Goodreads Librarians Group discussion

I have been marking Proof copies Invalid. As I understand it, marking them Deleted would allow the bot to import them again.

Is this correct?

Don't think so.
otherwise
If a Librarian attempts to add a book in the “Deleted” status using the Add a New Book page that matches an existing ISBN13/ISBN10 or ASIN, an error will prompt that the book already exists with a link to the “Deleted” status book edit page where the book can be set to “Public” to restore the deleted edition to the Goodreads database. would make no sense.
Quote from rivkas status info posting https://www.goodreads.com/topic/show/...

I had experienced that with bot made "deleted" editions. I could not add them but got the error message and could then set the book to public.
So contrary to the past it's now always the identifier in the system. Difference to invalid seems to be only what happens to user activity and if error or book page is shown.

Had seen several staff members setting the proofs to deleted in various logs.

Although, I would not be surprised if bot additions have special rules with the current bot surprises to be honest.

Thanks. I was under the impression that Invalid was like the old Not a Book, and Deleted merges with the default if there is one. In the old days, deleted actually deleted the edition from the database if there was no default with which to merge.

Some authors are saying the Proof editions are returning. I thought maybe that was the result of setting to Delete. Maybe it just means there are more Proof editions out there and they didn’t note any differences in identifiers.

Elizabeth (Alaska) wrote: "I have been marking Proof copies Invalid. As I understand it, marking them Deleted would allow the bot to import them again. Is this correct?"

As I understand the delete status process, it only work as intended if combined with a merge target. The manual mentions a warning msg if we try to delete something that isn't combined with another edition, but I never been able to trigger it.

Regardless, the 'deleted' records are never deleted as they were in the old delete process. The record remains, only invisible and unsearchable unless you have a direct url address to it, or find your way to it via user booksheves (if it still has shelvings attached to it).

Edit: the warning msg was mentioned in the announcement, not the manual

If the book edition is the last book in the work, it can not be deleted if there is any user activity associated with the book, which would be lost. Librarians will see an error message if they attempt a deletion. The last edition of a work can only be set to “Invalid” (not “Deleted”) if it has member shelvings...

Ok. I understand Deleted better. It seems, though, that Invalid accomplishes what we need to set out to do with these Proof copies.

Elizabeth (Alaska) wrote: "Some authors are saying the Proof editions are returning. I thought maybe that was the result of setting to Delete. Maybe it just means there are more Proof editions out there and they didn’t note any differences in identifiers."

Maybe they are just thinking they are returning.
I saw several records the last days with more than one existing proof for a record, mostly at least different format (HC and PB) but at least once it was two paperbacks. all with different ASIN.

So that might just be a "Amazon thing" as there can be more than one proof.

I actually just came across about a proof with 1 rating since bot added it 2 weeks ago. As the books were spread over different initials I had a merge target and as it had no written review I was allowed to delete.
So I tested it.
Deleted the Proof, waited until merge was through and tried to add it again with the asin that I just deleted.
Would have worked in the past but did not now.
Asin is taken by an existing book with the title "PROOF: Dark Comes Quickly" in 'Deleted' status. If you believe this is an error please double-check your data, then check the Goodreads Librarians group. ASINs are optional so you can also add this book without an ASIN.
(the link to the book goes to the edit page of the proof record and the book page itself is an error page as it should be. seems like at least one thing is working as it should)

Nice test, Lachgas! I'm still leery of marking books Deleted.

My understanding is that setting a book Invalid does not move shelves and reviews while setting Deleted does (if there is a proper target). If I am correct, for a proof that has reviews/shelves, setting it as Deleted is actually beneficial. It's not common, but I did see a couple already.

yes, I do understand it that way too David - invalid keeps the ratings/shelvings and the book on the shelves.
With a rating/shelving (hate that it's now more difficult to see and we have to rely on the combine page) I don't think it's even wanted to have those proofs on any ones shelves so I do delete to move the shelving. and as only the author gets them it doesn't make sense to me to keep the records accessible via link, so that might be the reason staff does delete them.
It's clearly a "not publicly available" record and usually without written review so delete should be well within the rules.

Elizabeth (Alaska) wrote: "Nice test, Lachgas! I'm still leery of marking books Deleted."

I've noted the ASIN and will test again in a few weeks - just to see if maybe something is messed up. (whe hatd that some time in the past already if I remember correctly that link NABed stuff got reimported so with all these bugs/weird behaviour rather testing again)

☕ Lachgas ♿ wrote: "yes, I do understand it that way too David - invalid keeps the ratings/shelvings and the book on the shelves.
With a rating/shelving (hate that it's now more difficult to see and we have to rely on the combine page)"

The statistics link on the book page will tell you shelvings for a specific edition, even if it has been combined.

I'm still confused. Are librarians NOT supposed to mark proofs as invalid? It seems like deleting doesn't always work.

annob wrote: As I understand the delete status process, it only work as intended if combined with a merge target.
...
If the book edition is the last book in the work, it can not be deleted if there is any user activity associated with the book, which would be lost. Librarians will see an error message if they attempt a deletion. The last edition of a work can only be set to “Invalid” (not “Deleted”) if it has member shelvings...

The only time I tried to delete a proof, it didn't go away. I didn't get an error message, but the proof was still there and still marked Valid.

So I've been marking proofs as invalid. Perhaps I should leave them alone and let more experienced librarians handle them? :D

Support have been using both Invalid and Deleted, so I think you can safely do either.
Personally, I'm marking them Invalid since it saves me the work of combining, except when they have ratings/shelvings, in which case I'll combine and mark Deleted.

if they have no ratings/shelvings you can use delete without combining before.
(although not sure if I would do that as a super as non-super you get the error page if there is a rating you didn't see but as super they probably go through)

Mesembryanthemum wrote: "The only time I tried to delete a proof, it didn't go away. I didn't get an error message, but the proof was still there and still marked Valid."
had that with setting records to invalid also - just happens sometimes because well ... buggy GR I guess.
(especially annoying when I left a note and that then exists but the record is still public while the note says why I marked it invalid...)

Hi everyone! Thanks for starting this thread. Can you please add some examples of proof copies to this thread as you come across them? Staff will use them to try identify patterns.

Actually never mind! I'll use the examples coming in via the Questions folder.

Jaclyn wrote: "Actually never mind! I'll use the examples coming in via the Questions folder."

There are dozens of requests daily. Mostly they get posted in the Book & Author Page Issues folder.

Jaclyn, here is one this morning, piggy-backed on a previous request.

https://www.goodreads.com/topic/show/...

Thank you!

Here is a big one I dealt with yesterday.

https://www.goodreads.com/topic/show/...

Latest (?) post:

https://www.goodreads.com/topic/show/...

Jaclyn wrote: "... Staff will use them to try identify patterns."

Jaclyn, here's the pattern:

Proof copies have a title that begins with "PROOF:" (in all caps) and there is almost no other information -- just an author, ASIN, and format.

Yes, a tiny number of real books also start with the word "Proof", but it's not in all caps. And there's usually more information, such as the ISBN, publisher, date, and description. IMO, the bot should skip anything that starts with "PROOF" (and log those items for a human to audit). Better to skip one or two real books than to add thousands of proofs to the GR database.

In case you want any more examples, here are some proofs that are old enough for the DuckDuckGo search engine to find today:

https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...

I found PROOF copies in French language these do not begin with PROOF but with ÉPREUVE. Examples:

https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...

I'm so confused! Why are so many of my books coming up as PROOF???? Help

Caroline wrote: "I'm so confused! Why are so many of my books coming up as PROOF???? Help"

Removed your proof copies. Librarian notes will only be seen when we are editing that edition, so you're best off putting requests in the Book & Author Page Issues folder.

Elizabeth (Alaska) wrote: "Caroline wrote: "I'm so confused! Why are so many of my books coming up as PROOF???? Help"

Removed your proof copies. Librarian notes will only be seen when we are editing that edition, so you're ..."

Thank you so much!

Tami wrote: "I found PROOF copies in French language these do not begin with PROOF but with ÉPREUVE. Examples: ..."

Good point, Tami. My comment shouldn't have been so English-centric.

For PROOF in my comment above, please substitute "proofword, where proofword is DRUKPROEF (Dutch), PROOF (English), ÉPREUVE (French), PROBEDRUCK (German), BONECO (Portuguese), PRUEBA (Spanish), and so on."

And another piece of the pattern: The title in the URL is just an ID number, a dash, and the proofword, as in these examples:

https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1...

FWIW, I marked about seven pages' worth of German proof copies (PROBEDRUCK) invalid this afternoon...and there are now two pages more PROBEDRUCK than when I started. Don't think I'll be continuing with that particular Sisyphean task!

Yup. Until GR stops the bots from running, the problem will just get worse.

Tami wrote: "I found PROOF copies in French language these do not begin with PROOF but with ÉPREUVE. Examples:

https://www.goodreads.com/book/show/1...
https://www.goodreads.com/book/show/1......"

Does seem to be also about the author's language not only the book's
https://www.goodreads.com/book/show/1...

(Found that as only record when searching for "probedruc" - not with "probedruck" which seems weird behavior from gr search.)

I see the "amazon_import" is now marking PROOF copies as Invalid. (Sorry, I don't know if it is doing the same with the other language words for PROOF.)

Also of note is that "amazon_import" is not the same as "amazon_catalog" which imported them in the first place.

Yes, we've started working on bulk 'deleting' these. Closing this thread as future proof urls should be added here.