#metabrainz

lol

zas, ruaok: I'm making a clone of our prod postgres cluster onto williams so I can test the 9.5 -> 12 upgrade against our prod setup & data

in case you notice disk space being eaten up there

Project bookbrainz-data-js build #1445: passed in 1 min 53 sec: https://travis-ci.org/bookbrainz/bookbrainz-dat...

MBS-10764: Convert historic Remove Label Alias edits to React https://tickets.metabrainz.org/browse/MBS-10764

[musicbrainz-server] reosarevok merged pull request #1464 (master…MBS-10765): MBS-10765: Convert historic Add, Move and Remove DiscID edits to React https://github.com/metabrainz/musicbrainz-serve...

MBS-10765: Convert historic Add, Move and Remove DiscID edits to React https://tickets.metabrainz.org/browse/MBS-10765

[musicbrainz-server] reosarevok merged pull request #1469 (master…MBS-10773): MBS-10773: Convert historic Change Release Quality edit to React https://github.com/metabrainz/musicbrainz-serve...

MBS-10773: Convert historic Change Release Quality edit to React https://tickets.metabrainz.org/browse/MBS-10773

Project bookbrainz-site build #2909: failed in 2 min 47 sec: https://travis-ci.org/bookbrainz/bookbrainz-sit...

!m chaban for your work scouring the blog for timeline suggestions! https://tickets.metabrainz.org/browse/MBS-9495

MBS-9495: Update timeline with new events

You're doing good work, chaban for your work scouring the blog for timeline suggestions! https://tickets.metabrainz.org/browse/MBS-9495!

ishaanshah[m]: hey

iliekcomputers hi

interesting stuff, but do we really want to listen to machine generated music?

ruaok: I don't see why not? :)

I mean, we do listen to nature-generated music already :D

Although I suspect they're trying to make computers make human-like music, which does seem more pointless

Project bookbrainz-site build #2911: failed in 2 min 49 sec: https://travis-ci.org/bookbrainz/bookbrainz-sit...

ruaok: Yes! Generative art is a new frontier, and all frontiers should be explored.

Now I'm not saying it's necessarily 'good' music, but I'm sure there will be some generated music I prefer to [insert some musical style or artist], which I guess is litmus test #1

I'm more excited about what new things could be produced than some rehash "in the style of", to be honest, but I suppose it's a step towards it. Also this: https://en.wikipedia.org/wiki/Oulipo

[musicbrainz-server] reosarevok merged pull request #1471 (master…MBS-10775): MBS-10775: Convert historic Add/Remove rel edits to React https://github.com/metabrainz/musicbrainz-serve...

MBS-10775: Convert historic Add/Remove Relationship edits to React https://tickets.metabrainz.org/browse/MBS-10775

ishaanshah[m]: hey, just wanted to chat about what you wanna work on next when you get the time

Mr_Monkey: I've listened to a few tracks and I think it is amazing what they've done so far. But, dragging that across the finish line and to make it good music, is going to be a stretch.

Yes, and also the interesting part

and we have enough talented, human musicians, already...

Well, I see it more as an instrument. After all it's being controlled by humans

iliekcomputers: I wanted to work on something related to the backend, as I haven't played with that part of the code yet

awesome, that's what I was thinking too

did you see the jira epic I made?

there's a task there about making the artist stats available via the API

i think that would be a nice starting point

ruaok: isn't that kinda like saying we shouldn't write more classical music because we have a ton of good stuff already? :)

Yes, that sounds good

great

I have an exam tomorrow

oh

yeah, you should focus on that the

then*

LB can wait :)

I will start working on it after that

Will I have to setup stats generation on my local environment

ping me before starting, i'm hoping to experiment with jsonschema or protobuf while doing this task too

ishaanshah[m]: i do not think so

Or should I just assume the schema

i would just assume the schema

i'm gonna try to document the stats stuff this weekend

(or at least improve the docs and dev environment)

I will get familier with the backend codebase and ping you

great.

jsonschema is for enforcing structure right?

if you need any help or have any questions, happy to help

ishaanshah[m]: yes

right now we just put in data willy nilly

which makes that part of the code very hard to change

as we're not sure what the schema of the stats is

I had written down the possible schemas while writing my proposal

I will write them in a document and share it over here

hmm, great thanks, this is more about adding the schema to the codebase and enforcing it.

Project bookbrainz-site build #2913: passed in 2 min 36 sec: https://travis-ci.org/bookbrainz/bookbrainz-sit...

iliekcomputers: hey, what part is this for? In our experience with AB, jsonschema was a bit hit and miss

alastairp: we store stats in the database in a single JSONB, and are gonna have a lot of them soon-ish, i guess.

at least, at the time we used it, it could be used for saying yes/no during validation, but it wasn't great to use its response to give feedback about what was wrong, so we ended up writing our own validator that gave nice error emssager

messages

i'm wondering if we need to move the schema into postgres

so it depends on why you want the validation

or maybe use some other way to codify the schema

seen stuff like https://pydantic-docs.helpmanual.io/ ?

it'll convert to/from json too

i haven't, thanks!

that does seem like it'll fit the usecase here

I believe that it'll integrate with dataclasses too

I'm becoming less and less enamoured with plain dicts. they're great for just creating and throwing data around with

but yeah, validation becomes hard very quickly

alastairp: yeah, true. it becomes a pain to actually change code too, because there's no easy way to know what the code is expecting

[listenbrainz-server] dependabot-preview[bot] opened pull request #815 (master…dependabot/pip/pytz-2020.1): Bump pytz from 2019.3 to 2020.1 https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] dependabot-preview[bot] opened pull request #816 (master…dependabot/pip/google-api-python-client-1.8.2): Bump google-api-python-client from 1.8.0 to 1.8.2 https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] dependabot-preview[bot] opened pull request #817 (master…dependabot/pip/sphinx-3.0.3): Bump sphinx from 3.0.1 to 3.0.3 https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] dependabot-preview[bot] opened pull request #818 (master…dependabot/pip/flask-socketio-4.3.0): Bump flask-socketio from 4.2.1 to 4.3.0 https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] dependabot-preview[bot] opened pull request #819 (master…dependabot/pip/click-7.1.2): Bump click from 7.1.1 to 7.1.2 https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] dependabot-preview[bot] opened pull request #820 (master…dependabot/pip/redis-3.5.0): Bump redis from 3.4.1 to 3.5.0 https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] dependabot-preview[bot] opened pull request #821 (master…dependabot/pip/numpy-1.18.3): Bump numpy from 1.18.2 to 1.18.3 https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] dependabot-preview[bot] opened pull request #822 (master…dependabot/pip/jinja2-2.11.2): Bump jinja2 from 2.11.1 to 2.11.2 https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] dependabot-preview[bot] opened pull request #823 (master…dependabot/pip/spotipy-2.12.0): Bump spotipy from 2.11.1 to 2.12.0 https://github.com/metabrainz/listenbrainz-serv...

alastairp: would you get a chance to look at the messybrainz PR today? I'm hoping to merge it over the weekend

ohh, I haven't looked at that at all.

anything in particular I should look at?

iliekcomputers: I saw that you opened all the tickets that I suggested. It's fine by me

would you like me to approve it?

alastairp: yes, thanks!

ruaok: it's just messybrainz code with some cleanup.

i also opened a ticket for removing the webserver part of the code completely.

ok, if you two already got it, then I'll wait until its merged and then I'll need to move my mappings branch from the old repo to the LB report.

sounds good!

how long does it take now to generate a LB dump, iliekcomputers?

I'll use the latest dump for testing, but for a test migration, we'll need to trigger a dump after we connect the timescale-writer queue....

i triggered a dump yesterday, there hasn't been a full dump in the month of april for some reason, wanted to debug, i'll have the exact time tonight or tomorrow I guess :D

ok, ping me before you trigger the next one.

off the top of my head, the listens dump should take 24-36 hours

wow.

yeah...