#metabrainz

okay :)

ruaok: i sent a support request about solr-1 down time

k

kartikeyaSh: yes, I think that is a safe assumption to make. otherwise we would never make any associations at all.

as you can see, we will need to make our tools a lot smarter as time goes by.

because data changes, we will need to adapt and create new clusters and collapse others.

Agreed otherwise we can't really cluster any releases. for now, our tools can first truncate cluster tables and create clusters maybe every month to tackle such a problem

possibly. that is a lot of computing work to do.

more incremental changes might be better, but that is still hard to tell.

we just need to go step by step for now.

yes, lets not worry about that right now

Woo!

NPR, the only quality news network in the US just signed up with us. :)

👏🏻

they are non-profit, so no money, but still.

And a japanese company signed up -- a first for us.

\o/

ruaok: please review this one. https://github.com/metabrainz/messybrainz-serve... I'll need this to create clusters for fetched releases

ok, I need to sign one contract and then I can do that.

zas: https://github.com/lcech/tumblr-checker-bookmar...

\o/ for NPR and the japanese company (what's its name?)

ZaphodBeeblebrox: yes, but still i'll boycott

hmm :)

ruaok: see hetzner answer, network issue leads to server down+reboot ?

I call BS.

agreed.

if that in fact is the case, then they have SERIOUS engineering issues that they need to address.

3 questions to support, 3 dumb (no-)answers

yep.

well said. typical corporate speak.

i guess we have what we pay for... ;)

yeah.

I'm not impressed.

the google stuff never behaves like that

https://www.reddit.com/r/PleX/comments/8xdxcq/w... :)

"I speak no hyperbole when I say that it is absolutely essential to the existence of my current music library." <-- ;)

this looks quite interesting: https://cloudplatform.googleblog.com/2018/07/in...

LordSputnik: https://travis-ci.org/bookbrainz/bookbrainz-dat...

I think tables are getting truncated but does not reset the primary key sequence

*it

turn on the sound after it starts playing: https://twitter.com/Soypinchewey/status/1014614... !! :)

This sounds scary though => `TRUNCATE musicbrainz.language CASCADE - deadlock detected`

zas: Can we setup a keras container for Leo_Verto?

bitmap: LordSputnik: iliekcomputers: yvanzo: Leo_Verto: Please make sure to get the GSoC evals done soon. 4 out of 5 students are done and only 1 mentor is done. 🙄

:)

was just doing it!

great

the fail all button has gone missing again. :)

that's it. I'm opening a ticket with google.

yvanzo: starting a new VM build and heading home.

wether we need it or not. :)

cool

I did test the previous VM (but reindex/replicate) and did not find any issue.

reindex is taking too much time here :(

also, I started coding a less hacky util for turning ports on/off

samj1912: I think you're getting trolled in https://community.metabrainz.org/t/search-resul... where the comment is "Yep, it seems the trouble is he is mostly using solr defaults and hasnt actually bothered to create set of tests to check the results against." I seem to recall you guys doing tons of testing and tweaking to get the best results.

[design-system] mwiencek pushed 3 new commits to master: https://git.io/fNUpi

design-system/master dfe7835 chhavi: Add dropdown button component

design-system/master edfa539 chhavi: Add story of dropdown button

design-system/master 8c0fb6e Michael Wiencek: Merge pull request #11 from chhavijustme/dropdown...

Ah I have stopped giving fucks to that particular person.

They are free to make PRs to improve things.

I've had it with his shitty attitude.

ahhh, now to balance the negative with some positive:

> Thank you again, from all of us at NPR!

> PS: Love the “Excel is not a database!” signature! : )

hehe sound like nice folks

samj1912: just ignore him -- he's been acerbic like this for years. in person too.

bitmap: yeah. I thanked them pretty hard for NPR.

LordSputnik: Also, max-lines-per-function rule might raise warning in a couple of existing functions, like updateAliasSet.

NPR! https://metabrainz.org/supporters/tiers/8

[design-system] mwiencek pushed 4 new commits to master: https://git.io/fNTfZ

design-system/master 60fa330 chhavi: Add progress bar component

design-system/master 5e48858 chhavi: add story of progress bar

design-system/master 9a9a664 chhavi: Add less for progress bar

lol NPR seems like kewl "our kidna" folks :D

you mentioned a japanese company. can you tell us which it is?

not yet.

:)

ruaok: neat

ruaok: is it okay? displaying MBIDs kind of seems necessary here https://www.irccloud.com/pastebin/jJGoTI66/

seems ok. I think we can fine tune them once we start fine tuning some of these algorithms...

okay

bukwurm: ping

LordSputnik: pong

mame/quine-relay: An uroboros program with 100+ programming languages (https://github.com/mame/quine-relay)

Super cool stuff

bukwurm: hey, will be with you in a minute

We can maybe close a PR or two tonight

LordSputnik: Cool

ruaok: will aim to do the eval tonight or tomorrow - please ping me if you see it isn't done by Friday morning

bukwurm: ok, so how have things gone yesterday and today?

Ok, so bb-data work is complete. Finishing up the PR right now.

Great :)

Upgrading and deletion part requires work, but importing should work now.

Mapping part is done, but I am unhappy with the code structure. Will try to refactor it up and send in a PR tomorrow.

BB Data PR is up, so what ever has been done can be reviewed.

I have worked on creating views and ORM object for the imports, but need to test all of them yet.

For that I've changed the `entityType_import` to `entityType_import_header`

And gave the view the name `entityType_import`

Seemed to be more consistent with the `entity_header` and `entityType` view

OK, that sounds smart

*`entityType_header`

For the data PR, am I looking at 112 or 113?

*122 or 123

Right, so 123 is what I'm working on

OK, which is ready for review?

122 was supposed to be alternative solution which I'll be discontinuing I guess

OK

So I should close 122?

Yeah, I suppose

It had mostly code from bb-site

OK, how about on -import?

Should I be looking at #9 or #13 or both?

#13 first

OK, thanks