#metabrainz

need to look over a bigger time period, https://stats.metabrainz.org/d/OGg5QUCGz/listen...

eg:see 10/23, 10/29

Yep

What are the usual symptoms if the spotify reader on our side is blocked/crashed?

As opposed to Spot API issues

the spotify reader logs or sentry usually have something to tell

there are a few false positives though, occasional read timeouts, service unavailable stuff.

those are retried so at worst expect a ~7 min delay.

these were the logs an hour or so ago. its iterating continously but finding nothing to import

Hence "Encountered errors while processing 0 users" ?

I get mixed messages there :D

uh yeah, that message needs to be improved.

sloppy error message from me instead of no errors, errors for 0. i'll open a PR to improve it.

we should work out why we're logging to console twice too, that's been bothering me for a while :)

me too

that's probably because we set up both flask logger and a normal python logger for listenbrainz package. i had tried a few suggested solutions but nothing worked.

[musicbrainz-server] 14reosarevok merged pull request #2160 (03master…MBS-11764): MBS-11764: Make user tags sortable by count https://github.com/metabrainz/musicbrainz-serve...

[musicbrainz-server] 14reosarevok merged pull request #2024 (03master…MBS-11537): MBS-11537: Collapse work attributes when there are too many (but for edits) https://github.com/metabrainz/musicbrainz-serve...

[listenbrainz-server] 14MonkeyDo opened pull request #1725 (03master…monkey-fetch-covertart-LB-904): LB-904: Best effort to show cover art in ListenCard https://github.com/metabrainz/listenbrainz-serv...

!m monkey

You're doing good work, monkey!

Meh, it's not perfect by any means, but I don't think I want to wait until I have a perfect solution :p

good enough :D

bitmap: I'll release beta in a bit, lemme look at the milestone again and see if there's anything else that seems ready

Oh. Not for today, probably? But https://github.com/metabrainz/musicbrainz-serve... might still show to you as approved but that was when I had 2 commits there, so you should re-check before we merge

And I wanted to ask about https://github.com/metabrainz/musicbrainz-serve... again - guessing not today either, but what's the plan with that?

lucifer: still around? just throwing around ideas for these schema numbers. should I swap your change and increase the correct value? (it has been a while since you made it, but perhaps better now than never)

alastairp: yes makes sense to do it.

great. will increase the correct one - leave the one that you chaged, as we don't want to go backwards

and will also rename them and add some clearer documentation about when to update

i don't think we use that one anywhere though so shouldn't matter

we will as of this PR :)

+1

ah right

lucifer: fyi, I've been commenting on a bunch of your PRs today

if you've not seen them yet

oh! i only saw one mail and replied to that.

ah i see the others now. will look

thanks!

no problem!

hopefully more tomorrow

finishing up here, I'll try and get another push to the dumps tonight, but still some things to do. if you want to pick it up tomorrow then no prob from me

sure sounds good. most of it looks ready already to me!

right. tests, sequence numbers, testing import, double-check ftp upload

first 2 are done

reosarevok: I think the perlcritic one is still ok to merge

Have you actually rechecked? :)

If so, I'm happy to

sounds good. i'll test the import and upload tomorro

reosarevok: I did a while ago whenever you added the new commits, dunno how recently you changed it besides rebasing

I didn't, AFAICT

I guess I should rebase again, make sure it still passes

reosarevok: re the other PR, I guess I was waiting to test it against test.mb...but I didn't want to update test.mb yet because I had dbmirror2 stuff running there

which was patched into the running container

I'll skim the perlcritic one again while you rebase

Ok :)

bitmap: remember there's https://github.com/metabrainz/musicbrainz-serve... too :)

I'll test it in a couple mins

if you wanna wait before updating beta

Yeah, I'm waiting for tests and whatnot anyway

[musicbrainz-server] 14reosarevok merged pull request #1868 (03master…MBS-11059): MBS-11059: Indicate which release groups have CAA art in listings https://github.com/metabrainz/musicbrainz-serve...

bitmap: perlcritic still passes after rebase on master it seems

bitmap: also, whoops, we've already merged the others that do this, re https://github.com/metabrainz/musicbrainz-serve...

I guess I can add a commit on top for all of those

If a page needs more than one manifest, just separate lines works?

oh, I thought I already commented about that on the other one

maybe it was on IRC

I guess if you did I missed it, whoops.

So move them to the page that *calls* the table with these as columns, then?

So anything using defineArtistRolesColumn gets this, etc?

if you can't include it with the table, or if the table is used multiple times per page, then yeah

Hmm, does <Table> accept children? I'm guessing not, in which case I'd need to return a fragment with the table in it

I forgot react-table makes that difficult

I guess I could make <Table> accept a conditional that then calls the manifest

If we wanted to go that way

probably just move the manifest.js calls to the page components then

well, I guess we'd be passing it from the page component either way

either directly or passing it to Table as children

So I guess rather than

return <Table columns={columns} data={events} />;

We'd have

yeah

finished re-skimming the perlcritic one and it looks fine, though stuff like $foo->{"bar"} can just be $foo->{bar} instead of changing to single quotes

Wanna send a quick extra commit? ;)

sure

but just in the one file it was being changed in, not the whole codebase :)

So that's also ok?

yes

actually this file isn't even used - admin/replication/ConvertReplicationPacketToXML

heh

I'll add a commit to remove it though, heh

I wasn't sure about all those old scripts

Updated the manifest thing

What happens if a page has a table twice, though?

Like, is it only a waste of bandwidth or will it actually cause issues?

normally it will execute the script multiple times which could cause issues, but webpack might prevent that from happening, not sure

bitmap: so, https://github.com/metabrainz/musicbrainz-serve... and https://github.com/metabrainz/musicbrainz-serve... seem ready now? :) If so, I'll merge and beta

alastairp: ruaok: i think we should trigger full dumps as is now. the listens dump is going to take a lot of time. we can patch cron once that is done to dump everything else but listens.

I didn't re-test it but the code looks good

and yeah, the script is executed multiple times, but webpack only runs the module definition part once

(so it works but it's a minor waste of time and space)

Ok. Well, hopefully it won't happen often

I just worried about us accidentally putting two tables in one page that call the same manifest

Without noticing

But if nothing serious happens, then

[musicbrainz-server] 14reosarevok merged pull request #2156 (03master…MBS-11756): MBS-11756: Collapse artist roles when there are too many https://github.com/metabrainz/musicbrainz-serve...

it should be fine

[musicbrainz-server] 14reosarevok merged pull request #2283 (03master…more-perl-critic-conditions): Enable more Perl::Critic conditions https://github.com/metabrainz/musicbrainz-serve...

Ok, releasing beta then

Done

lucifer: ok, lets do it.