#metabrainz

woa

apart that, overheating here ;)

fin. yvanzo ?

hey!

hia!

Last hot week I fixed some other issues in MusicBrainz VM and rebuilt images (already available on FTP, to be announced soon).

I tried to improve ISNI support for both artist and label in SOLR.

I updated beta.musicbrainz.org and fixed a few small issues.

Along with usual stuff, that’s all from me. Leo_Verto?

Thank you!

I've spent the last week suffering without non-mobile internet, preparing datasets and building models for SpamBrainz, currently trying to deal with a model that gets up to 90% accuracy and then finds some stupid way to optimize the loss function while ruining the accuracy to guessing levels. Yay for machine learning black magic!

Other than that it's just been unbearably hot here.

fin.

I feel for you guys. yo ugot the heat that was here last week

bukwurm, go!

We just try to do cool things when it gets hot.

hahaha

yvanzo: which "server"

go find a river, that worked for me

Hi everyone

hi bukwurm !

(People still up: Freso, chhavi_, kartikeyaSh – anyone else, please let me know ASAP.)

This week I wrapped up discard import entity and approve entity features

This took up more than expected time, as I had to build most of infrastructure for importEntities in parallel as of the regular entities

This required refactoring quite a bit of present code

and adding a lot of new features. I was saved with the existing code for a number of things.

*by the

The past three months have been a great learning experience for me.

I will try my best to wrap up edit and approve feature this week

Along with reviews and documentation for the existing work.

That's it for me for this week!

kartikeyaSh: ?

ruaok: "mayhemchaos" has 761M free over 416G/439G for /home

kartikeyaSh: Ping?

chhavi_: Go in the meantime?

hmm

Or maybe I'll go and we'll see if chhavi_ or kartikeyaSh shows their head. :D

Or LordSputnik !

Go!

chhavi_ first

hi chhavi_ !

Worked on a PR for table component and opened a couple more documenting my existing work

Apart from that just tidying up for the last week and started working on my blog

Fin

LordSputnik go

This week I worked on improving our entity handling code, and getting rid of a load of Promise chains and 300-line functions. As part of that, I also added in relationship editing, and have worked out how to do entity merging and reverting merge revisions 😃

\o/

I also met with bukwurm a few times, and reviewed some of his PRs. Finally talked with Monkey about some BookBrainz getting started tasks, and he's having a go at converting our JavaScript build system over to webpack

So far Monkey is looking well suited to the new BookBrainz role, and I'd like to work out next steps with ruaok and him

Nice.

That's it from me

hmm sounds good

Is Monkey_ still in Barcelona? Will he be at the summit?

Still no word from kartikeyaSh, so I guess I'll go.

🙋

:D

hah there kartikeyaSh is

I mostly had a spontaneous vacation kind of week this week, so other than a very minimum of IRC and forums, I've mostly just been keeping an eye on my e-mail to see if anything urgent came up.

So fin from me.

kartikeyaSh / kartikeyaSh__: Around?

I guess not.

So that's everyone for reviews (thanks everyone! theveryone!) and we have no other items on the agenda.

So thanks for your time! :)

thanks freso!

</BANG>

reosarevok, Freso some time this week i'd liek to ave the talk with you about the drums?

ruaok, iliekcomputers: there are especially large (196G) acousticbrainz dumps (28072018) on FTP.

CatQuest: 👍

:D

Thanks Freso

These files are hidden from FTP index as their names start with a dot.

yes, I think this might be alastairp doing sneaky backups.

yvanzo: really?

alastairp: ping

I put one file on ftp this week

It looks like daily dumps indeed, but only one (on 1st Aug) of them takes 196G, others are empty (like failed dumps).

yvanzo: I'll check

let's make sure we have one valid dump.

and then nuke the others from the ftp server.

The one I linked is perfectly valid

we can find other machines with free disk space.

Frank ;)

yvanzo: feel free to delete older dumps that are not the one that iliekcomputers linked to.

iliekcomputers: perfect. please move another copy to frank then.

yvanzo: which folder are they in?

the one you mentionned

Is that temporary file? It doesn’t even have the same size as the dump on williams.

which server is that?

mayhemchaos

i don't understand

we should give that server a better name. I'm moving out of it and turning it over to MeB.

name checks out?

heeh

iliekcomputers: it is a VM that has free bandwidth that I used to have.

now hosts ftp.eu.metabrainz.org

oh

the dump on williams is valid.

how does stuff get to ftp.eu, some rsync cron job?

it might be sync failures causing those partial files

zas can explain the sync that is happening. I don't know the details.

iliekcomputers: indeed, the file name looks like a temporary one and the size is slightly smaller than the dump on williams.

it might just be a straight up mirror of our public mirror.

the exception is the -vm stuff which lives no places else.

speaking of vm stuff, is the new image final?

yvanzo: i think they can be safely deleted.

Leo_Verto: unless too many bugs are found.

iliekcomputers: I’m worrying this might happen again in 4 hours and 21 minutes.

okay, just ping me if you need help seeding the torrent

Thanks, I am currently testing it.

yvanzo: let's remove them and see if it happens? then we'll know if its sync problems or not.

i'm gonna add a md5 file to williams.

hi

kartikeyaSh: 🙋

I just saw your /query

If you give us your review, I'll be sure to include it in the notes.

I don't know what went wrong https://usercontent.irccloud-cdn.com/file/xtqx5...

kartikeyaSh: You weren't identified with Nickserv.

Freso: here it is "This week I created PR for LB-384, LB-387 and I'll create PRs for LB-385, LB-386 after the meeting (these are very small bugs). Apart from that, I worked on writing the blog post of my work at MessyBrainz I think I can get it done for a review by tomorrow around this time. I'm left with the integration tests for LB-384, LB-387 (most probably I'll get it done before the final submission) and working on the reviews of the open PR.

And the GSoC project will be 'kind of complete'. Unfortunately, I didn't get time to work on any additional ideas."

LB-384: MessyBrainz: Create a script to cluster incoming recordings. https://tickets.metabrainz.org/browse/LB-384

LB-387: MessyBrainz: Add API endpoints to fetch MSIDs or MBIDs. https://tickets.metabrainz.org/browse/LB-387

LB-385: MessyBrainz: Create indexes to speed up clustering https://tickets.metabrainz.org/browse/LB-385

LB-386: MessyBrainz: Check corner cases for empty MBID fields. https://tickets.metabrainz.org/browse/LB-386

I forgot they had added that restriction last week.

ruaok: i sent a message to you and i guess it never reached. https://usercontent.irccloud-cdn.com/file/NVFOE...