I was just curious. Annie 15:21, 5 July 2017 (EDT)

Preferences for image source?

The Image URL section of the add/modify pub provides rules for what is acceptable for links. It does not provide any guidance as to preference or order. For example this pub is unverified and has an image from Bookscans. I will be making a transient verification and wondered if I should replace the image with an actual scan? Are ISFDB hosted images preferred to all others? I know that Amazon can change an underlying image, so there is a preference for ISFDB hosted scans over Amazon. If such preferences exist, can they be added to the help, please. Doug H 21:03, 6 July 2017 (EDT)

ISFDB-hosted images are generally preferred because we have more control over them. They are also more stable and they are publicly available as part of our public backups. Of course, quality also matters: we don't want to replace externally-hosted high quality images with ISFDB-hosted low quality images. Other than that, I don't think there is a hierarchy of external sites. Ahasuerus 23:35, 7 July 2017 (EDT)

ASIN for printed books with ISBNs

Amazon's rule is that for printed books with ISBNs, their ASIN is their ISBN-10 number (Wikipedia and links from there show that as well). E-books and audio-books get proper B-starting ASINs even when they have ISBN; no-ISBN items get a B-ASIN. So my understanding had been that the new ASIN field is only for cases where the ISBN-10 is not also an ASIN (as we already have the ISBN on the page) and I had been acting accordingly (edited one out today). Am I mistaken or do we want to make that part of the rules? Or do we want to leave that to every editor idea (I think having the number in 2 places is an overkill). Thoughts?Annie 22:53, 7 July 2017 (EDT)

As described in Bug 660, some Amazon records use a "B0" ASIN even though an ISBN is available. And, of course, 979 ISBN-13s can't be converted to ISBN-10s, so that's another scenario where we have to link by ASIN. I wouldn't enter an ASIN that is the same as the book's ISBN-10, though. Ahasuerus 23:38, 7 July 2017 (EDT)
Right - which is what I was saying above albeit a bit too complicated I think - if the ASIN is not ISBN-10, we record it (And all of those are B-ASINs). Otherwise, we do not need it. :)
And now comes the big question - do we leave the editor to decide if they want to repeat the ISBN-10 when it is the ASIN or do we make the rule (and try to enforce) that we only record as ASIN if it is not the ISBN. Annie 23:49, 7 July 2017 (EDT)
I would discourage editors from re-entering ISBNs as ASINs unless there is a specific reason to. Did the editor(s) who have done it have additional thoughts on the subject? Ahasuerus 00:08, 8 July 2017 (EDT)
The one today did not know that ISBN-10 was used for ASINs so not really(here is the discussion). The reason I brought it up here now is because a second one has 2 pubs in the queue now like that. I will put them on hold and go talk to the editor to ask her to come chime in here (of course I have misgiving for those two anyway because they are graphic novels but one thing at a time). Annie 00:13, 8 July 2017 (EDT)
I thought the whole point of the new field was so the links would work... ? so ISFDB gets the link credit? Susan O'Fearna 00:58, 8 July 2017 (EDT)
If you have the ISBN, the link is already there (on the left under other sites) :) The idea of the ASIN external identifier is to allow ISBN-less books to have an identifier (and while we have it, we throw the links in there because most e-books are available on all sites) and thus to allow them to be automatically submitted from our friendly neighborhood robot (as he does for books with ISBNs). The rest of the templates are indeed for the links mainly - so we do not have 20 versions of them, half of them not working. Ahasuerus can comment more - but that is my understanding. Annie 01:03, 8 July 2017 (EDT)
That's right, books with ISBNs already have links to Amazon and many other places under "Other Sites", so adding the same ISBN to the "External IDs" field would create a duplicate link. Ahasuerus 09:21, 8 July 2017 (EDT)
Personal tools