#metabrainz

Yea, that's the sad part about job interviews TBH

But also a reason why I in turn like GSoC so much. Selection based on past OSS contributions is such a brilliant idea

it also removes most biases from the interview process, since the selection is based on the applicants performance specifically inside your organization

True that, especially after removing the student-status requirement GSoC is the best opportunity to gain any experience for graduates

Bascially, one just need to know the stack and get in touch with the org early

yep that's what it seems like so far. Also I think culture fit, etc takes care of itself through the very long nature of the selection process

<akshaaatt> "yellowhatpro: definitely, you..." <- okii then I will be preparing a mock design asap ヾ(•ω•`)o

outsidecontext, zas: Hello, could we start already? I have electricity problems, so I can be up for as long as my laptop and phone batteries will last

The problems started like 20minutes ago, my luck in a nutshell haha

mayhem: i'll recheck it. the container is not running master atm but that has only been the case since yesterday and jams have been broken for a week.

I'm around

I've been tracking that down, but pylb 0.5.1 keeps getting installed and I have no idea why...

oh that's because building from source fails so pip fall backs to latest version from pypi.

that's the issue i faced last time and why we need a pypi release.

but we could also change how pylb is versioned and fix building from source.

alastairp and I did just that.

it should build from source now.

maybe something else broke now.

hmm, that's weird

outsidecontext: skelly37: zas fyi we're having an LB meeting now too

understandable

aight. lets get this meeting started.

lucifer: is this error reproduceable by building current master?

alastairp: yes.

ok

i tried installing `git+https://github.com/metabrainz/pylistenbrainz.git@v0.5.2#egg=pylistenbrainz` in isolation and it works

I did too.

outsidecontext isn't there anyway, let's decide what to do when he shows up

so some other issue probably.

so, lets talk about LB planning for a few minutes.

sure

first off, I'm not going to be around a whole lot in August. Starting July 27th.

monkey doesn't seem to be around.

I've also got conference last week of july + holidays first week of august

my point being, our schedules are going to be pretty fragmented.

skelly37, zas: I guess we should move, because two meetings in parallel won't work

makes sense

i'll be around for rest of july and august.

and I think we ought to not have another sprint before sept or even later sept.

I've been mostly doing CB maint, SoC work, and finishing off AB dumps the last month or so, so not much on LB either

how about near/during summit then?

monkey and I have this pending react sprint too, so there is something for us to work on together if we are around/have time

I think some hacking time around the summit could be nice.

outsidecontext: the same hour but tomorrow?

I wouldn't mind trying to do something in sept if possible

especially if that is something like oauth (so that we can talk about results at the summit)

the one thing that I feel I need to do soon is to have some page where users can see the playlists we generate for them.

skelly37: would do for me, zas?

yes, I think oauth would be a good thing to plan at the summit. and then to execute right after -- so that we all know what needs doing.

skelly37, outsidecontext : I can't tomorrow at this time, what about Monday one hour before usual meeting?

works for me

do you mean a page having text description of playlists ?

Monday evening is bad for me

+1 on oauth

lucifer: more like a page that shows off daily jams and other playlists that will be generated soon.

ok, looks like doing oauth at summit + after is a better idea than before + at summit

right now these playlists are too hidden in the collaborative playlists.

zas: Are you off the whole tomorrow's afternoon?

i see, sounds good.

mayhem: so a bit more styalised than the current recommendations tab?

nevermind then

stylised

Saturday or Sunday evening would also do for me. Tomorrow afternoon not

weekend is okay for me too

I hadn't really thought about it, yet, but those pages could be repurposed.

outsidecontext, skelly37: then Saturday same time?

fine

a new tab for "created just for you" would be nice.

zas: +1

ok :)

do we have some other playlist things to do? (server-side automatic write to spotify, finishing those tests of mine, ...?)

perhaps this is a mini-playlists sprint worth of work

alastairp: besides writing tests for the playlists code, you mean? :P

hey, I put it in the list!

outsidecontext, zas: To make our talk more smooth, I'll just mention that I got interested in PICARD-2519 (urls in command-line), so I'll be asking questions how you see it, how urls work now etc.

we'll manage :) Now let's have the LB guys have their discussion

PICARD-2519: Allow passing supported URLs on command line https://tickets.metabrainz.org/browse/PICARD-2519

:)

and until saturday then :)

alastairp: your react skills are sharp, right?

yes

we could cooperate on this tomorrow.

err next week.

skelly37: great :)

sounds good. tues/weds is good for me

perfect, lets do that.

I'll come to officebrainz

I've got to debug this pylb thing and some other cleanup, but I'll be ready for tue/wed.

nice.

perhaps I can knock out a second playlist as well, but we'll see.

lucifer: what are you planning on working on the next month or so?

I feel that we really need to be able to publish playlists to music services.

brainzplayer simply does not work for day to day listening like daily jams.

in the near term getting couchdb addition done, reviewing riks PRs and finishing up open PRs.

so I would like to be able to tick a box that says: "send my playlists to my music service"

the spotify metadata cache would likely need to get implemented before we do this.

but that is a longer project... and no one really seems all that interested in it.

i think next week i could do some tech debt tasks, like moving msb to TS, working on that listen table migration, trying out kombu in other places of LB to see if it helps.

Sorry I wasn't here y'all. I also see fragmented summer schedules making it hard to have a full team sprint, but happy to move the react router project with alastairp when we find the time

hi monkey!

Yo!

s/next week/next month

that seems like a good goal for aug.

lucifer: I was wondering how much sleep you were planning on getting...

great

mayhem: got a moment to follow up on some other stuff

?

hehe.

sure, if we're sufficiently happy with planning for now.

I'm also working on a few improvements to finish off the listening now feature, so that we can actually publicly release it

I think I'm OK for planning, happy to wait for anyone else to chime in

I think saying that Aug is to tidy loose ends and clean up some debt (or go on vacation) is ok by me.

So might rope everyone in for aome feedback on that

monkey: great -- I'm quite enjoying that one.

yes, hit me.

monkey: listening now will also require updating that metadata table, which is also a pending task which I've done a bit of but not finished

loads of debt to go around. I guess our plans are set then. finish shit. :)

Oh? What are we missing on the db side?

shall we move on to misc discussion?

monkey: automatic updates.

monkey: it doesn't update when new data goes into mb

mayhem: I'm going through my notes for making dumps of the mapping, and found that we're already dumping public.mbid_mapping_metadata and public.mbid_mapping: https://github.com/metabrainz/listenbrainz-serv...

Yeah ok that makes sense :)

I can't remember if these are an older version of the mapping.canonical* tables, or if they're still required

those are different tables -- that is the mapping itself. the other tables are the canonical tables.

alastairp: so both are required. but we may need to review all that for nomenclature, since things have changed quite a lot over the evolution of this data.

yeah, right. and if we want to put them in the same schema or split them

ding

I'll revise these tables + the documentation that I started writing so that we can talk through them next week in person

sounds good

what else should we cover?

I think I'm good

me too.

lets call it done then.

I'm looking forward to the mini sprint. it will feel good to get some team hacking in.

and to highlight these new features.

did y'all see kepstin's message from last week? (or earlier this week?)

!m LB team

You're doing good work, LB team!

yeah, that was nice

thanks kepstin

zas: hi, can we remove the zombie processes check in wolf? This is something that we're unable to fix when we use it for BB development