outsidecontext: much later, but thanks for the feedback!!
riksucks joined the channel
opal has quit
opal joined the channel
MRiddickW has quit
bitmap
reosarevok: sorry I forgot to reply re MBS-11602 -- I haven't seen those on sentry at least, and a search doesn't find any. so looks like a transient issue
reosarevok: Yes, thanks for the type definition of type options. I'm not sure about the ordering of properties though, my ESLint doesn't throw an error for that, and you left an unneeded console.log in the code. :)
reosarevok
Oh, whoops, I keep doing the console.logs
It's not an eslint error, it's a warning, but we try to avoid warnings too
There's more here, but some were already there earlier
Got rid of the console.log
There's one eslint error on that file, but it's years old, so I guess it doesn't affect this patch :)
yyoung[m]
I don't have those warnings here, they'll appear only in type definitions but not variable values. And what is that eslint error?
(since we only fully enforce the eslint issues we have fixed over the whole codebase, so we have two different eslint files)
yyoung[m]
And they didn't appear in CI tests, through we have set max_warnings to 0
reosarevok
Yeah, same reason :)
yyoung[m]
Yeah I guess that's the reason
reosarevok
The .unfixed file is ignored by the CI tests, because we know it has issues - we slowly try to fix them all, then move them to the standard file :)
yyoung[m]
The ordering warnings could be fixed automatically I guess
reosarevok
(I'm surprised the one about import groupings is still unfixed, but I guess there's some issues)
(I fixed that for these files, anyway)
Yeah, they could, but for example in URLCleanup IIRC yvanzo wants to keep them in the custom order, so we should probably just disable that rule there, heh
It's just useful to make sure each actual block is in the correct order (so, applemusic is before wikipedia), but annoying inside blocks
But it feels stupid to have 143 disable/enable lines, so maybe we'll just disable it for the whole object
Ignore those, in any case :)
yyoung[m]
BTW, I might have to squash your commit within the original 5 commits, but it contains fixes that are not in the scope of this PR, so I'm not sure now :)
MrClon has quit
MrClon joined the channel
reosarevok
Did I fix stuff that was already there? Maybe I did
yyoung[m]
We can set eslint-ignore at file level so we don't need 143 disable lines
reosarevok
Just put it with the flow bits I guess :D
Sure, or just have one disable line, it's just then it won't check the order of the link type blocks either, as I said
But that'd probably be the smart choice either way
yyoung[m]
OK so I'll squash LinkTypeOptionT into the last commit and merge the rest with the flow support commit
Sorting imports could be set in IDE configuration, auto-formatting rules also
sqlalchemy.exc.InvalidRequestError: When initializing mapper Mapper|CustomRecording|recording, expression 'RecordingFirstReleaseDate' failed to locate a name ("name 'RecordingFirstReleaseDate' is not defined"). If this is a class name, consider adding this relationship() to the <class 'sir.schema.modelext.CustomRecording'> class after both dependent classes have been defined.
Maybe I just have something outdated :)
lucifer
indeed, i do not like pydantic very much due its lack of validation. given the number of bugs we have had in the listen submission api, i think we may very well benefit from using a dedicated library that does validation.
Wait, actually none of those docker_whatever sound like "I'm a website"
Am I missing a container?
Maybe I should try a build again
Doing that
yvanzo
reosarevok: 'musicbrainz' is the MB website/webservice container
s/container/service/
(container's name is likely musicbrainz-docker_musicbrainz_1)
reosarevok
Yeah, probably removed it and never rebuilt it. Smart
Because sudo docker-compose build is doing a lot of stuff
I'll get used to this docker thing one day. Getting there.
btw, yvanzo, akshaaatt[m], bitmap: what do you think about slightly increasing our basic font size? From the forum, seems I'm not the only one consistently using MB at 110% or larger
yvanzo
Probably the most disturbing thing is that docker(-compose) is not building files were you run it, but under its data directory.
So when running the same project again from another place or with a more recent version, previous images, containers, and volumes may produce unexpected issues.
reosarevok
Ok, there's a musicbrainz-docker_musicbrainz_1 now, at least
Still not working though
Now for those logs
yvanzo
+where
reosarevok
That'd be then docker logs -f musicbrainz-docker_musicbrainz_1 ?
That said, it seems sensible to allow it and then trust that people who care about it will police it a bit if they get weird shit in their easy listening
CatQuest
I guess like "lounge music" ? would richard cheese and lounge agaisnt the machine be "easy listening" ??
reosarevok
We should have per-genre playlists or something, then people would be more likely to grumble and fix stuff :D
CatQuest
hopefully
reosarevok
"Easy listening music is often confused with lounge music, but while it was popular in some of the same venues it was meant to be listened to for enjoyment rather than as background sound." says enwiki
CatQuest
(tho i want a "sillyname" playlist too ofc)
reosarevok
lounge seems to be a subgenre of easy listening, to some degree?
I want a "fix-me" playlist :D
Then if I like something in it, I can go fix it
CatQuest
:D
not a bad idea actulaly
actually
reosarevok
So I'll add "easy listening" then, but feel free to downvote anything that doesn't seem to match the general idea :)
akshaaatt[m]
reosarevok We'll definitely have to re-adjust the font size
yvanzo
reosarevok: just an idea: sudo docker-compose rm --stop musicbrainz; rm /media/data/musicbrainz-docker/musicbrainz-server/lib/DBDefs.pm ; sudo docker-compose up musicbrainz
