Goodreads Librarians Group discussion
Policies & Practices
>
Alternative cover versions - which version takes priority?
date
newest »
newest »
Andrea wrote: "Now that we are talking about this specifically...If the original edition had be incorrectly updated with a newer cover (be it by ingram/onix or by a librarian), would the proper procedure be to undo this change and create an alternate edition with the newer cover?
I understand that covers shouldn't be changed, but that it would be OK to change them back (you'd be fixing the cover for the users that shelved before the first change even if the consequence is that those who shelved in between the change and the change-back, and will see their cover modified?) "
If a record is screwed up by anyone/anything then proper practice must be to correct the screw up. (Unless Rivka or PTB says different).
Sometimes these errors take some considerable time to come to light by which time many people will have shelved the book. The route in that case might be to start a discussion, like this one, and come up with a plan for that book.
Michael wrote: "...Is not all that trouble primarily caused by the fact that the ISBN ist treated as a "special" attribute?...Just curious - I do not want to start any discussion along "this is GR and you have to take it or leave it ad you can help yourself by using the following trick..." - but why does the ISBN attribute need to be unique? Is there any technical reason behind that? ..."There are occasionally some problems with isbn number being reused or otherwise not unique—but, that's usually the exception to the rule that isbn's are supposed to be unique editions. As an exception, well, that's likely to cause questions that wind up on this thread and may make it seem like happens a lot more than all the times was unique so not mentioned (and of course no help at all with books published before isbn system).
I'm not sure of the statistics, anyone have an idea how often isbn's are goofy (other than someone entering with typos)? Aren't they almost always unique? Plus can actually find the isbn number on back covers, spines, front pages etc. if you have book in hand which makes them useful.
A point, and then a question. Point: librarians should always add a librarian note when creating an alternate cover edition. The librarian note shows up in combine view, but if you only add a note to the book's description, this only shows up in book view.Question: Bots are not supposed to overwrite user-added data. Does this mean that if I undo a bot cover upload, my undo now counts as user-added data and the bot can't overwrite it again?
Lobstergirl wrote: "Question: Bots are not supposed to overwrite user-added data. Does this mean that if I undo a bot cover upload, my undo now counts as user-added data and the bot can't overwrite it again? "It shouldn't overwrite again, yes. (see http://www.goodreads.com/topic/show/9...)
Debbie R. wrote: ... I'm not sure of the statistics, anyone have an idea how often isbn's are goofy (other than someone entering with typos)? Aren't they almost always unique? Plus can actually find the isbn number on back covers, spines, front pages etc. if you have book in hand which makes them useful."They are not goofy. They are just occupied already. And they are by far not almost always unique, especially for the period '60s to '90s...
I have recorded 160 books so far (of roughly 4K to go in total), most of them English (US and UK), some German, a handful French.
Out of these, 120 (75%) could be reused from "GR stock" because cover AND ISBN did match (although at least 80 needed some brushing up with lacking info, e.g. editor, pub year, num pag, format, lang).
30 (19%) matched on ISBN but not on cover (or editor or pagnum or format) and required the creation of an alternate cover edition.
For 8 (5%) I could identify an existing alternate cover edition (whether the underlying ISBN is the same as mine I can only guess, but editor, pagnum and format match).
The remaining 2 are too old for carrying an ISBN.
So much for the stats.
Regardless of the ISBN being unique or not, I would have had to create the alternates anyway.
But the correct ISBNs of the alternates I entered (and not only mine but everybody's) are hidden in editor's notes (and my private notes). No way to share this knowledge with the community via the field where it would naturally belong because there can be only one of each...
Banjomike wrote: "One reason is that it is the only thing that prevents hundreds or thousands of identical editions being added to Goodreads.EDIT: we have MANY people adding books who are convinced tha GR does not have entries for Shakespeare or J.K. Rowling and that they must add them."
From what I have seen so far, those people will not be hampered to re-invent the wheel by enforcing the ISBN field to carry unique ISBNs. You can always enter a goofy one or leave it out completely, so you do not enforce better quality this way.
Michael wrote: "From what I have seen so far, those people will not be hampered to re-invent the wheel by enforcing the ISBN field to carry unique ISBNs. You can always enter a goofy one or leave it out completely, so you do not enforce better quality this way. "So, are you suggesting we allow duplicate ISBNs or that we just don't check them?
Banjomike wrote: "So, are you suggesting we allow duplicate ISBNs or that we just don't check them?"I suggest that you allow duplicate ISBNs (for the sake of realism cause that is happening in the real world) AND that you check them to be formally correct (because that is what computers are made for) [the algorithm to calculate the checksum on the last position can be found here: http://en.wikipedia.org/wiki/Check_di...].
The former allows a better identification of alternate versions, the latter a better protection agains goofy ISBNs.
Same applies for the ISBN13, of course.
Alternate cover editions do not equal duplicate/problem isbns. I'm not arguing the 120 book sample data with 19% needing alternate cover editions (I have a 120 book sample with none needed and someone else likely has a 120 book sample with 80 books needing alternate cover editions), but I'm not sure what those statistics are saying about duplicate/problem isbns and how isbn as an unique identifier for a book should get tossed out.
I'll argue to the death wiki anything being used as a definitive anything other than a good start to research that has to be verified by other sources.
I concede (well seems to be unargued consensus at goodreads) the way alternate cover editions are currently done is a jury-rig to work in current system; it's an already acknowledged issue both for searching on isbn and for members exporting their book data (exports minus isbn for alternate cover edition unless member noted in private note and then have some finagling to do depending on how data will be used), isbn not readily visible to members, etc. Not sure isbn system should be tossed because book rereleased with updated cover.
I'm still convinced that (except for the early years where all publishers weren't on board or using same standards or dealing with internet retail sales likely dependent on data searches) isbn re-use/duplicates/errors are rare and getting rarer. In its inception, despite "international" wording, not standard in use by all countries (I lived in Germany and Greece when first started being used—completely different book markets and even copyright standards depending on publisher).
Personally, the late 70s and early 80s irritated me the way books were reissued with different titles and or covers and sneaky about, if any, wording that had been published previously. I bought mail order or from bookstore new release signup sheets a few books I thought were new from favorite authors by not having book in hand to check copyright pages, particularly overseas where foreign rights sales created new titles.
Amazon asin numbers, on the other hand, are not always things that should even be in goodreads database (sometimes match up to just something in amazon marketplace).
I'm sure individual librarians have their own experiences and statistics (and I'm newish) — are there any stats on overall database or librarian change log that show what percentage of books have alternate cover editions or what books/librarian-notes have had problems with isbn not being unique?
Mostly just curious; original post about issues with imported data overwriting existing has been answered. Mostly curious; but, seeing the stats would give me an inkling of how likely goodreads staff would be addressing possible methods of handling alternate cover edition issues. If very rare, I expect there's a longer bugs, to-do or wishlist set of items that will be higher priority. I'd like to see a better alternate cover edition method; but, I don't think tossing the isbn numbers as unique identifiers is it.
Regarding those alternate covers and not-unique ISBNs, almost every book I have from Penguin, Oxford, BBC Books, Wordsworth, has been reissued with the same ISBN. Movie Tie-In Editions almost always have already taken ISBN. Some books were also much older than those that have already taken that particular ISBN, but I have to comply with the GR principle first come, first served, never mind the "age" of some particular edition.EDIT: This was sloppily-written comment on Debbie's remark of ISBN re-use getting rarer. Certainly it is for newer books, but for any older books getting republished by the same publisher unfortunately it isn't.
^I agree, particularly about re-issues, but it is not a duplicate isbn if still the same book (different edition, yes, duplicate isbn, no). I understand the current alternate cover edition policies; but, I would like to see that change as it separates cover and edition from where users expect isbn information—not sure if it needs to be (or what's easiest for goodreads to implement) an"alternate cover for isbn" field, some sort of multiple cover display, a "reissue of isbn" field or what.
Lots of thread posts on cover issues - changing, updating, alternate cover editions, etc. It's clear goodreads doesn't allow covers to get overwritten (I heartily agree as messes with user catalogues and just because author doesn't like an original cover doesn't mean it did not exist; but, boy do some new authors get yelled at when innocently asking about).
With isbns with different covers and reissues, particularly now that the primary edition can be set by author instead of by popularity, maybe goodreads will be making new policy on which edition gets the isbn. Current policy is the first edition entered; maybe will become the oldest edition or the primary edition should get but regular librarians and authors not being allowed to make the change (last thing needed is isbn numbers being messed with).
The isbn usually showed up when most popular edition displayed; may not now that authors can set primary edition. One big reason for authors to change primary editions is likely to display most recently available edition/cover, if they are changing to show a new cover they are likely changing to alternate cover edition so isbns not displaying.
Debbie B is right, when a book is reissued with a new cover, it SHOULD have the same ISBN because it is the same edition of the book. This is a valid and sensible use of alternate cover editions. The problem mentioned upthread is when a completely different book-formed-entity is issued with an old ISBN. I've noticed this a lot in books from around the 80s, but rarely now as publishers in general see the validity of the ISBN for identifying the specific book they are selling...To explain with an example, say I need to buy the 7th edition of Haralambos' Sociology textbook for the library at work. The content matters, not the looks of the outside, so I will search our supplier for the book using the edition's ISBN. By using the ISBN I know for certain that I have got the 7th edition in paperback, and not the hardback or the 5th edition by accident! (The first would be an expensive mistake, and the second would be a problem if the information contained had bee significantly updated).
Debbie R. wrote: "I would like to see that change as it separates cover and edition from where users expect isbn information—not sure if it needs to be (or what's easiest for goodreads to implement) an"alternate cover for isbn" field, some sort of multiple cover display, a "reissue of isbn" field or what..."Emy wrote: "By using the ISBN I know for certain that I have got the 7th edition in paperback, and not the hardback or the 5th edition by accident!"
I agree with both posts -- I follow the alternate cover policy, but rant because I do not trust very much automated imports regarding the covers. For an example here's The Horseman on the Roof, the cover is added by onix macmillan, as well as all the other data, but the movie was made in 1995. How that cover can be the right one for the book that went out in 1982?
I wouldn't like to change the way things work now, as Debbie mentioned, users do not like later to see the other cover than that one they've chosen first. But would like to see that what she said about "reissue of isbn", something like that, to make these ISBNs visible on these alternate cover editions too, without having to write in their description. Librarian note is helpful for librarians, not really for users.
And I agree with Emy, ISBN may be reused for the same book that changes only a cover (that's the point of alternate covers...) but not if the content differs in any way, like in those textbooks. We can only try to make distinction between every edition of them by writing in the descriptions and edition boxes :/
Maybe I start to repeat myself: The problem is with the restriction on imposing unique ISBNs.A distinct edition is defined by:
- cover
- author
- title
- publisher
- num pages
- format
- language
- ISBN
Each of these may vary in the real world - with everything else remaining fixed.
All of the above may be reused - except for the ISBN.
Imagine that e.g. page numbers would be imposed by GR to be unique. So, the first edition entered may show the correct number of pages, subsequently entered editions (being distinct editions, so again e.g. cover & ISBN are different) having by chance the same number of pages are perfectly legal - but their number of pages needs to be hidden (because of "uniqueness"). Does that make sense?
Debbie R. wrote: "Personally, the late 70s and early 80s irritated me the way books were reissued with different titles and or covers and sneaky about, if any, wording that had been published previously. I bought mail order or from bookstore new release signup sheets a few books I thought were new from favorite authors by not having book in hand to check copyright pages, particularly overseas where foreign rights sales created new titles."Indeed. And even if the situation has become better wrt. reuse of ISBNs, those 20 years of misuse won't disappear. We need a strategy to be able to catalogue those editions by preserving the maximum of pertinent information - omitting the ISBN (because it has already been taken and has to be unique) does not really help.
Zeljka wrote: "I wouldn't like to change the way things work now, as Debbie mentioned, users do not like later to see the other cover than that one they've chosen first. But would like to see that what she said about "reissue of isbn", something like that, to make these ISBNs visible on these alternate cover editions too, without having to write in their description. Librarian note is helpful for librarians, not really for users."The permission to re-use an ISBN number (or rather to abandon to insist it being unique) would not lead to users suddenly seeing different covers. In GR, a work is represented by different editions which in turn are NOT based on an ISBN number. A users link to a certain incarnation of a work would not change - whatever happens to its ISBN attribute.
Michael wrote: "The permission to re-use an ISBN number (or rather to abandon to insist it being unique) would not lead to users suddenly seeing different covers. In GR, a work is represented by different editions which in turn are NOT based on an ISBN number. A users link to a certain incarnation of a work would not change - whatever happens to its ISBN attribute...."Oh I realize that, it just frustrates me to have to go through 15 to 20 pages of books and data to make sure that some particular edition isn't already there, because its ISBN is used. Especially because not all the librarians add notes to the editions they add without ISBN. If only we could somehow at once see which editions belong to the same ISBN, it would make the "job" of the librarian so much easier :/
Banjomike wrote: "What would you do if the cover did not match the ISBN + Publication date? The same ISBN is obviously going to be reused in an alternative edition but the publication date or other edition data can..."ISBN plus Pub date are not good enough. There might be an US/UK overseas editon (differing on publisher/cover). If in doubt, I will always create an alternate (cover) edition but (thanks to GR which will enforce uniqeness) will not be able to enter the correct ISBN - that fact will be hidden in editor's and private notes...
Zeljka wrote: "Oh I realize that, it just frustrates me to have to go through 15 to 20 pages of books and data to make sure that some particular edition isn't already there, because its ISBN is used. Especially because not all the librarians add notes to the editions they add without ISBN. If only we could somehow at once see which editions belong to the same ISBN, it would make the "job" of the librarian so much easier :/"That is why I claim that the ISBN field should be reusable and not unique.
Goodreads Librarians Group
Books mentioned in this topic
The Horseman on the Roof (other topics)The Lantern Bearers (other topics)
That is indeed not what should be happening.
Now that we are talking about this specifically...
If the original edition had be incorrectly updated with a newer cover (be it by ingram/onix or by a librarian), would the proper procedure be to undo this change and create an alternate edition with the newer cover?
I understand that covers shouldn't be changed, but that it would be OK to change them back (you'd be fixing the cover for the users that shelved before the first change even if the consequence is that those who shelved in between the change and the change-back, and will see their cover modified?)