#metabrainz

If you want to do what we do, you should release every 2 weeks or 1 mo even if it's just one small patch :)

1 mo is what I would like.

(tiny improvements > no improvements anyway!)

reosarevok: how do you deal with incomplete stuff? Just not merge it?

Like stuff thats basically done but has small things that need to be fixed

or "get this merged, so this PR can go away. smaller fixes PR coming very soon."

Usually not merge it until it's fixed, but depends on what "fixed" means

Yeah, I think that makes sense to me.

Is it "this is an improvement but we could improve it even more with X"? Merge, send second PR

hmmm. maybe we should approve PRs and then make a new branch off the approved branch for the fixes PR

Is it "this is kind of a mess but we can improve it"? Then fix before merge :p

but not merge into master until follow up PR is complete.

Let's merge stuff to master only when we're completely ready to release it next cycle.

which is what I am trying with the lb-labs tests with pristine__ right now

agreed.

And aim for monthly releases?

oh, that brings about a natural next question: what is the least amount of stuff we need to do in order do do a release?

that should inform the goal of our next release.

I think having a schedule and not caring about the amount of work done is a better way to go here.

I dont think the two are mutually exclusive.

Keeping goals independent of releases.

my point is more "lets not start more stuff, but push a release soon"

That I agree with

that was my main point there.

Hmm.

but lets just agree on the "if it aint release for release, do not merge to master" right now, shall we?

+1

then we work towards the first release with the goal of soon thereafter doing monthly releases.

Yep, sounds like a plan.

which also means that 1 release = 2 work periods.

nice cadence, I think.

So right now, I have two things incomplete.

Stats and Incremental dumps

Incremental dumps is a cron job that we can just stop until we get around to fixing it.

So it isn't exactly blocking

what is the problem with the incremental dumps?

how broken are they?

There were some listens missing/more in a dump.

yes. anything else?

That's it.

I mean, this is a needle in a hackstack problem, but just tyring to get the scope.

Yeah

I spent a few days looking at it

But didn't really find anything

let take another evening and have you talk me through the process.

Plus it's a PITA to debug because it needs actual dumps which are created really slowly.

and I'll look at code as we go. I think a second set of eyes might make this a lot easier.

Yeah, having a new set of eyes on it would be good.

++

I hear that -- the MSID mapping has been frustrating because each full run takes 1-2 hours.

and there are a lot of little things to try -- at least now I can use partial data sets that run much faster.

ok, lets plan to do that walk through in the comings days.

user stats. lots of open items in the doc.

shall we make a Week of Dec 16 section and start looking for things to toss in?

Or put in trello?

I see bugs, completing features and needed features as three categories.

my brainpower is dwindling TBH -- that flight home was rough.

let's make a new section and then tomorrow I'll move them to trello.

sure, sounds good.

(50 screaming teenagers caused 20 babies to freak out, causing teenagers to scream louder because they couldn't hear each other)

I think I found a new worst case scenario for noise cancelling head-phones.

if you want, we can just stop here. the process discussion was really helpful and I could easily prioritize work for these 2 weeks myself.

noise cancelling headphones are ⭐⭐

that would be nice, but it feels incomplete to me. shall we do more tomorrow?

and pick a time that is earlier?

noise cancelling headphones were only so effective!

but what about noise as music genre?

hmm. sure! we can pick it up tomorrow. would 7-ish work?

might be closer to 8 for me, but round then, yeah.

that works better for me too, tbh. :D

thanks ruaok, the process discussion has been really helpful

for both of us!

oh, did a new volunteer pop in over the weekend?

if not, I'll follow up. did you see his email to support@?

Don't think I did

[musicbrainz-server] reosarevok opened pull request #1331 (master…MBS-9919-pt-2): MBS-9919: Convert the place map tab to React https://github.com/metabrainz/musicbrainz-serve...

MBS-9919: Convert the place public pages to React https://tickets.metabrainz.org/browse/MBS-9919

reosarevok: sure :)