Rules and standards discussions/Archive/Archive12

This is an archive page for the Rules and standards discussions page. Please do not edit the contents. To start a new discussion, please click here.
This archive includes discussions from January - December 2014‎.

Archive Quick Links
Archives of old Rules and standards discussions.

---

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

---

Expanded archive listing

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 Shannon Messenger'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 Melissa Landers. 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)

Done. Mike Christie (talk) 12:23, 8 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)

Based on the answers to my questions above, I'd like to suggest the following changes to the relevant help page:

Page - The page on which this item of content can be found. The page number refers to the pagination provided in the publication, not the actual count of pages from the start. Use the lower case form of Roman numerals, for pages in introductory material. This will happen, for example, for material on the inside cover of a magazine, since the pagination usually starts inside. For material after that last page, count forward; for magazines this can extend to the back cover, but for books the inside and outside covers should not be given page numbers. Note that if you are entering data for a novel, and a NOVEL title is the only title that will appear in this publication (i.e. there are no essays, or interior art records, and so on) then entering the page number is harmless but unnecessary, as it will not be displayed if the NOVEL is the only content in the publication. Note also that if you are entering page numbers for interior art, there are three ways to enter the data, leading to three options for entering the page number. See the entry type help section for more details.

My suggested changes are in bold. Any objections? Mike Christie (talk) 12:58, 8 February 2014 (UTC)

There are actually only two methods of entering INTERIOART records. The "third" one noted in the entry type section is just a variation of the second, and not even a distinctive variation. I also object to the implication that an editor should use Roman numerals for pages in introductory material even when they aren't present in the publication. I've never seen anyone use them for magazines in the manner described here. It should also be made clear that when a content starts on an unnumbered page, it is the standard to interpolate its starting page by counting forward or backward to the nearest numbered page. (I've actually seen records where the page numbers are bracketed because the content starting page is unnumbered.) Mhhutchins 16:44, 11 February 2014 (UTC)

I can cut the bit about Roman numerals; that's been there for a long time, so I'd like to hear others agree that it should go. Any other comments? Mike Christie (talk) 23:57, 12 February 2014 (UTC)

Ellipses and punctuation

As per Help:

• An ellipsis should be entered as the sequence "space", "period", "space", "period", "space", "period". If the ellipsis is in the middle of the title, it should be entered with a space after it as well, prior to the start of the following word.

Which means that, e.g., Elève de quatrième ...dimension" should be changed to "Elève de quatrième . . . dimension".

So far so good. However, what should we do when an ellipsis is followed or preceded by a punctuation mark, e.g. a quote or a parenthesis, rather than a word? Should we change "...For Here Shall They Reign Again" to ". . . For Here Shall They Reign Again" or to " . . . For Here Shall They Reign Again"? Similarly, should the "...)" in * * * (Gołębie wpisywały...) be changed to ". . . )" or ". . .)"? How Deep Is the Ocean . . .? to "How Deep Is the Ocean . . . ?"

My preference would be not to use an extra space in these types of cases, but one way or the other it needs to be clarified in Help. Ahasuerus 14:36, 11 February 2014 (UTC)

I wouldn't think adding a space to the beginning of a title would be desirable. Are there any in the database currently? The standard should be clarified to omit the space before the first period if the ellipsis appears at the beginning of the title, and that the space after the last period should be omitted if the ellipsis appears at the end of the title. Mhhutchins 16:34, 11 February 2014 (UTC)

Regarding your example Elève de quatrième ...dimension" , I tend to strongly disagree as it seems to me that the typographical rules of each langage should be kept (cf the case of capitalization), I don't understand the urge to transform a perfectly valid (and strictly identical to what's on the title page) title into something else because the Help (initially meant for english) says so. It's the same problem for the spaces before colons (not in the case of subtitles, of course), someone took pains to change all the french titles that I've entered by suppressing them, even if they were clearly present in the title. The net result is a database containing french titles with english rules of punctuation, which is quite annoying for french users (let's not talk of search problems), I suppose that it's the same for other langages. Hauck 16:42, 11 February 2014 (UTC)

I think consistency is important. If we continue to manually punctuate, then it will be inconsistent even within a language as pointed out above. If it's possible to distinguish between languages, especially between those that use a form of Roman alphabet, maybe a script could be written for each language concerned.--Rkihara 20:21, 11 February 2014 (UTC)

I was taught that when a sentence is broken at a punctuation mark, that it remains, and the ellipsis is inserted as usual. So if the ellipsis is broken at the end of a sentence, there's a period with no preceding space, followed by three, spaced, periods.--Rkihara 20:28, 11 February 2014 (UTC)

(unindent) When considering a standardization/regularization issue, I try to think of the reason(s) that prompted the creation of that particular ISFDB standard in the first place. Typically they boil down to avoiding duplicates (which could be created if different editors used different data entry rules) and making it easy for users to search for records.

When dealing with punctuation and capitalization I can't think of a good reason to apply the same rules to different languages, so upon consideration I agree with Hervé. What makes the case of ellipsis somewhat special is that the rules governing its use can be complex and uncertain, e.g. see this overview of its use in English provided by Wikipedia. (I believe that we went with The Chicago Manual of Style (" . . . ") rather than the "AP Style" (" ... ") when this issue was originally debated.) And there are similar uncertainties in some other languages, e.g. Swedish. The fact that we have 3,000+ titles entered one way and 4,000+ titles entered the other way is a good indication of how uncertain this area is.

Anyway, I would support clarifying the relevant Help template to state that language-specific rules should be used when they are unambiguous and that ISFDB uses the Chicago Manual of Style rules for English titles.

As far as the software goes, it could be modified to change " ... " to " . . . " automatically only if the language of the title is English. Given the number of possible permutations, I am not sure it would be easy to handle all possible scenarios, though. Ahasuerus 20:44, 11 February 2014 (UTC)

Hauck, is there a rationale for the French leaving a space before the ellipsis and, conversely, are there cases where there is no space for an equally valid reason? Or is just the general practice? The Wikipedia article (which admittedly could be incorrect) states that it is the French equivalent of "et cetera". That would indicate it is used exclusively at the end of a listing. But of the 500+ French titles in the database that contain the ellipsis (many of which are translations of original English works which contain an ellipsis in the title), that collection and its disambiguated introduction, are the only ones in which an ellipsis in the middle of the title has a space before it. (I only searched for French-language titles that contain "..." and not ". . ." of which there are less than a dozen titles.) Does that mean that the original French essay on page 141 of this publication or the last story in this collection should have been entered with a space before their respective ellipses? Mhhutchins 21:38, 11 February 2014 (UTC)

In the case of the Blackwood collection, it seems clearly a deliberate typographic effet (which explains why it's the only one of its kind), for the "natural" place of the ellipsis (at the end of a list where its means "etc." or at the end of a phrase or a title where it's a kind of "to be continued" or interruption) there shouldn't be a space before them (in fact and barring the problem of lists between parenthesis ellipses at the "end" of sentences in french are simply to be treated as periods with no space before them and with the next word capitalized). So the spaces before final ellipsises in french titles should be removed because they're either improper entries (I've done some the english way before full language support) or later "corrections" to conform to the english usage. Hauck 07:05, 12 February 2014 (UTC)

So you're saying that the space before an ellipsis without spaces is not a typographical rule of the French language as you had earlier implied. If it is atypical, then it should be made to conform to the standard regardless of the language. There are also English publications which clearly don't have spaces in ellipses, but have been made to comply to the standard. The standard was created to avoid the creation of variants based on a subjective perception of whether or not the presence of spaces or lack thereof was intentional by either the author, editor, designer, or printer. I have to believe that the same subjectivity is also a factor in non-English publications. Mhhutchins 07:52, 12 February 2014 (UTC)

Your description of the use of an ellipsis in French is identical to its usage in English. So I can see no reason that there should be an exception made for French. Typographically this record when printed as either "Deux planètes trop semblables..." or "Deux planètes trop semblables . . ." is identical. Giving the editor the option to enter it either way leaves open the possibility that another appearance of the same work might be varianted or would force the editor to merge them and reconciling the difference, a reconciliation that might bother a primary verifier of one or both records. A moderator would have no standard in order to mediate the dispute. Mhhutchins 08:16, 12 February 2014 (UTC)

I've not made myself clear, the lack of space before an ellipsis (or a period) is mandatory in french AND they are no spaces between the periods. The french standard is thus without space and "Deux planètes trop semblables . . ." or "Deux planètes trop semblables. . ." are simply not correct as opposed to "Deux planètes trop semblables..." which is the only correct form (see here for the french rules. Hauck 17:42, 12 February 2014 (UTC)

Yet it's OK to enter a space between the text and an exclamation point or question mark in French ?

Yes it's the rule.Hauck 12:41, 13 February 2014 (UTC)

But then, we're really not talking about punctuation rules here. We're talking about typography and the standardization of entering titles based on a non-subjective interpretation of how an ellipsis appears in printed form. If you're proposing that we enter them as they appear in the publication, then shouldn't that also apply for English language titles? And what's to prevent someone from merging or varianting titles based on their interpretation of how the ellipsis is formed? You've still not addressed that situation and that's the main reason why its entry was standardized at all. Mhhutchins 00:30, 13 February 2014 (UTC)

BTW, the page you link to states that the American usage of the ellipsis is identical to the French use. So I see no reason that they should be handled differently. Mhhutchins 00:35, 13 February 2014 (UTC)

Assuming that my understanding of the differences is not too far off, there is only one acceptable way of entering an ellipsis in French (...). In English, OTOH, different standards exist, so you should use " ... " if following the AP Stylebook and " . . . " if using the Chicago Manual of Style. Originally, we went with the Chicago Manual of Style, but that was a fairly arbitrary decision. Ahasuerus 03:44, 13 February 2014 (UTC)

Exactly. Hauck 12:38, 13 February 2014 (UTC)

The choice of which standard to use may have been arbitrary. The decision to create a standard was not. This is my last word on the subject. Mhhutchins 15:26, 13 February 2014 (UTC)

I agree with Mike that the decision to create a standard is the important thing. As a computer science professor, I regularly tell my students that having a standard (e.g. at a software company) is far more important than what the standard is. But with 4086 titles containing ". . ." and 3070 titles containing "...", we (1) don't have a standard; (2) the numbers are close enough that it doesn't matter which standard we use, we still have nearly the same number of corrections to make; and (3) the AP Stylebook format seems to have a tremendous amount of following here, based on how many editors have ignored our "standard" (probably unintentionally) and instead used what seems natural to them. I'll admit that while I've tried to consistently use ". . ." here, my personal preference is "...". As such I think we should give very serious consideration to changing our standard. Chavey 17:11, 13 February 2014 (UTC)

It sounds like we all agree that there should be a standard. The questions then are:

• Should our standard be broken up into language-specific sub-standards? and
• If so, which standard should we use for languages (e.g. English) which have multiple standards ?

To answer the first question, I have done a quick review of various language-specific conventions -- French, German, Italian, Romanian, Spanish, Finnish, Russian, Kazakh, Hungarian, Polish, Danish, Azerbaijani, Norwegian, Swedish, Slovak, Slovenian, Serbian, Croatian, Turkish, Thai, Arabic, Hebrew, Bulgarian, Esperanto, Lithuanian, Latvian, Dutch, Ukrainian -- and all of them use "..." rather than ". . .". Korean and Japanese apparently have more elaborate conventions, but they don't use ". . ." either. So it looks like the AP Stylebook format is in line with everyone else while the Chicago Manual of Style is an outlier. If my understanding is correct, then I think the easiest thing to do would be to change our standard to "..." and convert all ". . ."'s to "..."s, which can be done automatically.

Having said that, it looks like there is no international standard for using spaces before and after an ellipsis and even language-specific rules can be complex, so we may have to default to each language's rules in this area. That may make exact searches more difficult, but then we really need to add support for approximate/fuzzy searches anyway. Ahasuerus 21:51, 13 February 2014 (UTC)

Beating my favo[u]rite drum: I think it is better to have a uniform standard that requires neither editor nor moderator to be familiar with the language's specifics. That also covers situations where different authorities for the language promulgate conflicting styles -- if it happens for English, you can bet it happens for other languages. FWIW, I think the intra-elipsis spacing debate is further compounded by the relatively recent (funny how the older I get, the longer ago things seem "recent") trend to treat an elipsis as a single glyph/entity typographically; I suspect modern text-editing software helps reinforce the notion that an elipsis is a collapsed set of three dots. --MartyD 12:53, 15 February 2014 (UTC)

That was more or less the goal when we started thinking about adding non-English works to the database. Unfortunately, it turns out that it can be difficult for editors/moderators who are not familiar with "the language's specifics" to interpret the data correctly. For example, consider Золотое побережье, a Russian translation of Kim Stanley Robinson's The Gold Coast. The editor who entered the data noted that the text was "Translated by М. Пчелинцева (M. Pchelinceva)". If you check the FantLab record, you'll find "перевод [translation by] М. Пчелинцева" at the bottom of the page, which appears to confirm the spelling. However, if you follow the link to the translator's page, you will see that his last name is actually spelled "Пчелинцев" -- "Пчелинцева" is an inflected form of the name. I have run into similar problems with Czech and Polish translations taken from WorldCat, so it's not limited to Russian or even Cyrillic. Ahasuerus 03:25, 19 February 2014 (UTC)

Your Russian is probably much better than mine, but for our listening audience: The trouble there is "Пчелинцева" is a declined form (genitive) of "Пчелинцев" due to "перевод". It's "M. Pchelincev's translation", not "translator M. Pchelinceva". So, yes, for words one definitely may need knowledge of the language specifics. --MartyD 13:50, 19 February 2014 (UTC)

Proposal to update the definition of CHAPTERBOOK

The current definition of CHAPTERBOOK, as given here states:

CHAPTERBOOK. 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. Such a publication may contain an essay or essays, may contain interior art, and may have cover art listed. Content records should be used to record the shortfiction, and any essay or interior art, just as with an anthology or collection. This publication type is also used for an ebook or audiobook edition of a single work of short fiction. This publication type may also used for anything smaller or flimsier than a standard paperback. These are often, but not always, saddle-stapled; publications from conventions are frequently in this format. However, if such a publication contains multiple works of fiction, it is usually better to list them as anthologies or collections, with a binding type of "ph" (pamphlet).

Note that a CHAPTERBOOK commonly includes a work of SHORTFICTION of the same name - these two records should not be merged. See Help:How to merge titles for more detail.

This definition has both inaccuracies and contradictions, as well as several ambiguous points. I propose the following change:

CHAPTERBOOK. This publication type is a unique ISFDB designation for a separate publication of a single work of SHORTFICTION (q.v.) or a single POEM, regardless of the publication binding/format. In addition to the single SHORTFICTION or POEM content record, such publications may also contain one or more ESSAY and INTERIORART content records. This type may also be used for an ebook or audiobook edition of a single work of SHORTFICTION or POEM.

• Publications which contain a single ESSAY without a SHORTFICTION or POEM content should be entered as NONFICTION. Publications with more than one SHORTFICTION or POEM content record should be entered as ANTHOLOGY (for multiple-author collections) or COLLECTION (for single-author collections).
• Because the single content record is more often than not the same title as the publication itself, it is important that their respective title records not be merged. (The "Check for Duplicate Titles" function no longer matches such records for merging consideration.)
• This unique ISFDB designation should not be confused with the generally accepted publishing format called chapbook or the term used for books intended for intermediate readers called chapter book.

BTW, there is nothing in this proposal which actually changes the de facto working definition for CHAPTERBOOK. What is currently stated in the documentation is woefully outdated and should be corrected. Mhhutchins 00:41, 12 February 2014 (UTC)

Looks good. The only thing that I would consider changing is the first paragraph where it first says "regardless of the publication binding/format" and then "This type may also be used for an ebook or audiobook edition". Perhaps we could combine these two statements in a single sentence along the lines of "This type covers all bindings and formats, including short ebooks and audiobooks"? Ahasuerus 01:53, 12 February 2014 (UTC)

Good suggestion. I considered combining them when I copied the original statement of the current documentation. But I'm not sure that it's wise to add "short" to the description of ebooks and audiobooks as some novellas can be quite long, and "short" can be subjectively different for each user. Mhhutchins 02:38, 12 February 2014 (UTC)

True, "short" may not be the best term in this case. Perhaps something like "of similar length", but I can't think of a good way to phrase it at the moment (insomnia issues; the brain is not cooperating.) Ahasuerus 03:49, 12 February 2014 (UTC)

P.S. Also, we may want to clarify the relationship between CHAPTERBOOK publications and CHAPTERBOOK titles. Something along the lines of "A CHAPTERBOOK publication must contain:

• one CHAPTERBOOK title, which the software will create automatically when a New Publication submission is approved, and
• one SHORTFICTION or POEM title.

CHAPTERBOOK publications may also contain additional ESSAY and/or INTERIORART titles". Ahasuerus 01:57, 12 February 2014 (UTC)

Well, the explanation of the "behind the scenes" software functioning seems to be unnecessary when the purpose of the section is to describe this type of publication. Yes, the software automatically creates separate CHAPTERBOOK records for titles and publications, but the same is true of every publication type. I don't believe that the description of any other type goes into such detail. When you tell the editor that it "must contain one CHAPTERBOOK title" some new editors will assume it is their duty to add it as a content record. And the last line here duplicates what is already in the proposed change. Mhhutchins 02:38, 12 February 2014 (UTC)

I think the problem with my proposed change was that I was trying to cover both publication types and title types at the same time -- see my note further down. If we spread this information across the two affected templates as described below, perhaps the issue will go away. Ahasuerus 03:52, 12 February 2014 (UTC)

In the second paragraph I would suggest changing the first sentence to "Do not use this type for publications which contain a single ESSAY without a SHORTFICTION or POEM content. They should be entered as NONFICTION." Ahasuerus 02:04, 12 February 2014 (UTC)

Maybe a little too wordy, and perhaps semantically identical, but I have no objection if others find it clearer. Mhhutchins 02:38, 12 February 2014 (UTC)

And I think the second bullet point may look better if we moved the subordinate clause to the end, e.g. "Do not merge CHAPTERBOOK titles with their related SHORTFICTION/POEM titles. Although they usually have the same names, they are separate records and appear in different sections of author bibliographies."Ahasuerus 02:04, 12 February 2014 (UTC)

You're correct. That clause should not start the sentence, but I would suggest an even further simplification of that bullet's first sentence: "Do not merge a CHAPTERBOOK's title record with its identically titled content record." The fact that they appear separately on an author's summary page is irrelevant, at least to me. Perhaps some other editors will provide an opinion. Thanks for the suggestions. Mhhutchins 02:38, 12 February 2014 (UTC)

Do we perhaps want to add an additional bullet stating that a CHAPTERBOOK should never be added to a title series? --Ron ~ Rtrace^Talk 02:50, 12 February 2014 (UTC)

Another good suggestion. I doubt that that relatively recent standard (maybe three or four years, de facto at least) has ever been documented. Mhhutchins 03:06, 12 February 2014 (UTC)

It occurs to me that we may be trying to add too much title-specific information to this Help template, which, after all, only covers publication types. There is a different template, "Template:TitleFields:EntryType", which is supposed to cover all title types, but doesn't mention CHAPTERBOOKs at all. Perhaps we could move the information about not assigning series/synopses information and not merging CHAPTERBOOK titles to that template? Then we could link the two templates, e.g. "For more information on CHAPTERBOOK title records, see Template:TitleFields:EntryType". Ahasuerus 03:46, 12 February 2014 (UTC)

I had not realized there was a template dedicated to entry standards. In that case, I agree that we should distinguish each document with a link between each. Perhaps the same distinction should be made in the documentation of the other publication types? Mhhutchins 05:13, 12 February 2014 (UTC)

Yes, it sounds like a review of different templates (and whether the right Help pages link to them) may be in order. Ahasuerus 03:28, 13 February 2014 (UTC)

(unindent)I've been following these discussions with a good deal of interest, not least because I wrote nearly all the help, long ago. Some of the mistakes that are identified above are my fault, based on my understanding of the rules at the time; others are perhaps due to changes in practice or clarifications since I was active. I'd like to add one suggestion to the above, which is to consider adding example links to the help, so that a new reader can see what you're talking about. I'm a good test audience here since I knew this database very well at one time, but am having trouble following some of the ins and outs above -- Rtrace's comment about CHAPTERBOOKs not being added to a title series is completely opaque to me, for example. If I'm confused, *nobody* who doesn't already understand how to do these things will be able to follow the help.

One of the difficulties I had in assembling the help files was in coming up with a way to present the right help to the user at the right time. We never really solved that, I think; pages like Help:Screen:EditPub are helpful but I have to know what I'm doing already to be able to use them. Placing detailed help at the individual field level is probably the best answer. Perhaps the individual help paragraphs could pop up as tooltips from a question-mark button next to each field? Mike Christie (talk) 00:10, 13 February 2014 (UTC)

Check again, Mike. There are pop-ups on new entry forms. If you hover over the circled "?" for each field, there's a general overview of that field. Clicking on the ? will lead you to a more detailed help page for that field. Mhhutchins 00:18, 13 February 2014 (UTC)

I never use LOL, but I did laugh out loud when I read that. Just shows how little it is possible to notice. As usual, Ahasuerus is way ahead of me. That's exactly what I was suggesting. Mike Christie (talk) 00:30, 13 February 2014 (UTC)

Well, to be fair, not all data entry forms have had this feature implemented yet and Edit Pub is one of the stragglers. It's on my list of things to do, but there are so many other things that need to be done... Ahasuerus 03:21, 13 February 2014 (UTC)

I've made the changes incorporating the suggestions of those editors who participated in the discussion. Mhhutchins 23:54, 25 March 2014 (UTC)

Probable error in the documentation for INTERIORART

According to this help page (look for INTERIORART under the Entry Type subsection:

Illustrations not attached to a story are given a title of "Untitled".

It has always been my understanding, and the way I've instructed dozens of new editors, that if an illustration doesn't specifically illustrate a contained work (either story or essay), then it is given the name of the container (the title of the book or the magazine). This aligns with the general premise that all illustrations illustrate something. If that's the case, no illustration is truly untitled. Otherwise all illustrations that aren't explicitly captioned would be considered untitled. I belief the reason to avoid "untitled" as a title and to give the work the name of the container is to prevent the accidental (or intentional) merging of the works by the same artist. For example, look at this artist's summary page. It tells us nothing. But if the pieces were titled for the publication in which they appear, at least each work would be distinguished and the user would have a head start trying to figure out which is which. If more than one appear in the same publication, they would be handled just as we disambiguate different pieces by the same artist that illustrate the same work: add a bracketed number to the title. (See how this record handles "untitled" work.)

I propose that the help documentation be changed and that editors are advised how to handle such unattached illustrations. An advance search shows that there are more than 900 records for INTERIORART with "untitled" in the title. Some of them may actually be titled "Untitled" (like the ones in this publication), but most of them are probably the result of the instructions. Mhhutchins 03:21, 14 February 2014 (UTC)

I agree with your recommendation. Chavey 14:11, 14 February 2014 (UTC)

Me too. Ahasuerus 18:52, 14 February 2014 (UTC)

Me three. --MartyD 01:45, 15 February 2014 (UTC)

Varianting Uncredited Works

Based on a discussion with an editor who was confused by our help, I have made a changed to Template:TitleFields:Author. Under the "Anonymous or uncredited works" section, the last two lines were:

If there is external evidence (such as a collection of editorials from a magazine, making it clear who the author was) that identifies the author, then you can add a variant title to that item, using the real name. This will attach the work to the true author's bibliography, without giving incorrect data about what is actually in the source publication.

This is obviously contradictory as for it to show up on the true author's bibliography, the "untitled" work needs to be the varianted to the true author, not vice versa as was stated. Therefore, I changed it to:

If there is external evidence (such as a collection of editorials from a magazine, making it clear who the author was) that identifies the author, then you can make the untitled record a variant to the real name (see Help:Screen:MakeVariant). This will attach the work to the true author's bibliography, without giving incorrect data about what is actually in the source publication. The external evidence should be documented in the title notes.

If anyone has recommendations to improve the wording, feel free to edit away. Thanks. -- JLaTondre (talk) 13:28, 16 February 2014 (UTC)

Yes, it should be corrected. It's an error which would obviously lead to the wrong results by an editor following it verbatim. Good find. I would suggest changing "external evidence" to "a reliable secondary source" which is the more common bibliographic term. And instead of "untitled record a variant to the real name", I would suggest "uncredited record into a variant of a record crediting the actual author". Mhhutchins 18:25, 16 February 2014 (UTC)

Good suggestions. Changes made, but I used "source" instead of "secondary source" as the example given didn't seem like a secondary source to me (I'd consider reprints another primary source. When I think of a secondary source, I think of something not in a publication containing the content.), but that's minor so don't mind it being added if you feel it makes more sense. Thanks. -- JLaTondre (talk) 16:38, 17 February 2014 (UTC)

Works for me. Mhhutchins 18:26, 17 February 2014 (UTC)

Help:How to verify data#Replacing a Primary Verifier

The software was changed awhile back and users can no longer replace a verifier. However, our help wasn't updated to match. I've deleted (see [1]) the "Help:How to verify data#Replacing a Primary Verifier" section as it's no longer applicable. Thanks. -- JLaTondre (talk) 13:53, 16 February 2014 (UTC)

Thanks for removing the outdated verbiage. Perhaps we should start a project that examines all of the help documentation to find obvious errors and outdated advice? Mhhutchins 18:29, 16 February 2014 (UTC)

Yeah, our help could probably do with a complete restructuring and rewrite. However, that would be a lot of work so is unlikely to happen. Focusing on obvious errors and outdated advice might be the way to go. We might also want to look at areas where new people make frequent errors. I also think we should be more aggressive at implementing changes that are discussed here. We routinely get cases where several people agree and no disagreement is expressed. When that happens, we should make the change and move on. We're never going to get high rates of participation here. -- JLaTondre (talk) 17:04, 17 February 2014 (UTC)

True. But charges of unilateral changes to the help pages have been made in the past, and I try to wait for a consensus, not necessarily unanimity. Thanks. Mhhutchins 18:28, 17 February 2014 (UTC)

Multiple page numbers for the same Content title?

I am looking at FR 370, "Page number ordering fails with multiple page numbers", and I wonder if this is something that we want to support. Here is what the FR currently says:

• If an artist is listed as having artwork on, say, pages "3, 45", then this should be shown in the publication display as appearing after things on pp. 1-2 (and before items on later pages). An example is at: http://www.isfdb.org/cgi-bin/pl.cgi?404104

I am sure it's possible to modify the software to support this capability, but is this really desirable? I see that there are 9 pubs which use this "comma" format (7 of them entered at the same time), so the data will be easy to change if we decide that it's not a good idea. Ahasuerus 02:38, 20 February 2014 (UTC)

Why aren't there two separate content records for the works on page 3 and 45? That's the way it's always been handled. If it's the same work then the second piece can be disambiguated as "Title (reprint)". A content can only be or start on one page. So I don't understand the purpose of the feature request. Surely I'm missing something here. Please explain. Mhhutchins 04:08, 20 February 2014 (UTC)

There are some publications, primarily fanzines, where one person will put in a large number of small pieces of internal art -- often as page fillers. It seems worth noting them, but it doesn't quite seem worth having 20 items added to the user's bibliography because they did what really amounts to one thing: add some miscellaneous art to that issue. Chavey 17:52, 21 February 2014 (UTC)

Fantasy Art Catalogs

Apparently there is a policy that catalogs are inappropriate for the database. I can understand how most catalogs would be left off. But some should be included. One in point is Frank Frazetta, a publication sold by Frazetta Prints in 1996. It does not have an ISBN, but it has an LCCN, there are two editions, the illustrations are suitable for removal from the book and framing (since the art is one-sided) and while it's a catalog of sorts, it's really constructed as a book. This isn't like a auction house catalog with thumbnails of the artwork, it's every bit as much a book as the Betty Ballentine Frazetta books (actually, it's better quality). Offhand, I can think of at least one other suitable book of Frazetta work, a hardcover from the Alexander Gallery exhibition that should be in the database; it was the finest Frazetta collection prior to the publication of Icon. The artwork in these "catalogs" would be merged or varianted with the covers and contents of many other publications. These books are sold like any other books by booksellers; current prices for the first of the above is around $30 and up, and for the hardcover, $200-$300. Any thoughts on the inclusion of such "catalogs"? Bob 19:22, 9 March 2014 (UTC)

I think we should distinguish catalogs as published for museums' art exhibitions and those published for the purpose of selling a product. Allowing sales catalogs in the database would blur the line between an actual publication, i.e. an art collection, and a catalog published by a book dealer or auction house just to sell a book or a work of art. Having said that, I would agree that a book as you describe should be allowed, but that it should not be used as a precedence to allow other "catalogs". What was the actual purpose of the book? It really doesn't sound like a catalog from your description. Mhhutchins 19:58, 9 March 2014 (UTC)

I think (but I'm obviously guessing) that it was meant to make a profit by itself and to sell posters as well. It must have worked to some degree because of the second edition two years later. And I think there was an element of pride involved. It has very little in common with most catalogs I've seen. The Alexander Gallery catalog has very few paintings for sale, although there are some, including one oil painting that I remember ("The Moon's Rapture"); most of the artwork is for exhibition only. Bob 22:35, 9 March 2014 (UTC)

I've accepted the submission based on your description, as it appears to be more art collection than catalog. I still open to what other editors or moderators have to say on the subject. Mhhutchins 22:39, 9 March 2014 (UTC)

I agree with the conclusion you've reached: Fantasy Art books should be in; Art/Gallery catalogs should not be. Of course the rub comes when you try to find the line in the sand between those two categories, but I think the principle at least can then be used to try to apply it fairly to individual circumstances. Chavey 01:20, 10 March 2014 (UTC)

An interesting question indeed. I don't know much about art, but FWIW, a few (non-SF) exhibition catalogs that I have seen were exceptionally well done, better than most "regular" art books out there. I concur with the general sentiment expressed above although I can't think of a definitive "in/out" test. Ahasuerus 04:38, 10 March 2014 (UTC)

To me, one of the most common distinctions is the availability of the publication. Many exhibition catalogs, or studio catalogs, are available only through that gallery/museum, and not generally available. Those, IMHO, should be out. Those that are available only that way or through some type of self-publishing (Lulu, CreateSpace, etc.) are probably questionable. Those that are published through "regular" publishers should, I suspect, pretty much always be included. Chavey 07:50, 10 March 2014 (UTC)

Review the rules for page counts?

"The ISFDB standard for entering books (regardless of the format) is to give the last numbered page as the page count and if necessary make adjustments for unnumbered pages of substantial text."
On the one hand I'd like to know how Long the story, novel, etc is - and on the other I'd like to know the actual size of the book. In almost all cases the latter is a multiple of 16 but never-ever an uneven number such as 127... so where I'm coming from is this: Shouldn't a book's page count always be the total number of pages in the book, excluding the cover, even if the last x pages are blank or used for advertisement? And if a publisher does start counting from the cover "page" then that may be reflected in any Table of Contents but not in the official page Count.
Or am I over-complicating things here? - JLochhas 18:37, 11 March 2014 (UTC)

I'd say a cautionary yes to your question. The main emphasis for our db is to document the lengths of the texts. The easiest way to document the length of a given book (in my opinion) would be to add any additional pages to the notes. (See here for a possibility to account an unusual beginning of page count). Stonecreek 19:50, 11 March 2014 (UTC)

I agree with Christian. Any unusual pagination should be noted in the record's Note field. It's long been established that we don't count advertising pages in books (that's why there is a different standard for magazines.) I don't see much value in changing the standard now when doing so will affect thousands of records. Mhhutchins 20:46, 11 March 2014 (UTC)

I'm more interested in the length of the story than in the size of the book. So I don't want to include extra advertising pages, extracts of upcoming novels, etc. As such, the current rules are pretty close to what I'm looking for. Chavey 02:59, 12 March 2014 (UTC)

Clarifying the ISFDB definition of COLLECTION

The current definition of COLLECTION (as given here):

• 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.

First, a definition shouldn't contain the word it's defining. So, I suggest "a single-author collection" be changed to:

• A publication containing two or more works of SHORTFICTION or POEMs by a single author or authors writing in collaboration.

The last part of the current definition is a rather jumbled attempt to explain why publications containing collaborative stories should also be typed as COLLECTION. I propose:

• This type should also be applied to publications which contain collaborative stories by two or more authors. For example, The Wonder Effect contains SHORTFICTION works written in collaboration by Frederik Pohl and C. M. Kornbluth, and is properly typed as COLLECTION. An extreme example of this type is Slow Dancing Through Time which is a COLLECTION by Gardner Dozois containing works written with four other authors in various combinations.

Does anyone disagree with the proposed change or have suggestions or additional input into clarifying the definition? Mhhutchins 00:12, 26 March 2014 (UTC)

I do agree, but would like to direct the attention towards this (perhaps a mere possibility): a publication could contain stories written solely by different authors (say two stories written solely by Pohl & three written solely by Kornbluth) plus a bunch written in colloboration AND be stated as COLLECTION in the publication. I'd say to handle such a case also as COLLECTION here at ISFDB. Stonecreek 07:26, 26 March 2014 (UTC)

I agree in that case, as long as the book is mainly credited to the two authors. That's getting about as close to an ANTHOLOGY as you can get without being one. Does such a case actually exist? Should we be making rules for non-existing books? Perhaps such odd publications could be handled on an individual basis without having to document their entry, which would surely confuse most new editors. Mhhutchins 20:53, 26 March 2014 (UTC)

I can agree with Christian's addition, which is the opposite of the current/old definition which specifically states that the same hypothetical book with stories by Pohl, Kornbluth and both in collaboration should be an anthology. I'd suggest that we add a sentance to Michael's second bullet: "Such a publication may also contain works by written by either of the authors individually and still be classified as a COLLECTION". Rtrace^Talk 02:31, 27 March 2014 (UTC)

Again, why should we create a definition which would encompass a type of publication which doesn't exist? In fact, that part of the current definition is illogically stated and awfully confusing. That's why I proposed what I thought was a clarification of what the original author of the definition intended. Or perhaps he did feel that adding a story that was written by one of the authors to a book of collaborations kept it as a COLLECTION, but that adding another one individually written by the other author made it into an ANTHOLOGY? That's mindboggling! Or am I the only one confused by such an interpretation? Mhhutchins 03:12, 27 March 2014 (UTC)

I will offer two more examples: Pretty Maggie Moneyeyes which contains only stories by Ellison except for a single story by Bloch. I made this an ANTHOLOGY because of the Bloch story and I'm comfortable with that classification. The other example is the book that started two above discussions 1 and 2. Out of the Unknown contains stories by Van Vogt and stories by Hull, but none by both in collaboration. It was argued that publications containing works by two authors should be typed as COLLECTION even if no stated collaboration exists. Personally, I feel such publications should be classified as ANTHOLOGY, but I don't think we had a consensus and I seemed to be in the minority. If COLLECTIONs are to include 2 authors with no collaborative stories, we should probably document that too. --Ron ~ Rtrace^Talk 02:31, 27 March 2014 (UTC)

There are going to be disagreements about individual uncommon publications, and their type can be settled by discussion. The point of this clarification of the definition was due to the ambiguity of the current definition and is in reference to the 99% of the publications which are currently typed as COLLECTION in the database. It doesn't help anyone by including singular, and exceedingly esoteric exceptions in the definition. I personally feel that the Van Vogt and Hull publication is a COLLECTION because both are credited on the title page as the authors of the publication, without crediting an editor of the publication. I would reconsider that position if only an editor was credited, thus making it an ANTHOLOGY. I also believe the Ellison publication is a COLLECTION because it is solely credited to him. (I'm assuming this because you didn't credit Bloch in the author field of the publication record.) By typing it as an ANTHOLOGY, the record gives Ellison as the editor and not the author. An ANTHOLOGY is usually edited by someone who is not the author of 95% the publication. Mhhutchins 03:12, 27 March 2014 (UTC)

BTW, I could live with both the Van Vogt/Hull and the Ellison publications typed as ANTHOLOGY, if the author field of the publication is credited to "uncredited". Mhhutchins 03:19, 27 March 2014 (UTC)

I'm quite sure that I encountered a case, but can't lay my hands on it (or my memory, which would even be better), so I'd say it's not purely theoretical. And I'd say it's better to have a case covered that is quite likely to exist or come to existence. My idea would be to put up a collection The Best Fantasy Stories of Terry Pratchett and Stephen Baxter. Stonecreek 07:20, 27 March 2014 (UTC)

I would actually find a standard based on whether authors or editors are credited on the title page far more workable. The Van Vogt/Hull would then fit the definition, but Sometime, Never, which also lists only authors should be reclassified as a COLLECTION. The only ambiguity would then be publications that credit both author(s) and editor. With a single author and an editor (e.g. The Door Gunner, it's clearly a COLLECTION. With multiple authors and an editor (e.g. Three Times Infinity), or even Three Trips in Time and Space which doesn't list Silverberg as an editor, bur rather credits him with the foreword only) it's clearly an ANTHOLOGY. Such a standard would match what the software does (collections have authors, anthologies have editors). --Ron ~ Rtrace^Talk 13:01, 27 March 2014 (UTC)

Using the title page as one of the bases for determining type could be added to the definition. As you said, there would have to be exceptions for single-author publications which additionally credit the editor, e.g. The Door Gunner, The Best of Cordwainer Smith, The Science Fiction Stories of Rudyard Kipling, et al. But a publication with a title page which credits multiple authors and an editor should clearly typed as an ANTHOLOGY, just as should all of those paranormal anthologies which credit only the authors without an editor credit. Mhhutchins 15:12, 27 March 2014 (UTC)

[unindent] Since this discussion petered out because of the conflict concerning books with collaborative authors, I propose the following definition for COLLECTION:

• A publication containing two or more works of SHORTFICTION or POEMs by a single author or authors writing in collaboration should be typed as a COLLECTION. The typing of individual publications which contain works with various combinations of author credit should be discussed on the Community Portal on a case-by-case basis. The title page credit should be the major determining factor in typing these publications.

Regardless, we need to at least clarify the current definition. Mhhutchins 17:49, 29 May 2014 (UTC)

Sounds about right. Perhaps reference the comment about excerpts -- "However, sample chapters placed at the end of a book for advertising reasons do not make a novel into a collection" -- further down the page? Ahasuerus 21:38, 30 May 2014 (UTC)

Or "However, excerpts from other works placed at the end of a novel for promotional purposes do not make the publication into a collection." Mhhutchins 23:02, 30 May 2014 (UTC)

Free "previews" of ebooks

Looking at this submission, I am concerned about whether we should add such records to the database. These previews are probably more common than one might think. (Knowing next to nothing about ebooks, I can't swear to that.) I can see a similarity between them and those innocuous excepts stuck in the back of paperbacks to fill out the page count, and one could argue for including them because of their similarity. I see both as just cluttering an author's summary page with a lot of useless "data" (and I hate these excepts being placed into title series, but that's for another rant on another day!) Mhhutchins 02:09, 29 March 2014 (UTC)

I have been mulling this issue over for the last few weeks. For now, I am limiting Fixer's ebook submissions to AddPubs since they are easier to handle, but pretty soon I will start including NewPubs in the mix and this will become a bigger problem.

In general, we do enter cheap/free promotional books which contain the first few chapters of a novel, e.g. The First Part of Book One of The Wheel of Time. In the past there were very few books/series so popular that publishers found it cost-effective to do something like that, hence the effort to catalog them was minimal. However, the advent of ebooks changed the math dramatically and I have been seeing more and more of these "free previews" over the last couple of years, e.g. Reckless (Free Preview).

The thing that gives me pause is that although the number of "free previews" has grown, they seem to have plateaued at a manageable level. I suspect that publishers have found other marketing techniques -- like offering e-novellas and short stories set in the same universe (or even the first novel in the series) for free -- to be more effective. If we could be sure that free previews were guaranteed to stay at the current manageable level, I would support including them. However, there is no way of telling what publishers may do next. If we start listing them and then publishers decide that free previews are cost-effective after all, we may be stuck with an unmanageable volume of submissions and that would be a Bad Thing (tm).

In other words, I think we need to have a broader discussion before we go one way or the other. Ahasuerus 02:48, 29 March 2014 (UTC)

Personally, I'm all in favor of doing anything we can to limit ebooks, but I suspect I'm a bit of an outlier on this issue. Chavey 18:41, 30 March 2014 (UTC)

My stance is less about ebooks than it is about excerpts. I hate them in whatever form they appear. Mhhutchins 19:30, 30 March 2014 (UTC)

Since there doesn't appear to be much interest in a broader discussion, I will remove my hold on the submission. (Another thing I dislike: holding submissions in the queue for a needless length of time.) Perhaps this will prompt other moderators to join the discussion. Mhhutchins 21:27, 31 March 2014 (UTC)

I would be inclined to exclude. I'm not even sure what the publisher was thinking in this case. A free abridged ebook to market a print book? Hopefully this is a small case as I cannot see it being that effective. -- JLaTondre (talk) 22:35, 1 April 2014 (UTC)

Magazine 'Editor' credits

Further to the recent discussion on several editors' pages concerning Foundation, I'd like some clarification on which editors should properly be credited. I can't find earlier written guidance on this (please point me to it if it exists) and I seem to have just followed previously established patterns for different publications. It seems there are two standards: one for entering the chief Editor only, and another for entering all the various editors who contribute towards a particular publication, such as 'Features Editor', 'Reviews Editor', 'Consultant Editor', 'Production Editor', etc.

Looking at Foundation, there are anywhere between one and five editors credited for different issues, which suggests some inconsistency (for the issues of Foundation I've personally added, I see I have usually added only the 'Editor'). However, over at Vector I seem to follow a different standard when adding recent issues, ie. adding all credited editors for different roles (including Reviews Editor and Production Editor), as that was the pattern set by Bill Longley for the many earlier issues he indexed.

Do we need to standardize this? It seems to me we should at least standardize it for all records of a particular publication, if different publications are treated differently. PeteYoung 07:56, 10 April 2014 (UTC)

Yes, I'd think we need a standard. My personal standard is or would be to state all editors who have a considerable leading role (which also could be for a recognizable department such as reviews or features). Notable exceptions that wouldn't be entered into ISFDB would be 'technical', 'assistant' and 'advisory' (or similar named) editors, because I tend to think that those weren't part of the selection and major editing process. I also didn't enclose graphics editors. Christian Stonecreek 10:06, 10 April 2014 (UTC)

An afterthought: If we'd credit all editors this really can become quite tedious, as for example in the case of Asimov's Science Fiction. Stonecreek 10:30, 10 April 2014 (UTC)

My rule of thumb: For a fiction magazine, anyone titled "Editor-in-Chief", "Editor" or "Fiction Editor" should be credited, even if those roles are handled by three different people. For a review magazine (like Foundation) I substitute anyone credited as "Reviews Editor" for "Fiction Editor". I do not credit "Associate..", "Assistant...", "Contributing..." etc. I provide a breakdown in the Note field, if I choose to credit in the editor field more than one person if each have different roles: e.g. here. Every magazine is different, so it's going to be impossible to create a single standard to cover every situation. There are many magazine records in the database for titles from the 30s through the 50s where only the "Managing Editor" is credited. This question came up back around 2007-2008, when there was an influx of new editors who were working exclusively on magazines, and they tended to discuss things among themselves without involving other editors. It is their standards which are still be using today, even though a lot of those standards aren't even documented. People tend to see what was done before and "go with the flow". It's going to be impossible to change those credited editors now even if we do come up with a new "standard". Mhhutchins 17:02, 10 April 2014 (UTC)

That's right, this issue was discussed extensively back in 2007-2008 and then again in 2009 when development was restarted. I believe the consensus was that we wanted to enter "editor(s)-in-chief" (or similar) in the Editor field and relegate "other" editors to Notes. A suggested long term solution was to leverage the (yet to be implemented) "Roles" field, which is discussed here and here. Ahasuerus 22:52, 10 April 2014 (UTC)

Publisher credits

This is the transfer of an exchange initiated on the Moderator noticeboard page :

The publisher of this edition of Farmer's The Magic Labyrinth, primarily verified by Bill Longley, is entered as “Grafton”, which is confirmed by both copyright page and spine; but the title page only has “Panther: Granada Publishing”, which I mentioned in the notes. I was wondering whether it shouldn't be the reverse here ? Thanks for your help. Linguist 15:02, 22 June 2014 (UTC).

I agree that the publisher should be given as "Panther / Granada", but since there is no standard about which publisher credit has priority (that I'm aware of), we can only note the differences. Without a standard there are going to be identically credited books which appear differently in the database, based upon the verifying editor's subjective decision. Feel free to start a discussion on the Rules and Standard page. Mhhutchins 17:19, 22 June 2014 (UTC)

Personally, I would also consider that the data on the title page has precedence, and I was under the impression that this was standard practice. I realize that under certain circumstances (e.g. French Livre de Poche) this rule can't be followed, but I can't see any reason why it shouldn't be in this case. I also see the point of choosing a more prominent name, as the publisher template allows you to do, but all in all, the advice given on that page remains surprisingly vague — to my eyes at least — and the lack of clear rules account for the not so magic labyrinth of publisher notations in this database. Linguist 21:56, 22 June 2014 (UTC).

That standard was written in 2006, updated in 2008, and with only a few slight revisions since. I could easily update it to the current practice, but I'm not particularly up to dealing with the back and forth that is required to form a consensus of opinion. It seems strange that there never was an effort to create a standard before, using the title page as the source for the publisher data, even though that's been the standard for the title of the work since before I came here. Mhhutchins 23:56, 22 June 2014 (UTC)

I'm also for establishing a standard, and also would prefer the one proposed, that is the publisher as stated on the title page. Stonecreek 03:54, 23 June 2014 (UTC)

I agree on the need for a standard, and I agree on the title page as the primary source for that. However, I think there are some contractions of the publisher name that are appropriate. For example, the publisher we list as "John C. Winston Co." has, on every book I've seen, had the company name listed as "The John C. Winston Company". I'm all in favor of dropping "The", and of shortening "Company" to "Co.". "Houghton Mifflin" is actually "Houghton Mifflin Company"; "DAW Books" is "DAW Books Inc."; "Hodder & Stoughton" Is actually "Hodder and Stoughton Ltd.", etc. There are certain formalities in the "legal name" that I don't think we should have to include in the name as we present it. Chavey 04:50, 23 June 2014 (UTC)

[after edit conflict] I agree (and I think most of the active editors would as well.) At this point in the database's development, a canonical form of almost all major publishers has already been established. So that shouldn't be an issue. If a verifier uses a non-canonical form, even if it matches the actual title page of the publication (e.g. "Doubleday & Company, Inc."), we will use the canonical name regardless. The problem is when imprints change over the course of its lifetime. In those cases, we can easily track those changes. For instance, last year Del Rey was moved out from under the Ballantine division of Random House and is now a separate imprint. We can look at its history and see that all books published during a certain time period will be "Del Rey / Ballantine" and others "Del Rey". It will be easier to track these changes if we establish where the publisher credit should be taken from a publication, and every editor uses the same standard. Mhhutchins 05:05, 23 June 2014 (UTC)

Well, the reason that the "enter the title as it appears on the title page" standard is viable is that the software lets us create variant titles. If we had support for "variant publishers", then we could adopt a similar data entry standard for publishers and avoid publisher fragmentation (e.g. on Publisher pages.) It's not a trivial change, but if there is demand for it, it can be done. Ahasuerus 05:03, 23 June 2014 (UTC)

Oh, God. No! No! No! We wouldn't have had so much publisher fragmentation if a standard for publisher credit had been established from the start. Mhhutchins 05:08, 23 June 2014 (UTC)

Convention Souvenir Books

There doesn't seem to be a consensus regarding the publication type for Convention Souvenir books; just looking at pubs with the tag convention books they seem to be split between NONFICTION and ANTHOLOGY. Any suggestions? Albinoflea 02:27, 24 June 2014 (UTC)

I generally choose ANTHOLOGY when there is a significant amount of fiction published in them. If it's mostly essays, then NONFICTION. --Ron ~ Rtrace^Talk 02:29, 24 June 2014 (UTC)

It's documented in the help page:

A publication that collects both non-fiction and fiction together should be classified by whichever appears to be predominant.

It should be fairly easy to type them since most are predominantly one or the other. Mhhutchins 04:36, 24 June 2014 (UTC)

Thanks; I didn't know why I had it in my head that they should be one or the other. Albinoflea 04:49, 24 June 2014 (UTC)

Author's Birthplace

The "Birth Place" paragraph of theHelp:Screen:AuthorData page currently says:

• Birth Place - Place of birth of the author. There is no standard format, but city and country are a minimum to strive for. If you have more details (street address, county) feel free to add them.

I propose that we change it to reflect the current standard:

• Birth Place - Place of birth of the author. Use the "City, Municipality, Country" format, e.g. "Ante, Champagne, France" or "Annapolis, Maryland, USA".
  • If the exact location is not known, enter the part that is known, e.g. "Japan" or "Quebec, Canada".
  • For countries that belonged to multinational entities at the time of the author's birth, include the name of that entity at the end, e.g. "Ménil-la-Horgne, Duchy of Bar, Holy Roman Empire" or "Chipping Sodbury, Gloucestershire, England, UK".
  • For locations whose names have changed over time, use the name as of the time of the author's birth, e.g. for authors born in Lviv between 1773 and 1918, use "Lemberg".
  • For locations whose name is transcribed differently in different languages, use the English version, e.g. "The Hague" rather than "Den Haag", "La Haye" or "L'Aia".

Ahasuerus 16:42, 27 June 2014 (UTC)

I like this. My one (slight) concern is over the definition of "Municipality". For example, I was born in "Flint (City), Genesee (County), Michigan (State), USA (Country)", hence there are two intermediary "municipalities". I recently added the birth place for Jonathan Scott, where one source listed his birth place as "Shrewsbury (town), Shropshire (County), West Midlands (Region), England (Country)", and that's without getting into "United Kingdom (Kingdom)" (parenthetical elements added by me). Looking at what Wikipedia has to say about what constitutes a municipality, even restricted to English speaking countries, is somewhat confusing, and usually not what we want. I'm not sure what a good substitute would be, possibly "Region", where we then define Region as "State, Province, County, or equivalent governmental unit". Chavey 17:59, 27 June 2014 (UTC)

I like the proposed update which follows the current practice. The examples given should make clear the meaning of "municipality". Mhhutchins 18:08, 27 June 2014 (UTC)

In the case of different levels of municipalities, usage can differ from country to country : today, a town in France would normally be identified by the département it is situated in, more than by the region (modern regions are administrative late-comers, and often have little to do with historical ones). Thus, "Ante, Champagne, France", which is a pre-revolutionary localization, would be expressed today as “Sivry-Ante, Marne, France”, the rarely mentioned administrative region being Champagne-Ardenne until 2014 (it has just become Picardie-Champagne-Ardenne, but might change again…). What I mean is that in this case, the smaller administrative unit is usually preferred to the larger one (which is almost never mentioned in letter adresses, for instance). I think it would be helpful to make it clear whether such an indication, although judged unnecessary by most French people, would be expected to appear in this database. Linguist 21:39, 27 June 2014 (UTC).

In England, as in France, the smaller administrative unit (County) would be preferred over the larger administrative (Region; also an administrative late-comer). But in the US, it's the larger administrative unit (State) that would be preferred over the smaller (County). In Mexico, things are worse. The ordering of government units, from smallest to largest, is town/city (or "localidades") -- municipality -- state -- Mexico. What we would want for "municipality" is NOT what they call a municipality, but rather the larger administrative unit, i.e. the state. That's an example of why I'm concerned about our "Place" definition referring to "municipality". Chavey 05:27, 28 June 2014 (UTC)

I like it except I would recommend using "administrative division" in place of "municipality". Like Chavey, I am concerned municipality would be confusing (in the US, states are not municipalities). Administrative division is the generic term and the examples clarify the level we are looking for. Whereas, if we use municipality, then the examples contradict the standard. -- JLaTondre (talk) 11:23, 28 June 2014 (UTC)

I like the proposed update. It is certainly better than what is there now, regardless of the nuances of "municipality". --MartyD 11:40, 28 June 2014 (UTC)

On the municipality front, I like that better than region because I think it is helpful to draw a line somewhere to avoid what could be extreme but irrelevant detail. Governmental/administrative units seem to be a clear way to limit. For example, take someone born in Boston. You could use this: Boston, Suffolk County, Massachusetts, New England, USA. A county is a unit of government/administration, but in Massachusetts they play a nearly insignificant role; no one from Massachusetts would offer "Suffolk County" as part of the identification of Boston unless strongly pressed -- probably half the residents of Massachusetts couldn't even tell you Boston's county -- which is different from other parts of the US where counties are significant entities. New England, on the other hand, happens to be a strong regional identification, yet that has no governmental/administrative organization to it at all; it's simply a geographic corner of the US comprising the six states present there. But I think Suffolk County should be acceptable (even though I would never use it), while New England should not be. --MartyD 11:40, 28 June 2014 (UTC)

I'm all for your proposal, Ahasuerus. It would be best to have a native of the municipality (or perhaps of the greater region) to have a final word, but as we probably don't have one at hand for every purpose, the next best thing will be to ask somebody trustworthy if in doubt and if that doesn't seem possible go for your best (wikipedia) estimate. Stonecreek 12:12, 28 June 2014 (UTC)

(unindent) It sounds like we all agree on the direction of the proposed change. We also seem to agree that administrative divisions rather than historical regions (like "New England" in the US or "le Midi" in France) should be used for the "intermediate" level.

Re: "municipality", I couldn't think of a better term when I crafted the posted language, but I think that JLaTondre's "administrative division" may be more generic and less confusing. Ahasuerus 21:15, 28 June 2014 (UTC)

Agree. Chavey 13:09, 29 June 2014 (UTC)

Change made. Ahasuerus 18:15, 29 June 2014 (UTC)

Online publications

As currently stated in the policy about the Rules of Acquisition:

Out - Web only publications such as webzines. The problem for us is that they are not durable; anything we index could change, be reorganized, edited, updated with new versions or revisions, and so on. In addition, the URLs are not likely to be durable. So for now web only publications are not accepted into the ISFDB.

In - On a provisional basis, the Helix and Clarkesworld webzines have been added primarily because they received Hugo nominations, and Subterranean Online has also been added as an SFWA qualifying publication. Other webzines may be added based upon criteria yet to be determined, a primary one being the willingness of editors to update data in a timely manner.

This was brought up for discussion again a few years later (after the 2009 date of the last update), where it was determined (but not codified in the policy) that only online publications which are downloadable as electronic files (ebooks) were eligible for the database. The exception for non-downloadable stories would be their publication on a website which is a market that SFWA considers an author eligible for membership. Those markets are listed here. I believe the policy should be amended to reflect the current practice:

In - Works published on the Internet which are downloadable as electronic files in any number of ebook formats (ePub, Mobi, PDF) are eligible for inclusion in the database. Works published in a web-based publication (webzine) and available only as an HTML readable file are not eligible for inclusion with the following exceptions: 1) It is published by a market in which SFWA considers an author eligible for membership (listed here), or 2) It has been shortlisted for a major award. (The latter includes works which are self-published by an author on their own website.)

I'm not sure that the last proviso in the current policy (that webzines are only eligible if an ISFDB editor is willing to update the data) should be considered part of the policy as it sounds too egocentric. Suggestions for improvement and clarification welcomed. Mhhutchins 04:14, 3 July 2014 (UTC)

I think that's a good improvement. Chavey 18:41, 3 July 2014 (UTC)

I like the change and don't see that the last proviso is terribly necessary; I wonder if this could be consolidated with the last bullet entry under ROA #1:

"downloadable e-zines (periodicals in electronic format) which have been assigned an industry standard registration identifier (ISSN or ISBN) or have a history of containing reliably stable contents. Copying live webzine pages from the internet to a local computer does not qualify an ezine as downloadable."

I think the bit about ISBNs or ISSNs could be added to your proposed draft, while the "copying live webzine pages..." piece seems covered by your "available only as an HTML readable file" clause. Albinoflea 04:14, 8 July 2014 (UTC)

Chinese names

I was wondering whether there existed a home rule for the transcription of Chinese names. If not, I would be in favour of systematically adopting the pinyin spelling (when known) as a canonical form, and varianting all Western notations from it.

Another point that might deserve raising : the fact that the Chinese custom is to use the family name first means that the “first name” usually comes last (Mao Zedong's “first name” is Zedong). Shouldn't this be taken into account when considering the “last name” entry in the record of a Chinese author ? Linguist 08:22, 9 July 2014 (UTC).

The first name / last name problem is the same as in Hungarian, which we have resolved in favor of putting the family name into the "last name" field. I think this is a place where the original English-language chauvinism should hold sway, and that we should continue the policy used in for Hungarian. However, it would be a good idea to add that note to certain help screens, e.g. to the "Last Name" "?" popup in Edit Author -- specifying that the "Last Name" should be the "family name", and that the "First Name" should be the "personal name". Chavey 14:09, 9 July 2014 (UTC)

Would it be possible to change the field name from “Last Name” to “Family name” ? That would clear matters up a bit. Linguist 14:17, 9 July 2014 (UTC).

I prefer "anglocentric" to "English-language chauvinism" as it is a more objective word, and "chauvinism" has so many bad connotations. That makes it appear that the creators of the database had intentions of keeping out other languages, when it was simply a matter of English being their native language, nothing they can be faulted for. Mhhutchins 17:35, 9 July 2014 (UTC)

In many cases things simply "evolved" without anyone making a conscious decision. For example, why does our database use the Swedish collation sequence? Well, the MySQL software was originally developed in Sweden and Al never bothered to change the default setting -- or perhaps didn't realize what the default settings were until much later :) Ahasuerus 18:00, 9 July 2014 (UTC)

I second the change in field name from last to family name. This would also help with the Japanese entries where the family name is customarily first. I would also like to suggest the Hepburn transliteration for Japanese.--17:43, 9 July 2014 (UTC) —The preceding unsigned comment was added by Rkihara (talk • contribs) .

Chinese, Japanese, Korean, Cambodian, Vietnamese, and quite a few other East Asian names as well. As far as the Chinese pinyin spelling is concerned, it would be helpful if some opinions were expressed, one way or another, as I got stuck with my editing of Chinese authors. Thanks ! Linguist 19:49, 9 July 2014 (UTC).

I have absolutely no knowledge of the languages, so I have no opinion, one way or another. Mhhutchins 20:42, 9 July 2014 (UTC)

(unindent) OK, so it sounds like the consensus is that we want to change the name of the "Last Name" field to "Family Name". We can certainly do that, but keep in mind that this field's value is not displayed on authors' Summary pages, so the change will only affect editors and moderators. If that's OK, I will create an FR. Ahasuerus 21:36, 9 July 2014 (UTC)

OK, thanks. Just to make it clear : pinyin is the official system used in China for transcribing Mandarin into the Latin alphabet. European scholars had developed, throughout the 18th and 19th century, different systems of notation that never became unified, hence the different spellings, for instance, of Pu Songling's name, as P'u Sung-Ling or Songling Pou. For the sake of coherence, if no one objects, I'll pinyinize all the Chinese names I come across (there can't be that many), and variant all the other spellings. Not tonight, though (it's 1 a.m., and the shootouts have killed me). Linguist 23:00, 9 July 2014 (UTC).

Crediting the author of a single artist artbook

This may have been discussed before, but I can find or remember it. There's nothing in the rules about who to credit as author of a single artist artbook. So far I've seen credit go to the artist, the editor, the author of the text or any combination of these, but no consistency. The best example I've seen of what goes wrong is "Hannes Bok: A Life in Illustration". We have a set credited to Hannes Bok (the artist), and an identical set credited to Joseph Wrzos (the editor). I would prefer to credit both artist and editor/writer. Any ideas? --Willem H. 19:31, 17 July 2014 (UTC)

Hm, I thought that we used the artist's name, but I can't find anything to support it. I don't really have a preference since I know little about art and will defer to our art experts. Ahasuerus 03:48, 18 July 2014 (UTC)

First off, I'll admit that publications that I have verified don't appear to be consistent. However, in my opinion, we should reflect what is on the title page. In the cited Bok collection, the title page has "Hannes Bok" over "A Life in Illustration" over "Edited by Joseph Wrzos". I took this to mean the title was "Hannes Bok: A Life in Illustration" with Wrzos as the editor. This appears to be the way Worldcat handled it. Other bibliographies have handled these differently. For example, these other two collections of Bok's work ([2] and [3]) have similar title pages (artist's name is part of the title, editor is credited). Worldcat again assigns responsibility to the editor and the Library of Congress follows the same scheme. However, Locus1 lists these titles under Bok. Another thing to consider is whether there is significant writing about the artist as there is with the Wrzos edited book. --Ron ~ Rtrace^Talk 21:48, 19 July 2014 (UTC)

IIRC, a similar discussion several years ago resulted in no clear decision, except that it allowed that both the editor and the artist should be credited if the publication has an overwhelming percentage of interior art by the artist and occasional text by the author/editor who is credited on the title page, i.e. an art book compiled by someone other than the artist, especially if after the artist's death. (There are some books of Virgil Finlay's art which were compiled by Gerry de la Ree after Finlay's death, which probably should be co-credited to de la Ree, but I don't know how he's credited in the actual publications.) If there is an author other than the editor credited for the text, then he/she would also be credited. I believe we came to this decision because otherwise the title would not be found on both editor and artist's summary pages. (For an example of a text author credited in an art book see Nigel Suckling.)

A biography with only occasional pieces of art (more text than art) would only be credited to the author/editor with a content record for the INTERIORART credited to the artist. An art book compiled by the artist him/herself with minimal text, would be credited only to the artist. Essentially it comes down to how much is text, how much is art, and who is actually credited on the title page. So this means there is nothing set in concrete, and each book is handled on a case-by-case basis. I see nothing wrong with that and assume that any multiple verifiers will determine the credits in case of a disagreement. Mhhutchins 22:41, 19 July 2014 (UTC)

Thanks Michael! It looks like nobody else is interested. I'll delete the duplicate Hannes Bok entries and add Bok as author. I can live with a "case-by-case" rule. --Willem H. 09:40, 22 August 2014 (UTC)

Non-Latin characters

While editing records containing Chinese or Japanese characters, I noticed that some titles already in the db were accompanied by a transcription, such as 心理試験 (Shinri Shiken), and some were not, such as 暗黒星雲のかなたに. I looked for some information in the help pages, but didn't find anything conclusive.
I think it would be helpful to dispose of clear rules (or at least suggestions) about such matters, for the sake of coherence. For example :

• Title : non-Latin spelling + transcription (if known), either between brackets or separated with a slash (I would personally favour the latter : 心理試験 / Shinri shiken).

• Author : same thing, apart from the fact that putting the transcription first would allow the names to appear in the Author Directory (e.g. : Edogawa Ranpo / 江戸川 乱歩).

This also implies, as I see it, the use of certain standards in transcriptions (a recently discussed topic), such as pinyin for Mandarin, Yale for Cantonese, Revised Hepburn for Japanese, etc.

• Legal Name : I couldn't determine if it was allowed to enter (as I have just done) “芥川 龍之介 / Akutagawa Ryūnosuke” in this field on Ryunosuke Akutagawa's page; but I can't see why not.

It might therefore be helpful to mention (or warn against) such possibilities on the corresponding help pages, in the case of original titles in non-Latin characters. Linguist 12:37, 23 July 2014 (UTC).

There is a feature request (FR 375) to "add a field for romanized/transliterated form of each title" and it's on my short list of things to do. It's not hard to implement, but we also need to update the search logic so that when a user enters "Shinri Shiken" in the search box, 心理試験 would be found. We also need to add a similar field to author records. Ahasuerus 17:08, 23 July 2014 (UTC)

Thanks for your answer, and the promise of forthcoming changes. Yet, while waiting for these feature improvements, I still don't know whether it is acceptable to enter an author's name as “Edogawa Ranpo / 江戸川 乱歩”, for instance. Linguist 20:17, 23 July 2014 (UTC).

I would advise against using this format and any changes in an author's canonical name (but see no problem with it in the legalname field.) This would require a variant created for every work, since no work is published with this type of credit. If the feature that Ahasuerus speaks of is implemented, it would involve too much work to change these records back to their original publication credit. (The ISFDB policy is, and presumably always will be, to create records for works as they are credited in the publication and then make any variants when necessary to the canonical author.) Mhhutchins 20:33, 23 July 2014 (UTC)

Fine. Thanks for your answer. Linguist 20:52, 23 July 2014 (UTC).

The policy, as I remember it, has been generally to use the transliterated name in the "official" fields, and put original script forms into the notes. That keeps all the information available, but gives a consistency to searches, etc. -- except of course for the many variations of transliterations. But having most records structured this way should, we hope, make it easier to then adjust to a transliteration field when it becomes available. Chavey 07:39, 24 July 2014 (UTC)

Old e-mails addresses

Earlier today I noticed that we had Joel Rosenberg's e-mail address on file. I remembered it from the days when he was active on Usenet in the mid-1990s and my first thought was "Surely it's out of date now". My second thought, a millisecond later, was "Oh, wait, Joel died 3 years ago :-(". Which brings up two questions:

• What should we do with obsolete e-mail addresses?
• What should we do with e-mail addresses associated with dead people?

Re: question #1, on the one hand, amassing a collection of dead e-mail addresses seems counter-productive since they can obscure the current address(es). On the other hand, sometimes it can be useful to be able to confirm that an old message posted on the Web was authored by one of "our" authors rather than by some other "Joel Rosenberg", "John Smith", etc. Besides, many people, especially in academia, keep their old addresses indefinitely and forward them to their current address. Sometimes an ancient .edu address may work when a newer .com address wouldn't.

Re: question #2, the only reason to keep this information on file (that I can think of) is to allow associating old messages with authors. For example, it could help confirm that the Damon Knight who briefly posted on Usenet and the Damon Knight who hung out on GEnie was "our" Damon Knight. In any event, it seems harmless to leave them in the database.

Anything else that I am missing? Ahasuerus 17:25, 28 July 2014 (UTC)

Correct year of a title first released in a different language

There's no information in the help about the year to use if a book was written in one language, but a foreign language translation was published first.

According to this discussion at the Help Desk it should be set to the year the canonical title was released. This information should be added to Help:How_to_enter_foreign_language_editions.

Proposal for the third paragraph: "If a book was written in one language, but a foreign language translation was published first, then the original language title is entered as the canonical title and the translated title is entered as a Variant Title. The year of the canonical title has to be set to one of the canonical title, not to the one of the translation (though the latter one was released earlier)." Hitspacebar 07:37, 24 August 2014 (UTC)

Your suggestion is consistent with our previous discussion of this topic (in 2012). In that previous discussion, the last comment includes the statement "I guess we will want to update the Help page to make sure that the outcome of this discussion doesn't get lost in the mists of time." I guess we didn't do that :-) , so it's helpful that you're suggesting that we get back to that. Chavey 13:23, 24 August 2014 (UTC)

Recommend that be revised to "If a work was written in one language, but a foreign language translation was published first, then the original language title is entered as the canonical title and the translated title is entered as a Variant Title. The year of the canonical (i.e. parent) title should be set to publication year of the canonical title, not to the year of the translation (though the latter one was released earlier)." This applies to more than books & trying to make it a little more clearer. -- JLaTondre (talk) 16:33, 24 August 2014 (UTC)

I'd completely forgotten the original discussion and its outcome. I like the proposed changes, and will adjust the help page accordingly... before anyone forgets this time. I changed "entered as" to "considered as", since we don't really "enter" work as either parent or variant. It's only after the publication record and its title record is in the database that we consider its relationship to other titles in the db. Thanks. Mhhutchins 21:48, 24 August 2014 (UTC)

Help text for unumbered pages is confusing

I think the help about page numbers about how to enter unnumbered pages is very confusing, at least for new editors. The third bullet point says:

"Sometimes a publication will have unnumbered pages (...) 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."

And the fifth bullet point:

"It is fairly common for the last page of text in a book to have a different graphic layout which may not include a page number. The "last printed page number" rule would then use a page number before the end of the work. In these cases, count forward to the end of the text and use that as the last page number."

I understood bullet point three simply as "do not count forward to the end of the text but enter the unnumbered pages as [n]", no matter if the first unnumbered page is in the novel, last short story, afterword etc. That's what I actually did for lots of books. As a result I thought that the fifth bullet point was an additional rule especially for pages with "different graphic layout" (which isn't defined there any further, but I think a simply unnumbered but otherwise same as the others looking page isn't a different graphic layout).

Or is bullet point three only intended for unnumbered additional material (e.g. afterword) printed after the (fully numbered) main text (e.g. novel)? If so it should be re-worded to avoid further confusion.

(see also older "Change to help text for page numbers " discussion above). Hitspacebar 20:37, 25 August 2014 (UTC)

You're correct. The two appear to be contradictory and should be clarified. But you seem to have figured out what each mean. If a novel ends on the numbered page 245, and there is an afterword on unnumbered pages which follow that numbered page, the page count becomes "245+[n]" with "n" being the number of unnumbered pages. If a novel ends on the unnumbered page after 244, then it should be entered as "245" with an explanatory note about the unnumbered last page. I'll clarify the third bullet to indicate "additional material". Thanks. Mhhutchins 00:08, 26 August 2014 (UTC)

Thanks for your improvements. There are still some unclarities, though:

Question 1:

Here are some examples I compiled with my current understanding. Are these correct?

1.1: A novel ends completely on page 245 and has 2 more numbered pages containing acknowledgements: 245 (though the last printed number is 247 but because acknowledgements are generally not included and don't require a content record)

1.2: A novel's last numbered page is 244, there's 1 more unnumbered page containing the rest of the novel plus 2 more unnumbered pages containing acknowledgements: 245 (reason see 1.1)

1.3: A novel's last numbered page is 244, there's 1 more unnumbered page containing the rest of the novel plus 2 more unnumbered pages containing an afterword: 245+[2]

1.4: A novel ends completely on page 245 and has 2 more unnumbered pages containing an afterword: 245+[2]

1.5: A novel ends completely on page 245 and has 2 more pages containing an afterword, the first one numbered, the second one unnumbered: 247 Hitspacebar 20:24, 26 August 2014 (UTC)

1.1: "245" if you don't think the acknowledgement deserves a content record. (Some acknowledgements are more than just a "thank you" list and deserve a content record, so create it and enter the page count field as "245+[2]".)

1.2: Same as 1.1, but note that the last page of the novel is not numbered.

1.3: "245+[2]", and note that the last page of the novel is not numbered.

1.4: "245+[2]", with no note necessary.

1.5: "247"

Mhhutchins 21:21, 26 August 2014 (UTC)

I think I'm close to being fully enlightened now. There are two answers I don't understand:

1.1: in the second half of your answer you say "245+[2]", but the assumption was that all pages are numbered, even the two acknowledgement pages. If the acknowledgements are worth recording, why not just enter "247"? Is that again because catalogues like OCLC etc. would record it with 245 pages only (see your answer below)?

1.5: you didn't mention a note there. But it should contain a note that the last page of the novel is not numbered, shouldn't it?

Thanks for your patience. Hitspacebar 19:30, 27 August 2014 (UTC)

Sorry, I misread it as "unnumbered" in the 1.1 scenario. You're right, it would be entered as "247" regardless of whether you include a content record for the acknowledgement. And in 1.5, the last unnumbered page should be noted. This should be assumed in any case where significant text appears on unnumbered pages. Hope this clears it up. Mhhutchins 23:58, 27 August 2014 (UTC)

Question 2:

Like I wrote initially I had first thought that one should generally count from the last printed number to the end of relevant content and enter it as "+[n]". Now I learned that this can be wrong (bullet point three vs. five). But thinking of it now I wonder: what's the reason at all for the rule in bullet point three (entering unnumbered pages as "+[n]"). Why not simply count to the end of the content as well? Why not use "247" instead of "245+[2]" in the examples above? I'm not asking to change the rules, but I'd like to understand them. Hitspacebar 20:24, 26 August 2014 (UTC)

I see your point. I can only guess that the first editors of the database (before I began to edit here) thought that you should use the "N+[n]" if there is significant content on unnumbered pages at the end of the book. Their rationale, and again I'm just guessing, is that OCLC, the Library of Congress, and many such organizations use the last numbered page regardless of what is contained after that last numbered page. If a user sees our record as "247" pages and the LCCN record shows "245" they may think there's a discrepancy in one of the records. Using "245+[2]" helps reconcile the difference. Mhhutchins 21:21, 26 August 2014 (UTC)

Here's an example I came across yesterday that points out the difference in how we handle page count and OCLC handles it. The ISFDB record gives the page count as "427" with a note about the unnumbered page, while the OCLC record gives the page count as "426". I'm sure the same rationale could be applied to books that contain significant content on unnumbered pages which we at the ISFDB believe should be part of the record. Mhhutchins 21:28, 26 August 2014 (UTC)

Thanks for the explanation. That backgound informations helps a lot. Hitspacebar 19:30, 27 August 2014 (UTC)

An introductory paragraph explaining the intention or purpose of the pages field could probably help in understanding these rules better. It should explain what the page count is for what it is not for (I could try to write that myself as a proposal once I fully understand the page counting). Hitspacebar 20:24, 26 August 2014 (UTC)

I would hope that the words "pages" is self-explanatory, but if you feel up to it, please propose an introduction to that section. If I had my druthers, the field would be titled "page count", since some users could confuse it with "pagination" which is a horse of a different color. Mhhutchins 21:21, 26 August 2014 (UTC)

Question 3:

"different graphic layout" in bullet point five is undefined. Which books do apply, which don't? In Germany, lots of books have the last page of the novel unnumbered but with no change in layout (except for the missing page number). In my understanding, bullet point five would not apply for these books because they don't have a "different graphic layout". So, should the words "different graphic layout" better be removed from that paragraph? Thanks for your help. Hitspacebar 20:24, 26 August 2014 (UTC)

I'm not sure how else to explain the purpose for that paragraph to an editor. Perhaps it should say "pagination" and not "graphic layout"? Quiet often a book's designer (that's where the "layout" comes in) determines that some pages don't require page numbers, because they mess with the "look" he/she's trying to achieve. For example, most container-typed publications (COLLECTION, ANTHOLOGY, etc), don't provide the page numbers on the title pages of the contained pieces. So we count forward or backward to get the page on which the story or essay begins. When they happens at the end of a book, we just count forward to the last page of text. I just learned yesterday (from Christian) that many German publications don't number the last page. So it's good to point out in the help section, regardless of what we call it, that some books have different designs (or "layouts") that could cause confusion when working with the page count field. Mhhutchins 21:21, 26 August 2014 (UTC)

Ok, I see now. Like already offered above I'll try to come up with some proposals for an improvement of the help, based on this discussion. Thanks a lot. Hitspacebar

You're welcome. Don't hesitate to bring any other confusing or conflicting help pages to our attention. We've become so jaded that it's hard to catch the things that a fresher set of eyes can. Thanks. Mhhutchins 23:58, 27 August 2014 (UTC)

Dual Language Novels

Fixer added this pub which raises a question that I don't believe has come up before. When a novel appears in two languages in the same publication, should the OMNIBUS type be used? If not, should the novel type be maintained for the publication with two novel title records contained, i.e. one for each language? If not that, which language should be chosen, or do we need an additional "multiple language" option? My own preference would be to use OMNIBUS containing the two title records for each language, one of which would be a variant of the other. --Ron ~ Rtrace^Talk 12:55, 4 September 2014 (UTC)

This was discussed here and here, with some consensus but no apparent resolution. Albinoflea 17:23, 4 September 2014 (UTC)

I believe it should be entered as an OMNIBUS, which can contain two titles with different languages. (You will have to update the language of the titles after entering them as contents.) The OMNIBUS title record should be in the original language of the author. Mhhutchins 18:20, 4 September 2014 (UTC)

As I said then, my opinion is that the book should be recorded as OMNIBUS. My interpretation of those discussions at that time was that this was more-or-less a consensus. Only Bill Longley suggested otherwise, and he said "this isn't a strong preference". So that's how I've been entering those books since then. (And I think I've done a half dozen or so of those.) And I agree with Mike on how to handle the languages there -- although that means I should go back and correct some of those earlier bilingual entries. (I think I did one trilingual one also.) Chavey 01:03, 5 September 2014 (UTC)

Sorry that I didn't find earlier discussion. I had searched for "dual language" instead of "bilingual". I've updated the record in question as well as its original paperback printing. While I generally agree with Michael's suggestion to use the author's language (German here) for the omnibus. I've made it English in this case. My reason is that the introduction appears in English only and the book is clearly intended for a primarily English speaking market. --Ron ~ Rtrace^Talk 12:31, 5 September 2014 (UTC)

Definition of OMNIBUS

Our current definition of OMNIBUS requires that one of the works contained in an omnibus be a novel. As a result of a discussion of a request to change a publication that does not contain a novel, but is clearly an omnibus by the common definition, I'd like to propose that we alter our definition by striking the words ", and at least one of them is a novel" from the current definition. Alternatively, we could specify a longer list of works allowed (novel, collection, anthology, nongenre or nonficiton), but I'm not sure that's necessary. I am assuming that we would not want to include CHAPTERBOOKs (as was not done in the specific example). Do folks think this is a desirable change? Thanks. --Ron ~ Rtrace^Talk 12:38, 10 September 2014 (UTC)

I agree with the thrust of this proposal, but it may be better to enumerate the title types that should cause a book to become an OMNIBUS. If we don't, then it could be argued that the presence of a previously separately published ESSAY makes it an OMNIBUS. Ahasuerus 19:13, 10 September 2014 (UTC)

I think this proposal is an excellent idea. If you need another example to support the proposal, then look at this series: All 4 books in the series are collections of previously published anthologies. In theory, our current rules forbid these to be entered as OMNIBUS, which really does not make much sense. Apparently someone had the heart to bend the rules in this case, and thus has entered the titles as omnibus. Enumerating the allowed title types, as suggested by Ahasuerus, is fine, although someone else would have to come up with the definition. Patrick -- Herzbube ^Talk 19:36, 10 September 2014 (UTC)

Yes, we have been bending the definition for some time now. We just need to document the de facto usage, and the clean-up records should reflect that. Might I also add, that when a novel-length work is published for the first time in an ANTHOLOGY (like Fighting Madness in this publication) or COLLECTION (the two novels in this publication), we will continue to type these publications as ANTHOLOGY and COLLECTION, and not OMNIBUS. A question: should a publication of novellas which were previously published as standalone books be considered an OMNIBUS? For example, should this be typed an OMNIBUS or remain an ANTHOLOGY? Either way would be fine with me, the answer just needs to be part of the standard definition. Thanks. Mhhutchins 20:36, 10 September 2014 (UTC)

I would also caution care on broadening the definition, but I approve of the general thrust of the idea. One that I have always found odd are the Tor Doubles, of which some are anthologies, some are collections, and 5 of which are omnibus. And yet they are really all the same format. Chavey 14:15, 11 September 2014 (UTC)

I'll tread lightly so as not to stir any memories of the Great ISFDB Bookist-Lengthist War: Check the lengths of the story contents. You'll find that occasionally they'd reprint full-length novels so that would technically be an OMNIBUS. Sometimes there would be two novellas by two different authors, so that would be an ANTHOLOGY. Sometimes there would be two novellas by the same author, so that would be a COLLECTION. They may have used the same "format" (two works of fiction), but the contents determines the "type" of publication. (Now Ace Doubles are handled differently, and that was the shot that started the war.) Mhhutchins 16:49, 11 September 2014 (UTC)

Oh, yes, I completely understand the reasons for the differences, because that's based on our existing definition of OMNIBUS. That doesn't mean I agree with it. Chavey 02:19, 13 September 2014 (UTC)

I don't see how changing the definition of OMNIBUS to include other types of publications (like ANTHOLOGY, COLLECTION, or NONGENRE) will change the classification of any of these publications which contain either SHORTFICTION or NOVELs. Please elaborate. Mhhutchins 03:28, 13 September 2014 (UTC)

You're right, it won't change any of them. Chavey 00:45, 14 September 2014 (UTC)

Yes, please: broaden the definition of OMNIBUS but let's enclose a list of possible required title types. Christian Stonecreek 14:44, 11 September 2014 (UTC)

Of the nine current types, I can see five of them being included: ANTHOLOGY, COLLECTION, NONFICTION, NONGENRE, and NOVEL; three of them not: CHAPTERBOOK, FANZINE, and MAGAZINE. And of course an OMNIBUS wouldn't include an OMNIBUS (although it's possible that such a publication could exist, we'd still only need to record the contents of the original publications.) —The preceding unsigned comment was added by Mhhutchins (talk • contribs) .

(unindent) ANTHOLOGY, COLLECTION, NONFICTION, NONGENRE, and NOVEL sound reasonable. BTW, I don't want to sidetrack this discussion, which seems to have reached consensus, but just as a point of reference for a future topic, we may want to create a Help matrix showing which title type combinations result in which publication types. Something like:

Title Type	None	NOVEL by the same author	NOVEL by another author	SHORTFICTION by the same author	SHORTFICTION by another author
NOVEL	NOVEL	OMNIBUS	OMNIBUS	NOVEL w/bonus story	NOVEL w/bonus story (?)
SHORTFICTION	CHAPTERBOOK	NOVEL w/bonus story	NOVEL w/bonus story (?)	COLLECTION	ANTHOLOGY
ESSAY	NONFICTION	NOVEL	NOVEL	CHAPTERBOOK	CHAPTERBOOK
POEM	CHAPTERBOOK	NOVEL w/bonus poem	NOVEL w/bonus poem	CHAPTERBOOK	CHAPTERBOOK

Ahasuerus 21:10, 12 September 2014 (UTC)

As per the consensus above, the Help page has been changed to cover ANTHOLOGY, COLLECTION, NONFICTION, NONGENRE, and NOVEL titles. Ahasuerus 19:07, 14 September 2014 (UTC)

The ellipsis problem revisited

Back in February we had a discussion about the way ellipses should be entered. To summarize the findings of that discussion:

• In English, you can use either ". . ." (as per The Chicago Manual of Style) or "..." (as per the AP Stylebook) to represents the ellipsis. To quote Jane Straus, the author of Blue Book of Grammar and Punctuation: "Newspapers, magazines, and books of fiction and nonfiction use various approaches that they find suitable. Some writers and editors feel that no spaces are necessary. [...] Others enclose the ellipsis with a space on each side. [...] Still others put a space either directly before or directly after the ellipsis."
• At some point in the now-distant past we realized that different editors had been using different ways of entering ellipses. This was causing problems with searching and indirectly increasing the likelihood of accidentally entering duplicate titles.
• To avoid these problems, we decided to standardize our usage and picked ". . ." over "..."
• What we didn't realize at the time was that in most languages only "..." is valid. (A quick review finds this this to be the case in French, German, Italian, Romanian, Spanish, Finnish, Russian, Kazakh, Hungarian, Polish, Danish, Azerbaijani, Norwegian, Swedish, Slovak, Slovenian, Serbian, Croatian, Turkish, Thai, Arabic, Hebrew, Bulgarian, Esperanto, Lithuanian, Latvian, Dutch, and Ukrainian.)
• Although our standard calls for using ". . .", the actual usage has been inconsistent, with 4,457 ". . ."s and 3,284 "..."s as of yesterday.

Based on the outcome of another, more recent discussion on the Moderator Noticeboard, I would like to propose that we change the data entry standard for all titles/pubs to "...". The proposed change will synchronize the languages represented in ISFDB and hopefully make data entry and searching easier.

If the proposal is accepted, I can write a simple data conversion script which will find and automatically convert all occurrences of ". . ." to "...".

Please note that the proposed change doesn't address the question of spaces before and after the three dots as the latter is a separate issue and may require a separate discussion. One discussion at a time :-) Ahasuerus 20:05, 14 September 2014 (UTC)

How about using the keyboard's ellipsis key as standard? (alt-semicolon = …). It appears as slightly more spaced than typing three consecutive dots, and less spaced than interpolating spaces between the dots. I also get 26 search results from this, mostly French titles. PeteYoung 00:29, 15 September 2014 (UTC)

I don't believe that keyboard combination is terribly universal; I've only ever used it in InDesign. Also, from a Unicode perspective it's equivalent to 3 unspaced periods, so from a search standpoint not much would be gained. But it would be something that we would want to include in the clean-up script and with whatever validation/conversion we wind up performing in the future. Albinoflea 02:17, 15 September 2014 (UTC)

I agree there should be one standard, and it should be three consecutive periods, and not a single keyboard character. I don't even know how to create a single character (I'm guessing it's probably an AlT+ code), and I doubt 1 in 50 users would know how either. Mhhutchins 02:27, 15 September 2014 (UTC)

In my own writing I always use the single character … for an ellipsis. And I think that would be a terrible standard for us to adopt, for the reasons Mike suggests. I strongly endorse the idea of switching from Chicago to AP style for ellipses, e.g. for improved consistency with other languages. (Aside: Chicago and AP both require spaces before and after the ellipsis.) Chavey 04:46, 15 September 2014 (UTC)

The change to "..." (without spaces) would be fine with me. Christian Stonecreek 08:16, 15 September 2014 (UTC)

I am in favor of "...". -- JLaTondre (talk) 21:48, 15 September 2014 (UTC)

I am in favor of without spaces, too. --MartyD 03:11, 16 September 2014 (UTC)

I had never realized one could use spaces between the periods, and it does look odd to me. I always use the single character … (on my French keyboard, it is alt + period), which goes faster. Does it cause a problem if one uses it instead of ... (which I much prefer to . . .) ? Linguist 16:03, 16 September 2014 (UTC).

Searches for … won't find titles using ... (and vice versa). So we should settle on one, and since most people don't use the single character to enter an ellipsis, we should make ... the standard. Mhhutchins 18:26, 16 September 2014 (UTC)

Back when I started editing here it took me a long time to get used to the spaces. I'm all in favor of "..." --Willem H. 18:52, 16 September 2014 (UTC)

Here, too: Please use 3 consecutive periods without spaces in between. Patrick -- Herzbube ^Talk 18:51, 17 September 2014 (UTC)

(unindent) It looks like we have pretty much reached consensus re: changing the data entry standard from ". . ." to "...". I will go ahead and convert the database/change the cleanup software accordingly and then change the Help pages. Thanks to everyone who has contributed to the discussion!

As far as the issue of using keyboard shortcuts goes, the problem is that they insert a special "ellipsis" character rather than three dots. We have 26 title records entered that way and they cause searches to fail. Thankfully, there is a fairly simple way to fix this problem: we can adjust the software to convert this special character to three dots at data entry time as well as during searches. This will allow editors to continue using their favorite keyboard shortcuts while enforcing the data entry standard. Ahasuerus 20:23, 18 September 2014 (UTC)

The software has been changed and Help has been updated.

Now that that has been settled, would it be desirable to make the software convert ". . ." to "..." at data entry time and during searches? Or are there situations where ". . ." may be legitimate? Ahasuerus 23:07, 19 September 2014 (UTC)

I can't think of such a situation. Even when there are visible spaces in a title, it appears to be a choice of the book's designer. That's why a standard was needed in the first place. Mhhutchins 18:58, 20 September 2014 (UTC)

Agreed. I say make the software convert them. If someone comes up with a use case later, we can deal with it then. -- JLaTondre (talk) 22:40, 20 September 2014 (UTC)

Done. I will post the details on the Community Portal. Ahasuerus 02:42, 22 September 2014 (UTC)

Proposal: add info about page count for unnumbered last page which has no different layout

I think it'd be good to make clear that the "count forward" rule for unnumbered pages applies also for last pages that have just the page number omitted but apart from that the same layout as the numbered pages (which is very common in Germany).

Proposal (green text) for bullet point 5 on Template:PublicationFields:Pages:

• It is fairly common for the last page of text in a book to have a different layout (design) which may not include a printed page number. There are also cases in which the layout of the last pages is the same as on numbered pages before but only the page number has been omitted. In all these cases, count forward to the last page of text and use that as the page count. For example, if a novel ends on the unnumbered page after page 244, enter the page count as "245" with an explanatory note about the unnumbered last page. Hitspacebar 20:29, 19 September 2014 (UTC)

I may be misunderstanding what you're trying to say, but it seems to me that the text you want to add has the same meaning as the preceding sentence. By "different layout (design)" it's saying that the page numbers may not appear as they do throughout the other pages of the text. The phrase "may not include a printed page number" is the same as "the page number has been omitted." Or perhaps I'm unable to see the difference. Mhhutchins 04:42, 20 September 2014 (UTC)

Maybe the misunderstanding is a language/translation problem between German and English. German uses the same words "Layout" and "Design" as well, but maybe with a slightly different meaning. If I have a case like described in my proposal, I would not consider it to be a different layout. It's the same layout, simply unnumbered. A different Layout in my (German) understanding is something that has a different look by using other fonts, bigger fonts, other page margins etc. I was scratching my head when I was reading that rule about counting forward for the first time because I didn't recognize the mentioned case in it. After having searched the wiki for my case without finding anything else I came to the conclusion that "different layout" probably includes that case as well. But it's not obvious from my point of view and adding the proposal might save new editors some time and help them. It definitely would have helped me. Hitspacebar 09:38, 20 September 2014 (UTC)

I can pretty much underwrite all that you have been writing (me = new editor, me = confused by how the words "layout" and "design" are used). However, we should probably rephrase the first sentence to become more clear, instead of adding a second sentence that merely states the same thing with different words. What about this simple sentence? "It is fairly common for the last page of text in a book to be unnumbered." Patrick -- Herzbube ^Talk 11:06, 20 September 2014 (UTC)

I like it. It works for me. Mhhutchins 14:59, 20 September 2014 (UTC)

Yes. works for me too. Much simpler, precise and clear. Hitspacebar 18:20, 20 September 2014 (UTC)

Clarification made. Thanks for bringing situations like this to our attention. Some of us who have been around awhile will miss things that fresher eyes will catch. Mhhutchins 03:08, 22 September 2014 (UTC)

Proposal: add explanatory page with examples about "Pages" field

Proposal for a new page containing some background information and examples about how to enter "Pages". I think a page like this could help understand why the rules are the way they are, especially for new editors. It should probably not be added to the Template:PublicationFields:Pages because it would bloat it too much. It should rather become a separate page, though I'm not sure where to put it. Maybe a new page called How to determine the value for "Pages" in the Help:How_to? It should then be added as a link in the Pages template.

I put the proposal on a separate page in order to keep this page cleaner. I hope that was a good idea. Discussions, boos and cheers should be going on here and the seperate proposal page can be amended based on the discussion here. Or should I put it all here?

So here is the proposal: User_talk:Hitspacebar/Proposal/HowToPagesField. The page is based on what I've read and learned about the "Pages" here so far. I hope I got it all right. What do you think? Hitspacebar 20:53, 19 September 2014 (UTC)

To my eye, it seems that an average user will look at that page and be overwhelmed by the minutest detail. But I see no problem with creating such a page to point at if a new user needs it. Linking it to the Hot to list would be the best way to go about it. We could also link it in the "Pages" template. Thanks. Mhhutchins 04:47, 20 September 2014 (UTC)

It would also be a good idea to indicate that this help page only refers to book publications. For periodical publications the rules are different...and much simpler. :) Mhhutchins 04:49, 20 September 2014 (UTC)

That page would be a welcome addition. But there really should be a BOLD headline or something similar to point out that this is a page concerning hcs, pbs and tps (and no magazines/fanzines/periodicals). Thanks for the effort! Christian Stonecreek 09:59, 20 September 2014 (UTC)

That's right. I changed the proposal accordingly, adding a note and changing all mentions of "publication" to "book". The title of the howto should then be How to determine the value for "Pages" in a book. Hitspacebar 18:24, 20 September 2014 (UTC)

Regarding "where" this page should go, I think that the Pages template should be modified to have a "For further details, see HERE" link. In fact, I think this kind of "Further details" page could be useful in several places. For example, with "Author Names", having a linked page which talked in further detail about things such as romanization of names, name ordering for languages where the family name is placed first, diacriticals, etc., might be useful. Chavey 10:54, 22 September 2014 (UTC)

I have created the page How to determine the value for "Pages" field in a book now and added links in the How-to and the Page template. Hitspacebar 18:00, 23 September 2014 (UTC)

Content pages

Re this help documentation: I have cleared up some confusion concerning how to enter contents on unnumbered pages, and made other changes that clarify the standards based on current practice. Mhhutchins 03:03, 22 September 2014 (UTC)

Reviews of works not eligible for the database

A concern was raised in this topic, concerning the conflict in the stated standards and a cleanup report which finds non-conforming records. A proposal was made to adjust the standards for entering reviews as follows:

• For reviews of books and short fiction that are eligible for inclusion in the database based on the Rules of Acquisition, create a record in the "Review" section of the data entry form.
• For reviews of media products (films, TV shows, games, recordings, plays, etc), magazines and fanzines (regardless of their genre), and books that are not eligible for inclusion in the database (graphic novels, nongenre novels by authors that are below the threshold, nonassociational nonfiction works), create a record in the "Content" section of the data entry form typed as ESSAY. If the review is not titled, this title format is suggested: Review of "Work" by Responsible Person(s). Descriptions of the type of work are also encouraged. For examples, see Review of "Fungi of Yuggoth" by H. P. Lovecraft, read by John Arthur, Review of the graphic novel "Agatha Heterodyne and the Clockwork Princess" by Phil & Kaja Foglio, and Review of the nongenre collection "False Gods" by Louis Auchincloss.

I support this change. Mhhutchins 18:01, 25 September 2014 (UTC)

I agree with the overall approach, but we may want to clarify the meaning of the word "plays" in this context. I think that we want to keep Avram Davidson's and Graham Andrew's REVIEWS of J. B. Priestley's The Thirty-First of June, which was published as a book. However, if a reviewer were to review a stage production of this play, I would think that we would want to enter it as an ESSAY. Perhaps we could change the wording to something like "live performances", which would also cover operas, ballets and so on? Ahasuerus 19:41, 25 September 2014 (UTC)

How about "stage productions" which encompasses them all? (NYRSF has been reviewing spec-fic plays in the NYC area for the past few years, and I've been entering them as ESSAYs with a descriptor, like "opera" or "play".) Mhhutchins 23:44, 25 September 2014 (UTC)

"Stage productions" looks good. Ahasuerus 05:52, 27 September 2014 (UTC)

I support this change. -- JLaTondre (talk) 00:09, 26 September 2014 (UTC)

I agree with this change. I like it. --MartyD 14:42, 27 September 2014 (UTC)

Principal support from me, but it seems to miss on radio plays (or let's call them 'acoustic plays', because there are productions that were published on LPs, CDs and the like, were never broadcasted, but aren't live performances either). Christian Stonecreek 08:49, 27 September 2014 (UTC)

Wouldn't that be covered under "recordings"? But we do need to point out that reviews of audiobooks are entered as REVIEWs. How about "dramatized recordings" which would cover all non-audiobooks? So the first part of the second bullet would be:

• For reviews of media products (films, TV shows, games, music and dramatized recordings), stage productions, magazines ...

Mhhutchins 15:31, 27 September 2014 (UTC)

One thing that occurs to me -- and this may be more in the nature of a general observation -- is that we may want to avoid duplicating what's currently stated in the Rules of Acquisition when updating our Help pages. If we ever change the RoA, as we have been known to do from time to time, our Help will get out of sync and getting it back in sync may not be easy as we have discovered. It seems much safer to point editors to the RoA page for the details of what's "in" and what's "out". Ahasuerus 15:04, 28 September 2014 (UTC)

We would, of course, have to remove RoA #5. And since we don't need to state specifically in the RoA about how to handle reviews (the RoA is about publication records, and not about content records), we'd record the change we're discussing here in the "What to Include" section below the RoA. Mhhutchins 16:12, 30 September 2014 (UTC)

Another thought: What are we going to do with the NONFICTION titles that exist solely due to reviews in SF magazines? There are probably a few thousand of them at this point. Ahasuerus 19:37, 29 September 2014 (UTC)

Some are going to be cleaned up because they appear on one of two cleanup reports: the nonfiction listings on Titles Without Pubs and Reviews Not Linked to Titles. If a review of a nonfiction work by a "below the threshold" author is a) linked to a title record, and b) the title record has a publication record, it will not show up on any cleanup report. And I don't think we should write one to find them, even if it's possible. I say let sleeping dogs lie. Eventually a few will be cleared by editors who are sticklers to the standards and are aware of this change. The other reviews will remain in the database as REVIEW-typed records linked to titles with publications, which really does very little harm to the integrity of the database.

If there is no more input into this discussion in the next day or so, I'll document the change on the help page and post a message on the community portal about it. In the meantime, I've been using the cleanup reports to find such REVIEWs and converting them to ESSAYs. Mhhutchins 16:12, 30 September 2014 (UTC)

I have made the changes in the Rules of Acquisition and the What to Include section of the Policy page. In updating the latter, I rearranged the individual possible contents into three sections to make it easier to read and understand. There may be areas where I may have added too much detail, but it doesn't appear to severely overlap the separately documented data entry standards. Mhhutchins 20:01, 8 October 2014 (UTC)

Proposal: Add Wiki page with example diagrams regarding "Page" field of content records

After making many mistakes while recording the "Page" field of content records, I have now made an effort to illustrate the newly worded rules (Template:PubContentFields:Page) with a few diagrams. It seems that my mind is better able to grasp the rules in a clear and unambigous way when they are presented in this way. Here is a link to the page: User:Herzbube/ExamplesForPageFieldOfContentRecords.

I still have a faint hope that it's not just me who is slow-witted, but that these pseudo-graphical diagrams might also help someone else to better understand the rules (along the lines "sometimes a diagram is worth a thousand words ☺). I would therefore like to propose that we add the Wiki page in question as a supplementary page to the Wiki help, in a similar vein as was done for Help:How_to_determine_the_value_for_the_"Pages"_field_in_a_book a couple of weeks ago.

Any feedback, corrections, etc. are welcome. If you think we need more examples, feel free to edit the page and add them yourselves. Open questions from my side for which I would welcome an answer:

• Examples 1 and 2: Is the designation "bp" allowed for the physical pages H and I?
• Example 5: Is the designation "ep" allowed for the physical pages V, W and X?

Thanks, Patrick -- Herzbube ^Talk 17:19, 28 September 2014 (UTC)

Another approach to explain it - yes, why not? The diagrams look a bit intimidating at first glance, but once you get familiar with the way they try to explain they work for me. As for you questions, here's my interpretation (which might be incorrect considering I'm a pretty new editor...):

Examples 1 and 2: I'd say "bp" is not allowed because "H" and "I" belong to the range of numbered pages where the page number can be derived. "bp" seems to be only for the range of unnumbered pages, otherwise the mention of the square bracket notation at the end of the "bp" rule wouldn't make sense to me.

Example 5: same as examples 1 and 2, same reason.

Hitspacebar 19:13, 29 September 2014 (UTC)

Presenting a standard visually is always a good idea. And I like what you've done here. Concerning Hitspacer's believe that "bp" is not allowed in the examples given, the rules provide that "bp" can be used alternately to record any content which appears before a range of numbered pages begin. (See the fourth bullet under the "Special Designations" subsection.) So Herzube's example is a valid representation of the current standard. Give me awhile and I may be able to come up with other examples of pagination, but at the moment you seem to have covered most bases. Thanks for your effort. Mhhutchins 20:34, 29 September 2014 (UTC)

You wrote: "'bp' can be used alternately to record any content which appears before a range of numbered pages begin". Maybe there's a misunderstanding, but that's what I was actually trying to say as well. And then I think my answer was correct: if "bp" is for content before a range of numbered pages only, then the answer to Herzbube's two questions raised above regarding examples 1, 2 and 5 should be "not allowed", because the pages in question are all in the range of numbered pages. Hitspacebar 21:40, 29 September 2014 (UTC)

Sorry, I see now you were answering his original questions. I missed that. Thanks for pointing out the error. Mhhutchins 22:12, 29 September 2014 (UTC)

Concerning "bp" and "ep": As the rules are currently worded, I agree with Hitspacebar's logic, but I had hoped I was wrong. Because if we think this through for "ep", then the logical conclusion is that "ep" cannot be used, ever! Whoa, wait a second, why should this be? As stated in the discussion points of visual example 5, it is impossible to have a range of unnumbered pages at the end of a publication if that publication contains at least one range of numbered pages. Thus, no range of unnumbered pages ≣ no page numbers with square bracket notation ≣ "ep" cannot be used.

Putting all logic aside, this is probably not how "ep" is intended to be used. The problem, as Hitspacebar has already pointed out, is the mention of the square bracket notation at the end of the "bp" and "ep" rules. If we were willing to rephrase the rules to merely state "although generally you can derive the page number", then we would be fine. It would allow "ep" to be used as it probably is intended, and as for "bp" this could be used on physical pages H and I in visual examples 1 and 2.

What do you think? Or should we discuss this as a separate topic? Cheers, Patrick -- Herzbube ^Talk 19:31, 1 October 2014 (UTC)

Sound fine to me. I've never used either "bp" or "ep" in my life (except when it comes to oil companies and extended play vinyl singles!) They were added originally because the magazine people were using them. I agree that all unnumbered pages after a range of numbered pages should continue the same pagination and that number should be entered in squared brackets, both in books and magazines. A content which begins two pages after the final numbered page 100 should be entered as "[102]". That's what most editors are using anyway. You could even get rid of the use of "bc" (back cover of a magazine) by having the editor count forward from the last numbered page. But I'd grant an exception in that case. Mhhutchins 20:28, 1 October 2014 (UTC)

(unindent) You are saying "A content which begins two pages after the final numbered page 100 should be entered as "[102]"." Thank you for mentioning this, because it made me realize that the visual example 5 which I gave was totally wrong - once again I mis-interpreted a tiny but important bit of the rules ☹. I have now fixed example 5. With this corrected, the entire discussion about "ep" has become moot, and rephrasing of the rules is not necessary. Apologies for the confusion. Patrick -- Herzbube ^Talk 02:18, 2 October 2014 (UTC)

Rules of Acquisition: Myths and Legends

Proposal: Myths and legends: Stories which were believed to be true, either by the writer or by the readers, are not included ("OUT"). This includes stories about, say, Greek and Roman gods during the time that these were viewed as true. By the time of Ovid's Metamorphosis, for example, the Romans no longer believe in such stories, and hence they had become pure fantasy for them. Lester del Rey views The Epic of Gilgamesh (ca. 2150-2000 BCE) as science fiction because it features a flood scene that in some ways resembles works of apocalyptic science fiction, and Wikipedia views it as "fantastic literature". But current scholarship suggests that this flood story is likely to be oral tradition descended from the filling of the Black Sea when the Mediterranean sea flooded into in around 5000 BCE. As such, this story is not included. This criteria eliminates all religious works from consideration, even works such as the Biblical "Revelations" which shares much in common with apocalyptic fantasies. (This does not exclude works such as The Left Behind series, which has a religious basis, but is written as speculative, not as "theology".) Chavey 19:06, 28 September 2014 (UTC)

(copied from the Community Portal)

It sounds like there may be a few different issues here.

First of all, our standard of what we consider fiction vs. non-fiction has always been whether the text is presented as fiction vs. as a factual account. Whether an account presented as factual is an accurate representation of what may have actually happened is not something that we consider when deciding whether a story is "fiction". Thus traditional religious texts are not considered fiction for ISFDB purposes simply because they are not presented as such. We treat stories of UFO encounters, alien abductions, modern miracles, ghostly visitations, etc similarly: if they are presented as fiction, then we list them as fiction, otherwise we list them as non-fiction (if at all.) To quote Frank Sinatra, "Nice 'N Easy" :-)

Granted, there are some borderline cases that we have to deal with on a case by case basis. For example, consider the Shaver Mystery. The original text was apparently a personal account of Shaver's contacts with "Deros" ("detrimental robots") and the dangers that they present. However, the edited version that appeared in Amazing Stories in March 1945 was heavily fictionalized and presented as fiction, although the editors kept things deliberately vague along the lines of "Truth or Fiction? You Decide!" We list the original story as fiction because it was presented as such, but many subsequent articles (by Shaver, Palmer et al) are not listed as "fiction" because, once again, they were not presented as such.

The next issue is the distinction between "speculative fiction" and "non-genre fiction". This is relatively easy to do when dealing with modern fiction, but it gets increasingly difficult as we go back in time. As our Rules of Acquisition say, "The older the book, the more likely we are to include it even if it is borderline eligible. This is caused by the fact that there were relatively few works published prior to 1800 and by the difficulties with distinguishing between speculative and non-speculative fiction (or even fiction and non-fiction) when you are dealing with pre-1800 works."

Even something as recent as Edgar Rice Burroughs' Tarzan (1912) could be questioned. There were certainly speculative elements in later books in the series (lost races, immortality, etc), but what was speculative about the first book? The semi-intelligent giant apes that raised Tarzan are the only thing that stands out and, given the state of evolutionary biology in 1912, it wasn't completely out of the realm of the possible as perceived by contemporary pulp readers.

When we go back almost three thousands years, all of these problems become magnified. Were Hesiod and Homer writing fiction or non-fiction? Would they even understand the question the way we do? Would they distinguish between "religious and non-religious fantastical events" as mentioned by Darrah above? That's hard to tell. For example, here is Homer's description of a chimera, the first known account of this beast:

• ... first he bade him slay the raging Chimaera. She was of divine stock, not of men, in the fore part a lion, in the hinder a serpent, and in the midst a goat, breathing forth in terrible wise the might of blazing fire. And Bellerophon slew her, trusting in the signs of the gods. Next fought he with the glorious Solymi, and this, said he was the mightest battle of warriors that ever he entered; and thirdly he slew the Amazons, women the peers of men."

That's an impressive collection of mundane (Solymi), legendary (the Amazons) and more-fantastical-than-anything-in-the-Odyssey (the Chimaera) foes, all intermixed! And that's precisely why the Rules of Acquisition read "The older the book, the more likely we are to include it even if it is borderline eligible". Ahasuerus 01:10, 29 September 2014 (UTC)

(end of the Community Portal quote)

Based on the above, I don't think we need to have a special RoA clause for myths and legends. We just need to make it explicit that works that were not presented as fiction are not considered "fiction" for our purposes. It won't address all of the borderline issues mentioned above, but at least it will clarify the basics. Ahasuerus 01:10, 29 September 2014 (UTC)

Apparently, we DO need a special note about religious mythology, since we just had some editors change "The Iliad" to genre, based on the presentation in that book of religious stories that were, so far as we know, presented as religious fact. Chavey 04:48, 29 September 2014 (UTC)

Well, I spoke for this change because a fictional character, Paris, has contact with the gods, which in my opinion is enough to categorize this as a work of fiction (as does the parallel plot on Troja, which is probable pure fiction, but at least fictionalized by Homer). Also, it is the 'prequel' to the Odyssey, and as such part of the same 'universe'. Christian Stonecreek 08:40, 29 September 2014 (UTC)

I have no problem with calling it "fiction". I just disagree with calling it "speculative fiction". And just because it's in the same universe, or same series, as a genre book isn't enough to justify classifying it as genre. We have several mystery series where some books in the series are speculative while others are not. We don't list the ones that are not. Chavey 14:10, 29 September 2014 (UTC)

I should add, though, that my primary argument isn't whether you or I view it to be fiction; it's whether the *intended reader* was expected to view it as fiction. And when the Iliad is speaking of Paris, I believe that the intended reader viewed him as a historical figure. And when the Iliad recounts him meeting with the gods, such as in the Judgement of Paris, I believe that the Greeks of that time would have viewed that as a religious truth, much as Christians view the appearance of Jesus to Saul/Paul on the road to Damascus as religious truth. Chavey 14:38, 29 September 2014 (UTC)

The problem that I see with the "intended reader" criterion is that it's hard to be sure what the intended reader may have thought/expected, especially hundreds or thousands of years ago. Again, let's take the Shaver Mystery as an example. Even though the first piece was published as fiction, there were many people who took it as a factual account, sent letters to the editor sharing similar experiences, formed Shaver Mystery societies, etc. The editor, Ray Palmer, continued catering to these readers while preserving a degree of respectability by using the previously mentioned "Truth or Fiction? You Decide!" approach. So what did the "intended reader" think in this case? And that's an example from the recent past -- things get much murkier when we go back hundreds and especially thousands of years.

That's why I think that "presented as fiction vs. presented as non-fiction" is a better, more objective and easier to use, criterion. There are some borderline cases like certain Whitley Strieber books, which were re-classified by the publisher, but at least we will deal with an objective criterion.

As far as the distinction between "speculative fiction" and "non-speculative fiction" goes, it can be hard to make if the line between what you know to be true and what you do not know to be true is not well defined, as was often the case with, say, medieval romances. Who, ca. 1200, could be sure that wizards, giants, fairies, etc did not exist in some faraway place? Even as late as the 19th century things could be murky, e.g. is Jules Verne's Five Weeks in a Balloon, which describes a voyage across (then barely explored by Europeans) Africa, speculative fiction? And what if the heroes had encountered a "lost race", which, for all we knew 150 years ago, was not out of the realm of the possible? That's why the Rules of Acquisition state that "the older the book, the more likely we are to include it even if it is borderline eligible."

That said, it's always a good idea to add notes and/or synopsis information when dealing with obscure, misleading or borderline cases. If there are indications that a tale that is fantastical by modern standards was not presented as such when it appeared, then adding this information to the record (citing the sources, of course) would make our data more accurate and more comprehensive. And that's always a good thing :-) Ahasuerus 16:46, 29 September 2014 (UTC)

And to make the waters even more murky, almost all fiction published before the 18th century could be considered speculative fiction, since the concept of realistic ("mundane") fiction didn't exist, at least not in English. Add to that the fact that our definition of "book" did not exist before Gutenberg's invention. The emphasis on "published" is relevant, since that is an important ISFDB standard for inclusion in the database. (Handwritten books are another matter altogether.) So if a myth or legend was published in an explicitly credited book after 1440, and the author presented them fictionally then it should be eligible for inclusion in the ISFDB. Whether the myth or legend is or was or ever will be believed by its readers is not particularly relevant to our purpose. If the myth or legend is intentionally presented as an objective narration in a work of scholarship or theology, I would not consider it to be eligible for inclusion, unless the author is "above the threshold". Mhhutchins 21:38, 29 September 2014 (UTC)

In the case of The Iliad, there's no reason to believe that Homer was presenting it fictionally. We view it as fiction, but there's no reason to believe he did. Experts argue as to whether he believed his stories to be true or not, but what he presented was presented as "the legends of our people". Chavey 07:14, 30 September 2014 (UTC)

As there isn't even known if one person called 'Homer' actually existed and is the author of The Iliad and The Odyssey, I'd say that — assuming he/she is this one author — it is even more guesswork of what his/her intentions were. I personally can't imagine the author as one who didn't know that the works in question were not based on true happenings. I believe it to be impossible for any author (even authors of hackwork) to suspend the thinking on his/her works. This applies even more to an author of Homer's stature. And we seem to really don't know how the works in question were presented by Homer. The earliest version of texts we know may very well have been edited and presented differently from the author's intentions. In any case we should go along the rule "the older the book, the more likely we are to include it even if it is borderline eligible."

Homer certainly didn't think of The Iliad as fantasy or GENRE, but so probably didn't the authors of the last mid-century's Utopias (perhaps they would have categorized their works as NONFICTION, if they had that category). Christian Stonecreek 08:10, 30 September 2014 (UTC)

Use of commas in the Page Number field

We have 10 publications (e.g. [4], [5] and [6]) which use commas as delimiters in the Page [Number] field in the Contents section. We also have Feature request 370, which reads:

• If an artist is listed as having artwork on, say, pages "3, 45", then this should be shown in the publication display as appearing after things on pp. 1-2 (and before items on later pages).

However, I can't find a reference to this format in Help. The relevant part says:

• If each story in a collection, anthology or magazine is illustrated by one artist, the artwork can be indexed with a single entry for each artist and story. For example, the December 1956 issue of Infinity Science Fiction contains the story "My Sweetheart's the Man in the Moon" by Milton Lesser, which is illustrated by Stallman. The story starts on page 5 but the first piece of artwork is on page 9. In this case there is only one illustration for the story, but if there were multiple illustrations a single entry would still suffice. The page number given is the page number of the start of the story in this case; the title is the title of the content item being illustrated. An alternate option is to enter the first page where artwork appears. Page 9 would also be acceptable in the above case.

It wouldn't be hard to add support for comma-delimited page numbers, although we may need to change the database definitions to allow more than 20 characters in the Page field. Would this be desirable or should we change the 10 affected publications to use multiple INTERIORART records and reject the FR? Ahasuerus 04:37, 12 October 2014 (UTC)

There's no reason that I can think of that a content record would ever need to have a comma. The page number on which the content begins (or appears in the case of interiorart) should be entered into the field. If an artist has work on page "3,15" then there should be two content records: one for the piece on page 3 and one for the piece on page 15. If an editor chooses to use the optional method of creating a single record for interiorart, then the page is left blank. Do you know who made the feature request and ask that they join the discussion to explain why two works should be entered as one content record? It appears that they were asking to have the ability to sort a content record among others. If that's the case, the FR is moot, since we now have the ability to sort, regardless of the characters used: comma, colon, semicolon, and any character an editor enters into the field. Thanks. Mhhutchins 05:25, 12 October 2014 (UTC)

The request was made by Darrah, although I see that a couple of other editors have used commas the same way. Darrah is currently on hiatus, but should be back on Sunday or Monday. Ahasuerus 06:14, 12 October 2014 (UTC)

See Multiple page numbers for the same Content title? further up this page. Also, we had a this discussion awhile back about duplicate content (mainly artwork) appearing in a publication. Using commas seems like a much better approach to me than to creating multiple records for the same work. -- JLaTondre (talk) 11:55, 12 October 2014 (UTC)

Yes, I agree. That would be a good usage of commas...but that doesn't appear to be the purpose of the original request. There's nothing in the current software or standards (that I'm aware of) which disallows the use of a comma to list the pages on which the same work appears in the same publication. And with the pipe character we can now sort it despite what is displayed in the record or what characters are used in the field. (God forbid that anyone would need to have more than 20 characters to record such contents!) Mhhutchins 17:31, 12 October 2014 (UTC)

My feature request, and my use of, comma delimited page numbers was primarily for artists who have multiple untitled artwork pieces in a work. This occurs frequently in fanzines, where the art is almost never connected to stories. IMHO, giving an author 10 credits for 10 little sketches scattered throughout a fanzine, all listed as "Untitled" or as "Bogus fanzine, July 1978" is overkill in the credit department, and doesn't add any information. If the artist had separate titles/captions for their artwork, or if they were associated with different stories within an anthology or magazine, I'd want to give them separate credits. In a similar vein, if a novel has 6 pieces of art by an artist, I wouldn't (personally) want to list them all as 6 different content items, but it would be nice if I could add the page numbers for all 6 of them in a single content listing, which was then listed under the first artwork piece. (Just my opinion; and I don't object to be outvoted, but I think it could be useful.) Chavey 22:08, 12 October 2014 (UTC)

Nothing in the software currently prevents anyone from using a comma in the page number field, but the content won't display in order. That's where the new sorting function comes any. The problem with your proposal to use page numbers for multiple works is that it rubs against the standard: one record for each different work or one record to represent all works. In the latter case, you'll find it's going to be difficult to limit the number of pages to add to a single field. Consider publications that have dozens of different works. (That's why an editor is given the option of creating one record instead of one for each of the dozens of individual works.) Adding multiple pages in the page number field can be confused with the purpose which JLaTondre proposes. A user wouldn't know whether it means multiple appearances of the same work or the appearance of multiple different works. Mhhutchins 23:30, 12 October 2014 (UTC)

Rules for creating variants and pseudonyms for "unusual" characters

I recently attempted to create some variants and pseudonyms for titles and author names that appear in publications with "unusual" characters. I made a short list to keep this discussion together, but if anyone is interested he/she can find the relevant discussions here and here.

• "Stanislaw Lem" vs. the canonical "Stanisław Lem" (note the l vs. ł)
• "Jorge Luis Borges" vs. the canonical "Jorge Luís Borges" (note the i vs. í)
• "Kobo Abē" vs. the canonical "Kobo Abe" (note the ē vs. e)
• "The Oögenesis of Bird City" vs. the canonical "The Oogenesis of Bird City" (note the ö vs. o)

Apparently we do not create variants or pseudonyms when the difference in title or author name is merely an accented letter or some other diacritical mark. So this takes care of bullets 2-4 above. But apparently the difference between the letters "ł" and "l" is a different matter and thus warrants the creation of variants and a pseudonym.

I have difficulties distinguishing the two cases. Why do I need to treat "ł" any different from, for instance, "ē"? Can we please establish some rules (or at least guidelines) when pseudonyms and variants are desired, and when not? I am not sure how to get there, though, because its hardly practical (or possible) to enumerate all characters/accents/diacritical marks that warrant creating pseudonyms and variants. But maybe we can compile a list of those that occur most frequently?

I would also like to point out the following sections of the current rules which strike me as incorrect/inaccurate and should be changed:

• Template:TitleFields:Author
  • Contains this statement: "Accented characters. If you are entering a name such as "Philip José Farmer" that is printed with an accented e, that accented character should be reproduced in your entry of the name. Two versions of an author's name that are printed with and without accents are treated as variants; you should not convert one form to another."
  • In my opinion, this contradicts the current practice. If people were actually following this rule, why is there no pseudonym "Philip Jose Farmer" (it's very likely that there have been many publications without the accented "e")?
• Template:PublicationFields:Title
  • Contains this statement: "Symbols and punctuation. Strange symbols should be entered if appropriate typographical characters exist. [...] Other characters should be entered in Unicode if possible; this includes accented characters, and symbols such as em-dashes."
  • I understand the statement to mean that editors should enter characters exactly as they appear in the publication, which would logically entail creating variants - especially for accented characters which are explicitly mentioned here.

Any thoughts? Patrick -- Herzbube ^Talk 20:07, 16 October 2014 (UTC)

One other thing that makes the issue more difficult for me: In my native language, German, the umlaut characters "ä", "ö", "ü" are differently pronounced than the normal characters "a", "o", "u", so I would consider these to be entirely different characters. Taking away the diacritical mark "¨" would mean an actual difference in spelling, and therefore would require, in my opinion, creating a pseudonym or variant. Patrick -- Herzbube ^Talk 20:17, 16 October 2014 (UTC)

The current data entry rules are admittedly muddled, but there is a method to this madness. Well, sort of.

The underlying problem has to do with the way the ISFDB database and software are configured. There is a technical discussion here, but the short version is that the software doesn't distinguish between certain "regular" and "accented" characters. For example, if you enter a new book by "Philip Jose Farmer", the software will create a record under "Philip José Farmer" instead. Similarly, "i" and "ï" are treated the same way. On the other hand, "o"/"ö", "a"/"ä" or "l"/"ł" are treated as separate characters. This is a problem for author, publisher, publication series and series records because the software checks newly entered values against what's already in the database (in a case-insensitive manner, I should add) and uses pre-existing values where available. This means that if you have two authors, e.g. "Wolfgang Werlé" and "Wolfgang Werle", you have to disambiguate them using a disambiguator like "(II)".

Unfortunately, this is not an easy problem to fix. Ideally we would want to upgrade the database to version 5.5, convert the data from Latin-1 to Unicode and change the software to undo many of the assumptions about the way the data is stored that have been made over the years. It will also make the database much larger, but oh well.

As far as title records go, we don't have this problem because the software doesn't do any kind of title matching.

Anyway, hopefully this sheds some light on the origins of the current data entry practices. Ahasuerus 00:51, 17 October 2014 (UTC)

Yes, your explanations and the technical discussion you linked to have certainly been enlightening. Two questions that arise from this information:

• Let's assume we could magically fix all the technical problems, and the software would no longer implement any workarounds for these problems: Would we then record pseudonyms and variants for all differences in characters? For instance, would we create a pseudonym "Philip Jose Farmer"? If the answer to this is an unconditional "yes", then I withdraw my request for establishing rules. -- Herzbube ^Talk 10:09, 17 October 2014 (UTC)

I suspect that the answer will depend on the nature of the fix. For example, if the "fix" enables us to have two separate records for "Philip José Farmer"/"Philip Jose Farmer" while simultaneously preserving the current search behavior, i.e. the ability to find "Philip José Farmer" when searching for "Philip Jose Farmer", then it will make it easier to agree that we should enter author names exactly as they appear. Ahasuerus 20:46, 17 October 2014 (UTC)

• Is it possible to have a list of those "accented" characters treated specially by the software? Or is the list too long? Still evolving with frequent changes?

Thanks, Patrick -- Herzbube ^Talk 10:09, 17 October 2014 (UTC)

I don't think this list has been compiled yet, but whatever it is, it should be very stable since we haven't changed the relevant code in years. I hope to make some changes to this area in the next month or two as I enhance support for non-Latin (Japanese, Cyrillic, etc) author names, but I don't expect the changes to affect "the list". Ahasuerus 20:46, 17 October 2014 (UTC)

Verse included in a larger work

Well, I've gone and done it again. I included the verse of Voyage of Slave as part on the content, several years ago before I found out that this is frowned upon. I feel that it should be, as it is separate content, like an illustration, But, what do I know? Authors can later reprint these poems into chapterbooks, or collections separate from their initial appearances. We've seen this in the way of Lewis Carroll's works. And what do we do if an author reprints another author's verse in his work? Would this be an exception to the rule? I understand that this is a slippery slope. I've been asked to bring this to the experts, and now I've done so. So let me know. MLB 01:43, 18 October 2014 (UTC)

I didn't want to add my opinion until others have had an opportunity to join the discussion. Since there hasn't been one, I'll just link my take on the situation here. Mhhutchins 03:42, 20 October 2014 (UTC)

I'd opt for not including such poems unless they have been published separately. If they were later published, I'd be fine including them in the original work (so, I differ a bit from Michael's opinion). However, I don't have a strong opinion on either aspect. -- JLaTondre (talk) 00:20, 22 October 2014 (UTC)

I can't imagine separating out all of the songs from Lord of the Rings as separate content items, and I wouldn't want to do that even if someone published a collection of all of Tolkien's songs (and hence they all had separate listings). As such, I would thus be opposed to doing so for anyone else's work. (Hence, I think I'm in agreement with Michael.) Chavey 04:09, 30 October 2014 (UTC)

Proposal: add shortfiction designation - vignette

I have been editing content of small press magazines recently, and many times I come across 'short stories' that are less than 1,000 words long, most less than 500 words, and one was only 22 words long. Calling these "short stories" sort of leaves a bad taste in my mouth, so instead of flagging them as short story, I leave them or change them to the shortfiction state, since ISFDB does not make a distinction for any stories that are less than 1000 words. However, as the guideline states, story length for shortfiction (no flag) is for stories of UNKNOWN length, for instance, with unverified publications where data was collected from other sites. Since I DO know the length of a particular story when I am editing, I am not entirely correct calling them shortfiction. Locus1 uses a label for "short short" content called vignette (vi). I would like to see this incorporated into the drop down menu for shortfiction, giving these stories a label other than 'short story'. I do realize that doing so is not without its inherent problems. How do you change the thousands of 'short' short stories that are already in the database? How do you even search for them? Some, like Feghoot and Breadfruit are easy to find but what of the rest? I don't know if this is even viable (computer is not one of my strong points). With that said, I see three possibilities:

• Add 'vignette' to the dropdown menu for shortfiction (proposed)
• Leave 'short' short stories as "shortfiction" (proposed)
• Label 'short' short stories as "short story" (current practice)

I would like to hear how other editors feel about this or if they have other suggestions. John Syzygy 17:59, 24 October 2014 (UTC)

I'm against such an addition. Why stop at the 1,000 words mark ? We can also implement the drabble at 100 words, and so on (why not the haiku ?) this has potentially no end. Hauck 19:07, 24 October 2014 (UTC)

You say that you "know the length of a particular story" when editing, and that it is "not entirely correct calling them shortfiction". If you know the length, and follow the guidelines, then you wouldn't have to resort to leaving them as "shortfiction". You'd enter the length designation as outlined in the current guidelines. Those designations are standard throughout the field. Any story less than 7500 words is a "short story". I see no benefit in creating standards beyond those that have been accepted by the SF community for decades. It's difficult enough to ascertain word count. Adding another layer of detail would make it even more so. Mhhutchins 20:14, 24 October 2014 (UTC)

I'd be against the "vignette" proposition, too: converting the all right ss records to such a designation would take forever, which means there would always be a large remainder of records unconverted and therefore with incorrect designations. A vignette is still technically a short story, as is a drabble. Leaving 'short' short stories as "shortfiction" is also potentially confusing: we already have too many shortfiction records that are not described as ss, nt or nv because no editor has worked on those publications yet and the stories' lengths remain undetermined, so leaving 'short short' records as "shortfiction" would be a step back from the clearer picture we aim for. I'm all for keeping current practice. PeteYoung 01:00, 25 October 2014 (UTC)

It has now been a few days since I posted this proposal and I would like to take some time here to reply. From your responses, you are all in agreement that story length should remain as is without adding a designation below 7500 words because it adds a level of complication not only to editing publication records, but also with the records that are already in the database. This I understand, my inclination is toward accuracy and display of information, especially for the users who use such information for research and reference. If this is the limit that most editors and users accept, so be it, I will adjust and comply. But if it is simplicity and ease of editing you're looking for, why not call all fiction shorter than novel length short fiction and be done with it? No more counting pages, no more wondering how to edit entries, eliminate the story length menu altogether. Just because major awards and magazines parse short fiction down into novella and novelette categories doesn't mean we have to. Call it all short fiction. Is this a change we are willing to consider? I'm pretty sure this has about as much chance of success as this original proposal, since as Mhhutchins stated above, this is how it has been done for decades within the community. So, that's about it, I don't see any change in story length designation coming any time in the near future (if at all). Chalk this one up to my inexperience and naivety for posting this proposal. I'll consider this subject closed for now. John Syzygy 17:10, 29 October 2014 (UTC)

Well, there is a fair amount of history here. Back at the dawn of time, Al and I looked at the very detailed list of categories and sub-categories maintained by the Locus guys and said "Nah, that's too detailed". We didn't do any kind of detailed analysis, it was just a gut feeling, so we went with the current, rather minimalistic, set of supported values.

In the late 2000s, as more editors became involved, especially when development was resumed in 2009, there were various proposals to add new title types as well as "story length" values -- see this discussion and its 21 (!) offshoots. None of them got anywhere, in part because we underestimated how time-consuming making significant changes to our then-buggy software would be. Even now, 5 years and many hundreds of patches later, certain kinds of fundamental changes would be rather painful.

I expect that these issues may be re-visited in the foreseeable future as we start making long-delayed changes like revamping NONGENRE data. Ahasuerus 01:56, 30 October 2014 (UTC)

SERIALs in ANTHOLOGIES, COLLECTIONs and CHAPTERBOOKs

A while back some kind soul came up with a list of SERIALs whose title dates did not match the date of publication that they appeared in. I was going to create a cleanup report to replicate this functionality, but a prototype report found so many occurrences (400+) that I would like to clarify the current rules before I finalize the report.

Help:Screen:NewPub says that "Serial installments of a work are always given the date of the magazine in which they appear even if the work has been published previously in book or serial form." Similarly, the Date section of Help:Use of the SERIAL type says "Serial installments of a work are always given the date of the magazine in which they appear even if the work has been published previously (or subsequently) in book or serial form". However, under Multiple Serializations it says "...each serial tile record gets the date of its actual appearance and title records for different serial publications are not merged unless all installments are known to contain the exact same text."

A related issue is that the Current Practice section says "Reprints in anthologies, collections, or omnibuses, or under separate covers, are not recorded with the SERIAL type, but a reprint that is in fact serialized, or a reprint of a novel in a single issue of a periodical, would use the SERIAL type." However, we have a few hundreds non-MAGAZINE/FANZINE pubs -- mostly facsimile reprints entered as anthologies -- which contain original SERIAL records. Obviously, if a SERIAL record is found in 2+ publications, it can't have the date of "its" publication because there is no single publication date.

So I guess the question is whether there are circumstances (like the previously mentioned facsimile reprints) where we want to associate SERIAL records with multiple publications. If we do, then we should document them in Help to make sure that we all use the same rules when working on cleanup reports. Ahasuerus 00:31, 31 October 2014 (UTC)

Yes, I believe we should be able to associate serializations appearing in multiple publications, even if a publication is typed as ANTHOLOGY. One of the fundamental concerns here is the practice of typing facsimile reprints as ANTHOLOGY instead of MAGAZINE. As far as I can remember, this subject was never brought up for discussion by the entire group. It appears to have been agreed upon unilaterally by a small contingent of editors who worked on magazine entry. (They created other standards as well without group discussion, but I won't go into that here.) Regardless of that, I believe that we should still limit SERIALs to periodicals, facsimile anthologies and CHAPTERBOOKs. The latter type because of the recent practice of publishing novels serially in ebook format before a complete publication, e.g. The Human Division. Mhhutchins 02:32, 31 October 2014 (UTC)

I just remembered this relatively recent reprint of a serial in book form: Zelazny's "...And Call Me Conrad" in this collection. We have to allow for exceptions to the rule (as in this example), but it would be good to have a clean-up report to find all non-standard records so that a moderator can make the decision to allow it or not. Mhhutchins 02:36, 31 October 2014 (UTC)

Evolving magazine titles

The other day Darrah submitted the following Feature Request:

• When a magazine changes name over the years, we combine their titles into a single overall title series record. But then some of the names by which it was known are hidden from the "Magazine Directory" pages. For example, "Science Fiction Classics" (at http://www.isfdb.org/cgi-bin/pe.cgi?26070) changed its name to "Science Fiction Adventure Classics" in 1969, and then to "Science Fiction Adventures" in 1973. But those titles are not in the Magazine Directory lists of titles. So if you have an issue of one of the latter two titles, and try to look them up in the Magazine Directory, you will be unable to find them; i.e. it will look like we do not have that magazine indexed. This should be changed so that all of the editor titles under a series title are included in the Magazine Directory.

My first reaction was that this is a data entry issue rather than a software issue. Here is what our current listing for "Science Fiction Classics" looks like:

Sub-series of The Ultimate Publishing Reprint Digests
   3 Science Fiction Classics (View Issue Grid)
       Science Fiction Classics - 1967 (1967) [ED] by Herb Lehrman [only as by Ralph Adris ]
       Science Fiction Classics - 1968 (1968) [ED] by Herb Lehrman [only as by Ralph Adris ]
       Science Fiction Classics - 1968 (1968) [ED] by Herb Lehrman [only as by uncredited ]
       Science Fiction Adventure Classics - 1969 (1969) [ED] by Herb Lehrman [only as by uncredited ]
       Science Fiction Adventure Classics - 1970 (1970) [ED] by Sol Cohen [only as by uncredited ]
       Science Fiction Adventure Classics - 1971 (1971) [ED] by Sol Cohen [only as by uncredited ]
       Science Fiction Adventure Classics - 1972 (1972) [ED] by Sol Cohen [only as by uncredited ]
       Science Fiction Adventures - 1973 (1973) [ED] by Sol Cohen [only as by uncredited ]
       Science Fiction Adventures - 1974 (1974) [ED] by Sol Cohen [only as by uncredited ]

In this case my preferred approach would be to split this series into three: one for Science Fiction Classics, one for Science Fiction Adventure Classics and one for Science Fiction Adventures. That way they will all appear in the Magazine Directory and there will be no need for software changes. If we want to preserve the relationships between these three titles, we can set them up as sub-series of "Science Fiction Classics Reprint Magazines" or some such, which will become a sub-series of "The Ultimate Publishing Reprint Digests".

On the other hand, we have magazine series like Amazing Stories where the title has alternated between Amazing Stories, Amazing Science Fiction and Amazing Science Fiction Stories. If we were to create three separate series for these three forms of the title, it would only confuse our users.

After thinking about it, it looks like there are two ways to accommodate these different scenarios:

1. Change the Magazine Directory logic to use magazine titles when building the directory. It won't be a trivial change because the logic will need to determine where the magazine name ends and the year/month begins -- think of the various way we separate them with commas, dashes, the pound sign, etc. It may have performance implications, but I don't know that for sure yet.
2. Leave the Magazine Directory logic as is and resole that if a magazine title change merits a separate entry in the magazine directory (like in Darrah's example), then it also merits a separate series name. Thus we will create three sub-series for "Science Fiction Classics", but keep "Amazing Stories" as one series.

My personal preference would be #2. What do the other editors think? Ahasuerus 23:24, 6 November 2014 (UTC)

I don't use the Magazine Directory when looking for magazines. I use the simple search box with the Magazine drop-down selected. Problem solved. I don't see why we should create separate series every time a magazine changes its name. If we had separate series for "Isaac Asimov's Science Fiction Magazine" and "Asimov's Science Fiction", we wouldn't have a single magazine grid. This benefit outweighs any problems caused by using the Magazine Directory to search for a title instead of using the Search function. I would personally choose to have the Magazine Directory fixed rather than having separate series, since that was the original request (if I understand it correctly). Put me down for #1. 00:38, 7 November 2014 (UTC)

I hadn't thought of the impact on Magazine searches. Checking its behavior, I see that it uses the same series-based logic as the Magazine Directory. Thus if you search for "Science Fiction Adventure Classics" or "Amazing Science Fiction", you won't find anything.

On the other hand, the software does support a unified view of magazine sub-series. You can go from the Issue Grid for Science Fiction Classics to the Issue Grid for its parent series, The Ultimate Publishing Reprint Digests. As an added complication, the software doesn't consider a series without EDITOR titles a "magazine series", so a Magazine Search on "The Ultimate Publishing Reprint Digests" -- which contains no title records, only sub-series records -- won't find anything even though you can get to it from one of its sub-series.

Hm, this is not good. I will have to think of ways to make the behavior more consistent. Thanks for the feedback! Ahasuerus 00:59, 7 November 2014 (UTC)

What if the simple Search under the Magazine drop-down looked for EDITOR title records instead? That would pull up any matching records including "Science Fiction Adventure Classics" or any other "hidden" magazine titles. Mhhutchins 02:37, 7 November 2014 (UTC)

Yes, that's the alternative that I am currently researching. Presumably we don't want to display raw EDITOR records -- a search on "Amazing" would return hundreds of them -- so what we want the Search logic to do is find all matching EDITOR records and display the series names that they belong to. It looks promising so far, but I am not sure it will help with the Magazine Directory issue. More to come... Ahasuerus 03:01, 7 November 2014 (UTC)

I have changed the search logic on the development server so that it now checks series records as well as EDITOR records for matching values. The results are better in some ways and worse in other ways. For example, here are the first 10 magazine series that are displayed when I search on "science":

2001: Das deutsche Science Fiction Magazin
Aboriginal Science Fiction
Aboriginal Science Fiction Anthology
Absolute Magnitude
Amateur Science Stories
Amazing Stories
American Science Fiction
Analog
Apex Science Fiction and Horror Digest
Apex!

The reason that Absolute Magnitude, Amazing Stories, Analog and Apex! are displayed is that some of their EDITOR titles include the word "science", e.g. "Absolute Magnitude & Aboriginal Science Fiction - 2002" or "Apex! A Different Vista for Science Fiction and Fantasy #1". Ditto Journey Planet, Imagination, Niekas, Salarius and a number of other magazines. I guess we should display a note at the top of the page explaining that the results include EDITOR records. Ahasuerus 03:26, 7 November 2014 (UTC)

I knew it wouldn't be a trivial FR, but I certainly didn't expect some of these complications. I would think that with a series such as "Absolute Magnitude", where some of its title recs contain "Science" and some do not, that it would be more natural to only report the EDITOR records that contained the search term, e.g. to only report "Absolute Magnitude & Aboriginal Science Fiction" and not report "Absolute Magnitude" at all. That of course raises the challenge of what page to go to if the user clicks that title. Unfortunately, it would seem that the only place to go is to the parent series when the EDITOR rec isn't the same name as the magazine series. But then what we're going to isn't quite the same thing as was being shown to us. Ick. But that may still be better. Chavey 06:42, 7 November 2014 (UTC)

After sleeping on it I came up with a somewhat different approach to this can of worms. As of 5 minutes ago, if you do a search on "Science Fiction Ad", you will see 6 "Science Fiction Ad..." series, "Science Fiction Classics" and "Thrilling Science Fiction". The last 2 also have "[matching magazine title]" next to the series title to indicate that, although the series title does not contain the search string, one or more of the magazine titles do. Does this layout look better? Ahasuerus 20:47, 7 November 2014 (UTC)

I have made the same change to the Magazine Directory logic -- see the Community Portal for the details. Ahasuerus 03:14, 8 November 2014 (UTC)

I like the change with respect to a Magazine Search from the pull-down menu. For example, entering "Science Fiction Adventure Classics" now takes you directly to the correct series, i.e. the main title containing that EDITOR title. However, this title is still impossible to find from the Magazine Directory page. It correctly finds the parent title that contains the title we're looking for, but since there are 62 hits under "Sc", there's no reasonable way to identify which is the series title that contains "Science Fiction Adventure Classics". If including the actual Editor titles is not possible (or not easy, or otherwise problematic), one alternate solution might be to have a "Search within these titles" field & button on the Magazine Directory pages. Chavey 20:47, 8 November 2014 (UTC)

Oh, I see. Yes, that's a very good point. Let me see what I can do... Ahasuerus 21:43, 8 November 2014 (UTC)

(unindent) OK, I think I got it. Could you please take a look? Ahasuerus 07:00, 11 November 2014 (UTC)

P.S. Magazine Directory and Magazine Search have been synchronized to display the same data. Ahasuerus 07:21, 11 November 2014 (UTC)

The original Feature Request was about being able to find the magazine "Science Fiction Adventure Classics" from within the Magazine Directory. That remains essentially impossible to do. Chavey 16:00, 12 November 2014 (UTC)

Unfortunately, "Science Fiction Adventure Classics" is a particularly difficult case because the name of its series, Science Fiction Classics, starts with "Sc", so it is already displayed in the "Sc" directory. Currently, the algorithm ignores magazine titles if the magazine's series is eligible for display.

I have spent a few hours looking for a compromise solution, but I couldn't think of way to (reliably) determine that "Science Fiction Adventure Classics" should be listed, but, to use a random example, "Science Fiction Adventures, November 1952" should not be listed because it's already covered by "Science Fiction Adventures (1952-1954)". Just too many possible permutations for my brain, I guess. If anyone can come up with a better algorithm for function SQLFindMagazine in module common/SQLparsing.py, I will be all too happy to implement it :-) Ahasuerus 20:42, 12 November 2014 (UTC)

This sort of topic again begs the question of whether magazines should be pub series (I am not saying we should not have editorial title records and title series for such content but that the magazine itself is really a pub series). Pub series obviously do not solve this multiple name problem but series in general (be they obvious title series or pub series) probably need some sort of variant naming concept anyway. Uzume 16:12, 23 January 2015 (UTC)

Adding a new pub binding format

Because of the relatively recent rise of periodical/magazine publications using the "perfect bound" method favored by POD publishing, I think we should add "tp" as an option under the "Print magazines" subsection of this help section. Any objections, suggestions, etc? Mhhutchins 17:15, 17 November 2014 (UTC)

Sounds reasonable to me. Ahasuerus 00:27, 18 November 2014 (UTC)

The current magazine pub formats are all based on size rather than binding. If we were to allow tp, we would have to have clear rules on when it should be used instead of the existing size options. I assume that the intent is not to place pulps and digests into the tp category. Though I don't know how you determine the difference between, say, a current issue of Fantasy & Science Fiction which is perfect bound with stiff card covers, and a trade paperback. I also wouldn't want to go back and retrofit magazines that had been entered under the old standard (e.g. Weird Tales from the early 90s, late Whispers, late Pulp Vault) Those are all clearly perfect bound and would be tps if they were books. Other online magazine bibliographies do list "TP" among their abbreviations (or "B-format paperback", in the case of FictionMags and Galactic Central), but I don't know that I've ever come across a listing in those sites that use it. Ultimately though, my concern is that we'll have 9 options based on size and 1 based on binding type and I wouldn't know which to choose when 2 options apply. --Ron ~ Rtrace^Talk 03:00, 18 November 2014 (UTC)

The intention behind adding "tp" as an option for print magazines is based on the fact that there are now periodicals that look exactly like trade paperbacks, in both size and binding. They look nothing like digests and pulps of yore. Without knowing they were periodical publications, one might even assume they were anthologies or collections. I have no intention of moving, or even suggesting to move, pulp and digest records into another category. Your statement that "magazine pub formats are all based on size rather than binding" isn't true. Pulp magazines were of different sizes, sometimes as much as two inches different and are categorized here based on the quality of their paper.

As for any confusion caused by choosing one of the 9 options based on size, "A4", "bedsheet", and "quarto" are roughly the same dimensions, just as are "A5" and "octavo". The only difference is the country in which they are published! If we offer "pb" and "digests" as options for either books or magazines, why not "tp" as well? There are already many magazine records in the database (for example Clarkesworld). There are even records for magazines showing "hc" as the binding, even though that's not given as an option (see Pulphouse). I just think the help pages should match the practice.

BTW, and since you brought them up, I have copies of the Scithers/Schweitzer era of Weird Tales, and am frankly surprised that they're in the database as "pulp" when they're printed on the best quality paper of any magazine in that period. When I think of "pulp", and I'd bet nine out of ten database users as well, those issues would not come to mind. Mhhutchins 03:38, 18 November 2014 (UTC)

Make definition for recording subtitles that denote a series more precise

I think the second half about Subtitles in the help about titles should be changed because it creates unnecessary edits. It states: Note that the title page may show the series name, and sometimes the publication's position in the series. It is left to the editor's discretion as to whether this should be part of the "title" that you enter for the publication. This definition leads to scenarios like I just had: that one moderator approves my submission of a title containg a subtitle of this kind, and a few weeks later another moderator removes the subtitle on his own account. That removal was perfectly ok for me, but in "worst" cases this could lead to ping-pong editing: one editor adds a subtitle, the next one deletes it later on, a few months later someone adds it again etc etc. My proposal is to change that paragraph so that a series name is never recorded as a subtitle. Or are there cases where it should be recorded? Jens Hitspacebar 19:57, 2 December 2014 (UTC)

First: Series data and subtitles are not necessarily the same thing. There are many subtitles which have no relationship to the series. Sometimes subtitles that appear in one edition aren't carried over to other editions. That's why it's important to determine the work's canonical title.

Second: The title field of the publication record should not be confused with the title field of the title record. It is OK to enter series data in the title field of a publication record as long as you're recording the data from the title page of the book. Series data should not be entered into the title field of the title record since it reflects the name of the work, not the publication. The conflict occurs when you used the "Add New..." function, because the system automatically uses the data from the title field of the publication to create a title record. This means that once the new title is in the database, it should be "cleaned" of all extraneous series data.

Yes, the help page should be rewritten to reflect the current practice. The problem is once the discussion to change the help documentation starts, it leads to conflicts among editors about wording and interpretation, resulting in nothing being changed and a lot of upset editors. Be assured that if you add publications to an existing title, you have the option to go back and adjust the title field of the publication. You can also remove series data from the title record for newly created titles. And if you are the primary verifier of a publication record, other moderators and editors should first discuss any changes in the title field with you. Mhhutchins 20:48, 2 December 2014 (UTC)

Nature Magazine added as a publication?

When I try to add this on the main page, I get told that I don't have permission to edit pages, so I'm not sure where to put this. I hope this isn't too confusing.

It was simply that I noticed that Nature magazine http://www.nature.com/ is not included as a publication, although it has a once-a-week science fiction story http://www.nature.com/news/futures on the back page. I didn't want to simply add it, as it is not primarily a SFF resource. Nature Physics also has a Futures story in every issue, I believe both are covered by the http://blogs.nature.com/futureconditional/ blog.

Should this be added? —The preceding unsigned comment was added by AkaSylvia (talk • contribs) .

This periodical is in the database. Search for "Nature" under the "Magazine" drop-down menu of the search box in the upper-left corner of the database's front page. You come up with this result. There are some missing issues so feel free to add them. Because this is considered a nongenre magazine there are special rules for entering them into the database. Use the existing records as a guide. Mhhutchins 20:22, 11 December 2014 (UTC)

I add new issues of Nature about every two months via the web API, which is faster and easier than entering them one by one. Fsfo 17:03, 13 December 2014 (UTC)

The missing issues from 2000 to 2005 and in 2007, by the way, don't have 'Futures' stories, because these were originally done in two temporary series, and as a permanent series only from 2007. Fsfo 17:08, 13 December 2014 (UTC)

Thanks for the info. It would be a good idea to add that to the notes of the series, to prevent others from having the same mistaken belief that the issues are missing for some other reason. Thanks. Mhhutchins 01:44, 14 December 2014 (UTC)

I have added the information as suggested.Fsfo 11:47, 14 December 2014 (UTC)

I copied the data you provided in the new Series wiki page and added it to the Note field which is visible on the database Series page. (We've been trying over the past years to rely less on the Wiki as a source for data, keeping it primarily as a communication tool.) Thanks. Mhhutchins 16:09, 14 December 2014 (UTC)

Non-Genre Marking - Artwork

Now that we have non-genre support, it raises some questions in my mind regarding the marking of artwork & essays. This section is with respect to artwork; essays are discussed in the next section.

Here are some example records that got me thinking:

• The Hills Are Dancing is a non-genre publication with both a cover and interior art record. The artwork is unambiguously non-genre.
• Black Emperor cover is an interior art record for a re-print of a cover to a non-genre publication. This re-print is in a publication that is artist oriented, not necessarily genre oriented. However, the publication is entered as a genre publication.
• Death in Disguise is an interior art record for a re-print of a cover to a non-genre publication. This re-print is in a publication that is genre oriented.

As genre books can have covers without genre elements and non-genre books can have covers with genre elements, it doesn't seem to make sense to me for us to judge the "genreness" of the artwork. I would suggest the following standard:

• If artwork is only in a genre publication, it is marked as genre
• If artwork is both in a genre publication & a non-genre publication, it is marked as genre
• If artwork is only in a non-genre publication, it is marked as non-genre

Therefore, for the three examples above, the first would be marked non-genre, but they next two wouldn't. The other option would be to base it on the first publication (in which case all three examples above would be marked non-genre), but this seem problematic in the case of artwork reuse.

Any other thoughts? -- JLaTondre (talk) 16:10, 30 December 2014 (UTC)

Why mark them at all? I personally feel that this would have less value than the effort it would take to make it work well. I believe we need to focus our attention on the literature more than the art. About the third example: what if a work of art first published in a non-genre publication is later used in a genre publication? You can't variant a non-genre record to genre one. I know this because I'm fixing them when they show up on a cleanup report that looks for such cases. Also, there are thousands of books that have rather generic or non-specific artwork which is obviously not really spec-fic otherwise. Marking them as genre just because they were used on a spec-fic publication is just wrong. BTW, I've been pushing to have the language field removed from COVERART and INTERIORART records. I'd be willing to get rid of the genre/non-genre field for those records as well, even though both of these moves probably wouldn't be easy to implement. Mhhutchins 16:37, 30 December 2014 (UTC)

I'd be fine with dropping the genre flag from artwork, but if there is not general agreement on that, then it would be better to have some standards on its use. As for your "what if a work of art first published in a non-genre publication is later used in a genre publication", that's already covered by the second bullet. Yes, plenty of genre books have generic artwork which is why, if the field is kept, I'd prefer not to have debates over is this cover genre or not. I'm not sure why you say marking them as genre is just wrong. We already do that today by the inherit inclusion in the database. We don't say "that's a non-genre cover by a artist below the threshold so let's not record the artwork". -- JLaTondre (talk) 19:37, 30 December 2014 (UTC)

Mostly, I agree with Mike that it's probably not worth the effort to try to mark art as genre or not. But on the other hand, different editors have different things that are important to them, hence like to spend their time on different aspects of the database, and I generally think that's ok. In general, though, much of the issue can be handled by generalizing existing rules on cover art to interior art as well. The help page for Non Genre Magazines (search for the word "cover", it's the 3rd appearance) says: "Do not enter a cover artist, nor a cover image URL. Exception: if the cover art illustrates the SF content, or is by a well known SF artist, enter the credit, and [image if] available." In general, if the cover art is illustrating a genre story, we create a record for it and include the cover; if it's illustrating a non-genre story, we don't. The same criterion is easy to extend to art in any other book which contains both genre and non-genre material: If it's illustrating a genre story, include it, otherwise, don't. I only see two special cases: Art by a well-known SF artist which is illustrating non-genre work (IMO, include); and art that's independent of a story, i.e. there for its own purpose, not for illustrating a story (IMO, include only if it's primarily a genre work and the art appears to be genre). But usually, something that isn't genre art shouldn't be included at all, so there would be no reason to mark it as non-genre. Chavey 06:03, 31 December 2014 (UTC)

Removing cover art credits from non-genre book publications would be a pretty significant change. It certainly would simplify this discussion, but I'm not sure how well that would fly in general. "Art by a well-known SF artist which is illustrating non-genre work (IMO, include)" is exactly the second example above (Frank Frazetta). -- JLaTondre (talk) 13:31, 31 December 2014 (UTC)

Non-Genre Marking - Essays

Now that we have non-genre support, it raises some questions in my mind regarding the marking of artwork & essays. This section is with respect to essays; artwork is discussed in the preceding section.

How do we want to handle essays in non-genre publications? In many cases, they will be non-genre also, but there could be cases where the essay talks about an author's work as a whole including their genre works (I want to say I've seen this before, but specific example are not coming to mind).

Ideally, we should treat them as we would the fiction, i.e. the verifier should make a determination of it is genre or non-genre and set the flag appropriately. However, another option would be to have it inherit the publication type (similar to my artwork suggestion in the preceding section).

I'm assuming most people would be for the former (base it on the content). If so, then this becomes more of a procedural question than a policy change.

How do we want to handle essays in publications that are tagged as non-genre? Do we want to:

• Default them to non-genre if publication unverified?
• Default them to non-genre if publication verified & notify verifier that they should double check?
• Leave them as is if publication verified & notify verifier that they should double check?

We have a number of publications that were marked as "non-genre" in the notes fields. These are starting to get marked as non-genre. Ideally, verifiers would go back and take care of their own work, but that is unlikely to happen on its own. It seems to me that we should have some consistent approach to this. -- JLaTondre (talk) 16:11, 30 December 2014 (UTC)

Even though some haven't gotten through, we're not supposed to enter essays from non-genre publications into the database, only fiction. But there are plenty of non-genre essays in genre publications (such as the science articles in Analog), and those that are obviously non-genre should be marked. But that raises the larger question about what should be considered genre. Is a Startling Stories article by Willy Ley about space travel considered genre? How about those "fact" articles by Richard Shaver? Even though I generally agree with your premise, it's going to be harder to implement when it comes down to such subjective decisions. That doesn't mean we shouldn't try to mark the obvious ones. BTW, the same situation can apply to NONFICTION records. Think of the hundreds of non-genre NONFICTION books by Asimov! Mhhutchins

I think the first question that we need to answer is whether making the distinction between genre and non-genre ESSAYs and NONFICTION books -- which will relegate some of them to the bottom of the Biblio pages -- would be useful.

For example, we currently list 26 NONFICTION books by L. Sprague de Camp. Some of them are clearly about speculative fiction, e.g. "Science-Fiction Handbook" and "Time & Chance: An Autobiography". Others contain a mix of genre and non-genre material, e.g. "Rubber Dinosaurs and Wooden Elephants: Essays on Literature, Film, and History" and "Lost Continents: The Atlantis Theme in History, Science, and Literature". Then there are borderline cases like "Spirits, Stars and Spells", which was about magic and occultism in the real world, an area somewhat close to speculative fiction. And then, of course, there are de Camp's popular science, archeology, etc books.

Based on the above, it seems like it would be difficult to draw a line between de Camp's "genre non-fiction and non-genre non-fiction" books. On the other hand, Manly Wade Wellman's non-fiction books have nothing to do with SF and can be safely sent to the bottom of the page. Ahasuerus 18:35, 30 December 2014 (UTC)