#metabrainz

Project bookbrainz-data-js build #1301: passed in 2 min 17 sec: https://travis-ci.org/bookbrainz/bookbrainz-dat...

[picard:update-mutagen-1.43@b471025 - build #202] CI passed! (https://travis-ci.org/phw/picard/builds/613359977)

mo’’in’

Moin!

Thanks for the feedback CatQuest, I'll have a look

mo''''in!

Morning

Mr_Monkey: hi!

yea I was adding the cd attached to this book to mb, so i was going ot dually add the book itself to bb

CatQuest (Mr_Monkey): https://tickets.metabrainz.org/browse/BB-302

BB-302: Date inputs don't allow partial date (only year, year-month)

!

thanks for finding this freso

man it's an year old?

it needs fixing prnoto! :D

CatQuest: I reported https://tickets.metabrainz.org/browse/BB-335 which is somewhat related, so… :)

BB-335: Can't edit release date of edition

oic

There's been a lot of work on dates generally, so small issues found themselves part of a longer testing/release cycle

ruaok: moin

moin!

> so, given that you have data that has an array of mbids, it needs to be matched an artist_credit. this avoids the explosion problem.

I am still not sure how to do it. Can you give a lil example about what is there in your mind, I mean any pseudo query involving ar-ar JSON and ar-credit JSON?

here is what I am going to do to make this easier...

I'm going to have the MSID mapping output artist_credit_is, instead of an array of artist MBIDs.

_id

that should make this change trivial for you, yes?

ruaok: so I match the artist_credit_ids to get artist_names from artist_credit_relation? Can you make a gist a with a sample table of a few rows ( like I posted one before with A,B,C) so that we better understand everything and know what suits us best before you start on modifying mapping or stuff? So in particular a gist of MSID mapping output artist_credit_id with correct col names but random col values, I can then just

perform joins and stuff on that gist and inference results so that in order to make things better we don't miss out on some points and land on the point from where we started.

Also, you have been putting in so much time so let's just be sure.

ok, i'll get you something in the course of the day.

Yeah. Sounds good

[bookbrainz-site] MonkeyDo opened pull request #321 (stable…master): November 18th production deployment https://github.com/bookbrainz/bookbrainz-site/p...

[bookbrainz-site] MonkeyDo merged pull request #321 (stable…master): November 18th production deployment https://github.com/bookbrainz/bookbrainz-site/p...

Project bookbrainz-site build #2428: passed in 2 min 43 sec: https://travis-ci.org/bookbrainz/bookbrainz-sit...

ruaok: https://github.com/metabrainz/listenbrainz-labs...

I added a comment, can you have a quick look?

going forward, can we please add tests for as much new code as possible in our pull requests

iliekcomputers: yes, we will :)

Awesome!

CatQuest: ^

[listenbrainz-server] dependabot-preview[bot] opened pull request #668 (master…dependabot/pip/pytest-5.2.4): Bump pytest from 5.2.2 to 5.2.4 https://github.com/metabrainz/listenbrainz-serv...

pristine__: honestly, I can't even follow your comments right now.

:(

I wish I could explain you face to face with pen and paper. Okay give me some time

what is more important? the artist_credit mapping stuff or providing feedback on this PR?

I will prepare a gist with tables and every step

former

no, please don't.

Okay.

Mr_Monkey: Do you have a minute to talk?

I just want to ask you questions.

abhishekpanwar: I do !

but first, what should I focus on first?

cool. Sounds good

PR or ac question?

artist_credit IMO

ok.

Yeah So I think there is something wrong in hosts used in the config

:)

Like when I do what is mentioned in the new debug docs then debugging works fine but normal develop.sh dont

But If I follow normal setup procedure then debug doesnt work but develop.sh works

Regarding elasticsearch

I think the hosts are not matching

Somewhere elasticsearch:9200 is usedand somewhere localhist:9200

*localhost

Do you think there is a chance of some descrepancy?

I dont think this issue is only specific to me

Right. So if you're using develop.sh, everything is run using Docker, so the hostnames in the config should be the same as the docker container names (defined in docker-compose.yml).

If you're running the server locally with node, or through vscode, then you need to modify the config to point to `localhost` instead of the container names, which won't mean anything in that case.

Does that clarify things for you?

So we can make them the same so we dont have to switch the config file everytime we debug or develop?

So depending on which way you run the server, you'll need to change the config accordingly

I'm afraid we cannot make them the same

But I sure can try to document it better

I guess if someone has basics of docker clear, might be able to understand it. Ill look into docker I guess.

Thanks

If you want, you can create two configurations and just switch the file names depending on debug or develop

Yeah, good idea. Will do this only.

That's also what I do, but I agree it's cumbersome. I'll think of a way to start the server and pass a config file path to load. Ideally, develop.sh would work the same as today, and you'll also be able to run, say `npm start --config ./config/config.local.json`

Yes, would be great if it was an option

To make it more clear for you, assuming all dependencies are running in docker. If you run 'develop.sh', server will run inside docker, thus to connect to other docker instances (dependencies), we need to tell it their 'names'. If we run the server locally, it's outside of docker, and it doesn't understand those 'docker names' we have given before, so we pass the location they are running (localost with port)

BB-366: Allow overwriting config file path

There we go ! ^

I was curious to know that when schema of MusicBrainz, LB and other projects was built how did the developers interacted with each other to arrive on a sol and make each other understand there thought process? I mean was/is there a dedicated platform? might help.

Good, It would make the development process easier.

IRC?

:(

well, the MB schema was far more than that.

then I don't know how to explain :(

5 years of planning and 3 separate in person summits focused on the new schema.

wow!

this was the summit that finalized NGS.

hah

It looks like something out of a "garage tech bros" documentary

but in a super fancy flat in London. hah.

really amazing pictures

thanks.

that was a super intense long weekend.

my favorite part was when there was something that wasn't making sense to us.

we felt we needed to change something. we couldn't put our fingers on it.

we got hungry, so I went for sandwiches.

when I came back everyone was all smiles: "We fixed it!"

"what did you change?"

"nothing."

O_O

visuals help a lot :) I think we should do something like this in the next summit.