#metabrainz

okno_: (and maybe others) I'm Ritiek and I've been somewhat involved with the musicbrainz_rs project lately. I've submitted an application on the GSoC forum to this project. I realize it's pretty late given the official application deadline is today, but better late than never! I'd be interested if anyone has any feedback to share on the application.

Hi yvanzo, I haven't received your gist comments yet, but I've made some changes to the schedule according to our last discussion, please have a look at the proposal on the forum so we can finalize the proposal today, thanks!

[listenbrainz-server] amCap1712 opened pull request #1386 (master…deploy-image): Automatically build and push image to Docker Hub when new tag is pushed https://github.com/metabrainz/listenbrainz-serv...

[musicbrainz-server] reosarevok opened pull request #2058 (master…MBS-11581): MBS-11581: Don't try to display / offer darkened art as RG cover art https://github.com/metabrainz/musicbrainz-serve...

yvanzo: if you want to do some reviewing today, I'll be around :)

reosarevok: Well that was fast... it's like you knew the ticket was going to be created 😆

I hadn't started work on it before you created, in any case :) It's just hopefully a small patch, unless I missed something

It's been so long since I've looked at MBS code. The perl looks so foreign to me now :(

hey! everyone

I have a query,if i am not wrong bookbrainz dont directly store users/editors email adress it rather stores editors metabrainz id . how can one collect the email adress of user using this metabrainz id , if possible ?

Mr_Monkey: :)

RikkoM: you should post to our https://community.metabrainz.org about your proposal too

hey navap 🙋

CatQuest: I did post it in the GSoC subsection. Is it not visible? Or is there another place I should post it too?

endurance21: That's a good question, and I don't have a 100% sure answer for you. Most likely we will have to establish a connection to the MB database, and armed with the metabrainz_user_id column fetch the users we're interested in. The BB and MB databases are hosted in the same postgres instance so configuration shouldn't be too tricky.

okay ! sounds promising !

thanks :)

[musicbrainz-server] reosarevok opened pull request #2059 (master…edit-preview-relationship-warning): Fix Catalyst warning: don't try using undef to access hash https://github.com/metabrainz/musicbrainz-serve...

bitmap, yvanzo: while I'm fixing warnings, https://github.com/metabrainz/musicbrainz-serve... fixes one and adds a filter for artist works lists that is extra useful now that relationships view is paged, can we get that one reviewed?

[musicbrainz-server] reosarevok opened pull request #2060 (master…no-parent-format-jsonld-warning): Fix Catalyst warning: ensure parent format id is not undef https://github.com/metabrainz/musicbrainz-serve...

[musicbrainz-docker] alastair opened pull request #190 (master…search-readme): Improve documentation for setting up solr https://github.com/metabrainz/musicbrainz-docke...

[listenbrainz-server] mayhem merged pull request #1347 (master…dump-user-feedback): Dump user feedback https://github.com/metabrainz/listenbrainz-serv...

Take that stupid feedback and dump it where nobody can find it!

well, if FTP is removed and we dont add HTTP capabilities to our download, then we're set, reosarevok

_lucifer: 3 minutes with cache sounds great! Is that just unit tests, or both unit and integration?

just unit tests.

but it should be straightforward to run both of them together i think.

let mw try that as well.

that was kind of the branch that I pushed for you a while back

[musicbrainz-server] reosarevok opened pull request #2061 (master…empty-review-markdown-warning): Fix Catalyst warning: don't pass undef review text to markdown https://github.com/metabrainz/musicbrainz-serve...

yeah right. but i got caught up in other things after that so wasn't able to progress on it.

sigh. I guess an all out war on the FSF will be coming.

alastairp: user dump feedback deployed and working as expected.

\o/

pristine___: user feedback dumps are now happening on a regular basis. will you have time to work on this spark side of things soon?

_lucifer: yeah, no worries. we can always deploy it as-is with the 4 test processes for now

did we merge or close the final jenkins fix?

neither, i thinks its open.

I wouldn't mind still adding in the additional --filter to `docker images ls`

then merge it

so that we have it in the git history

outsidecontext, zas: I've merged https://github.com/metabrainz/picard/pull/1790, but didn't do anything with the associated jira ticket. I don't know if what's currently in master will all make it into 2.7 or a later release. I'll leave that to you guys :-)

alastairp: updated Jenkins PR with filter on image name.

alastairp: do you know how I can choose where CB is built? I want to fix a couple bugs I found, but the docker images seem to go to /var by default and I don't have enough space in that partition

reosarevok: yes, you can change your docker root

Oh, so not in CB conf but globally for docker?

That's probably sensible too

Ok, so create /etc/docker/daemon.json and put a "data-root" in there

Let's see

alastairp, should i merge both jenkins and github actions?

go for it!

one sec,

let me look at the gh actions one again

what did you do with the junit-xml output?

i haven't added that yet.

we can add that in the same PR or i can open a new one. what do you say?

Getting that with ./develop.sh test up

there might be an old test db container running

which did not get cleaned up from previous test runs

Mineo: Thanks. It would go into the 2.6.1 release currently. It's probably fine, if we decide otherwise we can take it out

Ok...

I guess I wanted normal build, not test build anyway :)

I didn't realize that's not "debug/dev mode" but literally "run tests"

the db_test_1 logs are normal

just `./develop.sh up` for dev mode

Can I set this up to use my local MB DB copy?

Seems to be downloading its own dump atm...

yes.

you'll need to add an external network in the docker-compose file and change the database uri accordingly

i have that somewhere let me find it

_lucifer: I added some comments to the gh actions pr

some things to check first

That's what I do for MB when I want to access our live r-o DB via an ssh tunnel, which hopefully I can do here too :)

yyoung: posted comment to the gist

reosarevok: something like the above patch should work

Mr_Monkey: I've updated my proposal, have a look at it whenever you're free :)

yyoung: The schedule looks a bit better but it should be more incremental, that is delivering a small feature that works every two weeks, instead of trying to create full-feature components independently and to connect them on week 9 and then realizing it doesn't work when it's too late.

_lucifer: ERROR: Network musicbrainzdocker_default declared as external, but could not be found. Please create the network manually using `docker network create musicbrainzdocker_default` and try again. - just do that?

okno_: There is a draft proposal you might want to review :) https://community.metabrainz.org/t/gsoc-2021-co...

reosarevok: you should replace with `musicbrainzdocker_default` with the name of the network your musicbrainz db containers are running on

So, localhost? It's not a db container, it's local psql

@yvanzo I am on it :)

well, the one I access via ssh is technically a docker container, I guess, but I just access it via ssh tunnel and exposed port

no, i don't think you need to add this then. just changing the uri might be enough.

Let's see

ok i looked and it seems to be a bit more involved. alastairp or yvanzo might be able to better help you here.

what's going on?

you're on linux? and you have postgres running natively, and want to access it from within docker?

or pg is running somewhere else, but you tunnel it to a port on localhost?

_lucifer: how does the setup of the new spark cluster?

ruaok, i was reading up some more about it over the previous. we can probably do that today.

I'm around for the next 5.5 hours if you need help.

nice, thanks!

alastairp: yeah

I tried MB_DATABASE_URI = "postgresql://musicbrainz:musicbrainz@localhost:65401/musicbrainz_db" but that doesn't seem to do the trick

no, that won't work. because localhost is local to the container, not your host

alastairp: both are true - I have a sample DB locally and a full db via localhost port tunnel

Oh, I see

yeah, which setup is it?

I was trying to configure it for the tunnel

For now

two options: mount the postgres socket into the cb container, that'll give you the sample db

or you need to work out the docker ip address of your host (172.something.something maybe) and use that in the config

alastairp: _lucifer ignore sentry errors on user similarity stuff. testing some fixes.

reosarevok: the sample DB is in your system or in mb docker compose project?

My system

And the other db is on a ssh tunnel from pink but the only change I make to access it via DBDefs is specifying the port

(and localhost)

I'll try and find the ip

possibly 172.17.0.1

Hmm

"As of version 18.03, you can use host.docker.internal as the host's IP."

alastairp: https://github.com/metabrainz/listenbrainz-serv... this is locally reproducible by running `docker-compose -f docker/docker-compose.integration.yml pull` in the repository root

Guess not