#metabrainz

but I do think that setting up a full python 3 image would be the right thing to do, so that we don't have to make changes in the future

so, this file: https://github.com/metabrainz/acousticbrainz-se...

copy it to `docker/Dockerfile.py3`

and delete https://github.com/metabrainz/acousticbrainz-se..., and change the `FROM` in the first line to 3.7

ok!

I'm around all day, so feel free to ask me any additional questions if you run into problems

Another question, is how do you suggest to load the `models.sklearn.the_relevant_method` to the `dataset_eval.evaluate`?

with relative imports for example?

so, we actually run the program in the /data directory

so, we don't do: cd dataset_eval; python evaluate.py

we do `python -m dataset_eval.evaluate`

(we don't do exactly that, we use worker_manage.py, but the principle is the same)

that means you can just do "import models.sklearn" in evalaute.py and it'll work

that's what I saw in the dockerfile

ok cool!

thanks Alastair!

what better way to get bug reports than to release a new major version :)

alastairp: this is very true ;) Also shows that we had two beta releases this time out for several weeks, but the real bug reports only came in after the final release

Data about Picard installs from the Windows Store https://usercontent.irccloud-cdn.com/file/8bXPG...

Picard - A software for old men :P

A software for people who don't stream all their music

So, yeah

Speaking of lemmy!!

Upgraded docker-compose to latest (1.26.2) on bono (in /usr/local/bin)

[musicbrainz-server] reosarevok opened pull request #1654 (master…MBS-11041): MBS-11041: Make ModBot leave a note on autoremoval edits https://github.com/metabrainz/musicbrainz-serve...

MBS-11041: ModBot should leave note on autoremoval edits https://tickets.metabrainz.org/browse/MBS-11041

Is there any custom data in musicbrainz-docker db on bono that should be saved?

I will recreate that database from a fresh dump.

check if there's a messybrainz schema, I believe this is the result of ruaok's matching code

I think that it's easily reproduceable if you don't want to save and reload it

Ding. Relations and mapping schemas exist. Feel free to blow away, I can recreate.

Ok, will do. Is MB search server useful on bono?

I don't need it. I don't think the mapping stuff uses it either, we just connect directly to the db

Ok, built new images, downloading db dump.

Not for me either.

Ok, I stopped it too.

ishaanshah: hey, i have a conflict at our usual time, let's postpone by ~30 min, ok?

shivam-kapila: hey, how goes it?

Sure

iliekcomputers: Last test left for frontend then only tests for backend left

cool, sounds good. thanks. if you need any help, ping here.

alastairp, iliekcomputers, ishaanshah, ruaok: I’m going to nuke MB DB on bono, unless you're currently working with it.

fine by me

alastairp: after docker-compose upgrade, you probably need to specify 'acousticbrainzserver' as COMPOSE_PROJECT_NAME in .env to preserve current docker containers/volumes on bono

thanks. is this a service that you saw running in docker ps?

(or alias to the old /usr/bin/docker-compose)

Ok!

I believe we use a launcher script to add -p to docker-compose call, so it should be OK

what's the new behaviour?

(having said all of that, there will be no problem to recreate all containers and volumes, so I don't think this is an issue)

alastairp: I saw some with docker system df -v

the old behaviour was to ignore '-' characters from the current working directory, the new one is to keep it

ah, interesting. thanks for the info

I have set that var to 'musicbrainzdocker' cause I don't want to download dumps again

we should double-check our develop.sh and test.sh scripts for this, because I believe we do some modifying of this value

This is one motivation for requiring Docker Compose 1.21.1 at https://github.com/metabrainz/musicbrainz-docke...

(next question: why does musicbrainz-docker require 1.21.1?)

(and Ubuntu 18.04 comes with 1.17.x but it is easy to install a newer version)

(did I just answered next question already?)

I wasn't sure if there was a specific reason that you updated it. (I have no specific reason for asking, any version is fine for me)

That and probably annoying bugs that have been fixed since then

cool, great

thanks

no problem

10k

(did I just answered next question again?)

Stopped MB DB on bono.

don't talk to me about that, talk to my boss

iliekcomputers: can you take a quick loo at https://github.com/metabrainz/listenbrainz-serv...

I want to merge it today

look*

pristine___: can't right now, will take a look today.

thanks

iliekcomputers: ping

or ishaanshah

sup?

I was writing jest tests

I was jest writing some tests

This test is failing at line 18

saying no of calls = 0

Any idea what I am doing wrong

could be because your fetchCoverArt is called in componentDidMount and you're using `shallow`

can you try wrapper = mount(<ListenCard />)

same error

not sure then.

oh wait you were already calling componentDidMount explicitly

huh

then the mount not changing anything makes sense.

that's weird.

does the assert on line 19 pas?

nope

solved

line 8 is not needed and causing error.

iliekcomputers: 1038 is ready

awesome, did you do the integration into the current listens page as well?

Nope

I mean the code is ready with me

could you do that as well? I'd prefer to do it in the same PR

So we will release it?

the card?

yes.

Shall I also add the listen count card impllementation code

the count card, not in this one

okk

I will push for this

sounds good, thanks

alastairp, iliekcomputers, ishaanshah, ruaok: MB DB @ bono is back, starting replication now.

yvanzo: i don't have anything on bono, no need to ping me.

thanks yvanzo

send the invoice to ruaok

iliekcomputers: it relates to MeB dataset hoster IUC

iliekcomputers: done

shivam-kapila: thanks, i'll take a look today.

alastairp, ruaok: mb mirror @ bono is set to query search.musicbrainz.org if needed. ('indexer' & 'search' containers are stopped.)

also set website hostname to fix bundling of scripts/static resources.

is it publicly accessible?

I'm not sure if we have anything that uses the webservice, we just use it for the db

iliekcomputers: hello

yes, it is

Hey, give me a sec

hey ishaanshah i don't think we need to do the entire jira shindig this and the next week.

stopping search container saved a few GBs of RAM at least

yep yep

just wanted to update you on what I did today

i assume you'll be continuing on the artist graph and then start incremental dumps as soon as the artist graph is done

yep

awesome.

I worked on the tests

couldn't complete them unfortunately

not a problem.

I would prefer if you have a quick look at the non tests part

when do you think you'll be able to finish?

ok, that sounds good, i'll add it to my list for today

I'll try tomorrow, but not sure, I have a ton of classes tomorrow