ISFDB talk:Data Consistency/Short Fiction-Novel Mismatches

Jump to navigation Jump to search

OK, a lot of these are down to people linking short-fiction titles in fix-up novels. Understandable - I like the links myself. We COULD solve that by demanding that the Shortfiction entries are NOT entered - they could be linked to in the notes still. And I MEAN "linked to" - not just referenced. Here's an example of what we could do in the notes: The Whole Man. Displays a bit funny though, so Lists may not be appropriate for all browsers. It looks a bit better with To Open the Sky though. Could trim it more, too many redundant author links.

There's also an odd short-fiction title tagged onto a Novel. Seems fairly harmless, but it does confuse the clean-up script.

Short-Fiction actually published as a stand-alone book is a bit of a mess: some titles get the Shortfiction and the Novel included to make them work like here and appear under the "Novels" section for the author, others just have a NOVEL title and Shortfiction contents e.g. here - you can find the book via the shortfiction but you'd never spot it if you assumed that all publications appear in the Novels/Collections/Anthologies. Converting some of the publications to Chapter-Book works to some extent, like here.

It's been an interesting time looking at these, but I don't have all the solutions. BLongley 17:57, 4 Oct 2007 (CDT)

Concerning the Shortfiction as Novel type: I've been changing the standalone books to CHAPTERBOOK type, which automatically (so far) merges the pub with the shortfiction title. (I may come across some that will have to be merged, but none so far.) I'm also removing the shortfiction that is included in the fixups, and placing the source pieces in the title notes, as you suggest above. (Though I have not gone to the trouble of linking each of the source pieces to the original title record.) There also a few collections that have been mistakenly entered as a NOVEL type, very easy to fix. I'm not touching the double novels (such as the Ace, Signet, and Tor Doubles) at this time. Need to research if they should be omnibus or anthologies. Mhhutchins 19:57, 14 Jan 2008 (CST)
A few of us discussed Ace Doubles: although it petered out with little participation the few of us talking did seem to reach consensus and Krang and I went through most of them to fix accordingly. I don't think anyone updated the help though, and it was only dos-a-dos Ace doubles that were really covered. BLongley 12:44, 15 Jan 2008 (CST)
And if you haven't seen it, this discussion also mentioned the quick way to get the relevant links via "View Source" - it saves a lot of work constructing the HTML. Only works if you do it BEFORE removing the contents, of course. BLongley 12:44, 15 Jan 2008 (CST)
This question about Anthology, Omnibus or Collection needs to have an answer sooner rather than later. Without it many of these Data Consistency mismatches will remain unresolved or inconsistently fixed. Also should we map out different sections so we don't overlap?
Yeah, I saw that we were both working on the same page, so I switched over to another set. If you'd like to stake a claim, post it here and I'll check back if I get a chance later on tonight. Mhhutchins 20:24, 14 Jan 2008 (CST)
I'll take the S-Y section. Any thoughts on how to deal with Dandelion Wine(the mother of all fix-ups)?Kraang 20:38, 14 Jan 2008 (CST)
Yeah, that was the one title that popped out. Here's my philosophy about fix-ups: if the pieces are published in the novel as separate chapters AND retains an individual title (whether new or keeps the original title) then it should be entered as a COLLECTION regardless of whether the publisher calls it a "novel" (see Catacomb Years). Because the pieces do not retain their titles in Dandelion Wine, it should be typed as a novel. And as Bill stated above, the source pieces should be included in the title record's notes. (And we know how many short pieces make up this novel!) I'm going to take a break for the night. If it's there tomorrow, I'll fix up Dandelion Wine (no pun intended!) Mhhutchins 21:02, 14 Jan 2008 (CST)
I added the stories into notes[1] for a start.Kraang 23:03, 14 Jan 2008 (CST)

What it is for?

Could someone please write a short introduction, either at the top of the page, or at the top of the talk, explaining this project and what the page is showing? TIA Marc Kupper (talk) 02:25, 17 Jan 2008 (CST)

Done! Ahasuerus 17:46, 17 Jan 2008 (CST)
I'm sorry - that explanation does not add much more than what could already be guessed from the title. I'm guessing that there was a discussion somewhere that amounted to that "As a general rule publications of type Novel would not contain Short-Fiction and so let's run a data consistency check for that?" Marc Kupper (talk) 21:40, 17 Jan 2008 (CST)
I don't recall any such discussion recently, but various people (including myself) have felt it useful to list various perceived discrepancies in the data. For example, when Al made the Author Directory dynamically-generated, I felt the need to create a page for Authors that give NO useful data when viewed there. And it's cleaned up a lot of authors, but not necessarily in the way I intended. Likewise, many of the Data Consistency reports are either a little vague as to what the problem is, or demand knowledge of SQL, or just invite people to go look at bad data with no guideline about WHAT is wrong. E.g. Invalid characters in Publication titles has a nice list of titles where I can't see anything wrong. BLongley 17:26, 19 Jan 2008 (CST)
I think we should include reasoning of what the perceived problem is, show how we generate reports of what the perceived problem Authors/Titles/Pubs are: I'm not so keen on telling people HOW to fix them, as we will eventually reduce the lists to the major problem titles eventually. When we do so though, I would like DECISIONS: e.g. the Bad ISBN list is down to 30 unfixed titles, which is shorter than the discussion about how to fix the remainder! BLongley 17:26, 19 Jan 2008 (CST)
I generated ISFDB:Invalid characters in Publication titles by searching for certain types of non-ASCII characters. I haven't looked at the list in a while (I created a whole bunch of reports that day), but it looks like the problem with all but the last two records is with the wrong kind of apostrophes being used. These bad apostrophes have been known to cause problems with Series records when the software couldn't match pre-existing Series names with the newly entered data. After spending many an hour fighting these baddies back in 2006, I am leery of them, although they may be harmless in Title records. Thankfully, they can be easily fixed by temporarily removing the apostrophe and then putting a shiny new one back in :) Ahasuerus 01:25, 20 Jan 2008 (CST)
yes, that's what I'm looking for, reasoning on what the perceived problem is. This is something I ask for all the time in software bug reports. "What did you see?" and "what did you expect to see?" The answer to the latter question nearly always lays the solution in my lap. In this case someone had left a note on my talk page saying "one of your publications showed up on this discrepancy report" but in looking at the report I ended up doing black box testing to come up with what it is I believe the report is trying to show and why my publication ended up on it. While something like "Invalid characters in Publication titles" seems to be self explanatory it also helped greatly to know that you "can't see anything wrong with them" as then I could instantly see that there were characters that normal enough to humans but would likely to be reported as "invalid" by code. Marc Kupper (talk) 01:49, 20 Jan 2008 (CST)

Prologue and Epilogues

I noticed that a number of publications I entered are in the list such as The Alton Gift where I had entered the Prologue and Epilogue as short-fiction for lack of a better data type. Help:Screen:EditPub seems to define "ESSAY" as being for non-fiction material but is silent on how to classify fictional supporting material for a story whether it be an Epilogue or a glossary of terms. I agree that these are not "shortfiction" stories but I don't consider them essays either.

Is there a recommended method for entering this material so that it won't get tripped up by data consistency checks? The main reason I have been entering this stuff is so that if a story gets reprinted that verifiers of that edition can check to see if the same supporting material was included. For example, the 1st hardcover may have a time-line that's omitted from a PB reprint.

I could enter the stuff in the publication notes and to only enter things in the publication contents that have been well defined in ISFDB. However, I had been thinking of coding a data consistency check that would look at the publications and figure out the # of pages for each story. If a publication has an unusually high or low number of pages compared with other publications it would get reported. Thus I have been entering a fair amount of detail into the Contents section to better bracket which pages contain the story. Thus if Epilogue is not a "well defined content type" and gets entered it in the publication notes then the logic to compute the # of pages would be off. Marc Kupper (talk) 21:40, 17 Jan 2008 (CST)

Posted to Marc Kupper's talk page
Your verified edition of this novel contains content entries for both the prologue and the epilogue. Aren't both, in fact, part of the novel? Are we setting a precedent in which thousands of other records will have to be changed to allow for this rather common literary practice? I'm only asking because the record shows up on the short fiction/novel inconsistencies listing. Thanks. Mhhutchins 19:29, 18 Jan 2008 (CST)
Actually, since nearly all of my books are pre-bloat prologue/epilogue are rare for me :-). I'd always considered them separate from the main story but also see them as falling into the zone between essay (which they are not) and shortfiction (which they are not). As they are fictional I have leaned towards entering them as shortfiction with the assumption being that some day new title types will get added to ISFDB to better accomodate these and that a search/update for "prologue," "epilogue," etc. will be a fast process. I don't particularly like entering them into notes as it's extra pain to do the HTML plus all of the cross referencing needs to be done by hand.
At the moment Help:Screen:EditPub is silent on the subject. I had though there was a rule to not enter About the Author pages but that page specifically mentions About the Author in the standard title types with the implication being that they get entered. I guess I'll start adding About the Author records though a gray zone is that in many anthologies the editor will write about the author or story as an untitled preface. I have not been entering title records for those but do make note of them in the publication notes.
It's my understanding that the only thing that does not get entered into ISFDB is advertising but that it's also up to individual editors to decide on how much they are willing to enter and/or to verify. The Data Consistency projects are obviously going to trip over some of this but I have no idea on the best course for a solution. Marc Kupper (talk) 16:11, 19 Jan 2008 (CST)
I personally feel that prologues and epilogues are an integral part of the novel, and as such, should not require separate title records in the database. If these pieces stand on their own, and were reprinted as separate pieces, then an exception could be made. I've occasionally seen novels that use a previously published work as a sort of prologue (e.g. Transfigurations) and At the Seventh Level, but even in these cases, the shorter works are not included in the novel's content listings. Mhhutchins 16:57, 19 Jan 2008 (CST)
I think that prologues and epilogues are generally considered a part of the novel, so we wouldn't want to list them separately. Non-fictional prefaces and afterwords are a different story, although I would consider them a part of the novel if they are fictionalized ("I found this manuscript at the bottom of a snake pit in rural Thailand...") Ahasuerus 18:07, 19 Jan 2008 (CST)
Or the "Introduction" I just entered for this Thieves' World anthology.
"Thieves World" and "Merovingen Nights" are two series I have run away screaming from before, maybe I should tackle one of those tomorrow. (No way can I do both in that time, there's a dozen pitfalls in each title...) BLongley 19:28, 19 Jan 2008 (CST)
Here's an example of an exception. "Brothers" has been republished in unrelated works, and was Award-nominated on its own, "Amanda Morgan" in related Dorsai works, and the extra interstitial material is long enough to be recorded, is fiction, is listed in contents, but would be useless without the accompanying work(s). Yes, we could classify the whole thing as "Fix-Up" but then I want the links to the shortfiction represented somehow, and the extra pages noted somehow. It's always going to be a pain with fix-ups, but the viewer should be informed as to whether he/she's just buying the same stories again or if there's something ELSE that they may consider important. And as we're the "Internet Speculative Fiction Database" I do like to record the extra fiction. The extra 20 pages MIGHT be important to someone. But I agree with Marc, I think we DO need a category for extra short pieces of fiction that couldn't be published on their own, but DO add fiction to the work they appear in: we shouldn't need to classify such as Essay to get past Data Integrity checks. BLongley 18:17, 19 Jan 2008 (CST)
The thrust of my viewpoint was towards novels and not collections. Bill's exception is a publication of related stories, thus a collection, and so technically not a fixup. In this case, providing separate entries for the Pro- & Epi- makes sense. In novels, it's pointless. Mhhutchins 18:38, 19 Jan 2008 (CST)
I was persuaded to switch it from Novel months ago: here's the relevant bit from my old talk page:
I originally approved the Sphere edition of The Spirit of Dorsai as a Novel. However, looking at the contents of The Dorsai
Companion (at Locus) it would appear that we are better off changing The Spirit of Dorsai to a Collection
since both stories that it collects have been reprinted in various places. What do you think? Ahasuerus 23:47, 17 Mar 2007 (CDT)
  Yes, seems wise - I wasn't aware 'Amanda Morgan' had been reprinted, it didn't seem worthwhile to make it a collection if only
  'Brothers' was. The linking material is nothing special. BLongley 09:18, 18 Mar 2007 (CDT) 
We really need to sort out the "fix-up" rules soon, we're not getting into many edit wars on the publications/titles themselves, but we certainly are clashing on each others talk-pages! BLongley 18:56, 19 Jan 2008 (CST)
Considering all the data that's put in without consideration for this standard I think the labor involved totally outweighs the benefits. In an overwhelming percentage of cases the prologues and epilogues are probably common to the various editions and should be considered an intrinsic part of the works. There can be many textual differences between editions, the analysis of which helps feed English professors. Perhaps only a passing reference to them would suffice in the notes rather than the HTML formatting.--swfritter 19:43, 19 Jan 2008 (CST)
It looks like we are ok - there were very few violations when I checked for Prologue and Epilogue titles. I'm not even sure if the help file needs updating given how few times an ISFDB Prologue/Epilogue title was included in a novel. Anthologies, collections, and some omnibuses have them but those are nearly always essays about the editing process. Marc Kupper (talk) 01:21, 20 Jan 2008 (CST)

(unindent) I don't think that Prologues and Epilogues of a fictional work typically need to have separate entries of any type, if the individual chapters do not, unless they ahve been separately published (If an edition of The Mote in God's Eye ever includes "Motelight" as the prologue it was initially written as, that would be such a case). However, I would like an entry type for Glossaries, Indicies, and Appendices. These are often note truly fictional, but do not really fit "Essay". I Just added the chapters to In Search of Wonder as essays, but would have liked a type of "backmatter" for the index and appendix, which are not essays in the normal sense. Similarly for the timeline present in many of the Bujold Vorkosigian books. -DES Talk 10:23, 25 Jan 2008 (CST)

This goes back to one of the original design decisions that we made at the dawn of time. Some bibliographers, e.g. Contento, use an elaborate system of "title types", up to and including "unfinished work" and "speech". We, on the other hand, opted for a minimalistic approach that covered 95% of the cases and the rest was handled via Notes. It worked well for a while, but perhaps we have grown so much (in terms of both breadth and depth of the coverage that we provide) that it may be time to revisit this issue. The Intended Order of Work discussion, section 3 ("Add features considered necessary for solid bibliographies"), is probably the best place to open this can of worms. Ahasuerus 10:56, 25 Jan 2008 (CST)

Prologue and Epilogue cleanup

(unindent) I deleted the Prologue and Epilogue from The Alton Gift that triggered the discussion and also updated the starting page #. A search for "Epilo" or "Prolog" finds a rather short list that may need attention

Looks OK to me. The new epilogue is what distinguishes this edition from previous Arrow editions. It's even mentioned on the cover (now uploaded). BLongley 08:18, 20 Jan 2008 (CST)
I just read it again and compared it with the "Author's Note" in 2010. They're very similar but the Epilogue is an expansion of the Note, dated 2 months later. However, I notice nobody has found it desirable to enter the Author's notes in any edition of 2010 yet, so perhaps people really don't want such essays? I can understand that, as I think they get revised every few editions and unless you have both copies it's difficult to spot the changes. BLongley 08:27, 20 Jan 2008 (CST)
Oh yes, most notably The Iron Heel. Before Adam is a tale of "racial memory", a popular concept in the early-mid 20th century. Ahasuerus 01:17, 20 Jan 2008 (CST)
Just so I can complete the set: The Science Fiction of Jack London seems to make it clear that SOMEONE says "Yes". OK, I've commented on every title raised in this project, can I go sleep now? MY projects take longer. Not always a good thing. :-( BLongley 19:48, 20 Jan 2008 (CST)
  • Fixed - Bored of the Rings - verified by BLongley - I have a copy and so can check this too.
Feel free to take over, I'm not sure where mine's been filed. I think this IS an example of "might be shortfiction, might be part of the main story" that started this thread, rather than the Essay Prologues and Epilogues we've wandered off into. BLongley 08:42, 20 Jan 2008 (CST)
I had a different edition. The prologue turned out to be an essay and so I left it in. I also added a publication note explaining the supporting material that's in my copy as there are an untitled poem (bottom of [2]), an untitled short fiction (found on [3] posted at 05-04-2004, 04:42 PM and starts "Do you like what you doth see...?"), etc.
I've left a note to my cleaner to see if she can remember where she filed it. Although as her husband was such a Tolkien fan that their daughter is named "Lorien", such a sacrilegious book may have been filed in the bin. :-/ BLongley 17:43, 20 Jan 2008 (CST)
Yes, it is, as the notes state. The book actually has a 'Prologue' section too but it's fiction and I counted that as part of the Novel. BLongley 08:01, 20 Jan 2008 (CST)
  • Epilogue - Someone did a title merge on all of the Epilogues for Poul Anderson!
Ouch! Ahasuerus 01:17, 20 Jan 2008 (CST)
Possibly not a problem - it seems Anderson DID write a novella called 'Epilogue', first published here. You may notice that a lot of the Epilogues in the paginated pubs are in the middle of the book. There may be a few stray essays that got mixed in though. BLongley 08:01, 20 Jan 2008 (CST)
The title record is not displayed since it matches the Publication record's type (Nonfiction). Ahasuerus 01:17, 20 Jan 2008 (CST)
I don't mind if the Prologue goes. But then I don't mind if most of all "uncredited" titles go. (Awaits outrage from magazine editors...) And I would understand if most Nonfiction went too, so long as people stopped entering reviews of such that lead to dead ends: let's finish off the Reviews/Authors project before we discuss that though. BLongley 17:43, 20 Jan 2008 (CST)
I think it needs to be added to the "limited edition" of Trumps of Doom that is mentioned in the Notes field. We just need to find out which one it was... Ahasuerus 01:17, 20 Jan 2008 (CST)
It might also be in "Manna from Heaven". I want it kept for now, as I want those words. Unless they're expensive, available online, or just crap writing. (Actually, at least one of those reasons is why I have none of the Betancourt Amber books.) BLongley 17:43, 20 Jan 2008 (CST)
I had a look at this title too, and have grudgingly admitted that even my British publication omits the "u" internally. OK, totally unrelated to the topic, but it shows I'm willing to conform at times! ;-) BLongley 17:43, 20 Jan 2008 (CST)

Marc Kupper (talk) 00:33, 20 Jan 2008 (CST)

Report from the January 20 backup

Except for the few that I fixed, all of the remaining appear to be either bonus short stories, excerpts from forthcoming books or facetious articles (wish we had this category as Locus1). A few standouts include Ray Bradbury's Fahrenheit 451 in which some editions contained extra stories. To my mind these are collections, but I haven't touched them (too sacred, perhaps?) Is there a way to mark these pubs so that any future scripts ran for this test won't include them? Otherwise, we're just running in place. (Don't tell my treadmill that, or it'll think I'm insulting it!) Mhhutchins 10:50, 23 Jan 2008 (CST)

Yes, there are ways around it. I will see what I can do before the next iteration. Thanks for fixing them! Ahasuerus 22:57, 23 Jan 2008 (CST)
There is even an empty Fahrenheit 451 COLLECTION title waiting :-) Perhaps it is time to either use it or delete it. --Roglo 03:33, 24 Jan 2008 (CST)