reosarevok, cc bitmap: If these details are the same for all edits, they should probably be documented in a single place if it isn’t the case already. Didn’t find it in WikiDocs. On the topic of edit system documentation, it is currently missing which edits can be auto-edits or not (See MBS-7509). I don’t know if these pages would be the right place to do so. Otherwise I’m fine with suppressing these pages but still keep the
hi there I was setting up listenbrainz.I have one question to ask...will the the call back url while setting up last.fm api account would be same which we used for OAuth?
vibhoo_24 has quit
vibhoo_24 joined the channel
Rishabh joined the channel
Hi there I was setting up..I have one question...while registering an app on spotify developer dashboard can we provide any app name and description to generate client id and client secret.
lucifer: does this also address if the canonical recording changes?
lucifer
alastairp: no. because all recordings canonical or not are present in the mb metadata cache table.
alastairp
got it
lucifer
there's also the issue that in some case canonical recordings are not stable.
alastairp
yeah, exactly
so this is just if something which was in the msid mapping table is no longer in the metadata table
lucifer
right
alastairp
there's a small edge case where the new table is created but there might be bad mappings? as they happen in 2 separate transactions?
however, as you mentioned in the PR, this just results in no data being shown, rather than an error?
lucifer
yes, these cases happen currently already as well.
jivte joined the channel
alastairp
great
lucifer
i am not sure what is the edge case you mention? can you explain?
alastairp
in version I of the metadata cache table, msid a maps to mbid a
but then mbid a is deleted or merged, so it's no longer in MB, but it's still in the metadata cache table, so LB will show some value + data + link to MB
lucifer
ah ok, yes makes sense.
alastairp
when cache table is updated to v II, mbid a won't appear in it, but mapper has the data still, so it'll [continue to show a link? no longer show a link because it's not in the cache?]
which might be the case for 30 seconds, before this query runs
lucifer
yes right. it won't error just not show the data.
alastairp
got it
this looks fine, just one tiny recommendation to the docs
lucifer
great, thanks!
BrainzGit
[listenbrainz-server] 14amCap1712 merged pull request #2279 (03master…mapping-redirect-invalidate): Invalidate mbid_mapper matches for redirects and deleted recordings https://github.com/metabrainz/listenbrainz-serv...
alastairp
I don't know how this query syntax works, update where not exists + subquery/join
neat
I guess I might have written it as WHERE id in (select id from cache left join .. where mapped_col IS NULL)
because that's familiar with my understanding of the db
lucifer
its a format of anti-join. i had asked in the pg channel to clarify the syntax.
yes thats another way but likely to be less efficient.
alastairp
yeah, I would expect that it'd be less efficienct. there's always that tradeoff between easy-to-read and too smart
but the line always moves
lucifer
makes sense
alastairp
I used to work with someone who insisted that we didn't use list comprehensions in python because it was a "complicated syntax that not everyone might know"
sure, perhaps n years ago that was true
lucifer
hehe lol. indeed
Rishabh joined the channel
Rishabh
as a android developer, can somebody. guide me how can i contribute
Rishabh has quit
jivte
monkey: Hey I would like to work on LB-1157 can u please tell whether I can work on it.
Always good to see interest :) It's not going to be an entry level task, as we don't currently know where we want this feature, or what it should look like, so there is a fair amount of design and UX work before implementation.
If you're comfortable with that I'm happy to discuss it with you
jivte
Yes I actually had interest in this issue and thought we can have it on the listens like the main listens page
monkey
I suppose that makes sense, that's also where I would look for it myself. Of course it would only show on a user's own listens page
The closest feature we currently have is the search dropdown to add tracks to a playlist. I'd suggest starting by creating a playlist to try that out.
jivte
can u please elaborate what starting task should I work on
monkey
We could possibly use the same search dropdown to search for a track, although I would love to see an additional feature (that could be added as a second phase) to be able to paste a track MBID or full MusicBrainz URL
I think first you should start making mockups of what the feature would look like on the page
jivte
like developing UI
One more thing I can do is make a proposal type for this feature and then u can suggest changes on this feature
monkey
The methods to call the API already exist in the codebase so that won't be much work at all. I think the main question is how to let users search for and select the right track.
On thing missing in the playlist thing is a way to confirm it's the right track. Just showing track name and artist sometimes isn't enough
Yes I think that's the way to go jivte since it's a pretty complex feature and we'll want more eyes on it than just mine.
We can get aerozol involved with the design, and we'll rope in mayhem and lucifer for the aspect of searching for the track (the API to hit to do the search)
And I'll help you integrate all that and find the relevant pieces of code in the codebasE.
jivte
so should I work on this feature or pick some other task for the time being as it involves some serious discussion I think
monkey
At the end, we'll have a separate component that we can finally include in the listens page
jivte
okk I understood what the workflow would we like
monkey
If you are interested in the task, you could totally get started with simple mockups and doodles about UI and UX
Try out the playlist page and see how you would improve upon it
jivte
ok sorry to bother you but jus not getting clarity what to start with playlist page you are asking
monkey
Regardless of the technical aspects of which API to hit and such, the bulk of the feature will be the same: a good way for non-technical users to search for and select a track (from MusicBrainz), a way to set a date & time for the listen and a way to display the selected track fro confirmation before submitting it.
Possible advanced uses: using MBIDs/MB URLs, submitting an entire album in one go.
It's not exactly what we want, but it's the closest thing we currently have and will give you an idea of how to improve the feature for use in the listens page
Rishabh joined the channel
jivte
okk understood we need to improve that playlist part from the current part
monkey
I guess it's just to give you an idea of the direction
We might end up using the new and improved component for the playlist page as well, if appropriate; I'm not certain yet how well that would work, and it mostly depends on what your feature would look like
Whether it would integrate well on the playlist page as well
jivte
I think will work on its UI and frontend and then inform you
monkey
Awesome :) Thanks for taking this on !
jivte
Thanks for being so patience and helping me out :)
do you remember that I was working on a PR that was on hold because I had exams? well jivte happens to be my junior and I was sharing it with him today. And he told me that he's also working on revamping the playlist page apparently
I showed him around the components and the search bar I made
Will you be okay with him working on the PR? I can chime in and help him when needed?
monkey
Of course !
riksucks
he also showed me a mockup he made, I think he will share it with you when he's online
I've been wishing for something like this. I've requested some pricing information -- maybe it is possible to have our own streaming interface. :)
monkey
riksucks: there is a user search endpoint, I suppose if the endpoint returns no results then it's not valid
Streaming from where mayhem ?
vibhoo_24
@lucifer I have completed the setup.
jivte has quit
lucifer
vibhoo_24: nice, try exploring the LB website on your local server and listenbrainz.org to get an understanding of what's it about. then you can try to find a ticket to work on from tickets.metabrainz.org.
monkey: we can add an endpoint for exact check as well. maybe augument the search one to accept a param for exact or fuzzy match,
vibhoo_24
okay i'll do some exploration to get some idea.Also one more question...I am not very familiar with docker....should I learn it from some tutorial
monkey
Makes sense. Not sure if it makes sense in this case to call a separate endpoint over jsut comparing lowercased top search result with search query
jivte joined the channel
lucifer
vibhoo_24: probably. but completely upto you. you should only need to know very basics, the develop.sh and test.sh scripts usually work well enough.
vibhoo_24
okay when I'll need I'll learn on the go.For now I will try to explore LB website
aerozol
reosarevok: re. spam, soundandvision did mention he'd sent you a looot of emails/reports :p
Welcome back monkey!!
monkey
Fanks !
vibhoo_24 has quit
aerozol
jivte: monkey: re. manually adding listeners, user flaky is almost done making a Picard plugin that does that. I think it would be useful to see how they do it and then we can align with anything there that we like