"Goodreads" changing existing edition of book with reviews

[Peter]

Looking at the Librarian Edits for Murder, She Meowed for this edition (http://www.goodreads.com/book/edits/1...), I see that the image was changed, instead of an Alternate Cover edition being created, as well as the associated data.

I thought this was never supposed to be done.

[❂ Murder by Death]

Originally thought that this might be due to the new import of WorldCat and LoC data, but I see that the date is from November, so that's not it. I'm guessing though that "Goodreads" refers to some sort of automated process they have in place - quite possibly even the Amazon import. If that's the case I can see how the cover would have been over-written instead of an alternate cover added as is policy.

Do you have a link to the alternate cover? We can add it now.

[Peter]

Jennifer E. wrote: "Originally thought that this might be due to the new import of WorldCat and LoC data, but I see that the date is from November, so that's not it. I'm guessing though that "Goodreads" refers to som..."

The data source lists as ::Publisher.

Shouldn't the automated processes obey policy?

I would recommend either the 1st or 4th image at http://www.amazon.com/gp/customer-med....

[Paula]

Unless you have other examples sometimes it is good just to leave alone so the tech team can review

[Jan]

I think they did a bunch of Publisher imports in November, so the Goodreads thing is probably referring to that.

[Peter]

Jan wrote: "I think they did a bunch of Publisher imports in November, so the Goodreads thing is probably referring to that."

So, shouldn't Publisher imports obey policy?

[Peter]

Paula wrote: "Unless you have other examples sometimes it is good just to leave alone so the tech team can review"

That's a good thought. It's best if they have some data to work with.

[rivka, Former Moderator]

Peter wrote: "So, shouldn't Publisher imports obey policy?"

Automated processes aren't discriminating enough to do that.

For a variety of reasons, having data direct from the publisher is far preferable to having data from Amazon or other booksellers.

Moreover, the publisher imports started about November (I think there were some tests and small groups before that), but they are ongoing and will continue.

I do understand the frustration of suddenly having a book on your shelves with a different cover. It's happened several times recently to me as well.

[Peter]

rivka wrote: "Peter wrote: "So, shouldn't Publisher imports obey policy?"

Automated processes aren't discriminating enough to do that."

As a retired software engineer, I can tell you that automated processes can be more than discriminating enough to detect that there is an edition already in place with that ISBN, and either

a) convert the existing record to a record without the ISBN, keeping its shelvings, or
b) create the new record as an Alternate edition.

It is simply a matter of writing the specifications for the software to do so, and the IT group will do the coding.

It is better to anticipate such requirements, but there is no reason it cannot be treated as a bug and fixed.

Also, it is more than covers that are being changed. All of the data in the record is subject to change.

I believe that one of the dangers in the current situation could play out like this: The old edition had 90 pages and you finished reading it. If the new edition has 100, then, even though you have finished the book, you could suddenly be only 90% complete.

Another case: Titles sometimes change between editions. So The Story of M could change to The Story of O. Now you have what looks like a book missing from your shelves, and another one that you don't recognize being added.

While I agree that "For a variety of reasons, having data direct from the publisher is far preferable to having data from Amazon or other booksellers.", I do not agree that changing shelved books to conform to the latest Publisher's editions is the right thing to do. As stated in the Librarians Manual

If you find that a book with a specific ISBN has already been listed with a different cover, then it is likely the ISBN was reused for a later/earlier reprinting. ... as of now, it is not possible to upload multiple covers of a book, and because changing the cover of an existing edition is not ideal. Some people may have added the existing edition with that specific cover because it is the one they own. If you want to have an edition shelved with your specific cover, then it is accepted procedure to create a new edition with your cover art. .... In the other case of duplicate ISBNs, it may be that an ISBN was reused on a completely different book. This shouldn't happen because ISBNs are supposed to be unique, but exceptions have been found.

User scans of books will be lost as well as user descriptions. Publisher advertisements will (and now do) show up instead of descriptions.

I would appreciate any comments that anyone, and especially Goodreads staff, would care to leave on this subject. (If this has turned into a rant, I do apologize, but I do feel strongly that, if a user shelves a particular edition, for whatever reason, changing the picture or data for that edition is as wrong as a bait-and-switch retail scam.)

[rivka, Former Moderator]

Perhaps I should have specified that our existing automated processes cannot do it. I'm sure it's theoretically possible.


Peter wrote: "User scans of books will be lost as well as user descriptions."

From what I was told, that should not be the case. User-provided data is higher priority than publisher data which is higher priority than bookseller data.

[Peter]

But it is happening. Uploaded images (not the imported by default) have been replaced.

Page numbering as well.

And we all know just how accurate the publisher's page count is.

[rivka, Former Moderator]

If the overall book record is coded as having come from a bookseller, then all data will currently be overridden with publisher data. I know that breaking that down into components (so description can be replaced separately from other elements, for example), is being worked on.

[Peter]

rivka wrote: "Perhaps I should have specified that our existing automated processes cannot do it. I'm sure it's theoretically possible.

It is actually fairly simple. I could write the pseudo-code with barely a thought, without even knowing your database system setup.

It would start with something like

IF exists(import.ISBN)
import.edition := import.edition + "Alternate " + import.ISBN
import.ISBN := NULL
ENDIF

That way, the imported record will not overwrite the existing one, and the new record will retain the ISBN in the edition field.

[Peter]

rivka wrote: "If the overall book record is coded as having come from a bookseller, then all data will currently be overridden with publisher data. I know that breaking that down into components description can be replaced separately from other elements, for example), is being worked on. "

So, if I understand you correctly, if you import a book and then update its data, the updated record still counts as having come from the bookseller???

[rivka, Former Moderator]

Looks at the "source" listed currently for each book on the edit page.

[Peter]

rivka wrote: "Looks at the "source" listed currently for each book on the edit page."

Which does not change, even if the record is edited, if I remember correctly.

Maybe that value should be changed to "Goodreads user" when the record is edited? That would help fix the problem.