It'd probably be different in BB, but the main thing you'd need is basically an Internet Archive project, so it'd need a lot of talking with them. bitmap from MB is the one who has dealt with that the most
2022-03-31 09020, 2022
Shubh
Ah does seem complicated but something like it listen to rabbitmq queues for CA changes and then decides to index them on internet archive servers.
2022-03-31 09040, 2022
CatQuest
maybe it would be useful if more thna jsut bitmap talekd to IA abut images
2022-03-31 09011, 2022
reosarevok
Well, I've talked to them in the past, but basically, bitmap is working on CAA improvements anywya
2022-03-31 09036, 2022
reosarevok
That's of course most important for the Event Art stuff inside MB itself, BB implementation would differ to some degree, but still
2022-03-31 09050, 2022
Leftmost
Shubh: I haven't looked at the code in a while; do you have a sense of how hard it would be to make it so that author credits were required for new entities, but existing ones wouldn't force adding them if you were editing?
2022-03-31 09042, 2022
prabaaaaal joined the channel
2022-03-31 09009, 2022
prabaaaaal has quit
2022-03-31 09039, 2022
monkey
Leftmost: probably not very hard, but we still have the same issue that certain Editions won't have a know author, so requiring it would need special purpose Authors
2022-03-31 09058, 2022
Leftmost
We probably ought to have those. I'm not sure if there's any special treatment for them in MB, though.
But creating an "Unknown" author just so we can make author credits mandatory feels a bit wrong
2022-03-31 09051, 2022
Shubh
Hmm agree i still don't understand why would we want to make it mandatory...
2022-03-31 09055, 2022
CatQuest
hm I think that quite a frw of the works on https://bookbrainz.org/author/fd47e471-a994-4ed9-… even should be diss-associated. I agree, for historical works where it is (heh) *known* that the author is unknown, then anonymou i a good source, but only for historically so
2022-03-31 09016, 2022
CatQuest
s/source/ solution
2022-03-31 09047, 2022
Leftmost
monkey: I think it's not just so we can make them mandatory. I'd say we need them even if we don't want to make credits mandatory. I think it's useful to distinguish between the various states of unknown data. "BB doesn't have this data" vs. "this data is generally unknown" vs. "this data is probably unknowable". (Missing data, [unknown], [anonymous].)
2022-03-31 09013, 2022
monkey
Fair enough, I could see the use for them. But I'm still unconvinced that we need the author credits to be required
In particular I'm wary of making all the existing Editions invalid (editing them would require adding an AC before saving)
2022-03-31 09035, 2022
monkey
Although maybe that's not the right way to think about it
2022-03-31 09008, 2022
Shubh
Yeah that would add extra overhead for the user if they have to create author entity before adding edition.
2022-03-31 09043, 2022
Leftmost
monkey: that's what I was trying to avoid with "if editing, not required; if adding, required". Make it so that editions without existing credits don't need to have one added to make a small edit, but enforce it going forward.
2022-03-31 09053, 2022
monkey
Right
2022-03-31 09030, 2022
CatQuest
problem there it's annoying enough already with what's mandatory, it was publicist before, now it's language (wich i still tihkn sohuldn't be mandatory fwiw)
2022-03-31 09055, 2022
Leftmost
But I think the arguments for/against having AC required in general would probably be better hashed out as a community discussion and I'm too tired to make any of those arguments right now anyhow. :P
2022-03-31 09048, 2022
monkey
Roger that. So let's try and get this merged and deployed to beta, so that we can push the discussion to the wider community
2022-03-31 09022, 2022
Shubh
Hmm, It would be better if we have community post for author_credits in general
2022-03-31 09023, 2022
monkey
Testing on tesT.bb now, looks quite good. I'm going to deploy current beta to prod in preparation for deploying ACs to beta
2022-03-31 09020, 2022
monkey
Shubh: Yes, I think we should have a post to introduce the feature, and ask the few remaining questions we might have
2022-03-31 09047, 2022
monkey
And then we can implement the feedback in another PR
2022-03-31 09013, 2022
Shubh
Yep!
2022-03-31 09036, 2022
CatQuest
we should definitely involve the community, and also tell peopel to test!
2022-03-31 09041, 2022
Leftmost
On the plus side of testing, we can't make the existing AC data any worse. 😁
that is fast, how come we have 3 releases in the same month :0
2022-03-31 09026, 2022
monkey
Well, one of them was a correction the day after the first one. One release ~= every month seems about right
2022-03-31 09012, 2022
Shubh
Hmm
2022-03-31 09018, 2022
CatQuest
i think it woudl be neat with metabrainz blog too?
2022-03-31 09008, 2022
CatQuest
and thanking people who write tickets too (and i'm not saying that just becase i know i've written many bb tickets :D)
2022-03-31 09048, 2022
Leftmost
I'm getting some not-friendly behavior out of the auto-append ISBN, made a bit worse by the fact that a UPC will be recognized as an ISBN even if I manually selected barcode (which probably needs a bug of its own) and a 10 digit number will be recognized as an ISBN even if it's an incomplete UPC/ISBN-13 I'm in the process of typing. Even if there's already an ISBN-13 entry, completing an ISBN will create a new ISBN-13 field.
2022-03-31 09050, 2022
Leftmost
I typed out 9780380789 with Barcode already selected, which turned it into an ISBN-10 and added an incorrect ISBN-13. I finished typing it out, which changed it to an ISBN-13 and added an ISBN-10, so now I have two ISBN-13s, one of which is wrong, and an ISBN-10. I changed the type back to barcode and removed the incorrect ISBN-13, then added dashes to the ISBN-10, which created another ISBN-13 field. I started adding dashes
BB-650: Barcode (and ISBN) insertion and storage improvement
2022-03-31 09004, 2022
monkey
I'll let Shubh have a look at the auto-ISBN issues, and as CatQuest mentioned we have a ticket for the irritating "it automatically changes my selected type" issue.
2022-03-31 09037, 2022
CatQuest
i noticed that aages ago but I was procastinating about making a ticket.
no prob, i just linked the first I found with a semi-rememberd title :D
2022-03-31 09009, 2022
CatQuest
i shoudl link thee (it's on the description)
2022-03-31 09049, 2022
CatQuest
there is a point here for *typed* barcodes
2022-03-31 09054, 2022
CatQuest
i havent actually tried doing that
2022-03-31 09016, 2022
CatQuest
but i think that the guessing definitly needs ot wait until there is no impt, maybe a second or so.
2022-03-31 09036, 2022
CatQuest
and no guessing what-so-ever if there is a selected type by the user
2022-03-31 09041, 2022
monkey
Agree with extra delay before auto-selecting
2022-03-31 09055, 2022
monkey
And also with the no overriding of selected type :)
2022-03-31 09004, 2022
reosarevok
I agree Language shouldn't be mandatory, ideally
2022-03-31 09029, 2022
reosarevok
Because I might have a book which I'm not sure whether it is in Romanian or Moldovan but I still want to add it :p
2022-03-31 09036, 2022
reosarevok
(I do not, mind, but I might :D )
2022-03-31 09002, 2022
reosarevok
I think that's why MB doesn't make it mandatory, although of course for BB it's a more central part of the data
2022-03-31 09001, 2022
reosarevok
I personally expect enforcing an author for every edition makes things easier to maintain and less exception-riddled in the long term, but I also don't necessarily think it's outright better in a data modeling way
2022-03-31 09045, 2022
reosarevok
I guess a question would be "is the author of an instruction manual for a vacuum cleaner unknown, or not applicable"? :p
