Rules and standards discussions

__NEWSECTIONLINK__

This page is for discussions about the rules and standards, such as whether certain kinds of publications belong in the ISFDB, or whether the help text defining capitalization should be modified. It also includes questions about interpretation, such as whether a SERIAL type can be used for sequences of short stories subsequently republished as a novel.

To add a new section at the bottom, [ click here].

For old discussions see Rules and standards discussions/Archive.

Also see R&S Example page.

Letter columns
As was pointed out to me by a new editor, there is a a section of the help page (under "Anonymous or uncredited works") concerning author credit for letter columns that up until now I'd not been aware of. It states:

If a work by its nature has no author or editor, use "N/A"; this applies to unedited letter columns.

I positive that almost no one else has seen this as well considering there are only two records in the entire database credited to "N/A". I've always used "various" as the author of letter columns, regardless of whether there are embedded editorial responses. The help goes on to say:

Letter columns with embedded editorial responses should be credited to whoever writes the responses, or to "uncredited" if this is not obvious.

Is there anyone currently working on the database who follows these standards? "N/A" is obviously not being used, but why attribute a column to just one person when it includes work written by several people? An editor chooses which letters to print, and may or may not respond to them, but he also chooses which stories to print and we don't credit him with those. (Yes, the analogy has been stretched, but shows the absurdity, in my opinion, of crediting editors with a letter column.) Is there anyone else who feels the help section should be updated, and if so, how should it be done? Mhhutchins 00:45, 4 October 2012 (UTC)


 * I enter the author of letter columns as "The Editor" or "The Editors", though on rereading that template, I've no idea how I arrived on thinking that was the standard . I suspect I arrived at this because the Weird Tales column, "The Eyrie" serves as both editorial and letters.  Clearly I shouldn't have extrapolated this and applied it to all letters columns.   However, I believe I've seen examples of this other than my own edits, but perhaps that is how the responses were signed in those instances. --Ron ~ RtraceTalk 02:01, 4 October 2012 (UTC)


 * I also believe that a credit to the Editor is most appropriate. It accepts the reality that many/most/all of these letters have been edited for content and clarity. Even my local newspaper carries the disclaimer that all letters are subject to editing. Kevin 03:03, 4 October 2012 (UTC)


 * But does the editor write the letters? Published stories are also "subject to editing" but we don't credit the editor as the author. Even more importantly, are the columns credited to the editor? Isn't the ISFDB standard to record credit based on what is stated in the publication? Mhhutchins 04:42, 4 October 2012 (UTC)


 * I'd say that, if the letter column has a distinct title it should be mentioned (as is the title of a review column) and that the editor is the logical person to be recorded as author. It's still debatable if something like 'various' should be entered as author for the letters also (maybe in general if the individual letters aren't totally submissioned?). Stonecreek 09:50, 4 October 2012 (UTC)


 * I'm trying to get a more clear consensus on this matter. It's obvious the current stated standard ("N/A") isn't a good choice, so there's only two options: letter columns, with or without embedded responses, should be credited to A) the editor of the magazine or B) "various" meaning that it's the work of several authors. Please respond below.


 * B Mhhutchins 19:32, 10 October 2012 (UTC)


 * C. Sorry. If the letters are published as complete "works" (as opposed to being explicitly excerpted), I would treat the letter as a published work and the letter writer as the author, and I would ignore minor ancillary editorial commentary (e.g., an editor's note or a brief response from editor or someone else).  MartyD 21:20, 10 October 2012 (UTC)


 * I agree, and that's what we all do, if we choose to create records for individual letters. But this discussion isn't about the crediting of individual letters. It's about crediting the letter column. Mhhutchins 21:42, 10 October 2012 (UTC)


 * I see this as similar to our treatment of an anthology or collection where someone has written a brief passage before or after each contained work: We don't treat that contained work as credited to the passage-writer, nor do we track the passages as distinct works/content. If the letters are quoted as part of a column whose primary purpose is the non-letter verbiage, then I think the column should be considered the work and the editor its author, and it should be credited to the editor.  I see this as similar to our treatment of a review: We don't treat the reviewed work's author as co-author of the review.  I think we should use "various" for the first situation, if multiple letters are published in a column/section and we choose not to document them individually.  --MartyD 21:20, 10 October 2012 (UTC)


 * So what do we do when we create records for both the column and the letters, as many of us do? The credit of the letter column should not be determined by whether the ISFDB editor chooses to create records for the individual letters. That is an optional choice for the editor. In that case, should we credit the column to "various" (which was choice "B") and then record the individual letters to their actual author? (Which is what I've been doing.)  Mhhutchins 21:42, 10 October 2012 (UTC)


 * I prefer A. I too see the letters column as similar to an anthology and like an anthology I think it is appropriate to credit the author of the letter column to the editor by actual name if credited.  Individual letters (by "significant" authors) may also be included and credited to the author of the letter.  This also handles situations like "The Eyrie" where the column combines editorials with letters to the editor. I feel that treating the column and the individual letters this way enables us to credit all the appropriate parties. --Ron ~ RtraceTalk 02:18, 11 October 2012 (UTC)


 * But Marty also sees it "as similar to an anthology", but comes up with a different interpretation (if I'm understanding him correctly): that interstitial editorial material/comments/responses are neither recorded as content, nor are they credited; that the editor be credited only if he includes excerpts from letters, and has created, by editing, an entirely original work. And I agree in that case. (And this looks like how you describe "The Eyrie".) But here we're talking letter columns with no obvious interjections by the editor to create an original work, only a collection of "various" letters. At least 90% of all letter columns are as I describe, and it's those types of columns which I'm trying to reach a consensus on how they should be credited. Mhhutchins 02:45, 11 October 2012 (UTC)

(unindent) I think you understand me correctly. I think a column that's a collection of letters should either be entered as its column header w/"various" as the author credit, or as individual letters with writer-as-author credit. It wouldn't have occurred to me to mix the two, but I could see wanting to go the "various" route while also wanting to call out a small subset of individual letters. It also makes sense to me to consistently use "various" for a recurring feature even if sometimes the column contains but a single letter. Regardless, I don't think the editor should be credited. Here's another parallel example: I have seen "columns" of reviews, which seem to be to be exactly the same scenario, with a column/section header, then the individual reviews. I don't think it would make sense to credit that to the editor, and if not there, then why for a letters column? --MartyD 03:14, 11 October 2012 (UTC)


 * Let's take "The Eyrie" and similar editorial/letters features out of the discussion, or at least table it, as I think Michael is suggesting. For the remaining collection of letters only, I can get behind crediting the column as a whole to "various".  I do still believe it is important to be able to enter both the column as a whole and individual letters by significant authors.  This is exactly analogous to my understanding of how review columns are to be handled, with the exception as to how the overall column is credited.  this help specifically asks for both an ESSAY title for the column and a REVIEW title for the individual reviews.  I've always entered the overall column as credited to all of the reviewers who contributed to the column. That help doesn't specify how the column should be credited and that seemed the most logical. It may be worth considering "various" as an author for both types of columns.  The help also specifies that a column ESSAY should not be entered for a single review, unless the column is a regular feature, and this makes sense for letters as well.  The only think that would be lost (for reviews) would be tracking reviewers of movies, games, comics, etc.  That could be solved by entering such reviews as ESSAYs.  I seem to have gotten into a digression on reviews.  Anyway, I support B, excluding letter columns with editorial content and provided that individual letters may also be entered.--Ron ~ RtraceTalk 15:16, 12 October 2012 (UTC)


 * I think the type of review column that you describe is quite analogous to the letter column, and I've been crediting such review columns (that contain reviews by more than one reviewer) to "various" while creating individual review records which credit the actual reviewer. For example, this periodical. (In this example, there were no significant letters, but if there had been, I would have created individual records. The three reviews were credited to the author, while the column was credited to "various".)
 * I believe the same rule should be applied to letter columns and review columns that contain contributions from more than one author: Create an essay record for the column title which is credited to "various", and then create an essay record for each "significant" letter, crediting the letter-writers (leaving the definition of "significant" up to the ISFDB editor) or a review record for each review, crediting those records to the stated reviewer. Mhhutchins 15:41, 12 October 2012 (UTC)

Untitled interiorart records
The current standard as stated in this section of the help page is

''Artwork. Interior art should have the same title as the fiction or essay it is associated with. If it is independent of other content, and has no apparent title, give it the title "Untitled".''

I was told (and have told those whose submissions that I've moderated) that an editor should not use "untitled" in the title field of an interiorart record. That if a work doesn't illustrate a specific work of fiction or an essay, it should be titled the name of the work which contains it. For example, the title of the magazine issue. I've used "Untitled" as a title field but only when that's the actual title given in the publication, such as this one.

An advanced search for "untitled" with an INTERIORART type returns more than 700 records. Many of these records are from early magazine entries. But there are dozens from Polish publications, which means there are current editors using this titling method. The result is author summary pages like this one and this one. There's always the possibility that some unsuspecting editor will try to merge these records, like the nine untitled records here.

So should we update the current standards, or just let sleeping dogs lie, and deal with the consequences on an individual basis? Mhhutchins 19:22, 10 October 2012 (UTC)


 * IMHO, the standards should be updated to essentially outlaw the "untitled" option for interior artwork altogether (and maybe even set up an auto-check to prohibit such submissions). My preference would be that if a piece of art was associated with a particular essay or story, it should get the title of that essay or story; but otherwise it should get the name of the book or magazine it's in. That would still result in some name duplication, such as when a fanzine has several pieces of art from the same artist, but would avoid some of the problems with the cases you've linked to. Chavey 18:33, 14 October 2012 (UTC)


 * Michael has raised this as a result of this discussion on my Talk page. I have no problem with "Untitled" (and similar such as "Cartoon: no caption") as they stand, but the Help section needs to be expanded upon.


 * Re. a piece of interior art sharing the title of the title it illustrates: it seems this is worth doing if it is actually connected to that title, as opposed to a piece of interior art where there is no connection to the story/article and is there merely as a fillo. Fanzines especially excel at this. Most fillos are untitled, just look at the hundreds – nay, thousands – of pieces of art by ATom, Bill Rotsler et al, that have appeared (and continue to re-appear) over the years, all untitled. It's probably not possible, or even a desirable course of action here, to accurately index the appearance of each one, first under any original titles they may have been attached to and then later any variants. That's quixotic, unless an editor already has access to books/databases of each artist's entire output and where each illustration first appeared. Much easier for cover art by feted artists, far less practical a task for interior art.


 * I've found "Untitled" to be useful (and dare I say accurate) when it simply appears alongside the name of the publication in brackets. IMHO this is what the Help is missing: advice on what else to include beyond "Untitled". If an artist does have a few items of art in a single pub that are worth disambiguating, just add "[2]", "[3]", etc. before the publication title in brackets. This pub covers most of what I'm talking about, but also, if unrelated/unconnected artworks or fillos are given titles elsewhere in the pub, (perhaps not alongside the artwork itself), use the titles they have been given (eg. here). All the above are offered as suggestions, but as I said in the previous discussion, I'm not doing anything that the rules say we shouldn't, and after quite a bit of trial and error this is what I've settled on as a means for accurate and practical data entry within the rules as they stand. Until someone points out to me why it shouldn't be so, it's worked okay for me. PeteYoung 01:18, 15 October 2012 (UTC)

Categories for title "Web Page" fields
For titles that are linked to web resources via the "Web Page" field, would it be useful to specify the type (e.g. full text, excerpt, review of the respective story) and format (web only, pdf, ebook, audio) of the linked resource? This could be implemented via one or more categories related to a specific title "Web Page" field. The suggestion is similar to what has been proposed for author "Web Page" fields here.

I'm bringing this up with another idea in mind (this as an example of one possible use of these categories): I maintain a web site which has several 1000s of links to free speculative fiction online. These links could be merged into the ISFDB as title "Web Page" links (now that most of the web magazines are covered here). With categorized links, I could then use a ISFDB database mirror directly as a data source; and new links for titles wouldn't have to be entered redundantly at two sites. Fsfo 10:31, 14 October 2012 (UTC)


 * This appears to be less than a rules and standards issue, than it is an issue of software design, display and implementation. Either way, I see no problem with creating a separate field which would then link the title record with a legitimate online publication of the title. (Many editors are already using the current Web Page field to do this.) BTW, "most of the web magazines" are not covered by the ISFDB, only certain ones which have shown to be "exceptional" and follow other criteria. This proposed field should not be used as an excuse to add web-published stories to the database. It should be used to link title records for stories that have already been added to the database under the current inclusion policy. At that point, the issue becomes a rules and standards one. Mhhutchins 20:43, 14 October 2012 (UTC)


 * That's what I had in mind - Adding categorized links to existing ISFDB entries, such as stories published by "exceptional" web magazines ("most of the web magazines" was definitely not the correct phrase here). These categories could be especially useful when a title has multiple web links, maybe including links to secondary resources (such as reviews or discussions). These kinds of resources are mentioned here - but apparently not included as links right now anyway?


 * If the suggestion is not primarily a "rules and standards" issue, should I additionally bring it up elsewhere? Fsfo 19:53, 15 October 2012 (UTC)


 * Sound like a feature request, which can be added to the Sourceforge page which the developers use to track software change requests. Go here for a list of current requests. If you do a search for "web page" you'll find several requests for similar upgrades. If you believe yours is a unique request there is a link to "Add new" requests. I'll post a message on the Moderator page which links back to this topic so that we don't have to start a new discussion. Hope this helps to get the attention of the developers who have the power (and the expertise) to implement such requests. Mhhutchins 20:46, 15 October 2012 (UTC)


 * It would appear that this is a request for two additional fields to be made available for each title-specific Web page, one for "Web page type" and the other one for "format". In both cases what can be entered in these fields should be limited to pre-defined lists of values, right? If so, then it's certainly possible to implement and sounds at least moderately useful. Ahasuerus 09:31, 17 October 2012 (UTC)


 * I have added a feature request [] (messed up the back-link to this discussion though). Fsfo 08:02, 18 October 2012 (UTC)


 * Looks good, thanks! Perhaps "full text" or "complete text" rather than "full title"? Ahasuerus 16:50, 18 October 2012 (UTC)


 * I have seen a few title "Web Page" links to images (for interior art), where something like "full text" would look a bit strange (maybe also for links to audio versions). "Full title" is not much better, but at least consistent with the use of "title" for the respective entries. Or maybe this is just a question of how to display a category (we wouldn't necessarily have to use the internal name). Fsfo 17:20, 18 October 2012 (UTC)


 * How about "complete work" which would apply to both text and art? Mhhutchins 17:44, 18 October 2012 (UTC)

Creating pseudonyms for crediting publications in a non-arabic alphabet
I don't believe that this is the proper way to handle authors with arabic-alphabet names who have works published in a non-arabic alphabet. It just doesn't feel right. There has to be a better way of handling it. Mhhutchins 05:13, 15 October 2012 (UTC)


 * I'm not disagreeing with you, just looking for amplification: Do you see a difference between that transliteration (I assume -- I can't read Japanese at all) and this mis-/alternate spelling?  Also, if it weren't Isaac Asimov, but rather a Japanese author, and the non-Arabic form were the native/original spelling, would you see that situation as any different?  --MartyD 10:31, 15 October 2012 (UTC)


 * (Aside) Here's another one. --MartyD 10:43, 15 October 2012 (UTC)


 * I'm saying that there has to be a better way of handling the transliteration of an author's name rather than creating pseudonyms. I don't know how exactly, but there must a separate "relationship" created between the two names rather than using the pseudonym relationship. Again, we're squeezing a new type of data into a hole for which it was never intended. If the name "Isaac Asimov" is literally transliterated as アイザック アシモフ in Japanese, then that is not a pseudonym. I don't want to see everybody jumping into the boat before we know it's sea-worthy. I believe we sank the "variant" boat when it started being used for translated titles, and don't want to see that happen again. Mhhutchins 15:39, 15 October 2012 (UTC)


 * There's a site that I catalog my music on, discogs.com, that distinguishes between two categories of name variation; one is called "Alias" which would be the equivalent of a full-blown pseudonym, and the other is simply referred to as "ANV" short for "Artist Name Variation".


 * The ANV includes minor changes to spelling as well as translations and transliterations. They have some good, clear documentation about it.


 * And you can see how they display Aliases and ANVs using Muddy Waters as an example.


 * I think it would be great to see ISFDB move towards this sort of approach, which would address Michael's concerns (which I share) as well. Albinoflea 20:51, 15 October 2012 (UTC)


 * Damn! I like the discogs approach. How difficult would it be to implement such a system in the ISDFB? Mhhutchins 22:17, 15 October 2012 (UTC)


 * That is a substantial improvement. The overloading of fields for multiple purposes is causing us significant problems. If it's possible to implement this system, it will take a lot of work to convert our records to it, but I think it would be worth the effort. (And yes, that means I'll volunteer to be on that project.) Chavey 15:55, 16 October 2012 (UTC)


 * I like it. FWIW, you could see a similar concept applying to titles as well for punctuation and perhaps minor word differences (e.g., "The" vs. no "The", or "5" vs. "Five").  It's probably worth considering that as well, lest we go one way with author names only to find that approach wouldn't work for titles at some point down the road.  --MartyD 16:20, 16 October 2012 (UTC)


 * That's a big can of worms you are opening, folks... Before we go too far down this road, though, let me point out that we currently use Latin transliterations when entering author names -- see any number of Japanese and Russian authors currently in the database. The reason for this practice is that we want to allow all of our users to search for author names, not just the subset(s) who can enter "安部 公房" (or even know what to enter.) Admittedly, it's less of a problem when the author's works have been translated, so we have the transliterated version on file anyway.


 * I should also point out that we faced a similar problem with titles records, but it was decided that we should enter them verbatim. The consensus was that (unlike authors) only those users who had appropriate keyboard drivers installed would want to search for original titles. It's obviously inconsistent, but it is what it is at this time.


 * The medium term solution to these twin problems, as we discussed in the past, would be to add a new field for "transliterated/romanized name" to the Author table and a "transliterated/romanized title" field to the Title table. We would then modify the search logic to check both the regular name/title and the transliterated name/title when running searches. And all would be well with the world... The good news is that it's not all that difficult to implement software-wise although it will take a fair amount of time to populate the new fields in the affected records.


 * As far as the "ANV" proposal goes, I will have to think about it. Depending on how it is implemented, it may be either a great deal of work or only a limited amount of work. Ahasuerus 08:30, 17 October 2012 (UTC)


 * I've been thinking about this, and I'm not sure I see the need to create a separate field for transliterated names, unless I'm missing something.


 * For instance, to use the example above, suppose a user has a copy of a book by "安部 公房"... if they don't know how to enter it (or what to enter), how likely are they to know how to transliterate that name into a Roman form that they can easily enter? And if they did know how to transliterate it, wouldn't they just search on the transliteration anyhow?


 * If we just enter the original non-Roman version of the name as a pseudonym (or ANV) then if a user does enter "安部 公房" then they'll be directed to Kobo Abe through the pseudonym/ANV page. And if they search on Kobo Abe directly, then they should see "安部 公房" listed under the list of pseudonyms/ANVs. (That's assuming of course that the variant link between the two has been established.)


 * The whole situation could be made additionally clearer if the Pseudonym column of search results was actually populated with the name behind the Pseudonym rather than just the indicators "Pseudonym" or "Has pseud. titles", which of course would be some work. But that is something that could be implemented separately from the creation of a mechanism for distinguishing between Pseudonyms and ANVs. Albinoflea 03:21, 19 October 2012 (UTC)

e-ARCs redux
I have Greg Goss's "Shadow of Freedom" submission on hold. It's a Baen "e-ARC" and Baen is selling it now, in October 2012, even though the book won't be officially published until March 2013. e-ARCs are a beast which, as far as I can tell, we last discussed here and there was no consensus as to whether we want to include them. Checking the database, I see that we have a number of them on file, e.g. 's Storm from the Shadows.

Based on the above, I have two questions:


 * 1) Is there a stated or unstated policy that we currently follow re: Baen's e-ARCs?
 * 2) If there is none, do we want to update the Rules of Acquisition and Help to state that "true" ARCs, that is ARCs not sold to the public, are generally "out", but "e-ARCs", i.e. electronic copies of forthcoming books that are sold to the public, are "in"? They don't appear to be any different than any other book first released electronically and then as a dead tree edition. Aside from some last minute copy editing, of course, but that has been known to happen in between editions as well.

Ahasuerus 16:49, 22 October 2012 (UTC)


 * I suggest not changing the standards, but only allowing exceptions of certain titles on a case-by-case basis. For the moment, this practice is only done by Baen, but if more publishers start doing it, we could come back to the issue. I personally feel that selling these "publications" is like selling copies of an author's manuscript. Mhhutchins 17:26, 22 October 2012 (UTC)


 * I agree that it's an unusual way to sell books, but I suspect that Baen may be (or may have been) thinking that they were doing their customers a favor by making new works available sooner. Ahasuerus 01:34, 24 October 2012 (UTC)


 * How many publishers do you know who are doing their customers a favor by selling them something? How big a "favor" can it be to sell an author's fans an overpriced draft? Mhhutchins 02:04, 24 October 2012 (UTC)


 * Well, sure, they are also making money selling these e-ARCs, but how many time have you heard fans say something like "Can't wait to read the sequel! Too bad it won't be out until December. I'd pay good money to be able to read it now"? So Baen presumably figured out a way to satisfy this pre-existing demand and make a fair amount of money in the process. As Kevin wrote in 2009: "They have been known to squeeze $15.00 (e-arc) + $3.75 (1/4 $15.00 for month of e-books) + $26 (HC) = $44.75 out of my wallet before they are done on some (apparently addictive) works." They may also, as Kevin speculated, "continue to use e-arc sales to justify or re-adjust expectations before some books even hit the shelves", but that's a whole different can of worms. Ahasuerus 04:53, 24 October 2012 (UTC)


 * In other words, the work has not really been published (which would make the e-ARC ineligible for inclusion). Looking at the listing on the Baen website, I could not find an actual "publication" date of the e-ARC. Except for the dates of some of the reader comments you wouldn't know when this became available for sale. And the submission gives the date as 2013, which wouldn't work if this is allowed into the database. Why not just add a note to the record for the final and official version of the work that it was previously available (in a somewhat different form) as an e-ARC? Mhhutchins 17:26, 22 October 2012 (UTC)


 * Well, that would be another way to handle these cases, but consider how much information the editor who created the Storm from the Shadows e-ARC record put into Notes. If we were to migrate all of that info to the Title record, would it overwhelm the latter? Ahasuerus 01:34, 24 October 2012 (UTC)


 * Back to my original argument: it's not been "published" so most of those notes are superfluous. I could pare them down to one line. Mhhutchins 02:04, 24 October 2012 (UTC)


 * About the last part of your remark, a personal response: judging on the readers' comments on the Baen website, Weber should consider rewriting this novel. And these comments are from people who have paid $15 for an electronic file, which gives Baen at least $14.95 in profit. This practice gives off a stronger stench than you'd normally find in the publishing industry. Mhhutchins 17:26, 22 October 2012 (UTC)


 * If we are talking about the same discussion on Baen's Bar, then the complaints are primarily about "monotonically increasing filler-to-story ratio of the series" going back to 2009 or thereabouts (echoed by multiple readers on Usenet, BTW.) I suppose it's possible that, since the trend reportedly started around the time when Baen created their e-ARC program, the two may be related, but I don't think I have seen similar complaints about other Baen authors' work. Still, it's something to think about... Ahasuerus 01:34, 24 October 2012 (UTC)


 * I've never been to Baen's Bar, and rarely go to Baen.com because I'm not interested in 95% of what they publish. I'll check out the Bar, but if they're serving the usual stuff, I'll have to leave for the local watering hole to get a better quality libation. Mhhutchins 02:04, 24 October 2012 (UTC)


 * I don't read Baen's Bar on a regular basis, but after you responded I wandered over and scanned the relevant threads to see what Weber's fans were saying about this installment. I stopped reading Weber after "Heirs of Empire" (1996), but I still scan Usenet once every few days (primarily to see if there are any ISFDB-related queries that need to be answered) and there have been growing complaints about his books for a number of years. Given this background, I suspect that the quality of the latest volume is not related to the availability of his books as e-ARCs. Ahasuerus 04:53, 24 October 2012 (UTC)


 * BTW, I wish more editors would get involved with these discussions. Mhhutchins 02:04, 24 October 2012 (UTC)


 * Well, this is at least the third incarnation of the "e-ARCs: In or Out" discussion (see Kevin's post linked above which provides additional links), so other editors may have grown weary. I try to participate when I have an opinion, but these discussions take time and most of my ISFDB time is spent on software fixes and Fixer. Ahasuerus 04:53, 24 October 2012 (UTC)


 * I'd been avoiding this discussion as I'd prefer to someday see a mechanism developed to include ARCs generally (as this is a personal collecting interest, and also because it is not information that is consistently documented elsewhere); so FWIW, my inclination would be to leave the rule the way it is and leave all ARCs and pre-published material "out", while acknowledging that interested parties can always document such items in Title notes. The fewer the exceptions, the more straightforward it is to apply the rules, and the less confusion there is all around. Albinoflea 05:13, 24 October 2012 (UTC)

(unindent) Sorry if this is a dumb question, but how is any sort of ARC that is produced and sold not a form of publication? What makes a work "published" instead of merely "printed"? I've been avoiding this discussion, too, but my inclination is that any ARC produced explicitly for sale/money/public distribution should be treated as publication, and so be "in", no matter what the publisher pretends is going on. --MartyD 11:03, 24 October 2012 (UTC)


 * As you've pointed out, the operative word here is "published". Baen doesn't consider the work to be published. Otherwise why would they give it a publication date of March 2013, but sell an electronic file of the author's latest draft in October 2012? It seems to me that both the author and the publisher wouldn't want this to be reviewed as a published work. It's still an ARC, no matter if it's sold or not. Just because most publishers don't sell their ARCs doesn't keep copies of them from showing up in dealer listings. They are being sold, but they're still ARCs, thus not published works, and not eligible for inclusion in the ISFDB. The fact that Baen chooses to sell them doesn't change the fact that they're ARCs. Eligibility for inclusion in the ISFDB isn't determined by whether the work is being sold or given away. Should there be a record for a copy of an author's manuscript that Lloyd Currey is selling? The answer is obviously no, but I don't see Baen's practice to be that much different. They've determined there's a market for such items and are driven for monetary reasons to sell an author's manuscript in the form of an electronic file. Pushing aside all factors involving ethics, we should continue the de facto practice of only entering exceptional ARCs into the database: works that were never published, or have marked differences (like ISBNs) from the published work, "Special" editions published for one-shot occasions (like conventions), or any reason that makes them exceptionally different from a run-of-the-mill ARC. Mhhutchins 16:41, 24 October 2012 (UTC)


 * I don't have any objection to the current practice. --MartyD 02:16, 25 October 2012 (UTC)

Doc Savage series
A lot of the Doc Savage titles are on the new "Variant Titles in Series" cleanup script. The variant is often in another series than the main title. The only discussion about Doc Savage I could find dates from 2008, before the publication series were introduced. I.m.h.o. it would be better to have the various Bantam, Black Mask and other reprints in publication series and the title records in a title series. I experimented with the first 10 titles according to the Wikipedia list, resulting in this publication series for the Bantam singles, and some order (chronological) in the title series. Is this the right way, or am I overlooking something? --Willem H. 19:27, 28 October 2012 (UTC)


 * I think it would be best to have all novel Titles appear in a regular series. Various reprints could then be organized as publication series. Once that has been done, we may want to consider putting omnibus Titles into one or more sub-series. Ahasuerus 20:46, 28 October 2012 (UTC)


 * I agree with Ahasuerus. Awhile back I tried to do some organizing with the series and found the co-mingling of the various publication series with the title series to me a major stumbling block. I did the separation and creation of the subseries for the Bantam double and omnibus series, as well as the Black Mask reprints. The only way to fix the single novel title series is to remove all series data and start from scratch (the omnibus records could remain as is). Not being a fan of the work, I felt my time was put to better use elsewhere, and I left the work to someone more familiar with the series. Mhhutchins 21:20, 28 October 2012 (UTC)


 * Thanks! That means I'm on the right track. I'll do the title series for the novels, the publication series for the Bantam singles and clean up the Kenneth Robeson page. See how it looks then. --Willem H. 16:00, 29 October 2012 (UTC)


 * Sounds like a plan -- thanks for undertaking the project! It's been on my list of things to do for a few years, but there are always higher priority items to work on... Ahasuerus 16:13, 29 October 2012 (UTC)


 * Finished. I think it looks better now. --Willem H. 10:05, 3 November 2012 (UTC)

Stated publication date vs. actual publication date?
If a book has a stated publication date, e.g., "First printing: April, 2010", and it has a known-from-the-publisher actual publication that that is different, e.g., "March 30, 2010", which date should be used in the "Publication date" field? I thought our standard was to use the date stated in the publication, relying on non-publication, secondary source information for month or month-and-year if not stated. But it seems we sometimes have been recording the other date. There are also cases of editors supplying more complete dates (to the day) from secondary sources (notably, Amazon). We should at least be consistent.

While acknowledging my "bookist" leanings, I'll propose that we standardize on using a publication's stated date, falling back on other sources only if incomplete or unstated. The corresponding help could be something along the lines of:


 * 1) The month and year (and day, if applicable) as stated in the publication, unless the date stated is known to refer to an earlier printing.
 * 2) * If the publication states only a year, the month may be supplied from a secondary source. See below.
 * 3) * If the publication does not state a date, or if the stated date is known to refer to an earlier printing, the date may be supplied from a secondary source. See below.
 * 4) * If the publication does not state a date and no date can be found from a secondary source, you may attempt to derive a date from other information in the book, such as dated ads. Note that ads and announcements may be reprinted as-is in later printings.
 * 5) * It is good practice, but not required, to record the publication's statement of publication date (or lack thereof) in the publication notes.
 * 6) When using a source other than the book's stated publication date:
 * 7) * If the publisher's website or catalog provides a date, use that.
 * 8) * Otherwise, prefer a date from one of the Secondary Verification Sources, if available.
 * 9) * Failing that, a date from an online bookseller may be used; it is strongly suggested doing so only if multiple such sources corroborate the date.
 * 10) *  Always  document secondary sources of date information in the publication notes.
 * 11) DO NOT GUESS.  If no reasonable source of a date can be found, leave the date unknown (0000-00-00).

Exact wording TBD. I'm just using it to illustrate what I'm thinking. We should no doubt agree on "publication's statement" vs. "actual publication date" first before worrying about the finer details. --MartyD 19:21, 24 November 2012 (UTC)


 * "The month and year (and day, if applicable) as stated in the publication, unless the date stated is known to refer to an earlier printing." Good work, Marty! But this can fail when printings from different countries are involved. From about 1960 until 1970 there are numerous instances where a later printing will differentiate between a first US and a first CDN printing. Most often it's only a month, but with Ballantine editions in particular the difference can be years! Do we update the earlier printing from the later? Even if the earlier is quite specific about a date?  In some instances I would say absolutely [Ballantine did not start printing in Canada until the early 60s, no matter what the copyright pages may say]. Fact over "fact" ??¿¿"?? --~ Bill, Bluesman 23:51, 26 November 2012 (UTC)


 * This situation is covered in the proposal: "When using a source other than the book's stated publication date..." use the source and note it, even if that source is a later printing of the title. This also covers the possibility that the publisher's stated date (or printing number) is wrong. Ever entered an Ace book from the 1970s? Mhhutchins 00:14, 27 November 2012 (UTC)


 * Several hundred, but Ace never printed in Canada. --~ Bill, Bluesman 16:59, 27 November 2012 (UTC)


 * It was a rhetorical question. I was referring to Ace's notoriously bad efforts to indicate printing and date data, which was the second point ("also") of my response. It was not referring to country of origin data. Mhhutchins 17:27, 27 November 2012 (UTC)


 * I agree that this should be the standard and incorporated into the help documentation. Mhhutchins 19:47, 24 November 2012 (UTC)


 * I also agree. I particularly agree that a statement in a book should overrule an Amazon date or a "It is known that this book was being sold by" (some date). (Although that would make a fine "Note".) Chavey 05:18, 25 November 2012 (UTC)


 * I agree as well. I don't know if the inquiries were spurred by this discussion, but shortly after the original post here, I received a request to correct an Amazon date on one record and add a month on another based on a gutter code by adding one month to date indicated by the code.  Both the books in question have the year only on the title page.  I suspect that the Amazon date was added prior to my verification.  I've no problem noting the gutter codes in the notes, but it seems to me that using them for release date by adding 1 month is close to guessing.  Do we want to use gutter codes for refined (month) dating?  Can we work gutter code dating into the above rules?  Thanks.--Ron ~ RtraceTalk 21:07, 25 November 2012 (UTC)


 * Dating publications by gutter codes may be a guess, but it is a well-established and thoroughly-researched guess. Adding six weeks to the week of printing works on more than 90% of all editions I've ever come across. It's as accurate as date codes printed on dustjacket flaps. So, most undated and gutter-coded books published by Doubleday for its trade or book club divisions between 1958 and 1987 can be given a reasonably accurate publication date. Of course, this source should be noted, and that's covered in the fourth bullet of Marty's #2 note, making it unnecessary to mention gutter codes specifically in the help documentation.


 * I also agree to Marty's proposal. Fine work! Stonecreek 10:02, 26 November 2012 (UTC)


 * I too agree with the gist of the proposal. Ahasuerus 04:21, 27 November 2012 (UTC)

(unindent) Sounds like we have consensus. I will propose new wording when I have a few minutes. I'll be sure to incorporate gutter codes and later printings as example sources, too. --MartyD 17:40, 27 November 2012 (UTC)

Updating standards for audio books and magazines (podcasts)
I've been cleaning up a lot of miscellaneous audio formats in order to give the most popular usages some kind of uniformity. Because most of the records for cassette audiobooks were given as "audio cassette", I've changed the 10% that were given as "audio (CAS)" to the more common usage even though that is supposedly the standard format as stated in our help documentation. This led me to change those few "audio (CD)" records to just "audio CD", again despite the current "standard". I believe the help documentation should be changed to reflect these de facto usages (even though, admittedly, I'm the one who made most them conform to a standard which wasn't stated. Mea culpa. It was just easier to go with the majority of records.)

Those formats are now old hat, and there are serious challenges to the latter from such newcomers as digital downloads (such as audiobooks from Audible.com and Amazon), and digital audio players (like the relatively new Playaway system in which the consumer buys a player pre-loaded with the digital file of the audiobook, for example.) Those records that I've found in the database for those new formats have been changed to "digital audio download" and "digital audio player", respectively. I think we need to update the help documentation to add these new formats.

This brings me to the other purpose of this post: the format that's now given as audio (MP3) and a smaller number of records for mp3 audio. There were records that were for actual compact discs that contained MP3 files (instead of standard audio files) and they have been changed to "audio MP3 CD". Every record that is now given as "audio (MP3)" and "mp3 audio" are either downloaded audio files from LibriVox and Project Gutenberg, or podcasts. The question at hand is whether there should be a separate "binding" or "publication format" for these two kinds of publications. Shouldn't the the LibriVox and Project Gutenberg audiobooks be considered the same as a "digital audio download" (like an audiobook downloaded from Audible.com and Amazon)? Should podcasts have their own format name, perhaps something as simple as "podcast"? I have no dog in this fight, as I've entered very few audiobooks into the database. But I feel there should be separate name for podcasts, similar to the way in printed books we separate books from magazines. If in the end of this discussion we decide this change would be beneficial, and because of the great number of podcast records in the system (thanks chiefly to the fine efforts of Desmond Warzel), is it possible to make a universal change to keep editors from having to update each record manually? Thanks. Mhhutchins 01:52, 3 December 2012 (UTC)

After working on conforming the bindings/formats that are present in the database (from more than 150 down to about 30), I've come to the conclusion that there shouldn't be a separate format for podcasts. After all, they're digital audio downloads, no different from what you'd download from any website for a fee. And typing them as a MAGAZINE with a binding of "digital audio download" is perhaps the best way of distinguishing them as podcasts. I won't ask Ahasuerus to do a universal change yet, but if there have been no objections in the next few days, I will. I can understand why there is so little interest in this particular format, supposing most of us are more interested in reading than being read to. But there's a lot of great recorded material that should be part of the database. Mhhutchins 04:30, 4 December 2012 (UTC)


 * Is it to be "digital audio download," then? I'd be game for that.  That it's a podcast (as opposed to Librivox/Gutenberg) is probably evident to anyone who's actually interested in the matter, and I for one have always tried to put explanatory notes in each editor series (i.e. "Pseudopod is a weekly horror fiction podcast").  As I sit here considering the matter, I like "digital audio download" more and more.  It's just specific enough to differentiate them from physical audio media, and just general enough to encompass any necessary variations in format (i.e. I assume there are other audio formats apart from mp3...I honestly wouldn't know).


 * Yes, there are other audio formats that are downloadable. Almost all of the LibriVox titles can be downloaded in ogg vobis. I've also seen sites that offer FLAC files, which are lossless, so they're quite large. But for audio files, you really don't need the high quality necessary for music or video. Mhhutchins 20:12, 4 December 2012 (UTC)


 * I started out doing podcasts because it's something I can listen to with little effort while I'm writing or out walking (though it turns out to be a lot of effort to track down places of first publication for some of the stories; Drabblecast and early Pseudopod are almost aggressive in their failure to provide this information) but having listened to entire runs of podcasts like Drabblecast and Cast Macabre, there is, as Mhhutchins says, a lot of great stuff there; reprints (well, repods) of important classics, and--as podcasts have become a more acceptable venue--exclusive stories from important contemporary writers that aren't available elsewhere. Speaking as a latecomer to almost everything modern (including the Internet itself), I find the whole business quite interesting. Dwarzel 18:42, 4 December 2012 (UTC)


 * So there's at least two of us who feel "digital audio download" is the way to go. I'll give it another day or so to see if we get any other opinion before I ask Ahasuerus to change the existing "audio (MP3)" records to the more generic, but also more correct, name. Mhhutchins 20:12, 4 December 2012 (UTC)


 * I agree that MP3 is unnecessary specific, maybe as "download" since it would re-introduce doubts about podcasts, which are not properly downloads (you can listen online to podcasts). I believe that "digital audio" with the source would suffice, encompassing both downloads and podcasts. --Pips55 21:01, 4 December 2012 (UTC)


 * It's my understanding that only downloadable podcasts are eligible for the database, although that is not confirmed by the stated policy which says nothing about them. I believe they "snuck in" when we started to allow records for ebooks and other files from Project Gutenberg and LibriVox. Strangely, there isn't any explicit policy concerning podcasts. Desmond, do you know if all the podcasts that are in the database downloadable? Mhhutchins 00:17, 5 December 2012 (UTC)


 * Everything on the linked table is downloadable. Actually (not that I want to start a whole big thing), though Pip55 is correct that podcasts can be listened to in one's browser, I would suggest that downloadability is in fact one of the defining characteristics of a podcast ("pod" being short for "iPod")... Dwarzel 05:59, 5 December 2012 (UTC)


 * After a week with no further input, I have updated the Publication Format standards to apply the results of this discussion concerning audio books. In the process, I edited the other formats for clarity and added other formats based on the usage as recorded in this discussion. I made no fundamental changes to the various magazine formats. Mhhutchins 05:43, 12 December 2012 (UTC)

Counts of current bindings
[This was removed from the above topic and given its own header.] Very nice! A bit more work and we may be able to narrow things down to a relatively small set of binding codes and then convert the field to a drop-down list. Here is where we stand as of a few minutes ago:

++---+ | pub_ptype             | count | ++---+ | pb                    | 88419 | | hc                    | 65144 | | tp                    | 55767 | | [none]                | 12616 | | digest                |  8389 | | ebook                 |  4856 | | quarto                |  2445 | | Pulp                  |  1965 | | ph                    |  1618 | | audio (MP3)           |  1478 | | audio CD              |  1345 | | A4                    |  1249 | | webzine               |  1234 | | A5                    |  1165 | | bedsheet              |   799 | | audio cassette        |   688 | | dos                   |   293 | | audio MP3 CD          |   257 | | octavo                |   216 | | digital audio download |  137 | | digital audio player  |    71 | | tabloid               |    56 | | email                 |    27 | | CD-ROM                |    17 | | audio LP              |    13 | | Half Foolscap         |     7 | | portfolio             |     7 | | broadside             |     6 | | audio DVD             |     4 | ++---+

Ahasuerus 03:31, 3 December 2012 (UTC)


 * Thanks for the list. I should have it pared down within the next day or so. At least 90% of those formats with less than 25 records can be changed or combined to create a drop-down menu of less than 20 formats. Mhhutchins 04:06, 3 December 2012 (UTC)


 * I'm thinking it be might more reasonable to have only a dozen or so choices in a drop-down menu, with the last option being "Other". With the last choice, the editor would be instructed to enter a description in the Note field. Then the moderator would have the ability to place the record into another dozen or so formats. It just seems to me that once you pass ten or twelve choices in a drop-down menu it begins to lose its effectiveness. Is it possible to design such a system that would limit editors choices while giving moderators more flexibility? Mhhutchins 16:56, 3 December 2012 (UTC)


 * I've made corrections that should have removed more than half of the 150+ formats. Ahasuerus, can you run the script again and update the list? Thanks. Mhhutchins 22:41, 3 December 2012 (UTC)


 * Done! I think we are at a point where we can start making mass changes, e.g. how about I change all occurrences of "mp3 audio" to "audio (MP3)" for starters? (We can always change them to another code later.) BTW, the pub whose binding code is set to "p" is Terre, Planète Impériale. Ahasuerus 22:58, 3 December 2012 (UTC)


 * Yes, please go ahead and make that change. Hopefully the previous post will get some kind of response. (And thanks for finding the single "p" record. There's no way I could have found it on my own!) Mhhutchins 23:01, 3 December 2012 (UTC)


 * Sure, give me a few minutes to test the change on the development server (and make dinner.) Also, I think it's safe to remove all "unk" codes since they don't seem to add anything. Ahasuerus 23:31, 3 December 2012 (UTC)


 * Also, I'm having trouble finding those 6 that are just "audio". I'm not sure what "large s/s" and "digest s/s" mean, but I've left them as is until I do more research. Mhhutchins 23:09, 3 December 2012 (UTC)


 * Hm, something to do with Saddle stitch stapler, perhaps? Ahasuerus 23:31, 3 December 2012 (UTC)


 * That's it. Miller/Contento uses it to mean "saddle-stapled". Mhhutchins 23:59, 3 December 2012 (UTC)


 * Forgot to mention: don't get rid of the "unk" yet. I'm using that the search for the actual bindings. We started with more then 500 and I've got them down to less then 300. Mhhutchins 00:01, 4 December 2012 (UTC)

(unindent) OK, all occurrences of "mp3 audio" have been changed to "audio (MP3)". The list above has also been updated. Should I also change "audio MP3 CD" to "audio (MP3)"? Ahasuerus 01:11, 4 December 2012 (UTC)


 * No. I pulled those out of the group and intentionally created a new category. These are compact discs that contain mp3 files instead of standard audio files. An audiobook on CD can run 6-12 discs or more ("audio CD") but the same audiobook on an MP3 CD could only need 1 or 2 discs. As I mentioned in the previous topic, all cases of "audio (MP3)" are either podcasts or audiobook downloads from LibriVox and Project Gutenberg. In other words, they don't exist in a hard copy (a compact disc), but only as a downloadable digital file. Thanks for running the script again. So I posted the above topic to get some kind of discussion started about how those records should be treated. After all, they're all a "digital audio download". Mhhutchins 01:34, 4 December 2012 (UTC)


 * Oh, I see. (Sorry, I apparently missed a part of the original discussion.) I haven't been following the audio scene closely, but is it a safe bet that "audio (DVD)" is likely to become more common in the future? Ahasuerus 02:55, 4 December 2012 (UTC)


 * It might make sense to think so, but I wouldn't count on it. The future of audiobooks is going to be 99% digital downloads. Even libraries, the last holdouts for CDs, are getting into downloads. You can even purchase a separate file called Whispersync with an ebook for your Kindle, which gives you the option of having the book read to you if you get tired of reading it yourself! And it starts up immediately where you stopped reading. Mhhutchins 03:13, 4 December 2012 (UTC)


 * BTW, are you able to pull up those 6 records that only give "audio" as the binding? Mhhutchins 01:34, 4 December 2012 (UTC)


 * Yes, here they are:


 * Holly and the Silver Unicorn
 * Planet of the Apes
 * Perfection
 * South of Salem
 * Garro: Sword of Truth
 * Blood and Gold


 * Ahasuerus 02:55, 4 December 2012 (UTC)


 * Another format bites the dust. Thanks. Mhhutchins 03:13, 4 December 2012 (UTC)


 * I'm surprised we have so few LPs in the database. I seem to remember quite a number of recordings by Caedmon Records and other companies of authors (and celebrities) reading their stories. Harlan Ellison even had a club where members would get a record every few months of him reading his stories. There has to be some information out there somewhere on the internet. Mhhutchins 03:20, 4 December 2012 (UTC)


 * Found it! Mhhutchins 03:28, 4 December 2012 (UTC)

(unindent) OK, so we are down to 40 formats, 13 of them with fewer than 10 pubs. Nicely done! Should we change the two "audio mp3" pubs to "audio (MP3)"? How about changing "pb+audio cassette" to "pb" and updating the record's Notes to the effect that the book comes with a cassette? Ahasuerus 03:41, 4 December 2012 (UTC)


 * I agree that it's not necessary to give both in the field. The lesser of the two formats can be noted. If there are no objections in the next few days, I will ask that you change all "audio (mp3)" records to "digital audio download". (Those two remaining "audio mp3" pubs are a couple of Bill Longley tests from 2008. I've left him a note about them in case he's forgotten about them.) Mhhutchins 04:36, 4 December 2012 (UTC)

Re: limiting the list of default binding codes to ~10 and then allowing moderators to add another ~10 less common bindings, well... it's technically possible, but it seems a bit awkward. How about making the drop-down list more descriptive, e.g. "tp - Trade Paperback format, taller than 7" (18 cm)", instead? We would still store the value as "tp" in the database, so it won't affect disk space or performance. Ahasuerus 03:41, 4 December 2012 (UTC)


 * Sounds fine to me. We may have to do some tweaking once it's implemented, but we seem to be on the right path. Mhhutchins 04:36, 4 December 2012 (UTC)

(unindent) Another update as of midnight (server time). Down to 32 unique codes. Ahasuerus 06:13, 4 December 2012 (UTC)


 * Congratulations on the progress on this front guys! Chavey 05:13, 5 December 2012 (UTC)


 * Updated the list above, which is now down to 28 unique codes + "[none]". BTW, if "quarto" is the same as "bedsheet", should we then merge the two (if not now, then eventually)? Ahasuerus 23:27, 5 December 2012 (UTC)


 * Yes, "quarto", "bedsheet", and "A4" are all basically the same format. I see no reason to keep them separate, but can anticipate there being a fight about which name to retain. So I propose we chuck them all and use the term used by the SFE III which defines them as "Letter Size": also called "Large flat" or, erroneously, Bedsheet, measuring 11.75 x 8 in (298 x 216 mm), and thus the approximate equivalent of Quarto or A4. This size also applies to large pulp or large slick. Issues are usually saddle-stapled but can be perfect bound. It's a rather drastic move, but it's about time we made a decision. Mhhutchins 23:43, 5 December 2012 (UTC)


 * Once the field has been converted to a drop-down list, we can easily change the displayed value to whatever we want, e.g. "Quarto/bedsheet/A4". And if we don't like it, we can change it again. Ahasuerus 00:09, 6 December 2012 (UTC)


 * Please go ahead and change those "audio (mp3)" records to "digital audio download". I'm going to start a new topic about the merger of "quarto", "bedsheet", and "A4". Thanks. Mhhutchins 17:40, 6 December 2012 (UTC)


 * Done. Ahasuerus 22:41, 6 December 2012 (UTC)

Proposed merger of publication formats
I propose that the records which are currently given with the formats "quarto", "bedsheet", and "A4" be merged into one group bearing a single name. Also, that "octavo" and "A5" be merged under one name.

Here are the dimensions and the current record count for the first group: "quarto" - 8.5" × 11" (2445 records) "bedsheet" - 8.5" × 11.25" (799 records) "A4" - 210 mm × 297 mm, or 8.27" × 11.69" (1249 records)

The Science Fiction Encyclopedia calls them all "Letter Size". At the most, there's less than an inch difference between the three. Publications given as the first and third terms are almost always saddle-stapled (folded sheets with center staples). I'm not absolutely certain of what is meant by "bedsheet", but it usually refers to perfect-bound publications (cut sheets or signatures which are either side-stapled or glued with a full sheet as the cover hiding the binding, i.e. National Geographic or a 1964 issue of Analog.) The SFE calls it an "erroneously" applied term.

And for the latter group: "octavo" - 5.5" × 8.5" (216 records) "A5" - 148 mm × 210 mm, or 5.83" × 8.27" (1165 records)

The SFE calls this size "Review", a term derived from its use for academic journals, which are usually perfect-bound. Most of the ISFDB records using either "octavo" or "A5" are for fanzines or semi-pro magazines, which are usually saddle-stapled. They appear to be slightly largely than a "digest" (even though the current issues of Asimov's are closer to this size than the standard digest.) Strangely, most of the records for Foundation and Extrapolation are entered as "tp" or "digest".

I realize that most of the non-American world is familiar with the terms "A4" and "A5", but I would wager that most of the ISFDB users are not. So my suggestion to change these terms may be considered by some to be American-centric, even though the terms might make very little sense to the average user. Of course, this could also reflect my own personal point-of-view. But then, the British SFE also avoids using the terms as well. Mhhutchins 18:43, 6 December 2012 (UTC)


 * Indeed, the proposal to replace an international norm (the AX series see here) by a typical american term is effectively simply American-centric. Hauck 21:05, 6 December 2012 (UTC)


 * The term "Letter Size" is not a "typical American term". Please re-read my post to see that it comes from the British editors of an internationally well-respected Hugo Award-winning publication, the Science Fiction Encyclopedia, Third Edition. I knew such charges would be leveled against me as they've been made in the past, so I purposely made a point to address that. So please try to be more constructive and give reasons for or against the change. Mhhutchins 23:10, 6 December 2012 (UTC)


 * The term "Letter Size" may not be a typical American term but it is certainly perceived as such in the UK - the number of times I've found a printer demanding "Letter" paper be loaded, instead of using the A4 it's full of already, is probably a contributor to my high blood pressure. If it's not Americans (or Microsoft alone?) at fault I don't know who is. However, I can live with standardising on "Letter" as A4 is the size of most UK letters. Although confusingly, the UK Post Office now calls that size Large letter and charges extra for postage unless you fold it in two. :-/ BLongley 00:09, 7 December 2012 (UTC)


 * Tell that to your fellow Brits who edit the SFE III. (I guess Clute would be an American as well since he was born in Canada.) Mhhutchins 00:22, 7 December 2012 (UTC)


 * Since the "binding" field is about to be converted to a drop-down list (similar to "publication type"), we won't be replacing one term with another but rather combining them. The end result will look something like this:

+-+ + hc - Harcover books                                                            + + pb - Paperback 7 in x 4.25 in (18 cm x 11 cm) or smaller books                 + + tp - Trade paperback larger that 7 in x 4.25 in (18 cm x 11 cm)                + + ph - Pamphlet (short, unbound, staple-bound or other flimsy binding)           + + dos - Dos-a-dos (back-to-back) paperback                                       + + digest - Magazine 7 x 4.5 in (17.8 x 11.4 cm) or 8.25 x 5.125 in (21 x 13 cm)  + + pulp - Magazine 9.5-10 in x 6.5-7 in (24-24.5 x 16.5-18 cm)                    + + A4/quarto/bedsheet - magazine 11-11.7 in x 8.27-8.5 in (27.9-29.7 x 21-21.5 cm) + + A5/octavo - magazine 8.25-8.5 in x 5.5-5.83 in (21-21.6 x 14-14.8 cm)          + + ebook - All electronic formats                                                 + + audio (CD) - Compact Disk                                                      + + audio (CAS) - Cassette tape                                                    + + audio (DAT) - Digital audio tape                                               + + audio (LP) - Long playing record (vinyl)                                       + + digital audio download - File with digital audio recording                     + + etc                                                                            + +-+


 * The same descriptions can be displayed on the Publication Listing page so that there would be no ambiguity as to what each format means. Ahasuerus 22:13, 6 December 2012 (UTC)


 * However, first we need to decide whether we really want to merge the 5 formats listed by Michael into two. Are any of them sufficiently different to make it desirable to preserve their separate identities? Ahasuerus 22:13, 6 December 2012 (UTC)


 * I don't believe that the country of origin should be a factor in determining what we call the format. If two publications laid side by side are almost identical, within an inch in dimensions, they should be given the same name. The 8" by 11" magazine is the most common standard size. Call it "standard" if you wish, but naming them all "A4" just for the sake of political correctness is not the solution. I don't need to take a poll to know that most users of the database have no idea what "A4" means. Mhhutchins 23:10, 6 December 2012 (UTC)


 * As far as I know, A4 is well known in the UK, but I am not sure how widespread its use is in other countries. However, as long as we list all of the names, in this case "A4/quarto/bedsheet", in the drop-down list and explain what the terms mean, it shouldn't be a problem. Ahasuerus 23:28, 6 December 2012 (UTC)


 * How will they be displayed on the publication record? If they're displayed as "A4/quarto/bedsheet", I don't see the point in merging them. So why not spell out "paperback" and "hardcover" on publication records? Mhhutchins 23:55, 6 December 2012 (UTC)


 * The current plan is to display everything to the left of the hyphen -- "hc", ""pb", "A4/quarto/bedsheet", "digital audio download", etc -- on the Publication Listing page. However, it's easy to change the way this information is displayed, e.g. we can display "Hardcover" instead of "hc", "Paperback" instead of "pb", etc. We can even add a "question mark in a circle" which will show you the gory details about the format when your hover your cursor over it, e.g. "Magazine 9.5-10 in x 6.5-7 in (24-24.5 x 16.5-18 cm)". Ahasuerus 02:43, 7 December 2012 (UTC)


 * As I see it, the big question is whether these formats are close enough to make their merging painless. For example, the "8 by 11" format and the A4 format differ by only 0.25-0.75 inches, but does this difference make it difficult to store them in certain containers/bookcases/etc? Ahasuerus 23:28, 6 December 2012 (UTC)


 * It certainly does. Certain parts of the Stableford collection are being bent out of shape as Brian (and most UK citizens) store them in boxes formerly used for A4 paper. It's one of the few metric standars we have adopted! BLongley 00:16, 7 December 2012 (UTC)


 * So you use the ISFDB to determine what box to use in order to store a magazine? We're talking about merging records, not magazines. Mhhutchins 00:22, 7 December 2012 (UTC)


 * No, I just bung them in A4 boxes. It's about the only size available now, A4 is ubiquitous. BLongley 10:28, 7 December 2012 (UTC)


 * I honestly can't see that as a factor in determining whether the records should be merged. When I need a database to tell me how far apart to build my bookshelves, just put me away. I'm hoping someone will come up with a better reason for not merging them that we've overlooked in our zeal to create a drop-down menu. Mhhutchins 23:55, 6 December 2012 (UTC)


 * I have adjusted my bookcases to fit more paperbacks in - I can get two more shelves in if I restrict myself to pb size. But we've already lost one opportunity to help me (and editors) by separating A, B and C format paperbacks. Two of them are lumped together under "tp" and the logical "hc, tp, pb" sequence of printings can't be relied on any more. :-/ BLongley 00:28, 7 December 2012 (UTC)


 * But I'd bet you didn't boot up your computer to tell you which format a book was in order to shelve them. Mhhutchins 00:36, 7 December 2012 (UTC)


 * No, but if somebody does enquire about one of my books I do check whether it's a tp rather than a pb, as that would mean it's not shelved with the rest of that author's works. I keep one bookcase of outsize overflow books. Well, did keep - almost all my books and bookcases are in storage now. BLongley 10:28, 7 December 2012 (UTC)

[unindent] More to the point: I'm sure there are records in this database of publications categorized as "A4" which differ in dimension by an inch or so, just as I'm sure there are differences in publications for which the records are called "quarto" and "bedsheet". My point is that this difference doesn't make an issue of Interzone sufficiently different from an issue of Locus such that would require them to be categorized differently. Regardless of what we call it, the records should be merged into one format. Is there anyone with an argument counter to that? Mhhutchins 00:36, 7 December 2012 (UTC)


 * I guess all format codes that are based on size only are somewhat arbitrary and, if you examine their origins, based on usage. After all, why draw the line in one place and not some other place? For example, why do we distinguish between "pb" and "tp"? Well, we presumably do so, at least in part, because the 7" (18cm) format is so common that bookstores, grocery stores, libraries and even individuals have specially built bookshelves to accommodate it. Some people even limit their book buying to pbs for shelving reasons. Thus it's important for us to inform our users whether a given book is a "pb" or a larger paperback since some may base their book buying decisions on this information (at least a few people on Usenet have indicated as much.) Ahasuerus 02:08, 7 December 2012 (UTC)


 * This all may be true, but it is outside the scope of this debate. We're talking about magazines, and not paperback books. Mhhutchins 03:15, 7 December 2012 (UTC)


 * Sure, but the same principle applies -- the formats that we use are driven by what publishers put out and how consumers react to them. Ahasuerus 05:15, 7 December 2012 (UTC)


 * So why don't we have more formats for trade paperbacks then? Anything more than 18 cm is such a wide range. Why not make a distinction in hardcovers which come in a wide variety of sizes? I'm saying that the difference of less than a half-inch in a magazine is not sufficient enough to consider it as being a different format. Mhhutchins 06:28, 7 December 2012 (UTC)


 * The issue of capturing different flavors of trade paperbacks has been raised a number of times -- Bill's comment about "B"-vs.-"C" tps above is not the first one. However, for better or for worse, it was first raised relatively late in the process when we already had thousands of tps in the database and it would have been difficult to go back and re-categorize them. Ahasuerus 07:44, 7 December 2012 (UTC)


 * That said, I agree with the gist of your argument, Michael. We don't want to have dozens of formats since it would make the average editor's (and user's) life difficult. On the other hand, we don't want to go to the other extreme and merge formats that are identifiably different. I see it as a question of balance and I think that the people who use these formats on a regular basis are the best judges of where the lines should be drawn. If there is no consensus, I believe it's better not to merge formats until after the dust has settled: it's very easy to perform a merge, but undoing it would be require a very time-consuming manual review of all affected records. Ahasuerus 07:44, 7 December 2012 (UTC)


 * Surely there are other editors who feel the same? Where are you? Mhhutchins 06:28, 7 December 2012 (UTC)


 * It's only been 36-48 hours since this topic became "hot", give them time :-) Ahasuerus 07:44, 7 December 2012 (UTC)


 * And I find it hard to take seriously anyone who makes purchases based on the size of the publication. Mhhutchins 03:15, 7 December 2012 (UTC)


 * The Usenet posters that I had in mind, e.g. David DeLaney, are serious readers/collectors, but space limitations and other issues have forced them to concentrate on certain formats. Ahasuerus 05:15, 7 December 2012 (UTC)


 * The question then is whether any of the formats that are currently documented in our data entry standards can be profitably merged without losing the kind of granularity that users find useful. Based on what Bill said above, "8 by 11" and A4 may be sufficiently different for practical purposes to make their continuing existence as different formats desirable. Ahasuerus 02:08, 7 December 2012 (UTC)


 * So you must be sufficiently satisfied that the publications of all records in the database have been measured to within a half-inch accuracy to insure that they have been placed into the right category. If enough editors truly believe that I'll withdraw the proposal to merge the formats. Mhhutchins 03:15, 7 December 2012 (UTC)


 * Few things in this world are perfect -- and I am generally not in a position to judge since my collection of British magazines is limited to older digests -- but if the editors who have been entering and maintaining these pubs believe that the differences are identifiable and consequential, then I think we will want to keep the formats in their present state. Granted, it will mean that the drop-down list will be over 20 lines long (which is unwieldy), but IMHO accuracy is more important. Ahasuerus 05:15, 7 December 2012 (UTC)


 * You can not say "accuracy is more important" without knowing how accurate the current listings are. I can unequivocally state that I have never measured a publication in all my years working on this database. Mhhutchins 06:28, 7 December 2012 (UTC)


 * Let's see...  A cursory review shows that some (trimmed) pulps are just a tad over 9.25" while untrimmed pulps can be almost 10" long. Come to think of it, this is a good example of a noticeable difference which we have chosen to ignore even though some sources distinguish between "trimmed pulp" and "untrimmed pulp". Which is not necessarily a bad thing as long as it's a conscious decision. Ahasuerus 07:57, 7 December 2012 (UTC)


 * No one can say that all records that give "A4" as the binding is exactly 8.27" × 11.69". I have issues of Interzone that are of varying dimensions and all are listed in the database as "A4". I wish more editors would weigh in on the matter to get a more "accurate measure" of the group's feelings. Neither of the two editors who have commented have actually given reasons why the formats shouldn't be merged, or have given evidence that there is a distinction of more than a half-inch that should preserve the separation. Mhhutchins 06:28, 7 December 2012 (UTC)


 * I am all for applying 'Letter Size'. When verifying issues of 'Interzone' I was aware that the format was changed some time from A4 to a smaller size. I wasn't acknowledging the change because it really didn't feel right to call them 'bedsheet' and I really didn't know about 'quarto'. There are differnet sizes of letters in the various countries but I do think it is really constructive and encompassing to call it 'Letter Size'.
 * A similar case is already fact in the use of 'digest', which really ranges from the almost paperback size to more than the A5 size (as with the booklets common in Europe, or Germany at least, Perry Rhodan being the most prominent example). So, yes I am for the reduction of formats.
 * No one is eager to have measuring instruments at hand when entering or veryfing a publication, and it's much better to have relatively simple rules that most editors can and will understand and use accordingly than a theoretically more exact, but de facto more chaotic system. Stonecreek 10:24, 7 December 2012 (UTC)


 * Thank you Christian. It's good to see that someone gets my point. I don't really care what we call the format. I just believe that there is no feature of either of the three formats that distinguishes them from the remaining two other than a relatively small difference in size. Mhhutchins 16:41, 7 December 2012 (UTC)

[unindent] Quoted from Ahasuerus' last statement (it's way up there but I wanted to respond to it here): "On the other hand, we don't want to go to the other extreme and merge formats that are identifiably different." That's my problem. These three formats are not identifiably different. Even if their dimensions were more than one full inch different, they would be no more different than a 9" hardcover is with a 12" hardcover. Your mention of the different sizes of "pulp" magazines goes further to prove my point. I would concede that point if we could prove that all "bedsheet" records are perfect-bound instead of saddle-stapled. I would even go through the trouble of researching each magazine title and distinguishing those that are perfect-bound. (As far as I know, no title given as "A4" or "quarto" is perfect-bound.) My basic point is that saddle-stapled magazines of roughly 8" x 11" are the same format. Mhhutchins 16:41, 7 December 2012 (UTC)


 * For shelving purposes, I have five different sizes of "hard cover" books. I wouldn't expect them to be listed differently in the ISFDB. So I don't think we should obsess over maintaining detailed descriptions of the paper size for pb's; we should use categories for "about this size", but with substantial clumping, e.g. of "Letter/A4/bedsheet/quarto". I would prefer NOT to standardize on A4, because I think this is too precise a description. It means exactly one size, and I would rather use a word that implies "about this size" (as "pb" and "tp" do). IMHO, "letter" has a vagueness about it, because I would interpret that to refer to any reasonable stationery size that one would use to write a letter. However, Bill Longley's comment seems to imply that this may not be the case -- that "Letter" would also be read by many users to mean "Exactly American letter size". If that's true, than this would argue against its use. In this case, I would argue for "quarto", because that is very specifically an "approximate" size. (At least according to the American Library Association, where specific sizes should be listed, if desired, as "crown quarto", "medium quarto", "royal quarto", "super quarto", or "imperial quarto".) With a "mouseover" tool tip that tells users what "quarto" means, I think its unfamiliarity would not be a substantial problem. My conclusion: (1) We should reduce the number of bindings by using "about this size" as a deciding factor. (2) We should use "Letter" if other editors think (as I do) that this implies a sense of "approximately". (3) We should use "quarto" if other editors disagree with #2. Chavey 23:07, 7 December 2012 (UTC)


 * Speaking as a Brit, I would only use the term "Letter" size when talking about correspondence. It does still have a vaguely American ring to it, although these days that specific size when referred to by Brits is usually called "US Letter", not plain "Letter". I do recall once researching online precisely what sizes quarto and octavo were for the purpose of data entry here, because I had absolutely no idea, and found answers to be frustratingly general rather than accurate to the fraction of a millimetre – I guess I will always be attached to the metric accuracy inherent in "A" paper sizes that the UK uses. Most Brits have no idea what sizes bedsheet, quarto and octavo are, even without all their 'royal' and 'imperial' variants.


 * The drop down list above looks good to me, and I appreciate the greater specifics, although it could be argued that it's still US-centric: my preference would be to see the inch measurements bracketed instead of the centimetre measurements. The US, Burma and Liberia are the only countries that still exclusively use imperial over metric. The UK confusingly still has a foot in both camps but paper measurements are usually in metric ...and we don't have a great many Burmese or Liberian editors here. :-/ PeteYoung 10:05, 9 December 2012 (UTC)


 * I don't have any magazines, so personally I don't care whether the magazine formats under discussion are merged or not. From a general point of view, though, I would agree that fewer formats are better, because less detail usually also means fewer errors (especially in a database maintained not by machines but by people). As for the terms to be used, I quite agree with Chavey's conclusions. "Letter" is ok for me, because although in some contexts I would certainly associate "letter" with "American letter size", I don't think that this would be the case while browsing through the ISFDB. Just my 2 cents, Patrick -- Herzbube Talk 00:50, 8 December 2012 (UTC)


 * I agree with most of the above. The number of formats should be reduced, and I dont really care how we call them, as long as the help text is clear about what to use when. --Willem H. 09:21, 8 December 2012 (UTC)


 * It sounds dangerously like a consensus on 'Letter' to me. Can we agree on that and take the octavo/A5 question to a separate thread? BLongley 02:42, 9 December 2012 (UTC)


 * If we could set aside what to name it, I'm sure the majority of us are coming to the conclusion that the three formats should be merged as one. In response to Pete Young: "quarto" generally means the same dimensions as US letter-size: 8.5 x 11 inches, but like "A4" and "bedsheet" there are slight variations in these dimensions.  All three are used interchangeably within the database already. We have US publications that have been entered as "A4" for no apparent reason (e. g. Marion Zimmer Bradley's Fantasy Magazine even though most of the data is from either Locus1 or Miller/Contento both of which classifies the title as "quarto".) There are a couple issues of Interzone which are typed as "bedsheet". Up until a few days ago, more than two dozen records of Interzone were given as simply "8½ in x 10¾ in". Some issues of Dragon are given as "quarto" while the majority are "bedsheet". There were more than two dozen records for issues of New Worlds which were typed as "large s/s" and I learned from the primary verifier that they were exactly "A4" dimensions. The reason for these disparities within the same title comes down to what one editor felt at the time he/she was creating the pub record or the source they were working with (Locus1 and Miller/Contento generally agree, but Tymn/Ashley has a different set of formats and Galactic Central seems to be a mixture of both. Only the SFE III has recognized these as one format.) There should be a definition for this format that provides the wiggle-room necessary for small deviations. We do it for hardcovers and paperbacks. Why not for magazines? Mhhutchins 17:56, 9 December 2012 (UTC)


 * I do not have magazines with the discussed dimensions (I'd go with a single, approximate measure): could someone explain why these slight differences should be relevant, ISFDB-wise? --Pips55 21:31, 9 December 2012 (UTC)


 * Not logically. :) Mhhutchins 21:50, 9 December 2012 (UTC)


 * I think that merging these formats makes a lot of sense; and I think that mouse-over help tips to assist users when they're choosing a format (as someone mentioned aboye) would be a good thing too. Albinoflea 08:24, 10 December 2012 (UTC)


 * If there's any confusion possible than why not simply include an explaining note in the rulebook that Letter Size does include various formats (American, English and other), and is not restricted to American Letter Size? That should be clear enough, I think. Stonecreek 10:53, 10 December 2012 (UTC)

Date for interior artwork
What date should we use for interior artwork: date, when this art was used for the cover for first time, or date, when this art was published as interior art? Denis 17:35, 10 December 2012 (UTC)
 * I'm not sure I could find a stated policy, but I've always used the date the artwork is first published. Usually, and I can't think of a counter example, that is the date it appeared as cover art.  Though if it appeared as interior art before it was used as a cover, I'd go with that. --Ron ~ RtraceTalk 00:24, 11 December 2012 (UTC)


 * There can be another question. Do we have any variant with another date? Denis 11:13, 12 December 2012 (UTC)


 * Yes. When a work is translated into another language, the record gets the date of the first publication of that translation. Also, although it may not be documented, when a work changes titles it may have a new date for the first publication under that title. Mhhutchins 16:43, 12 December 2012 (UTC)

Nonfiction books about UFOs
It seems right now that for a book about UFOs, either the author must be (somewhat) prolific, or the UFO book in question was reviewed in a SF magazine. I'd like to generate a discussion about if UFO nonfiction books in general be included in isfdb. I have one that I have submitted to be included, but it is being held for approval right now because it was neither reviewed and only one of the authors wrote a SF novel (only one). I would like to get it approved because it deals with UFOs. (It is also the only UFO book I have out of all my UFO books that is not in isfdb. The book in question is The New UFO Sightings by Glenn McWane and David Graham with a foreword by Brad Steiger)  What are the pros? What are the cons?--Astromath 04:01, 18 December 2012 (UTC)


 * UFOs are a subset of what can be loosely called "speculative non-fiction", so the first question that comes to mind is whether you would propose to limit the change to books about UFOs or whether it would cover all speculative non-fiction, e.g. "ancient astronauts", New Age, crystals, "did Neil Armstrong really walk on the Moon" and so on and so forth. Ahasuerus 01:57, 19 December 2012 (UTC)


 * I would argue against including ANY such works. There are far too many books about people who believe in things that are either "speculative non-fiction" or "pseudo-science", and I think we should avoid any such books. Do we want to include all of the books by folks who believe they've been abducted by aliens? Who believe they've talked to ghosts of their ancestors? Who believe they've had a near death experience and saw God? Who believe in cold fusion? Who believe in the healing power of crystals? Who believe in the truth of Dianetics? We do NOT include books about science (except by SF authors "over the limit", or of other particular interest to SF). To include books about "speculative science" or "I believe this is science" in a Science FICTION database seems dangerously close to saying "What you believe isn't really science, so we're going to put it in with the fiction." If the author believes that what they are writing is non-fiction, then IMO it should be treated as if it were, and it should NOT be included here (unless it's from someone over the threshold, etc. etc.). Chavey 02:33, 19 December 2012 (UTC)
 * +1. Hauck 14:58, 19 December 2012 (UTC)
 * I am also for not including those books, with the known exception of an existing review etc. Stonecreek 20:50, 19 December 2012 (UTC)
 * Ok. Let's discuss thresholds.  So far we have a vote for UFO pubs that have an existing review.  What should the threshold be?  I've seen threashold mentioned here and there, but it has never been 100% clearly defined.  Stonecreek's threshold would not only toss out my pub, but many others that's already in isfdb.  If we go by the undefined threshold, then there's a possibility of my pub being included.  It was written by an author that wrote fiction (although only one) and it has a forword by another author that already has UFO non-fiction listed in isfdb.  It looks like, to me, that the threshold is up to interpretation depending on the moderator.
 * P.S. I like Stonecreek's threshold suggestion. It is clear and simple.--Astromath 04:40, 24 December 2012 (UTC)


 * Yes, the threshold is a subjective judgment up to the moderator. And when we moderators aren't sure, we ask other moderators. The threshold, however, is NOT just some number of books published and listed in the database. There are authors here that have lots of books listed, but are not particularly important to the field of SF. For one example, Warren Murphy does not meet the threshold (IMO), because his 79 books are fairly "borderline" SF, and fit more into the "Action/Adventure" genre than squarely into the SF/F genre. We've debated the "threshold" for years, and never been able to come up with an objective statement as to what qualifies. If you want an idea of what qualifies, here's my own opinion (again, every moderator would have their own opinon): An author is over the threshold if they have at least 4 books that are owned by, and verified by, at least 4 different editors. In other words, they have to have published several books that many people think are important enough to own. (But there would still be other authors that would qualify; if Mary Shelley had never written any science fiction other than Frankenstein, she would still qualify.)
 * I suspect that you are right that many of the other UFO books we have listed should not be here because their authors are not over the threshold. If you feel comfortable doing so, I would invite you to suggest some for deletion. Adding books is important to our efforts; but deleting books that don't belong is important as well. Chavey 05:58, 24 December 2012 (UTC)


 * Quite a few UFO books are in the database because they were commonly reviewed in SF magazines in the 1950s-1970s, so please be careful with deletions. Ahasuerus 09:17, 24 December 2012 (UTC)


 * I believe I don't have the authority to do any deletions. I'll leave that up to the moderators.--Astromath 13:34, 24 December 2012 (UTC)


 * Any editor should be able to create a "Delete Publication" submission. When you pull up a pub, it should say "Delete This Pub" under "Editing Tools" on the left. Can you see this option? Ahasuerus 06:32, 25 December 2012 (UTC)

[unindent] Books about UFOs are currently not eligible for the database, just as Westerns and mysteries are not eligible. But there are exceptions. There should be an ISFDB publication record if the work is written by an author "above the threshold", i.e. an author chiefly known for his science fiction (emphasis on fiction). If it is reviewed in a science fiction magazine, there are two options: 1) create a publication record and link it to the review, or 2) change the REVIEW to an ESSAY. I personally prefer the latter to avoid creating records for non-genre authors in the database. About the book in question, are you certain that the author is the same who is  included in the database for his one work of science fiction? It's a very common name. I would be against any decision to change the current rules to allow UFO books into the database regardless of their authorship. As Ahasuerus points out, this would open the door to other pseudo-scientific speculative nonfiction. Mhhutchins 14:57, 26 December 2012 (UTC)

Page numbers in boxed sets
[Copying from my Talk page and adding a comment. Ahasuerus 02:05, 19 December 2012 (UTC) ]:

[Proposal to u]se the page number field of the contents to record the book's volume number. This allows the books in the boxed set to be listed in volume number order. This is what I did when I imported the contents to The Chronicles of Narnia. I supposed the volume numbers in the page number field could be clarified with "Vol" in front of the volume number, but it would look a little messy.--Astromath 20:12, 18 December 2012 (UTC)


 * It's an interesting thought, but keep in mind that proper order is only guaranteed if you use legal Roman or Arabic numerals. Also, page numbers that are not printed in the pub should be enclosed in square brackets, so in the case of a multi-volume boxed set it be would something like "[1]", "[2]", etc. Ahasuerus 02:05, 19 December 2012 (UTC)
 * Information noted and implemented. At least on the boxed set I have.--Astromath 13:35, 24 December 2012 (UTC)

Suggestion for unpaginated e-books
I notice that many (most?) e-books are flagged with a "Missing page count" flag. This is understandable, since they don't seem to have pages. Why not simply give them a page count of 1? This would get rid of a flag that isn't telling you anything that you didn't already know. It might create a problem when the contents of the book is 'created' by assigning pages numbers to stories to get them into sequence, but I don't know if the database checks to see if the highest numbered page listing in the table of contents exceeds the book's page count. If so, and if that is a problem, perhaps you could give the page count as the highest numbered story in the table of contents instead. Sjmathis 16:31, 18 January 2013 (UTC)


 * The problem with giving "1" as the page count is that some ebooks actually do have pages, depending upon on the format and the reader. And the problem with giving it "the page count as the highest numbered story in the table of contents" is that the story goes beyond the page on which it starts. The real problem is not with the ebooks themselves; it's because the program was written for print books, and it hasn't been adjusted to work with electronic books. Personally, if I were able to program, I would have the software disregard the page count field of ebooks altogether and not generate an error message. Just as its programmed to disregard the ISBN field of pre-1960 records. Mhhutchins 17:35, 18 January 2013 (UTC)


 * Good point., which covers missing ISBNs and page counts in e-books, has been created. Ahasuerus 19:37, 18 January 2013 (UTC)


 * (Just did a check and that date has been moved up to 1968. Non-ISBN 1969 books still generate the error. I hadn't realized the date had been adjusted. Thanks, Ahaseurus.) Mhhutchins 17:39, 18 January 2013 (UTC)
 * (Wait. I see some 1969 and 1970 books don't give the error while some do. Strange. So I guess some other factors are involved in generating the error.) Mhhutchins 17:44, 18 January 2013 (UTC)


 * There are two ISBN/catalog checks at this time. The first one warns you if the catalog ID/ISBN field has no value and the book was published after 1971. The second one warns you if the book is *not* a hardcover, was published between 1951 and 1971 and has no Catalog ID/ISBN. The idea behind the distinction is that most paperbacks had catalog IDs in the 1950s-1960s, but hardcover didn't. Ahasuerus 19:37, 18 January 2013 (UTC)


 * Got it. That explains why there's a warning for the 1969 trade pb but not the 1968 hc of this title. Mhhutchins 20:01, 18 January 2013 (UTC)


 * We're getting better. If we can actually agree on some new standards, we could even use this field for the number of cassettes/CDs an audiobook comes on. Or record the total running time, so that we can spot an abridged version from an unabridged one. Or fine-tune the logic for ISBN warnings to allow for the fact that SBNs were around a lot earlier in the UK than in the US. (Yay! We were once a world-leader!) I'm waffling again aren't I? To get back to the original post, "No, the database DOESN'T check to see if the highest numbered page listing in the table of contents exceeds the book's page count." We have issues with unnumbered pages after the story ends, where some people include an "About the Author" page or two and some others don't, or 'excerpts' from further books that may be included by some editors and ignored by others as adverts, or even dos-a-dos books where we've fudged the numbering by putting one side in Roman Numerals and the other in Arabic ones. And the Page count field can't cope with too many digits/letters, so some omnibuses where every constituent novel starts on a huge new page number, or each starts again at "one" just can't be dealt with yet. It's another big can of Gagh (my favourite new word for today) but I hope I've answered your question, while giving you a headache over all the other issues we've got. (Not your problem, of course.) BLongley 08:29, 19 January 2013 (UTC)

"Language" for artwork
Currently in the Mod's submission queue, there are a number of "Make Variants" for Interiorart and Coverart. No problem there. However, the newer title tends to have a Language code of "English" whereas the older one tends to have no language set - which means they're assumed to be "English" by default. Does this actually make sense to anyone? Aren't pictures usually free of language? Or does the fact that Coverart tends to be marred by pesky things like title and author mean "language" is actually needed? Discuss! BLongley 07:29, 20 January 2013 (UTC)
 * If really pressed, I'd say that the language of Interiorart and Coverart is the one of the publication they're in. Hauck 10:10, 20 January 2013 (UTC)
 * After sleeping on this for a bit, I still think "a picture is worth a thousand words" still doesn't imply a language. There are bi-lingual books - e.g., where I couldn't possibly decide what language the cover actually is. (Well, in this case I guess I'd be forced to "English", as we don't actually have "Klingon" as a choice!) BLongley 12:14, 20 January 2013 (UTC)
 * I think we should do with the art what we would do with the illustrated work. In essence, we already do this, since we variant artwork records solely based on the words (and, therefore, language) used in titling them, and those words usually come from the titles of the illustrated works.  --MartyD 12:29, 20 January 2013 (UTC)
 * Good point! Thinking about COVERART variants of COVERART records, the different titles and languages do make sense, if you for instance are recording French reuse of American Art. I'm still mildly dubious about INTERIORART variants of COVERART where the language is the same, but only by default. I guess someone could look into coding a mass update to set the Language for a COVERART record, but that only works when the actual pubs have a language definitively set. COVERART records are a bit of a nightmare anyway: they get created automatically fairly well, but certainly aren't properly updated when the work they illustrate is edited. Something else to discuss, maybe... BLongley 12:52, 20 January 2013 (UTC)


 * For some interior art, the language is a factor, e.g. for cartoons, maps, and captioned art (esp. in children's books), so I wouldn't want to discard "Language", although "Not applicable" (in some form) might be a useful option. Chavey 02:04, 17 February 2013 (UTC)

Title Page for stories/essays
This came up as a result of this discussion. In collections of short stories (or essays), there is occasionally a page preceding the start of the story which contains only the title of the story. This is frequently followed by a blank page with the story beginning on the next page with the title preceding the text of the story. The help page defines page as "The page on which this item of content can be found.". For publications that I've entered, I've used whichever convention is used in the table of contents, i.e. if they give the page numbers for the "title only" page, I use that. If they give the page with both title and story text, I use that. Do others always use one or the other in these situations? Do we have or want a standard as to which page is should be listed in publication records? Thanks. --Ron ~ RtraceTalk 03:27, 24 January 2013 (UTC)


 * I don't know if there's anything explicitly documented about this practice. I usually try to give the page number on which the title appears regardless of whether the text starts on the same page. Mhhutchins 03:44, 24 January 2013 (UTC)


 * I can't think of any documentation. I usually go for 'however it is reflected in the table of contents' like Ron does. (Unless the ToC has misprints on, which is why I treat ToCs as secondary sources.) For the sake of such a small difference, I won't argue - this has just never seemed important to me. BLongley 07:21, 24 January 2013 (UTC)


 * Yes, I agree to Micheal's way of doing it and I would add a note of that praxis in the respective publication. Stonecreek 10:39, 2 February 2013 (UTC)

Capitalization of "or" in titles?
Why is "and" in the should-be-in-lower-case list, but not "or"? Is that an oversight? Should "or" be added to the list? --MartyD 12:25, 27 January 2013 (UTC)


 * Probably. I don't capitalize it, and I also don't recall other editors doing it. Mhhutchins 16:03, 27 January 2013 (UTC)


 * I think "or" shouldn't be "Or" unless it starts a title. BLongley 12:30, 1 February 2013 (UTC)


 * Yes, please: my vote also is for a micro-sized o. Stonecreek 14:09, 1 February 2013 (UTC)


 * I will add it to the list. --MartyD 12:32, 2 February 2013 (UTC)

Putting letters into series
Please, take a look at this. Am I false to have rejected and would my suggestion be out of question? Additional: Shall we clarify the rules on the matter (or are they already clear? - In the help for editing new pub.s there is only mentioning of letters to the editor)? Stonecreek 12:14, 1 February 2013 (UTC)


 * I think the rules are NOT clear. If there's a large enough number to warrant a series, then consider a series for BY the author, and if necessary sub-series for who they went TO. BLongley 12:33, 1 February 2013 (UTC)


 * I personally believe that letters would not qualify as a series at all. It's my understanding that records in a title series share a common element: setting, characters, etc. By the same understanding, magazine columns wouldn't qualify either, but at least in that case there is a thematic connection. We seem to have wedged both of them in while expanding the definition of "series". We do have a penchant for pushing a function to the point where its original meaning disappears (re: variants). This series is another use of the function which I find odious. That is clearly a case of a sub-type of publication (poetry collection), not a series. Mhhutchins 17:33, 1 February 2013 (UTC)


 * Misuse of current functionality has been around since at least the practice of merging Editor records by year to save overlong pages. BLongley 20:41, 1 February 2013 (UTC)


 * But then, that misuse led to the creation of magazine series (based on the editor records) and the database magazine grid display. So at least something better came out of it. (Although it's possible that the magazine grid would have eventually come along on its own.) Now if we could only get editors to stop creating one or two issue annual records, or even worse combining multi-year records into one! Mhhutchins 22:11, 1 February 2013 (UTC)


 * We could probably code things so that multiple-year merges are banned. Raise a feature request or bug and we'll get round to it in the next decade or so. BLongley 22:58, 1 February 2013 (UTC)


 * A software-enforced ban on EDITOR merges spanning multiple years? That's a rather specialized case which I think is best handled by moderators rather than by the software. Ahasuerus 17:57, 2 February 2013 (UTC)


 * If we'd had more developers at the time, we might have addressed the underlying problem, i.e. "some pages are getting too long", with some sort of display changes so that sections of a biblio page show a brief summary with a "MORE" button on the overlong sections. Unfortunately, we seem to be LOSING developers while gaining more editors. :-( BLongley 20:41, 1 February 2013 (UTC)


 * It's not the best documented feature, but I think perhaps tags would be a better way of tracking both the letters and poetry series by a specific author. I wasn't even aware that individual letters are in.  If we are to allow their inclusion, we should probably come up with a standard as to how they are titled, since they don't usually include a title page.  However, that's a separate discussion.  Anyway, is this one of the intents for tags? --Ron ~ RtraceTalk 22:44, 1 February 2013 (UTC)


 * Tags seem to be a bit of an afterthought. I know they've been used to fill in the gaps in the software - e.g. there's lots of tags about awards that are now obsoleted by reenabling Award editing. But so far we've not got very close to deleting obsolete tags, however much the rest of the system has improved. BLongley 23:02, 1 February 2013 (UTC)


 * Tags are generally used to help users search for/identify subjects, e.g. "Mars" or "aliens". I don't think users would want to search for "letters", would they? Besides, Bob's problem was that, as he put it, "real essays were [...] hopelessly lost in the wash of letters", and tags wouldn't help with that. Ahasuerus 03:19, 2 February 2013 (UTC)


 * More generally, I think that regular series, although not perfect, are the best fit for letters. We already use them for "collected works", anthologies and other beasties that do not share settings or characters. Ahasuerus 03:19, 2 February 2013 (UTC)


 * And the other case is also quite common: that the occasional letter is lost in the wash of essays. So, I am quite in support of putting letters into a series. After all they are a type of ESSAY (no doubt about that), but they are a special type of ESSAY also - a type that allows more personal views than a normal ESSAY. Stonecreek 10:44, 2 February 2013 (UTC)


 * I think the organizational benefits outweigh the abuse of the true sense of "title series". What harm comes of placing them in a series?  --MartyD 12:37, 2 February 2013 (UTC)


 * Because each time we stretch a function's purpose, it kills any incentive to create one that better fulfills the demands of the one that has been pigeon-holed into another. (Like making translated works into variants...) The only reason the editor created this "series" of letters is because it cleans up the display on one author's summary page. Otherwise there is no reason whatsoever to group these titles separately from other essays.  What if there were an editor who was as enthusiastic about Isaac Asimov as this one is about Robert E. Howard, and decided that all of Asimov's introductory essays should be placed into a series?  We can't argue with any conviction if we've already allowed such a precedence.
 * Perhaps this situation about making letters into series might have been better handled if a LETTER was a type of publication, such as NOVEL, COLLECTION, ANTHOLOGY, NONFICTION, etc. Each of these types have been categorized to be displayed separately on an author's summary page. (And that was the sole purpose of this editor's reason for creating the series.) Currently we're asked to type a letter as an ESSAY. There had been discussions in the past to create other types to relieve the strain that is placed on that one type. Did you know that the Locus database has more than 30 types which we currently enter as just ESSAY. Maybe that's too many, but I could come up with at least ten solid ones. Mhhutchins 03:22, 3 February 2013 (UTC)


 * IIRC, the last time we discussed the issue of new title types, the thorny issue was whether we wanted to create additional types (which would have their own sections on Summary pages) or whether we wanted additional sub-types similar to the way "storylen" values function as sub-types for SHORTFICTION titles. Would you say that letters are a regular type rather than an ESSAY sub-type? Ahasuerus 08:58, 3 February 2013 (UTC)


 * But as long as we're getting by with stretching functions, there will be very little progress toward creating ones that are designed with a clear purpose in mind. Mhhutchins 03:22, 3 February 2013 (UTC)


 * I think there are two somewhat separate problems here. The first is that age old question of triage and prioritization. Higher priority issues get addressed first and anything that can be handled with a band-aid gets moved down the list. It's not that there is no incentive to improve things, it's just that there are bigger things that need to be done first. In the meantime, if you don't apply a band-aid to a minor wound, it won't be miraculously healed -- it will just continue bleeding while more serious wounds are addressed :-) Ahasuerus 08:58, 3 February 2013 (UTC)


 * What kind of bleeding is caused by allowing works that have been entered as ESSAYS to be displayed along with the author's other ESSAYS? What we're doing is allowing the whim of one editor decide the display of an author's work by manipulating one function. Has there been a great clamor among those of us who have worked on the database for years that such a display/categorization is needed? I haven't noticed any blood caused by the lack of such. Mhhutchins 16:35, 3 February 2013 (UTC)


 * A number of other editors have used series to organize letters, e.g. Letters: James Blish (verified by Swfritter), 's Letters to Crypt of Cthulhu (verified by Grymsayre), Letters by Joanna Russ and Letters by Philip K. Dick (mostly verified by PeteYoung). Ahasuerus 17:52, 3 February 2013 (UTC)


 * The second problem is that we are trying to model that infinite variety of possible permutations that we call "the real world". Since we can't possibly create a data model that would cover all eventualities, we are forced to settle for approximate solutions in borderline cases. For example, we use the concept of "series" to group related titles even though the nature of the relationship between grouped titles is not always the same.


 * As far as the use of VTs for translations goes, I think it was a near-optimal solution rather than a band-aid. It still needs to be expanded to add support for translators and address some display issues, but otherwise it reflects the real world relationship between canonical titles and their translations quite well. We have plenty of other band-aids which still need to be redone, of course. Ahasuerus 08:58, 3 February 2013 (UTC)


 * I'm trying not to have this discussion go more off-topic, but I continue to believe that the entering of translated titles as variants severely set back the creation of a sorely-needed relationship function. Mhhutchins 16:35, 3 February 2013 (UTC)


 * Michael, you are probably right in assuming that this series was created to clean up the display on one author's summary page, Stonecreek 13:04, 3 February 2013 (UTC)


 * It's not an assumption. It's based on the editor's response to your message. Mhhutchins 16:35, 3 February 2013 (UTC)


 * but - I dare to ask - what's wrong about that? It boggles the eye to stare upon an endless list of essays (Okay, in this special example it still will be an endless list of letters). Stonecreek 13:04, 3 February 2013 (UTC)


 * I've explained the reason in my above diatribe. I'm never someone who says "I like it" or "I don't like it" without giving the reason why. Mhhutchins 16:35, 3 February 2013 (UTC)


 * I like Ahasuerus's idea of setting up something like storylength for essays. I don't think it practicable to have numerous, really different types of nonfictional contents. After all, there are numerous examples where essays were drawn from letters: to variant them would be an unfortunate attempt - but it should be possible with the subtype idea, or am I wrong? And it would pay some respect to letters being something special. Stonecreek 13:04, 3 February 2013 (UTC)


 * And I as well very much like the idea of setting up sub-types. But if everyone starts using the series function to separate types, it will slow down (de-prioritize) efforts to actually create them. To use Ahaseurus's metaphor, not only do such band-aids slow down the bleeding, they slow down the healing too. Mhhutchins 16:35, 3 February 2013 (UTC)


 * The thing about sub-types -- at least the way they are currently defined in ISFDB -- is that they do not affect Summary pages. Think of the way "novella", "short story", etc appear on individual Title pages: it is just an extra line of information displayed at the top of the page, which doesn't affect the way short fiction titles are displayed on the Summary page.


 * Having said that, it may be advantageous to use this information on the Summary page in some fashion. Perhaps a User Preference to display "[nv]", "[ss]", etc next to short fiction titles? Ahasuerus 17:59, 3 February 2013 (UTC)

Non-Genre Round Number Something or Other
Mhhutchins is insisting that "the NONGENRE category is restricted to NOVELs and no other type". When I pointed out that the policy doesn't say that, he directed me to this discussion. However, that discussion: I don't care what the decision is, but I don't believe we should be giving editors direction that is contradicted by our policy. If we are going to limit non-genre to novels and non-fiction, then the policy should be updated to specifically state that. Our conventions can be confusing enough without them being directly contradicted by our written policy. I recognize that software issues play in a larger role in this than perhaps pure policy desires, but that can be reflected in the policy (as DES did in his one suggested re-wordings). -- JLaTondre (talk) 23:34, 8 February 2013 (UTC)
 * 1) is only about non-genre shortfiction; and
 * 2) is not actually reflected in the policy.


 * In this case, software limitations play a more important role than policy, stated or not. If I only gave advice based on written policy, there would be more issues than just this one. It may have taken me a few hundred thousand moderations, but I think I've pretty much figured out how this database works. I'll leave the documentation to those want to play around with the wiki while I'm over on the database side doing the heavy lifting. Mhhutchins 23:48, 8 February 2013 (UTC)


 * The software doesn't work well with non-genre collections, but it does allow them. We have many records in the database that have a publication type of collection or anthology and a title type of non-fiction non-genre. It causes display issues, but otherwise works. You may wish to consider that if our wiki instructions and our database practices matched, then you would hopefully have less work on the database side. Not everybody reads the instructions, but many do. -- JLaTondre (talk) 00:17, 9 February 2013 (UTC)


 * The problem I've had with non-genre (stuff that's not novels) is that we then lose links from a publication back to its title. That's not just "display issues", that's a serious problem. It would be nice to have a software solution. But until then, I think adding a title note to a non-fiction book, collection, or short story that they are non-genre is the best "work-around" we currently have. And then, assuming we eventually get full "non-genre" support, we would be able to easily find such works to correct them. Chavey 02:23, 17 February 2013 (UTC)
 * I consider the missing title references an inconvenience and not a serious problem (the title record can be accessed via the author page, search, etc.). If it's a serious problem, that gives weight to the argument of not including them, but you seem to indicate the opposite. As for shortfiction, the previous conversation shows general support for not including non-genre shortfiction, but the proposed language wasn't incorporated into the documentation. -- JLaTondre (talk) 12:56, 18 February 2013 (UTC)


 * I still do think that NONGENRE collections, nonfiction, anthologies and magazines shouldn't and needn't be part of this problem. It is really helpful to put them in under their type but list only the contents of speculative interest. Otherwise we would really slip into a discussion of treshold: exactly when is a COLLECTION Genre? We really should avoid this topic because it is irrelevant, I think. Stonecreek 13:30, 18 February 2013 (UTC)
 * I'm sorry, but I'm not following you. They should shouldn't be part of what problem? Since the basis of the originating question is "how are non-genre collections to be handled?", I'm struggling to understand how they cannot be part of the question. Given a collection unambiguously containing no speculative content, by an author that is unambiguously above the threshold, there are only three possibilities:
 * Don't allow it
 * Allow the collection, but not the contents - what the previous discussion recommended
 * Allow the collection & the contents - what the current rules & documentation state
 * Given the last two, then there is a follow-on question as to whether/how the entries should be marked, but we haven't even resolved 2 vs. 3. -- JLaTondre (talk) 17:23, 18 February 2013 (UTC)

Suggesting slight change in definition of the CHAPTERBOOK type
The current wording on the help page (under "Publication Type") states "This format is primarily used for separate publications of a single work of short fiction, even if bound as a standard paperback or hardcover, or in any other format." The "primarily" is misleading in that it may suggest there are other uses for the type. (A recent editor misunderstood it to mean that a stand-alone essay publication could be entered as a CHAPTERBOOK.) The only other use I can think of is the publication of a single poem. Should not the wording be made clearer? Mhhutchins 01:50, 11 February 2013 (UTC)


 * What would we do with standalone publications of "in" essays that also appear in other publications? Full disclosure: I perpetrated this chapterbook once on a piece I really struggled to categorize and ended up settling on essay for.  --MartyD 03:44, 11 February 2013 (UTC)


 * If "Eureka" is an ESSAY, the publication of it should be NONFICTION. There are many examples of standalone essays entered as NONFICTION, like here, here, and here. Mhhutchins 03:59, 11 February 2013 (UTC)


 * And standalone poems entered as CHAPTERBOOK, like here, here, and here. Mhhutchins 04:09, 11 February 2013 (UTC)


 * Thisis the essay publication which prompted the discussion. Mhhutchins 04:11, 11 February 2013 (UTC)


 * Well in this case I'd say it clearly is NONFICTION since it contains another ESSAY. And in the meantime I have come to terms with the use of CHAPTERBOOK for SHORTFICTION (or a single POEM). So, yes, the wording should be specified, I think. Stonecreek 05:59, 11 February 2013 (UTC)


 * If NONFICTION should be used on any book not containing fiction/poetry and is allowed to have true content records the way the other containers are, I don't see any problem with clarifying the CHAPTERBOOK wording.  --MartyD 11:40, 11 February 2013 (UTC)

Clute/Nicholls Verification
I don't recall seeing this answered before. But does a mention in the online 3rd edition of The Encyclopedia of Science Fiction merit a Clute/Nicholls verification when the mention is not present in the 2nd (paper) edition? The same question could be asked for the revised edition of Currey. My own opinion is that these newer updated editions should still count for verifications. --Ron ~ RtraceTalk 18:17, 17 February 2013 (UTC)


 * There's nothing that precludes it, but the descriptions of the verifications would have to be changed if we decide there's no reason to keep the editions separate. It appears that someone has already added SFEIII to the Clute/Nicholls description. I personally feel they should be kept separate because of the tremendous difference in the content and delivery between the last printed version and the current online version. Also SFEIII is not Clute/Nicholls. It is Clute/Langford/Nicholls/Sleight/et al. And because of its ongoing and everchanging nature, a reference today could possibly disappear tomorrow. (I've made contributions that have changed the data in the online version, so I'm aware of its ephemeral nature.) Any reference to the SFEIII could be made in a link to it from the title record so creating a new verification for it seems unnecessary. This actually can make it just as valuable as for example a Tuck verification, because all the user would need to do is just click on the link. There was even some talk earlier about creating a dedicated link to SFEIII, similar to the current link in title records to Wikipedia. Mhhutchins 19:07, 17 February 2013 (UTC)


 * I'd vote for keeping the dead-tree version separate. That makes it a finite task that we may actually complete some day year. As to links to SFE3, I'm in favour of having author links, which should remain fairly stable. BLongley 20:43, 17 February 2013 (UTC)


 * I agree that the 2nd edition and the 3rd edition are very different beasts and we shouldn't be using the "Clute/Nicholls" slot to record SFE3 verifications. I suggest that we remove SFE3 from the description page. Ahasuerus 00:04, 19 February 2013 (UTC)

Reviews of magazines and non-genre material
Discussion moved from Community Portal, existing thread copied over. New comments begin with the first "Unindent":

In verifying an issue of The New York Review of Science Fiction, Greg Cox has an essay on "Vampire Fiction" in which he reviews 7 books or stories. One review, of The Vampires, by John Rechy, notes that the book has nothing genre in it, other than the title, and says it "is included here solely because of its title". Our help page says of reviews like this that "Non-sf works should be entered". So I created a review of that title, added a title record (as "Non-Genre"), and did not include any of the publications of that book under that title record. Seems reasonable to me, but that means (1) the title record might get reported by some data validation scripts; (2) the author (linked above) now exists in the ISFDB with only a single work, that work being non-genre. Was this the right way to handle this review? Chavey 05:57, 12 April 2013 (UTC)


 * That seems to be the wrong approach to the situation, in my opinion, even though the current rules may allow it, depending upon which of the conflicting standards you follow. (I follow the one that says you don't create records for non-genre works by authors who aren't above the threshold.) I guess I'm a purist, but when a publication would not otherwise qualify for inclusion based on the current policy, creating a title record for an obvious non-genre title opens the door for other such works and creates a precedent. In these cases, I just create an ESSAY type record which covers all the bases, and I recommend that approach to others. You wouldn't be expected to create a REVIEW record for a movie, or a music recording, even though it appeared in an sf publication. What about all of those reviews of stage performance that have been published in the past few years in NYRSF? So why should this case be different?  And what's going to stop an editor from coming along and adding all of the publications for this non-genre novel? Mhhutchins 06:12, 12 April 2013 (UTC)


 * It seems that this approach (which I think I agree with) would require changing the current help pages. The current NewPub help page description of Reviews says:

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. 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.


 * A possible replacement for this would be (new text in red, removed text struck out ):

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


 * Is this an appropriate change? Should this conversation be moved to "Rules and Standards"? Chavey 07:35, 12 April 2013 (UTC)


 * I'd move magazines to the ESSAY section - due to the merging of editor records these are actually MORE likely to be unlinkable. BLongley 15:26, 12 April 2013 (UTC)


 * Agreed, as I'm having to working my way back converting several fanzine/magazine reviews into ESSAY records. Also it wouldn't hurt to give an example or two, for newer editors, on how to properly phrase an ESSAY record for a review of a below-the-threshold non-genre title, or film, fanzine/magazine etc. PeteYoung 16:57, 12 April 2013 (UTC)


 * Yes, Darrah, this should be moved to the Rules & Standards discussion page. But I like the changes you suggest, and like Bill, believe that magazines should be moved to the essay part of the instructions. And yes, Pete, there should be a standard titling in the conversion of a review to an essay. We'll go into that once the discussion has been moved. Thanks to all. Mhhutchins 19:24, 12 April 2013 (UTC)

(Unindent) So, with other modifications, the previous proposal is updated as below. I have also strengthened the phrase "don't need to" to "should not", as this seems to be in keeping with the discussion so far:

Note also that only genre books, magazines, and short fiction are entered as reviews ; if the column reviews magazines, fanzines, or non-genre material , you don't need to should not enter the review records for these, only the ESSAY record s. 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.


 * I think as it is worded here, there may be a possibility of an editor thinking he/she should not enter the review at all, and that's not the point of what we (or rather I) was trying to say. There should be a distinction made between which type of entry should be used for the review, e. g.:

Reviews of spec-fic titles (and non-genre titles by spec-fic authors above the threshold) should be        entered in the Reviews section of the entry form which automatically types the content record as        REVIEW. All other reviews, those for non-genre titles by non-genre authors, graphic novels, films, and other performances (whether audio, video or staged) regardless of their genre, should be entered into the Content section of the entry form using the ESSAY type.


 * Mhhutchins 23:42, 14 April 2013 (UTC)


 * The proposed change looks OK from the substantive standpoint, but there are a number of if-else-and-or clauses, which can make it inherently hard to make the guidelines unambiguous. Perhaps we could break it further down along the following lines:

Reviews of the following types of titles should be entered in the Reviews section of the entry form: 1. Speculative fiction titles 2. Non-genre titles by speculative fiction authors above the threshold Note that reviews entered in the Reviews section of the entry form will be automatically assigned the title type of REVIEW. All other reviews should be entered into the Content section of the entry form using the ESSAY type. This specifically applies (but is not limited) to reviews of the following types of works: 1. Non-genre titles by non-genre authors, or by authors below the threshold 2. Magazines or fanzines (even when spec-fic) 3. Graphic novels 4. Films 5. Other performances (whether audio, video or staged) regardless of their genre.


 * ? (Note that in one case I used the word "works" instead of "titles" because films and performances are not really titles in the sense that we use the term.) Ahasuerus 04:23, 15 April 2013 (UTC)


 * Yes, I think that's even better and hopefully prevents readers from just reading the first sentence(s). Stonecreek 08:32, 15 April 2013 (UTC)


 * Some additional suggestions in red. Chavey 22:07, 15 April 2013 (UTC)


 * Looking better with each iteration. Thanks, guys. Mhhutchins 00:07, 16 April 2013 (UTC)


 * I'm curious... the sense I get from this wording is that reviews from the second list of items should always be entered. For some publications reviews of this type are more numerous than those in the first group, and in those situations I will omit entering these reviews and make an explicit note about the omission in the note field. I assume that would still be acceptable? If so how best to change the wording?


 * Also, for example, if there is a section of film reviews as opposed to a stand-alone film review, is the expectation that only the top-level section or column name receives an ESSAY entry, or that each film would receive its own separate entry as we do for items from the first list?


 * Seems like we're moving in the right direction; this is an area that has tripped me up a few times, so the clarification is much appreciated. Albinoflea 21:29, 17 April 2013 (UTC)


 * Often, a single column (listed as an essay) contains several book reviews. Those reviews of genre books should be pulled out as additional review entries, so that the same textual material is actually being referenced by two items in the "Contents" listing. (That, I suspect, is why those reviews are indented more than the standard items.) That same review column may review items on the second list, but those items should not receive such special treatment. If a column is titled "Recent Movies", then that would be entered as an essay, but no special treatment of the individual movies reviewed would be included. If the essay was "A review of Avatar", then it should be entered as an essay by that title, but again no separate "review" record should be created. Chavey 17:27, 18 April 2013 (UTC)

Standardizing spaces in names
Following up on this discussion in the Community Portal, I propose that we specify that names of authors be standardized with respect to spaces in the name when they are entered into a bibliographic record. Specifically, I propose that we change the template giving the rules for Author Names as follows. I have shown some of the text around the proposed location of this rule for purposes of context. The changes proposed are in the new Spaces in Names rule:


 * Author - The name of the author of the publication. The name should be entered exactly as it actually appeared in the publication. This includes pseudonyms, abbreviated names ("I. Asimov" instead of "Isaac Asimov", "Robert Heinlein" instead of "Robert A. Heinlein"), etc.  As with the title, take the name from the title page in preference to the cover or spine of the book.
 * Case. Case should be regularized...
 * Initials. Initials should normally be entered followed by a period and a space ...
 * Spaces in Names: Spaces within a name should be regularized. If a name differs from a canonical name or existing pseudonym only by the lack of, or addition of, blank spaces, it should be entered as the existing name or pseudonym. For example, a book credited to "Ursula LeGuin", "Lester DelRey", or "A. E. VanVogt" should be listed by including the missing space, e.g. "Lester del Rey". Conversely, if a book were credited to "John De Chancie", the extra space should be removed, and the book credited to "John DeChancie". One effect of this rule is to avoid subjective judgements when there appears to be a "partial space" in a name.
 * Spaces in Names: Spaces within a name should be regularized. If a name differs from a canonical name or existing pseudonym only by the lack of, or addition of, blank spaces, it should be entered as the existing name or pseudonym. For example, a book credited to "Ursula LeGuin", "Lester DelRey", or "A. E. VanVogt" should be listed by including the missing space, e.g. "Lester del Rey". Conversely, if a book were credited to "John De Chancie", the extra space should be removed, and the book credited to "John DeChancie". One effect of this rule is to avoid subjective judgements when there appears to be a "partial space" in a name.

Chavey 05:20, 4 July 2013 (UTC)


 * Hearing no objection, I'll add this to rules for names. Chavey 02:41, 17 July 2013 (UTC)


 * No objection here. I would have shown approval earlier but didn't catch this when it was first posted, and took a week off afterwards. I believe there are cases in the db where pseudonyms and variants have been created. When I come across them, I'll follow this standard (but will notify editors when there is a primary verification). There will be some obscure cases where establishing the canonical name may be difficult. But those can be dealt with as they arise. Thanks. Mhhutchins 03:00, 17 July 2013 (UTC)

Currency information in prices, reprise
A question on an editor's talk page pointed out that this discussion: Rules_and_standards_discussions/Archive/Archive09 died out without incorporpation into the help. Since that was 2.5 years ago, I am raising it again, and I will follow up on the final conclusion.

The proposal is to standardize recording of prices:


 * Period is used as the decimal separator.
 * Currency symbol is placed before the number.
 * A symbol may be a glyph (non-letter), and should be used where available: $, £, €, ¥, etc. Do not include a space between these and the number.
 * A symbol may be a set of one or more letters: DM, Fr, Lit, Lt, zł, etc. Include a space between these and the number.

Now that Ahasuerus has converted all "L" to "£", I also suggest we strike the help text that says to use "L" if you don't know how to make "£".

So, for example: $1.85, £1.85, €1.85, DM 1.85, and zł 1.85.

Please comment. If you already voiced your opinion in the earlier discussion and haven't changed your mind, no need to repeat yourself. --MartyD 11:36, 29 July 2013 (UTC)
 * Note that there is ISO (4217) standard set of abbreviations used for currencies (mainly in the financial sector see here), I think that we should use it. Hauck 17:38, 29 July 2013 (UTC)


 * I strongly disagree about changing currency symbols to this standard of three-letter abbreviations. I believe we should adopt the standards that Marty has proposed. Mhhutchins 18:29, 29 July 2013 (UTC)


 * The ISO standard is for currency codes, not currency symbols. See, for example, http://www.xe.com/symbols.php.  The ISO standard also does not cover currencies that disappeared prior to the standard's adoption.  Of course, symbols run afoul of ambiguity (you can see from the list, even our A$ and C$ attempts at disambiguating are themselves ambiguous), but I suggest we leave dealing with currency code vs. currency symbol for another day.  This is really just meant to clarify placement and punctuation of the currency symbol and the number.  --MartyD 18:54, 29 July 2013 (UTC)
 * I was just speaking as a Frenchman and for others in the situation of not having a symbol for our old currency. In this case, an abreviation seems to be in order and the "Fr" abbreviation is really hurting my eyes as you generally find "F" on books (when present of course). As there are numerous other "Francs" (from Switzerland to Madagascar), the ISO seemed quite pertinent and having the big advantage of pre-existing (and thus freeing us of the task of devising our own set of abbreviations) but I've found here a certain resistance in using true international normalizations (cf. the paper size debate) in favor of the local anglo-saxon usage. Hauck 05:56, 30 July 2013 (UTC)


 * Three letter currency codes have become somewhat more popular lately, but I think we are better off using the symbols that are actually printed on books/magazines since that's what most people are familiar with. To give one example, "GBP" would look most peculiar if we entered it next to a "3/6" price.


 * As far as disambiguating popular symbols goes, that's a whole 'nother can of worms. We currently use disambiguating prefixes for "C$", "A$" and "NZ$" (which is how they are usually printed), but Help doesn't say anything about disambiguating "£" ("Lebanon pound" in Lebanon) or "$" when it is used to mean "peso" or "colon" in Latin American countries. Ahasuerus 09:18, 30 July 2013 (UTC)

(unindent) Sounds like the best approach to me. I believe it should be adopted and then documented in the Help pages before we forget again. Mhhutchins 17:14, 29 July 2013 (UTC)


 * Everything considered and well done. Thanks! Rudam 17:32, 29 July 2013 (UTC)


 * Agree completely. There was also discussion about 4 digit numbers, e.g. $1500.00 vs. $1,500.00 This comes up occasionally in US & British currency, but quite regularly in Italian and Japanese currency. There are, IMO, two reasonable options: Use a comma for all numbers 4 digits or more (e.g. $1,500.00), or use a comma for all numbers 5 digits or more ($1500.00, but $10,000.00). My personal preference is the former. Chavey 22:33, 29 July 2013 (UTC)


 * The Help should also list/note how to enter the unusual symbols £, € etc. [I still can't figure out how to do the Polish 'l' or the Yen symbol]. Not every user is going to know how and copious links to Help pages that are written for more knowledgeable people really don't "Help", as has been proven a few times quite recently. --~ Bill, Bluesman 16:29, 30 July 2013 (UTC)


 * As a courtesy, we could give a link to a page which gives ALT codes, e.g. here, but I don't see why it's necessary to include them in our help pages. (The page I link to even has a video tutorial, for those less knowledgeable people.) The Yen symbol (¥) is ALT+0165. About Polish and other non-Latin alphabets, I'm assuming someone who is entering books published in a those alphabets should know how to enter them using their keyboard. The ISFDB help pages is not the place to educate them. There are too many alphabets for us to give tutorials on how to enter them on a standard keyboard, if that's even possible. (I know this will bring out the usual accusations of ethnocentrism. Fire away.) Mhhutchins 01:14, 31 July 2013 (UTC)


 * It seems likely that a reference page would help editors working with secondary sources. We could either link to a third party site or create a separate Wiki page similar to ISFDB:Foreign Language Abbreviations. Or perhaps both in case the third party site(s) go away like so many of them have over the years. Ahasuerus 01:48, 31 July 2013 (UTC)


 * (Slightly off-topic) On Windows, there is the "Character Map" application (Start -> [All] Programs -> Accessories -> System Tools -> Character Map), where you can select any special character without needing to know the Alt sequence (and it tells you the Alt sequence, if there is one). And on any computer, one can copy symbols from another page (such as the one I cited above).  Those two pieces of information would probably be sufficient for most ISFDB contributors.  --MartyD 11:18, 31 July 2013 (UTC)


 * For Darrah's comment about thousands separators "period for decimal" would then imply/dictate comma (or nothing) for thousands. So how about some more wording to that effect: "Period is used for the decimal separator and comma(s) for thousands separator(s) in any number greater than 999." (to be word-smithed appropriately)?  --MartyD 11:18, 31 July 2013 (UTC)

Dating publications
Since this situation has arisen with a couple of new editors lately, and has been discussed with others in the past, I propose that the Help section for the Year field contain the following addition/amendment:


 * If the publication has a stated month and year of publication (e.g. "Published in September 2006"), the Year field of the ISFDB record should reflect that date exactly ("2006-09-00") even if a secondary source, such as Amazon, is able to provide the day of publication (e.g. September 21, 2006). You have the option of adding the day-date in the Note field, providing its source.


 * If the publication has only a stated year of publication (e.g. "Published in 1976"), you have the option of refining the date in the Year field if there is a reliable secondary source. This source should be credited in the Note field.


 * If the publication has no stated date of publication, you have the option of providing the date in the Year field if there is a reliable secondary source. This source should be credited in the Note field. If you are certain that the publication is a first (or only) edition, the copyright year can be used, but only when all secondary sources have been exhausted. Without a secondary source, 0000-00-00 should be entered into the field, which is displayed as "unknown".

I believe this is the de facto standard, but has never been fully documented in the Help pages. Any suggestions, corrections, clarifications? Mhhutchins 15:20, 30 July 2013 (UTC)


 * I like this clarification, it guides the editor through the decision making process while clearly indicating the additional documentation required in various situations. Albinoflea 18:45, 30 July 2013 (UTC)


 * I agree. I suggest going one step further, just to be clear: If publication states a full date of publication (day, month, and year), that date should be used.  Do you think it's worth a general statement of the ISFDB's date-capture "goal"?  I.e., we would like to have full dates on daily or weekly periodicals, month and year for all other publications; a publication "day" is not necessary for anything published less frequently than weekly. ?? --MartyD 11:26, 31 July 2013 (UTC)


 * Yes, I agree. That would be a good idea. I had a discussion recently with a new editor who insisted on entering the newsstand date for a magazine when there is a stated month only. I would have added the possibility of the statement of an exact day of publication to the above scenarios, but I was only thinking of book publications and have never seen a day-date on a book. In the case of periodicals, there should be a separate statement of concerning publication dating. Thanks for the input. Mhhutchins 16:41, 31 July 2013 (UTC)


 * I'm not certain that I am understanding the proposal correctly, or if I am I don't think I agree. Assuming that we have a reliable source for the full date, as I read the first bullet, we should never add the day (day of month) to the year field for a publication, when the year and month is present in the publication.  Whereas the second bullet implies that we may add the month and day of publication in the year field when only the year is present.  The third bullet allows us to add the full date when no date is present.  What seems arbitrary to me, is having different standards on whether we may add the day of month, based on how complete the date is as printed in the publication. RtraceTalk 01:43, 1 August 2013 (UTC)


 * It isn't arbitrary if we outline how each of the three possibilities are handled. There is a single standard for each of the three possibilities. How one handles one possibility should have no bearing on how one handles the other. Arbitrary implies that the choices are random or at an editor's whim. That's why the choices have been delineated, to make sure that we're all following a standard based on a publication's statement (or lack thereof). Mhhutchins 05:34, 1 August 2013 (UTC)


 * Ideally we'd have separate fields the stated date of publication and the actual date of publication. The only way I can see to have a single field server for both is to allow the actual date of publication trump the stated date and to document it in the notes when this occurs.  I'll also mention that having the actual date, accurate to the day can be useful in trying to determine which publication is the first edition or printing. --Ron ~ RtraceTalk 01:43, 1 August 2013 (UTC)


 * We've always had the option of correcting a publication statement when there is a reliable secondary source, but a month or so difference between the "appearance" of a book and its stated date of publication is not great enough to correct the statement. Book publishing standards have always allowed us some leeway when recording the publication date, because we know they almost always "appear" in book stores before the publication date. (Now if it's more than a three or four months, I see we have the option or even imperative to make the correction.) We can't go back and change thousands of records based on date of appearance. Imagine all of the magazines that appear months before the stated date. Since we don't have two fields, my proposal is documenting the de facto standard of how the Year field is being used in the db. Mhhutchins 05:24, 1 August 2013 (UTC)


 * While I'm in general agreement with Mhhutchins' proposal the choice of wording caused a bit of mental stumbling on my part. I suspect Rtrace also stumbled.


 * If I'm constructing a new publication record then I follow Mhhutchins' practice. However, if I'm primary-verifying a record that was previously constructed from Amazon or other sources I usually don't zero out the day field. Instead, I check Amazon and if the date matches what they have I add a note such as "the publication states 'Published in September 2006' and that the source of the 2006-09-17 date used for the ISFDB publication record is amazon.com." There have also been times where I ran across a record with a date and I can't figure out the source. In that case I usually add a note that the source of the date is unknown but is assumed to be accurate. If the existing previous date does not seem plausible I'll zero it out and add a note explaining what the old date was and my reasoning. I always document in the notes what the publication states in terms of its printing date and number.


 * My practice assumes people read the notes to understand the content of the Year field. Mhhutchins' proposal is likely better from a machine readable database and reporting perspective.


 * FWIW, Amazon's date is nearly always what publishers call the "On Sale By" (OSB) date. It's the date book stores are allowed to put the title up on the shelves for sale. Sometimes you will see an OSB near the end of a month and the publication states a printing date of the following month and year. If the publication states "September 1976" and the OSB is in August then it's more likely I'll update the publication record to have 1976-09-00 and add a note. I can't say I do that 100% of the time though. --Marc Kupper|talk 02:50, 1 August 2013 (UTC)


 * Correction to the last paragraph. I thought about it and realized that if the OSB is for the month previous to what's stated in the publication then I'll update the publication record to use a yyyy-mm-00 that matches what's stated. Mhhutchins' proposal allows adding extra detail from secondary sources for publications that have no date or state the year only and says you must not add extra detail if the publication states both a year and month. I'd object to that stated prohibition. I'm fine with someone that fills in the day of month from a secondary source as long as the publication record's YYYY-MM matches what's stated in the publication. Using that, if the OSB is in late August 2006 and the publication states September 2006 then ISFDB should have 2006-09-00.


 * In thinking about if I've seen a day date in a book I recall seeing one on the last page of a print-on-demand book. --Marc Kupper|talk 03:27, 1 August 2013 (UTC)

[unindent] My proposal for clarification follows the ISFDB standard to record the date of publication in the Year field as stated in the pub. If a pub states "August 2013", but Amazon says it was published on "July 30, 2013", the editor should enter "2013-08-00" in the Year field. The editor has the option of recording the Amazon date in the Note field, but not in the Year field.

The proposal also allows the possibility that there is no specific month stated in the publication. This "opening" allows the editor to do further research to determine if there is a reliable secondary source which gives a more specific date, whether it be for a month or a day. The editor then has the option of recording that month or day in the Year field, but must Note that the information is from a secondary source and not stated in the publication.

If there is no stated date of publication, the proposal outlines how it should be handled, following the basic steps as explained in the second possibility.

These are three different and distinct possibilities, without the chance of ever overlapping, as far as the book-in-hand is concerned. Creating records entirely from secondary sources is not part of this scenario, because it is (or should be) ISFDB policy that the editor enter into the Note field the source for his information, whether that date of publication is for a year, a month, or a day. In that case we don't know what is actually stated in the book, otherwise we wouldn't need to use a secondary source and would be following one of the three possibilities of dating a publication from the primary source.

I think what may be puzzling some of you is the first scenario. You may be asking "If there is a reliable secondary source that gives a day date then why should we settle for the stated date which only gives the month?" In the example I gave above, if you used the Amazon date, the ISFDB displays the publication month as July 2013 while the book states it was published in August 2013. I've seen too many submissions by editors who want to "correct" the dates of records based on Amazon dating, while other editors come right behind them and correct the records based on the stated date. I need documentation that I can point to and say "Here is the policy." Marty was dealing with an editor's use of Amazon dating just a few days ago. And when another editor I was working with started doing the same thing, I thought the documentation needed to match what he and I have been telling these editors. Anyone who isn't handling new editors' submissions may not be aware of the situation. Mhhutchins 04:41, 1 August 2013 (UTC)


 * You're supposition on what is puzzling (for me, at least) is mostly true. However, I wasn't considering cases where the the actual publication date, from the secondary source predates a stated publication date.  In those cases, I agree with your proposal, though I'd add:
 * "Dates from secondary sources that are earlier than a date stated in a publication should not be entered in the Year field, but may be added to the notes. This should be followed regardless of how much of the date is stated in the publication."
 * That would eliminate problems when the secondary source date is from an earlier year.
 * The part that still escapes me is that when the publication date from the secondary source is simply more precise than the date stated in the publication, we are proposing different standards depending on how much of the date is stated. I have handled adding books where the date is preloaded from Amazon in exactly the same manner Marc states (and have for years). I feel like this proposal is asking me to change that.  If I have a book-in-hand and there is already a record entered from Amazon has a date of 2013-07-15, this proposal seems to require that I leave the date as-is if my book has only 2013, but that I change it to 2013-07-00 if my book says July 2013. --Ron ~ RtraceTalk 12:07, 1 August 2013 (UTC)


 * Your proposed change creates an exception to what in my initial proposal is a simple rule, and only adds another level to the editor's decision. My proposal falls into the ISFDB standard of recording data as is. And only makes exceptions if the publication month or year is not stated, which also goes along with the ISFDB standard of recording additional data when it isn't stated in the primary source.


 * If the date of the secondary source is "simply more precise" then why use the statement of the primary source at all. What this boils down to is accepting data from a secondary source over that data stated in the primary source. It has been the ISFDB standard to only do this when there is obviously incorrect data in the primary source. A book stating "August 2013" as the publication date is not incorrect if it were available a few weeks earlier in July, or wasn't available until the last week of August. As I pointed out above, using street date over the stated date would lead to chaos in magazine records. Or do you propose that another exception be made for that? Mhhutchins 19:36, 1 August 2013 (UTC)


 * I actually do think periodicals should be handled by stated date alone and I find it jarring when I see a seasonal issue (e.g. Summer 2013) appearing in the middle of the generated grid. My objection applies only to books.  I don't understand the logic of a rule that states that when I have a reliable secondary bibliography with the full date of publication (e.g. 2013-07-15) whether or not I can add the day month depends on whether the date in the book in hand is July 2013 or simply 2013.  In the former I may not, in the latter I may.  In neither case, do I remove the data as stated in the book, and in both cases I am making that date more precise.  However, the proposed rule only allows me to make it completely precise when no date or only the year is provided.  If the year and month is provided, the day of the month must be excluded.  If simplicity is one of the goals of the rules, then surely not having an exception for books where the month is stated would be better.  Simply: "Enter the date as stated in the publication.  If you can enter more of the date based on a reliable source without changing the parts of the date from the book, do so in the year field and note your source and the parts for which it is used.  If the date from the secondary source disagrees with any part of the date stated in the book, it should only be reflected in the notes."  That is the de facto standard that I have been operating under and publications that I have entered have been done in that manner. I'd just prefer to understand the reason for the different rule when month is reflected.  --Ron ~ RtraceTalk 00:30, 2 August 2013 (UTC)


 * I've explained before the reason for not adding the day when only the month is stated in the book. Here I go again. The street date may be in the previous month and the ISFDB will display it as published in the previous month, even though the book has a stated month of publication which will not match the ISFDB record (if the editor chooses the street date or the stated date.) And the whole idea of using a street date is nonsense when it depends upon what street you live on! (Some stores receive stock earlier than other stores.) I've had my say, you've had yours, and this discussion will fizzle out as most usually do without any firm decision having been made. Nothing new, nothing changed, except the frustration level somewhat higher. I give up. Mhhutchins 01:28, 2 August 2013 (UTC)


 * Maybe it's simple enough to not try to exclude adding a day, but to emphasize that neither Amazon nor any other book seller is a reliable bibliographic source of date information. So dates from those sources should not be used in the presence of dates from a more reliable source.  I think if we define a "complete enough" publication date up front: Year + Month for non-periodicals, Year + Month + Day for periodicals (say), then it could be pretty simple: Only augment "complete enough" dates with information from reliable secondary bibliographic sources (which does not include Amazon or Barnes & Noble); more precise and conflicting date information from other sources may be recorded in the notes.  Other sources may be used to augment incomplete dates, as long as the information from those sources does not conflict with any date the publication does state.  And then have some exception for demonstrably/known-to-be wrong dates.  --MartyD 11:13, 2 August 2013 (UTC)


 * Overall, no objections to Mhhutchins' proposal except for the first bullet. I feel it should be:
 * If the publication has a stated month and year of publication (e.g. "Published in September 2006"), then the YYYY-MM part of the Year field must reflect that date ("2006-09-00"). If a secondary source, such as Amazon, provides the day of publication within that month (e.g. September 21, 2006) then you may record the full date, 2006-09-21, as long as you document in the Notes that the YYYY-MM part came from the publication and the DD is from Amazon. If a secondary source reports a YYYY-MM that conflicts with what's stated in the publication then you may document this in the notes. The conflicting data should not be entered in the Year field.
 * That allows people to fill in additional detail from secondary sources, as we have always allowed, while shutting down the practice of using a secondary source in lieu of what's stated in a publication. --Marc Kupper|talk 04:25, 21 August 2013 (UTC)


 * I agree with the Mhhutchins / Marc Kupper recommendations. I suggest adding a phrase like "The date in the book always trumps dates from other sources with respect to the formal publication date." Chavey 17:12, 21 August 2013 (UTC)

Unusual suffixes
Valancourt Books has been doing a good job reprinting horror/weird classics. The latest addition to their list is John Palmer, Jun.'s The Haunted Cavern: A Caledonian Tale. According to the publisher, it's a "“tale of shrieking spectres and bloody murders” (The Critical Review) featuring “gloomy caverns, dismal dungeons, ghastly skeletons, pale ghosts, fierce combats, [and] horrid murders”", so presumably it's "in" unless we later discover that the ghosts and specters are fake.

The question then is what should we do with the suffix? Should we keep "Jun.", which is used by the publisher, or should we change it to "Jr."? Ahasuerus 21:50, 7 September 2013 (UTC)


 * I'd go with it as presented by the publisher under the "as it appears in the publication" doctrine. Punctuation normalization is difficult enough for people to get right, but once you know the rules, it's not so bad.  For text, however, one would (a) have to know the "standard" text and (b) recognize the text being used is both non-standard and which standard text it is a variation of.  IMO, too much burden on editors and moderators.  --MartyD 10:50, 8 September 2013 (UTC)


 * And I'd say normalize it to "Jr.". It's an ISFDB standard to normalize suffixes in other cases. If it were "Junior", or "Senior", we'd abbreviate it. If it didn't have a period, we'd add it. If it were uncapitalized, we'd capitalize it. We add commas before "II" and "III", although stories and books are hardly ever credited that way. (How did that ever become standard?!?)Mhhutchins 21:39, 14 September 2013 (UTC)


 * I agree with Mike. The rules for standardizing suffixes are in the last section of the "author field" template (included in many help pages). It's pretty explicit that variations of "Junior" should be regularized to ", Jr.". For details, see here. Chavey 03:53, 15 September 2013 (UTC)

Compilation of Interview Excerpts
I have added the book "Dark Thoughts: On Writing" edited by Stanley Wiater. It is a compilation of excerpts from interviews conducted by Wiater over the years with 50 different authors. It is organized by topic rather than by interviewee, meaning the excepts from any one author are interwoven with others and scattered throughout the entire book. My thought is to add 50 separate interview entries to the record. Any suggestion on how I should title each entry, such as simply "Dark Thoughts On Writing"? Nowickj 20:04, 14 September 2013 (UTC)


 * Personally, I would urge you not to go through the effort. But that's just my opinion. There are no rules that I'm aware of which prevents you from doing it. But if you choose to do so... About the titles, since they're not separate interviews and individually titled, you'd have to make it clear that they are untitled. It would have to be something like "(Stephen King)" and "(Dan Simmons)". The parentheses indicate that this is not the actual title. You would have to explain the format of the interviews in the record's Note field, and you would leave the page field of the content record blank. Mhhutchins 21:37, 14 September 2013 (UTC)

Advertising images as interior art
Spurred by [this] discussion [the latter part]. I never even considered that anyone could be so far off base as to treat images in advertisements as interior art, never mind then varianting such. While, as Michael points out, there is nothing expressly forbidding such, there ought to be. The one noted record itself stretches the line of acceptable contents to the limit [and way beyond as far as I'm concerned] but to add advertising images as interior art is ridiculous in the extreme. Unfortunately, this page is the only means we have to apply the virtual bitch-slap the perpetrator so desperately needs. [Sugar-coating just ain't me ...] --~ Bill, Bluesman 00:59, 17 October 2013 (UTC)


 * As the verifier in the example you give, I can assure you that the images you object to are not advertising. They illustrate the essay on the Ballantines beginning on page 20. There is no content entered in that record that could in any way be considered advertising.  I have to say I am a little dismayed at your desire to have my moderator's flag "shredded" or you assertion that I need to be "bitch-slapped".  While tying to have a civil civil discussion with you regarding what content should be included in a book that we both own, you eventually accused me of overwriting your verifications admitting that your belief that I had done this affected the tone of your argument.  I've denied your accusations and to date, you've chosen not to respond to my denials.  I've made no changes to the publication in dispute.  I'm not sure how to deal with someone who holds a grudge with me for an act that I didn't commit and then goes on to mischaracterize content I've added in other pubs.  I'm not even sure what the point is of misrepresenting the content.  I agree that advertising should not be allowed.  The content you cite is not advertising.  --Ron ~ RtraceTalk 02:08, 17 October 2013 (UTC)


 * As far as verifications go, once a pub has been marked as primary/Currey/Worldcat/etc-verified, it should be impossible for another person to "unverify" it or to claim the slot. Only the original verifier can remove his/her verifications -- or at least that's how the software is supposed to work. If you find a way to undo another person's verifications, please report it as a bug. (Edit: The change that made it impossible to remove another editor's verifications went live in February 2013.)


 * I would also suggest that we refrain from using terms like "bitch-slapped" in the future. That way lies trouble. Ahasuerus 02:58, 17 October 2013 (UTC)


 * Creating content records for advertising is prohibited as stated here. There is nothing that forbids creating interiorart records for book covers which illustrate an essay, but I would warn others that it's a slippery slope. I've always made it clear that I believe we've allowed too much leeway in the creation of content records. But I still believe a primary verifier should have the last word in the matter. Sadly, there's always the possibility of conflict when there are two primary verifiers coming from opposite ends of the discussion. Recently I have approved submissions which Biomassbob has made to add interiorart records for incidental works in issues of Fantasy Review that each of us have primary verified. It sort of bugged me at the start, but it wasn't worth arguing over. (And believe me, we've had our share of arguments.) You choose your battles and you move on. If something isn't outright forbidden, why take the chance of losing an active editor?
 * On a personal note, I find the tone of Bill's accusations to be overly hostile, and hope that any further discussions would be carried out in a more civil manner. Looking over his original discussion with Ron was somewhat upsetting, especially considering Ron's appeal was more than polite and followed ISFDB rules of etiquette. There appears to be an underlying hostility that should be cleared up before it becomes a rift that could cause greater damage. Both Bill and Ron are valuable to the ISFDB and losing either over such a minor matter would be devastating. I would ask that Ron step aside in the case of the publication record created by Bill which was the subject of the original discussion, and that Bill allow Ron to create records to the minutest detail as allowed under the current rules.  I know "first come, first served" isn't an ISFDB policy when it comes to primary verified records, but in order to keep this argument from growing bigger, I hope that it will be considered in this case. Mhhutchins 04:35, 17 October 2013 (UTC)

Leading Illustrations in Publications Other than Magazines
I have tended to follow the rule for leading illustrations in magazines for other types of publications, that is, citing the page number of the leading illustration as the first page of the story or article. even if the title of the story or article appears on the following page. It is difficult for me to find a justification for have different standards for magazines, and say convention program booklets or anthologies where each story has an up-front illustration. I know that many of these latter publications will cite the title page in the ToC, ignoring the illustration, but then so will some magazines. For me, the illustrations are important, and should be considered part of the story or article. And there is value in a consistent policy. Any thoughts? Bob 17:38, 21 October 2013 (UTC)

Bilingual books
We have a modest number of bilingual books, e.g. a language learning book with English & Spanish versions of "Dr. Jekyll and Mr. Hyde". We do not have explicit policies as to how to handle such books, but I believe that we have, at least, implicit rules about such books. I have (tentatively) added the following policy to the Help page on How to enter foreign language editions. I move this for adoption as a formal policy, recognizing that others will probably be able to improve on aspects of the wording.


 * Bilingual books, e.g. one with parallel texts in two languages, essentially contain two different titles: One in each of these languages. These should be treated as two separate titles, each merged with the respective single-language title record of that work (where available), and (generally) one title record being a language variant of the other. Since such books contain two works, a bilingual novel would be listed as an Omnibus, and a bilingual novella, novelette, or short-story would be listed as a Collection. The Omnibus or Collection should be assigned the language of the canonical version of the story included.

Note that this policy does NOT specify which "language" should be assigned to the Omnibus or Collection, and it might be useful to specify this. Personally, I favor using the "non canonical" language for the combined book on the following argument: Usually, the original language has substantially more publications in that language than there are translations, so it's the translation that becomes most interesting. Of course if both contained works are assigned to titles reflecting their language, the language assigned to the work is probably unimportant. But we should probably still decide. Chavey 04:54, 9 December 2013 (UTC)


 * The omnibus/collection approach sounds like a reasonably compromise or at least I can't think of a better way to handle multilingual books. However, I am not so sure about using the "non-canonical" rule because some books contain more than one translation. To use an extreme example, consider the Heineman edition of P. G. Wodehouse's "The Great Sermon Handicap" -- a single volume can contain a dozen+ translations of the same story. It seems like using the language of the canonical title is a safer bet. Ahasuerus 05:24, 9 December 2013 (UTC)


 * That's a good point. You've convinced me. Chavey 06:42, 9 December 2013 (UTC)


 * I have included Ahasuerus' recommendation in the proposal, in red. I have updated all of the books in the database which contain the word "bilingual" in their notes to meet this proposed standard. (Only 2 had been verified: One needed no changes; the other was from an editor no longer active.) See Ghost Fever / Mal De Fantasma as one example. Chavey 00:00, 10 December 2013 (UTC)

Anthologies with No Editor and Few Authors
Our current rules are fairly clear that when a publication contains stories by more than one author (unless all stories are collaborations) that it should be considered an anthology. That would indicate that we have this publication incorrectly classified as a collection. Assuming that we change this to an anthology, this other rule states that for anthologies, we should "use the name of the editor". Since no editor is credited for this work (at least for the FPCI printing), that would leave us with an anthology edited by "uncredited". I believe that this would make this particular publication more difficult to find. I'd like to propose that the rule be changed for anthologies so that if no editor is listed and the the authors are listed on the title page, those authors may be listed in our records. Thus, in the case cited above, we would still have A. E. Van Vogt and E. Mayne Hull listed as authors of an anthology. Thoughts? Thanks. --Ron ~ RtraceTalk 00:07, 15 December 2013 (UTC)


 * I believe the de facto standard, at least for the recent flood of urban fantasy/paranormal romance anthologies, has been to use the authors as "editors" if no editor is specified, e.g. see this anthology or the half dozen anthologies listed on ' Summary page. I agree that we should document this practice in the Help text.


 * As far as the collection-vs-anthology issue goes, I seem to recall that at one point some editors argued that a book that includes stories by only two authors would be better classified as a collection, but, unfortunately, I don't remember the details. Ahasuerus 00:44, 15 December 2013 (UTC)
 * That's also my position, IMHO the collection format is more pertinent when dealing with "couples" of authors (AEVV/Hull, Pohl/Kornbluth, Kuttner/Moore) even with texts credited to only one author (partcurlarly when credits are imprecise and/or erroneous).Hauck 10:40, 15 December 2013 (UTC)


 * I agree with Hauck on this one. I read somewhere that it is believed van Vogt contributed to this wife's stories in this collection and possibly vice versa.  IMHO this one should remain a collection. Bob 23:05, 17 December 2013 (UTC)


 * It appears that we have agreement to the question of how to handle an anthology with no discernible editor. I'd suggest that we change existing help text to read as follows (added text in bold):

Editors, authors, translators, etc. If the book has a known author (or authors), use that name (or names). If it is an anthology, use the name of the editor. If no editor is credited or the editor cannot be determined, the names of the authors included in the anthology may be used. If the book is a single-author collection or omnibus, but has an editor, as occasionally occurs, the editor does not appear in this field, the author does. An omnibus should show the names of the authors of all included fiction. There is currently no support in the ISFDB for translators, or photographers; this information should just be entered in the notes field.


 * Please feel free to suggest alternate wording if desired.


 * I wasn't really asking whether to change the existing definitions on anthologies and collections. To me, that seems a larger change to our standards and I'll start a new discussion below with a heading that more clearly describes the proposed change.  My hope is that it will attract a wider participation in the discussion. --Ron ~ RtraceTalk 19:22, 29 December 2013 (UTC)

Disambiguating cover art
I've been doing quite a bit with cover artists the last few months. One time-consuming hassle is checking that a cover art title merge is correct when an artist has done more than one cover for the same title, eg. Joey Hi-Fi for Lauren Beukes's Moxyland, 1 and 2. While the DO NOT MERGE messages that we add are of course helpful, once it's been established that there are two different covers for the same title by the same artist it may be worth adding a disambiguation to the title record for the second cover, such that the records would read "Cover: Moxyland" and "Cover: Moxyland [2]" on the artist's Summary Bibliography. This would automatically prevent a merge possibility when checking for duplicate titles, thereby saving diligent editors several valuable minutes. The possible downside to this is that when adding a new pub edition/impression, an editor would have to add the "[2]" to the new record before doing a title merge of that particular cover, but I think it will save us all time in the long run. So, people, useful suggestion or bad? PeteYoung 16:35, 20 December 2013 (UTC)


 * I'd say quite useful (should save time in the long run). Stonecreek 17:03, 20 December 2013 (UTC)


 * Would it not also make an editor, confident that any different work had already been disambiguated, believe that a new record should be automatically merged without actually checking the image? It doesn't actually change the procedure that should be taken before a merge is submitted, i.e. checking the image. I can't see how disambiguating makes this procedure any easier. An editor would still have to check each image, and doing a search for duplicates, he would never find the one which has been disambiguated. Mhhutchins 22:43, 20 December 2013 (UTC)


 * Perhaps when performing an advanced search for COVERART or INTERIORART the images (if present) could be displayed in the results; likewise for the Title Merge confirmation page (tv_merge.cgi)? Albinoflea 01:58, 30 December 2013 (UTC)

Definition of Collection and Anthology
In a discussion on a separate matter (above), some editors suggested that our definitions of Collections and Anthologies be altered. As I understand the suggestion, titles containing only stories written by one or another or both of two authors who write in collaboration should be classified as a collection and not an anthology. The current rules state precisely the opposite: "a collection of stories by Frederik Pohl and C.M. Kornbluth would qualify as a collection. However, if such a collection included one story by Frederik Pohl alone, it would be regarded as a single-author collection of Frederik Pohl's. If it also contained one story by Kornbluth alone, it would be an anthology." Personally, I'm not in favor of such a change. I think it is difficult to come up with a standard that can be easily explained. I also worry about the slippery slope. Should books by three authors who collaborate be considered a collection? (e.g. ). If we decide to change the policy, we should come up with new wording for the help pages for both ANTHOLOGY and COLLECTION and should provide new examples for both. Actually, the example cited in the definition for ANTHOLOGY should be replaced in any case, as it doesn't fit the description (there are not any Willhelm stories in the collection). I'd suggest the book that caused me to ask the original question but that will depend on the outcome of this discussion and whether or not it remains a collection. If we do change the definitions, we should probably provide examples for authors who collaborate and those that don't. --Ron ~ RtraceTalk 19:22, 29 December 2013 (UTC)


 * I think any standard that requires the editor (and moderator) to have knowledge beyond the publication at hand should be avoided, if possible and practical. I don't have any complaints about the current classification scheme, but if it were to change, I'd prefer to see the standard based solely on the aggregation's contents.  Something like: If it primarily contains works by one collaboration of authors, it should be a COLLECTION credited to the collaboration even if containing works by subsets or supersets of the collaboration.  If it primarily contains works by one author, it should be a COLLECTION credited to that author, even if containing works by collaborations between that author and others.  All other cases should be classified as ANTHOLOGY.  The spirit would be: What's the aggregation's focus, the collaboration or the individual author, and do all of the stories fit within that focus?  --MartyD 01:14, 30 December 2013 (UTC)


 * I wish I had the time to go through the Help pages. The section you point out is absolutely illogical. Here is how it is currently stated:


 * COLLECTION. A single-author collection. See notes under ANTHOLOGY, above. Note that the "single-author" can be a collaboration: e.g. a collection of stories by Frederik Pohl and C.M. Kornbluth would qualify as a collection. However, if such a collection included one story by Frederik Pohl alone, it would be regarded as a single-author collection of Frederik Pohl's. If it also contained one story by Kornbluth alone, it would be an anthology.


 * I have an idea what the author (Mike Christie) was trying to say, but the way it is worded is almost the opposite of its intended meaning. (Also, a definition of a word shouldn't include the word it is defining.) Surely, everyone would agree that it should be corrected. I suggest:


 * COLLECTION: a publication containing two or more works of less-than-novel-length fiction by a single author. This definition also includes collaborations which are considered a collective author, i.e. both authors are credited on the publication's title page, and the contained stories, for the most part, were written in collaboration. For example, a publication of stories credited to "Frederik Pohl and C. M. Kornbluth" would be typed as COLLECTION, even if it contained a story which was written by only one of the authors individually.


 * This is closer to how the term is defined by its current usage. To answer Ron's question about whether a publication of stories by a three-author collaboration should be typed as COLLECTION, I would have to answer "yes." (Although, the book he cites as an example is not a true collaboration. If all of the stories were credited to REH with or without a collaborator, then it would be a COLLECTION.) An example of "subsets" of collaborations to which Marty refers, I offer this publication typed as COLLECTION. Mhhutchins 06:00, 30 December 2013 (UTC)


 * I forgot to add that the definition of ANTHOLOGY, and the examples provided, is just as bad. Mhhutchins 06:12, 30 December 2013 (UTC)


 * I do agree, but would like to emphasize more to which author(s) the COLLLECTION should be credited: I'd suggest that the author(s) credited in the publication should be considered, so even if a publication would contain nine stories by Pohl & Kornbluth and one by Pohl alone, but it is credited to Pohl alone on the title page, we would mirror this in our entry. Stonecreek 14:19, 30 December 2013 (UTC)


 * I doubt that would ever happen, but I agree that if it did the ISFDB record's author credit should reflect the title page. "Stories by Frederik Pohl" and "Stories by Frederik Pohl and C. M. Kornbluth" are different, but only determine the record's author credit, and not the type of publication. Both would be typed as COLLECTION based on their contents and not their title page. This publication is typed as a COLLECTION based on its contents, despite the fact that the title page is credited as "Damien Broderick with Barbara Lamar and Paul Di Filippo". It has eight solo stories by Broderick and one each in collaboration with the other two authors. Currently only Broderick is credited for the publication. Although the current discussion is about how a publication is typed, and not how it's credited, I suppose that would be a tangential matter. Mhhutchins 19:29, 30 December 2013 (UTC)

Template:AuthorFields:WebPage
Template:AuthorFields:WebPage currently says:


 * Web Page - If the author has an official Web site, enter it here. The full URL should be listed, including "http://", e.g.:

http://www.meganlindholm.com http://www.jasperfforde.com


 * If there is more than one official site -- including author-authorized sites, publishers' pages for the author, etc -- enter the one with the most comprehensive bibliographic and biographical data. If the author is deceased and has no official or authorized sites then alternate web sites with significant bibliographical and biographical data may be entered. If you need to enter additional Web pages, click on the "Add Web Page" button and a new "Web Page" field will appear.

I propose that we change the text along the following lines:


 * Web Page -- Enter the URL of a Web page about this author. Eligible pages include:
 * Author- and fan-run sites
 * Biographic and bibliographic pages from online encyclopedias
 * Author-specific pages provided by publishers
 * Author-specific pages at social bibliography sites like Librarything and Goodreads
 * Reviews and interviews should be avoided unless there is no other online information about the author. Bookstore pages should be avoided. If you need to enter additional Web pages, click on the "Add Web Page" button and a new "Web Page" field will appear.

What do you think? Ahasuerus 17:43, 15 January 2014 (UTC)


 * Looks good. I would suggest moving the last sentence (starting with "If you need...) to a separate instructional point. And it might be a good idea to point out to editors to not link to Wikipedia articles in this field but in its own separate field.  Mhhutchins 19:00, 15 January 2014 (UTC)


 * That's a good point, although I should mention that I am currently in the process of implementing FR 361, "Merge Web pages, Wikipedia links and IMDB links; enhance their display". Once the feature has been fully implemented, the current Wikipedia and IMDB fields will be going away and multiple Wikipedia links (EN-, DE-, FR-, etc) will be supported. Ahasuerus 21:19, 15 January 2014 (UTC)


 * How about including, as one of the options:
 * * Obituaries from authoritative sources.
 * Chavey 03:50, 16 January 2014 (UTC)


 * I personally don't like such webpages linked to the summary page. There are several editors who add such links from "findagrave" and newspaper obits. I believe such links should be provided in the wiki bio page as a source for the data, not as a webpage about the author and his work. Mhhutchins 04:17, 16 January 2014 (UTC)


 * I tend to view obituaries as "sources of last resort", similar to reviews and interviews. If there is no other information available, at least online, they are better than nothing. Ahasuerus 02:48, 17 January 2014 (UTC)

(unindent) I have updated the Help page based on the discussion above. The obituary issue is still outstanding. Ahasuerus 05:14, 18 January 2014 (UTC)


 * I have no objection to adopting the philosophy that an obituary should usually be considered a source for the biography wiki page. Chavey 07:01, 19 January 2014 (UTC)


 * I mildly disagree. How is an obituary different from, say, a publisher's author page?  An obituary is always about the author and may spend some significant time on the author's works and their significance.  I think unless we want to have stricter limits on website links (e.g., to the author's site and/or blog, to fan-run sites dedicated to the author, and to designated on-line reference sites: Wikipedia, IMDB, SFE, ...?) we should try to be inclusive.  As I write this list, I am wondering why we have Wikipedia and IMDB links.  Not that I don't want them, but rather whether the thinking that leads to their being considered valuable to have might not apply to other sites/pages.  Or maybe we should have stricter limits, and everything except the author site(s) and fan-run site(s) should be relegated to the bio page.  --MartyD 11:57, 19 January 2014 (UTC)


 * I have no objection to linking obituaries to an author's summary page just as long as they are substantial and speak as much about their work as about their death. That's probably a judgement call for the editor, so I'm not proposing the removal of all such links. I just think the editors should be more discerning about what they link to the ISFDB. (Findagrave? No.) Plenty of obituary links have turned out to be to newspapers which have subsequently closed off access to the pages and available to subscribers only. I've had to remove those links when found. While I don't see much value in the IMDB links, I'm not going to suggest they be removed. But I strongly believe that Wikipedia pages are valuable and should remain linked on the summary page, not the bio page. Sometimes they're the only outside webpage available and the source for much of the author data fields. Mhhutchins 18:41, 19 January 2014 (UTC)


 * It's been my experience that obituaries rarely add much to our knowledge of an author's SF. If the deceased was a major author, then better sources typically exist. OTOH, if he or she was a minor author or better known for other things, then the obituary writer is unlikely to spend much time on the SF angle. Of course, there are exceptions to every rule.


 * However, obituaries can provide biographical information which may not be otherwise available. They can also be useful when looking for heirs who may own the rights to the author's work. On balance, I see them as "second tier sources" and would prefer them mentioned in Help, but I don't feel strongly about it.


 * As far as "newspapers which have subsequently closed off access to the pages and available to subscribers only" go, I wonder if the data may still be available via Wayback Machine. Ahasuerus 23:20, 19 January 2014 (UTC)

Ebook prices
Michael and I have been exploring the wonderful (or perhaps not so wonderful) world of ebooks for the last few days and I thought I should share some of the findings.

Consider the ebook edition of 's Let the Storm Break. Its ISBN is 9781442450462 and it is sold by B&N and Amazon.com (but not by Amazon UK, at least not yet.) Amazon.com doesn't display the book's ISBN, but if you search on "9781442450462", the right page will come up.

So far so good. However, one thing that puzzled me when I first reviewed the data for this ISBN was that B&N's listed price was $10.99 while Amazon.com's was $9.78. The latter appeared to be some kind of discounted price, but there was no indication of what the "list price" was. It said "Print List Price: $17.99 Kindle Price: $9.78 You Save: $8.21 (46%)" at the top of the page, but that's not something that we would be interested in. So I ran a few more Google searches and found the publisher's page for this edition, which said "$10.99 List Price". In other words, it looked like in this case B&N was using the list price while Amazon was not. Additional searches suggested that although in some cases Amazon used the same "list" price as B&N, in many other cases it used a discounted price and there was no indication of what the "list price" was, e.g. Pillar to the Sky is $11.04 at Amazon and $12.99 at B&N.

The bottom line appears to be that at this point B&N is a more reliable source of list prices than Amazon. When B&N offer a discount, they state what the original list price was, e.g. this book is on sale ($12.49), but the list price ($21.99) is clearly stated. Amazon does the same thing with regular books, but not with ebooks.

The question then is whether we want to:

(a) use B&N's list prices as definitive and enter them in the "Price" field for ebooks, or

(b) leave the field blank and enter B&N's price in the Note field

? (And yes, I realize that UK books are a whole different headache.) Ahasuerus 20:58, 30 January 2014 (UTC)

P.S. Now that I have spot checked how Amazon UK handles ebook prices, it looks like it's the same story on the other side of the pond. For example, their Shadowmaster record says "Kindle Price: 	£5.59 includes VAT", but doesn't indicate that the list price is £6.99 as per the publisher. Other online stores like Hive and The IndieeBook Shop do show list prices, so they are better sources for our purposes. Ahasuerus 22:52, 30 January 2014 (UTC)


 * I'm all for finding the most reliable secondary sources available. If that means a little data from here and a little data from there, and you're able to program Fixer to create a single publication record, then I'm not going to argue where it came from. If you can confidently say that B&N has the actual list price, I say go for it. Mhhutchins 00:07, 31 January 2014 (UTC)


 * I concur. --MartyD 12:51, 31 January 2014 (UTC)


 * It certainly looks like B&N has the more reliable data, and that we should go with them whenever possible. Special cases where we cannot use B&N, or it's not viewed as reliable, would then be commented on in notes. Chavey 15:08, 31 January 2014 (UTC)


 * What do we do when the price changes... and the price alone changes? Kevin 17:23, 31 January 2014 (UTC)


 * Well, we are generally only interested in the "list price", i.e. the price listed by the publisher. If the list price of, say, the ebook edition of Damnation Road Show is $5.99, then it doesn't matter that B&N will sell it to you for $5.49 and Amazon.com only asks $4.61. As long as the stated list price is "$5.99", that's what we enter. The problem with Amazon.com/UK is that some of their e-records, including Damnation Road Show, show what it calls "Digital List Price", but others don't, while B&N is much more consistent.


 * And BTW, Fixer has been updated -- the pub record linked above was automagically generated a few minutes ago. Ahasuerus 04:02, 1 February 2014 (UTC)


 * I'm more stuck on the situation where the book is not 're-published' with a new date, but rather the list price is simply changed. What you verify today, can (and will) be verifiably different at a later date. (That later date may be years away, but it's already happened to Baen Books. Only because they had an announced date (15 December 2012) do I know when their backlist increased from the historical pricing to new pricing. They began selling single books for $4.00. That increased to ~$6.00 over 12 years IIRC. Now all new releases are priced depending on format. When the cheaper format comes out... the price is supposed to drop to the lower format price. Eventually (inflation/deflation/alien invasion) those prices will need to shift again as a jump step. I guess I'm questioning... do ebook prices need an associated 'date'? Kevin 17:19, 1 February 2014 (UTC)


 * Oh, I see. Right, that's exactly why Fixer adds notes like "Price from Barnes & Noble as of 2014-02-01" to e-books. Perhaps it should be changed to "List price from Barnes & Noble as of 2014-02-01", though.


 * And BTW, I have run into a case where Amazon was a better source of list price data than B&N. Consider the ebook edition of Alienated by . Our record says "$9.99" because that's what B&N claims. However, the publisher's site says "ELECTRONIC BOOK 9781423186991 $16.99", which matches what Amazon says about this book: "Digital List Price: $16.99, [...] Kindle Price: $9.99". So in this case Amazon's data is clearly better than B&N, although, of course, the Amazon API doesn't tell Fixer what the "digital price" value is. Most unfortunate.


 * Hm. Perhaps a better order of checks should be as follows:
 * Fixer's Amazon API
 * Amazon's Web page (only if it states the "digital list price")
 * B&N's Web page
 * I'll see what I can do... Ahasuerus 17:41, 1 February 2014 (UTC)

Change to help text for page numbers
I'd like to change this help text to change the third bullet to say
 * "Sometimes a publication will have unnumbered pages before page 1. If there is any material in these pages which needs to be entered as part of the contents of the book, you may record this by entering the count in [brackets]. For example [6]+320 would be a publication with six unnumbered pages and then 320 numbered pages. There is no need to record these unnumbered pages if they contain no content that needs to be recorded.  At times you will need to count backwards from the first numbered page to see which is page 1 and then would count the unnumbered pages that are before this. Likewise, you may record the count of unnumbered pages at the end of a publication. For example, [6]+320+[4]. As before, only do this if there is content in these pages that requires entry.  When counting pages in hardcover and some trade paperback publications do not count the endpapers. These can be identified as they are usually a stiffer paper than the standard pages."

The bold text identifies the new information. Let me know if there are any objections. Mike Christie (talk) 20:59, 3 February 2014 (UTC)


 * Wonderful. I agree wholeheartedly and have no objection to the clarification. There would need to be a single proviso added: this wouldn't apply if the pages are Roman numbered which should be included in the page count even if they only contain matter which wouldn't require a separate content record (title/copyright/dedication pages and tables of content.)
 * This latter is required because most authoritative sources like the LoC and OCLC include such pages in their records. A discrepancy between our record and the others would raise unnecessary questions for primary verifiers. Thanks. Mhhutchins 23:58, 3 February 2014 (UTC)


 * Good point. The prior bullet point in that help file covers Roman numerals, so here's a revised version of both that takes that into account.
 * "When a book has a section with Roman numeral page numbers for introductory material, followed by Arabic numerals for the main text of the book, enter the page number as e.g. viii+320, where viii is the highest numbered page with a Roman numeral page number. Pages without numbers that fall between the two types of page numbering can be ignored. Note that you should include the enumeration of the pages in Roman numerals even if there is no material that requires a separate content record (such as an introduction or preface) in those pages.  This is in contrast with the situation with unnumbered pages prior to page 1; see the following bullet point for what to do in that case.
 * "Sometimes a publication will have unnumbered pages before page 1. If there is any material in these pages which needs to be entered as part of the contents of the book, you may record this by entering the count in [brackets]. For example [6]+320 would be a publication with six unnumbered pages and then 320 numbered pages. There is no need to record these unnumbered pages if they contain no content that needs to be recorded.  At times you will need to count backwards from the first numbered page to see which is page 1 and then would count the unnumbered pages that are before this. Likewise, you may record the count of unnumbered pages at the end of a publication. For example, [6]+320+[4]. As before, only do this if there is content in these pages that requires entry.  When counting pages in hardcover and some trade paperback publications do not count the endpapers. These can be identified as they are usually a stiffer paper than the standard pages."
 * How does that look? Mike Christie (talk) 00:31, 4 February 2014 (UTC)


 * Very good. Mhhutchins 00:56, 4 February 2014 (UTC)

Recent changes to the help files
I recently made three edits to the help files, on the basis of some discussions at the help desk. I gather from Mhhutchins that it's usual to suggest changes here first, so I'll do that from now on. In the meantime, here are the three edits I made -- if there are any objections to any of these, please let me know.


 * What to do when there's no cover art
 * Author field for fictional essays
 * Entry type for fictional essays

-- Mike Christie (talk) 21:03, 3 February 2014 (UTC)


 * This changes reflect the de facto standards and I have no objection to them being documented as such in our help pages. (But I really wish we had a non-fact essay type!) Thanks. Mhhutchins 00:00, 4 February 2014 (UTC)

Variant Dating
With respect to this discussion: Our help still states that variant titles are to have the same date as the parent. However, this is no longer standard practice. I recommend we update the help to reflect that variants should have the date the variant appeared.

The one question though is what is to be done in cases where the variant was the first printing. I would recommend the parent always have the date of the first appearance even if that appearance was under a different title. -- JLaTondre (talk) 23:21, 3 February 2014 (UTC)


 * I agree on both counts. This has been the practice from at least when the software changed to allow the display of the dates for both variant and parent in publication records. And a even greater reason to allow dating of variants to be unique came about when translated titles as variants were allowed into the database. This shows how woefully out-of-date our help documentation truly is. Mhhutchins 23:46, 3 February 2014 (UTC)

Two more questions about page numbers
Sorry to be a pest, but I'm willing to update the help pages with the answers to these questions, so I hope answering them won't be a waste of folks' time. It appears that the standard is not to enter a page number for the novel title within a novel pub, at least if there are no other recorded contents. Is that correct?

Second, if there is uncredited artwork following the last numbered page of text of a novel, does that get recorded? I would have thought so, but there's no entry in this pub; I don't have that pub but I do have a paperback reprint and I would be surprised if the art is not in both. I would add it as uncredited art with a [477] page number, but I wondered if, since it's uncredited, it could just be regarded as the part of the main novel and not require separate indexing. (It follows three appendices, for example, which are not separately indexed, and I assume should not be.) If it's part of the main novel, however, the pagination should go to 479, not 476, since 476 is the last page of text.

Any help with either of these would be appreciated; and if I'm asking too many questions let me know and I'll go quiet for a bit. Mike Christie (talk) 02:23, 4 February 2014 (UTC)


 * The answer to the first question is irrelevant. If there are no additional contents in a record, the NOVEL title record isn't displayed, so adding a page number to it would not make a difference.


 * The answer to the second question: yes, but only if you're adding content records for each of the publication's INTERIORART pieces. Editors have the option of creating an unpaged content record that represents all of the INTERIORART works or to create content records for each of the works. If you've chosen to do that then you would have to give a page number (which you would anyway if there's only one piece in the whole book.) The artwork would be considered part of the whole work, and so should be accounted for in the publication record's page count field. So add numbers to the last numbered page even if you don't create a content record for it (regardless of whether it's credited or not). Every so often the last page of text of the book is also not numbered, but should be added to the page count as well. All of this should be documented in the record's Note field to avoid future verifying editors (and you) the trouble of having to message you.


 * And don't feel you're being a pest. We really need new eyes to look over the help documentation. Most of us have pretty much disregarded them for so long (once we've learned the standards) and don't even think about how they would look to new (or returning) editors. So please continue to bring these questions and the inconsistencies of the help pages to our attention. Thanks. Mhhutchins 02:46, 4 February 2014 (UTC)


 * I'm clearly spending too much time on Facebook lately, 'cuz I was trying to find the "Like" button for Mike's last paragraph. Chavey 16:47, 4 February 2014 (UTC)


 * Good. So I'm not the only one who has done that! Mhhutchins 17:22, 4 February 2014 (UTC)