ok, let me chat with Mr_Monkey for a minute, then I'm on that.
CatQuest
*may-be
🐒
reosarevok
CatQuest: improvement works
pulkit6559
alastairp: i figure we could have a new artist page in acousticbrainz, should i open a ticket for it?
alastairp
pulkit6559: what would the artist page show?
(there's an artist page on musicbrainz already!)
reosarevok
alastairp: if anything, I'd show exclusively recordings for which you have actual submitted data, but not sure if needed
yvanzo: so I'd suggest renaming MBS-9926 to "Convert subscribers pages to React" and then adding a new "Convert user subscriptions lists to React" ticket - given nothing from that ticket is in beta yet, I think it's not really that confusing and it makes more sense than waiting to merge until the entire subscriptions/ folder is converted. It'd help us merge this now for the next beta, and then work on the other thing post schema change
because a single query can with right inc parameters can get us all data we need in a single query, I suppose
gr0uch0mars: ^
gr0uch0mars
yes, I was thinking about `inc` params. We should use them
amCap1712
or we should wait for tabs to open and query each separately
gr0uch0mars
I don't think they will delay too much the response
amCap1712
I think one query is better
because I read that viewpager keeps 2 offpage fragments on hold as well
and we have 4
so in any case there will be 3 calls
so going with one query
adhawkins
ruaok: Last replication packet took about 25 minutes to process. Not sure if that's better or not to be honest! Could be disk I guess...
reosarevok
yvanzo: it is, if we split the "root/user/subscriptions/" part mentioned in the ticket description to a different ticket :)
(that's what I'm asking about)
(I have no PRs nor code written for that ATM)
gr0uch0mars
yes, and in the future I was thinking of caching that single response. If you re-enter the same artist, you already have all the necessary info
requesting multiple queries would make caching more difficult
amCap1712
retrofit does auto caching i suppose
also in searchviewmodel , we first check if uery had changed and only then we uery again otherwise display the same
gr0uch0mars
between sessions? I meant if you close the app and open it tomorrow
amCap1712
no not that
gr0uch0mars
ah ok. I was thinking too much into future :)
yvanzo
reosarevok: IIUC, only subscribers lists have been converted so far (if we include this PR), the rest for this ticket is subscriptions lists, right?
amCap1712
as long as its in memory/background
we can do that in a util class
gr0uch0mars
yes, your implementation is great, too. We should query only when strictly necessary
reosarevok
yvanzo: yes, but while all the subscribers lists are sharing code and very related, the subscriptions lists are not connected to them :)
(so we'd definitely be adding them as separate tickets if we were doing it now, with our smaller PR approach)
yvanzo
reosarevok: got it :) sure, go ahead!
amCap1712
I'll try to do it here as well with MBIDs
reosarevok
yvanzo: yay, thanks
CatQuest
huh. why do we not have "session" ability for performance (instrument/vocal) roles
reosarevok
CatQuest: what do you mean? :)
gr0uch0mars
amCap1712: I was thinking of using the Repository to query first local DB (for cache) and, if not, then API
amCap1712: just implement another data layer (cache in DB) in the Repo. So your ViewModels query are not touched
reosarevok
yvanzo: then I think there's only two sets of half-done things for me to finish sending for merge RN, the stats code and the ES6 changes. The stats has only one or two PRs left (the open one + one of the other two is where you had issues with a sample DB, haven't fixed that one yet)
iliekcomputers: On my side, I'm changing the follow page playlist functionality a bit, and that will only leave implementing refresh tokens for the player.
Prashant1223
hi, I am getting an error when importing the sql database in psql using command : sudo -u postgres pg_restore -e -C -O latest.sql -d postgres . Error showing is pg_restore: [archiver] input file appears to be a text format dump. Please use psql. Is there any other way to do it?
Mr_Monkey
I'd love to talk about it with you at some point
iliekcomputers
Mr_Monkey: I'm here right now if you want?
Mr_Monkey
As for ruaok , he's out at the moment, so won't speak for him.
Sure :)
iliekcomputers
Prashant1223: what project is this?
Prashant1223
bookbrainz-site repository
CallerNo6
yvanzo, Cyna: hi, I'd like a link to that doc too please
Mr_Monkey
Sooo. Currently the user's spotify token we get from the db is used to connect to the web playback api, which returns a token that expires after an hour. Not ideal. How did you set it up on the server side? Is there any way I can use the same method ?
CallerNo6
(tt → react that is)
Mr_Monkey
I.e a private API endpoint or something
iliekcomputers
there isn't an endpoint yet.
but i could create one.
Mr_Monkey
Prashant1223: I'll be with you very soon.
Prashant1223
Okay
CatQuest
iliekcomputers: i am annoyed how norway is not in this list
but i pay like 79nok for 1 GB so
Mr_Monkey
iliekcomputers: Well, that would be great.
iliekcomputers
i don't think LB has any private API endpoints yet.
Mr_Monkey
I suppose that means you have it implemented in the server at the moment
iliekcomputers
Mr_Monkey: yep, there is a function to refresh tokens.
iliekcomputers finds a link.
pulkit6559
hi, is there a guide somewhere to the musicbrainz api?
pulkit6559: normally redirects will just get redirected, I think? :) Why?
iliekcomputers
we only refresh tokens during the import loop
pulkit6559
i was trying to work this out in AB "Use the musicbrainz mbid redirect tables to find more duplicates", so i think ill need to access endpoints for redirect tables