[listenbrainz-android] 14dependabot[bot] opened pull request #633 (03dev…dependabot/github_actions/dev/actions/upload-artifact-6): Bump actions/upload-artifact from 5 to 6 https://github.com/metabrainz/listenbrainz-androi…
Yes, u can open issues on GitHub issues itself in case you face any issues. You may ping us here in case of any doubts or queries.
2025-12-17 35159, 2025
Kladky joined the channel
2025-12-17 35131, 2025
anuj_ joined the channel
2025-12-17 35152, 2025
ApeKattQuest joined the channel
2025-12-17 35153, 2025
aerozol[m]
Hey MB team ( I guess mainly yvanzo reosarevok bitmap ), I've been posting MB server updates to the socials when I see them. Would you like me to always do them 'officially'? I think I already asked and you said yes, but I'm thinking of adding a nice graphic like I do with LB, so I wanted to check again, and that an image would be OK with everyone :)
2025-12-17 35131, 2025
reosarevok[m]
Well, is the image going to have a graph like "LOOK HOW LITTLE THEY DID THIS TIME, LAZYASSES"?
2025-12-17 35137, 2025
reosarevok[m]
If not, I can't see why it would be a problem
2025-12-17 35149, 2025
reosarevok[m]
(if yes, it's also not a problem for me but it might be for others 😝)
2025-12-17 35123, 2025
aerozol[m]
Aw man I was just going to do a static image, but now I'm thinking about that graph
2025-12-17 35144, 2025
aerozol[m] uploaded an image: (189KiB) < https://matrix.chatbrainz.org/_matrix/media/v3/download/matrix.org/VWOsgbDqhlPOyoykCRNsBZot/image.png >
2025-12-17 35153, 2025
aerozol[m]
That's is the LB one I do
2025-12-17 35115, 2025
aerozol[m] uploaded an image: (251KiB) < https://matrix.chatbrainz.org/_matrix/media/v3/download/matrix.org/hWbCNbwJXRSSVNRvNWmEvcll/image.png >
Not much more work, and sets it apart from other posts a bit
2025-12-17 35156, 2025
reosarevok[m]
Works for me
2025-12-17 35113, 2025
anuj_ has quit
2025-12-17 35145, 2025
op3kay[m]
reosarevok: regarding the Areabot - currently almost everything is stored in notebooks. As of right now, the notebooks are able to fetch countries subdivisions and cities from wikidata and even able to compare them to the data in the musicbrainz to tell which ones are not present in the db yet.That being said, there is no single file that runs this code in sync and seem unorganized. And there seems to be no logic for linking
2025-12-17 35145, 2025
op3kay[m]
relationships. Im thinking we should first create a pipeline to see a list of diff between the mb db and the ones we get from querying wikidata along with some error handling for duplicates and so.
2025-12-17 35115, 2025
ApeKattQuest has quit
2025-12-17 35115, 2025
ApeKattQuest joined the channel
2025-12-17 35103, 2025
ApeKattQuest has quit
2025-12-17 35154, 2025
monkey[m]1
lucifer: If you have time, can I ask to review ansh's PRs relating to YIM in priority, so we can recalculate the data and add the missing elements on the page and do more testing?
2025-12-17 35113, 2025
lucifer[m]
monkey: will do
2025-12-17 35111, 2025
monkey[m]1
I've been going over the design with aerozol, I think we're getting to a good point now with only minor improvements.
2025-12-17 35138, 2025
monkey[m]1
After adding those elements and calculating the data I'll ask for beta testers on the LB channel
2025-12-17 35147, 2025
monkey[m]1
I mean, I guess I could ask now
2025-12-17 35107, 2025
lucifer[m]
sounds good
2025-12-17 35101, 2025
derat[m] has quit
2025-12-17 35129, 2025
_BrainzGit
[listenbrainz-server] 14anshg1214 opened pull request #3454 (03master…ansh/improve-yim-stat-calculation): feat: improve new release for top artist calculation https://github.com/metabrainz/listenbrainz-server…
2025-12-17 35109, 2025
FaizanAkhtar[m] uploaded an image: (1890KiB) < https://matrix.chatbrainz.org/_matrix/media/v3/download/matrix.org/tqrkpWJzJfBPauvpbOqvWOrJ/Screenshot%202025-12-17%20at%208.04.42%E2%80%AFPM.png >
2025-12-17 35129, 2025
FaizanAkhtar[m]
monkey: Hi.... I have implemented all the improvements you suggested....the album page now shows the track count followed by the total duration on a single line uses humanize-duration handles singular/plural correctly and matches the existing style I have also manually verified the total duration by summing all track lengths ..........please let me know if this looks good now....
2025-12-17 35110, 2025
monkey[m]1
Thanks for working on that, I'll have another review when I can
2025-12-17 35123, 2025
bitmap[m]
aerozol: I don't have any problem with a graphic. that sound fine, thank you!
2025-12-17 35136, 2025
bitmap[m]
reosarevok: yvanzo: hi, around?
2025-12-17 35143, 2025
yvanzo[m]
Hi!
2025-12-17 35145, 2025
reosarevok[m]
Hi!
2025-12-17 35151, 2025
monkey[m]1
Hi! 🙈
2025-12-17 35108, 2025
bitmap[m]
I don't have much more to report on the oauth/meb account side since yesterday, but maybe we can discuss the digest auth changes I mentioned last week?
2025-12-17 35158, 2025
reosarevok[m]
I have nothing to report either else from what we already know
2025-12-17 35139, 2025
reosarevok[m]
yvanzo: anything from your side?
2025-12-17 35102, 2025
reosarevok[m]
If not (or after that) we can move on to the discussions
2025-12-17 35106, 2025
yvanzo[m]
bitmap: Just submitted a partial review of your PR.
NOTICE: ● MBS-9209 (https://tickets.metabrainz.org/browse/MBS-9209): Allow individual users to opt out of HTTP Digest auth
2025-12-17 35141, 2025
NOTICE: ● MBS-9384 (https://tickets.metabrainz.org/browse/MBS-9384): Digest authentication fails for accounts where the username has been changed
2025-12-17 35105, 2025
reosarevok[m]
Oh, wait, this is a third way, on top of oauth or just entering your username and password
2025-12-17 35135, 2025
bitmap[m]
nonetheless it's a breaking change and I don't have any statistics on how often it's being used
2025-12-17 35150, 2025
reosarevok[m]
Is this something zas can get stats on?
2025-12-17 35100, 2025
bitmap[m]
well, this also uses your username and password
2025-12-17 35106, 2025
reosarevok[m]
In any case, if MeB does not implement it, we clearly should drop it
2025-12-17 35126, 2025
reosarevok[m]
Yeah I mean "in addition to the one I know where the browser gives you a username + password form" :)
2025-12-17 35157, 2025
bitmap[m]
depending on how much usage it has we could consider allowing users to opt-in by supplying a separate password from their main account password to calculate the ha1 value
2025-12-17 35144, 2025
reosarevok[m]
I mean, potentially if it needs to avoid breakage
2025-12-17 35156, 2025
reosarevok[m]
But we probably want to eventually remove it anyway, right?
2025-12-17 35157, 2025
yvanzo[m]
right
2025-12-17 35156, 2025
yvanzo[m]
Implementing the opt-in solution would probably be much more work than gathering usage stats.
2025-12-17 35124, 2025
bitmap[m]
yes but we should likely keep some kind of compatibility option for /ws/2, since it was the only authentication method for a long time and likely used by a lot of libraries still
2025-12-17 35155, 2025
yvanzo[m]
That changes everything.
2025-12-17 35158, 2025
bitmap[m]
e.g. python-musicbrainzngs uses digest auth
2025-12-17 35107, 2025
yvanzo[m]
So usage is high anyway.
2025-12-17 35141, 2025
bitmap[m]
it could be. it sounds like we should collect stats first anyway
2025-12-17 35141, 2025
reosarevok[m]
but if we add an opt-in method, that's also a breaking option
2025-12-17 35158, 2025
reosarevok[m]
In the sense that the library won't work without the user doing extra unexpected things
2025-12-17 35114, 2025
reosarevok[m]
I guess if the user cannot choose to move to oauth, maybe we still need to support this though
2025-12-17 35120, 2025
reosarevok[m]
And deprecate it with ws/2
2025-12-17 35124, 2025
bitmap[m]
it would force the user to enter a different password, but it would allow the application to function without any changes
2025-12-17 35123, 2025
yvanzo[m]
Your plan is sound.
2025-12-17 35102, 2025
reosarevok[m]
Yeah, it doesn't seem ideal but it does seem like the least bad of all the available options given the situation
2025-12-17 35100, 2025
reosarevok[m]
The idea of a second password is that if that one got broken via ha1, the evildoer could only access the few things accessible via API?
2025-12-17 35151, 2025
aereaux joined the channel
2025-12-17 35118, 2025
bitmap[m]
yes, and since it would be opt-in, the user would be made aware of the risks
2025-12-17 35143, 2025
reosarevok[m]
Not that they would have much of a choice depending on their library
2025-12-17 35146, 2025
reosarevok[m]
But yeah
2025-12-17 35103, 2025
aereaux
Is this the proper place to come for chatbrainz issues? When using the IRC bridge, I see `{"errcode":"M_NOT_FOUND","error":"Not found"}` for any attachments. I think that's related to something like this: https://mementomori.social/@rolle/113732824034474… . Is this something that can be fixed?
2025-12-17 35118, 2025
bitmap[m]
let's start with collection metrics, anyway, and see how many user agents are still requiring it
2025-12-17 35122, 2025
reosarevok[m]
Seems good :)
2025-12-17 35145, 2025
reosarevok[m]
I understand if we want to avoid breakage we'll need to have this alternative ready by the time MeB Oauth is released though?
2025-12-17 35156, 2025
bitmap[m]
technically we would still be storing the ha1 value associated with their current MB account password, and that won't break
2025-12-17 35128, 2025
bitmap[m]
at the same time it won't updated at all if they change their MeB username or password
2025-12-17 35138, 2025
bitmap[m]
* it won't be updated at
2025-12-17 35157, 2025
yvanzo[m]
Users should be informed about the MB separate password for ws/2 (and any other usage?) through the MeB username/password editing form.
2025-12-17 35101, 2025
reosarevok[m]
Oh. I kinda expected us to drop that ha1 to avoid issues in case of a leak
2025-12-17 35124, 2025
bitmap[m]
I mean we'll blank the field once the separate password can be set, but the risk of a leak isn't any greater than it has been before
2025-12-17 35129, 2025
reosarevok[m]
Fair
2025-12-17 35103, 2025
reosarevok[m]
So what happens if they change their pass before we can set a new one?
2025-12-17 35110, 2025
reosarevok[m]
(re: informing the user)
2025-12-17 35116, 2025
yvanzo[m]
(?)
2025-12-17 35114, 2025
yvanzo[m]
I don't understand your question, or I guess there is some misunderstanding that led to this question.
2025-12-17 35112, 2025
reosarevok[m]
"technically we would still be storing the ha1 value associated with their current MB account password, and that won't break. at the same time it won't be updated at all if they change their MeB username or password"
So if they change their MB password or username, I understand the log in option would break since we do not update the ha1 anymore?
2025-12-17 35157, 2025
reosarevok[m]
(until the time we allow them to set a separate pass for this)
2025-12-17 35145, 2025
yvanzo[m]
The MB ha1 should be valid at the time we move to MeB OAuth.
2025-12-17 35154, 2025
bitmap[m]
yes, but I'll aim to have something implemented that we can release at the same time
2025-12-17 35121, 2025
bitmap[m]
was just pointing out that the switchover won't immediately break every ha1 in case something isn't ready
2025-12-17 35127, 2025
yvanzo[m]
Absolutely, the solution must be implemented before the switchover.
2025-12-17 35159, 2025
yvanzo[m]
Does that answer your question reosarevok ?
2025-12-17 35117, 2025
reosarevok[m]
Sure, if we agree to postpone the switchover if needed until this is ready then I have no worries
2025-12-17 35155, 2025
aereaux
Jade[m], thanks! Is there anything I can do now to get the message by changing the URL, or is there nothing I can do but wait? Specifically for a DM I received that just said to go to a URL to see the full message
2025-12-17 35145, 2025
v6lur joined the channel
2025-12-17 35155, 2025
bitmap[m]
zas: would it be trivial to have openresty log all /ws/2 requests using 'Digest' in the authorization header to a separate file?
2025-12-17 35119, 2025
bitmap[m]
alternatively I can implement something on the MBS side
2025-12-17 35142, 2025
v6lur has quit
2025-12-17 35117, 2025
v6lur joined the channel
2025-12-17 35122, 2025
yvanzo[m]
Anything else about digest auth?
2025-12-17 35123, 2025
reosarevok[m]
What is the level of use where we'd be willing to just drop it?
2025-12-17 35148, 2025
v6lur has quit
2025-12-17 35154, 2025
yvanzo[m]
There is no doubt that it is used for the API.
2025-12-17 35157, 2025
reosarevok[m]
(I'm assuming if we want to know the usage is because it will potentially affect whether we implement this or not)
2025-12-17 35117, 2025
reosarevok[m]
(otherwise, we should just implement it anyway and knowing the usage is mostly just out of curiosity, right?)
2025-12-17 35125, 2025
yvanzo[m]
So it is out of question to just drop it.
2025-12-17 35135, 2025
v6lur joined the channel
2025-12-17 35142, 2025
v6lur has quit
2025-12-17 35122, 2025
reosarevok[m]
In that case I would just concentrate on implementing the alternative and I don't see them need to implement data collection at the moment
2025-12-17 35127, 2025
yvanzo[m]
See above (17:19) about compatibility option for /ws/2.
2025-12-17 35130, 2025
reosarevok[m]
(of course if it's trivial, sure, never hurts to know)
2025-12-17 35157, 2025
v6lur joined the channel
2025-12-17 35117, 2025
bitmap[m]
reosarevok: I would like to find out what the major applications or user agents using it are so that perhaps we can see if they are still maintained and can be updated to use oauth. but I plan to implement these changes regardless
2025-12-17 35137, 2025
yvanzo[m]
As you said, we will discontinue it in the future with /ws/2, so it would be nice to see if there is more than /ws/2 usage for future deprecation.
2025-12-17 35146, 2025
reosarevok[m]
Ok, gotcha, so these are two things that can happen at the same time, not a prerequisite
2025-12-17 35107, 2025
reosarevok[m]
I just wanted to make sure we didn't wait too long for the data analysis before working on this
2025-12-17 35119, 2025
v6lur has quit
2025-12-17 35132, 2025
bitmap[m]
nothing else from me on the digest auth topic :)
2025-12-17 35110, 2025
reosarevok[m]
Ok, thanks!
2025-12-17 35119, 2025
reosarevok[m]
No further topics from me at the time either
2025-12-17 35122, 2025
reosarevok[m]
yvanzo, you?
2025-12-17 35136, 2025
yvanzo[m]
Me neither.
2025-12-17 35143, 2025
reosarevok[m]
Then thanks all!
2025-12-17 35149, 2025
yvanzo[m]
Hi aerozol, I used to post to socials when releasing MBS updates. Thanks for having taken it over in my absence. It would be great to have identifiable graphic indeed, thank you.
