#metabrainz

mo’’in’

moin

aidanlw17: that's not fair. brb, moving back to Quebec

Mornin' !

I am giving a talk in my school for GCI on 9th.

If anyone wants to share any resource/link, I will be more than happy.

mo'in

thats great pristine__.

:)

Give more highlight towards open source and then mention GCI.

Thanks

pristine__: also make sure to remind them that it's not all about code, and that if they're into design, or docs writing, or even outreach and talking to people, all of those things are also covered (a lot of open source projects have a lot of good coders but nobody who can design to save their life, I feel, for example :p )

Yes. True

oh yes! we'd love that

design, doc writing and such are very wanted tasks/people

I trained a model on 7 days data, generated candidate sets of 50 tracks, these two are the constants.

Using these I generated recommendations.

Which are same on each run (deterministic)

Whilst writing tests we just need to keep in mind that these tracks muct appear in training set. Tracks which are not in training set but are in candidate set will not be given a score and simply discarded.

due to obvious reasons, if the algorithm has not seen a track before, Or not seen a track whilst training data, it cannot predict a score for it.

bitmap: reosarevok I'll facing problem with `Unicode::ICU::Collator` when setting up mb-server in a linux machine. I've using Pop OS 19.04 and I also have libicu-dev installed using apt. Tried reinstalling too but it didnt make a difference

*I'm

Cyna: what's the error?

```Running Makefile.PL

OS unsupported: No icu-config --cppflags found

-> N/A

-> FAIL Configure failed for Unicode-ICU-Collator-0.003. See /home/shubhamjain/.cpanm/work/1570458021.28905/build.log for details.```

There are the lines where error was found

Freso: I slipped on the MeB Summit draft for Sunday, but I have an outline started in the MeB blog if you want to take a look.

pristine__++

pristine__: Good luck with the presentation!

yvanzo but to mock wont i need some data ? i figured i should just write a test that imports from the updated mbrng lib instead of the mbrng from pip.

reosarevok: +1 to FOSS/GCI being more than only code :)

jwf: <3

Cyna: did you check if icu-config --cppflags works?

Pac23: see hard-coded examples in https://github.com/metabrainz/sir/blob/master/t...

Pac23: if you feel it would be simpler to do another way, please do so!

zas: have you looked at notarization for picard for osx catalina? https://developer.apple.com/news/?id=09032019a

nope, and I don't manage mac packages at all, outsidecontext does though ^^

alastairp: no, wasn't aware of this. I think we are still happy that we have managed to get signed builds on mac and now Apple comes with the next BS, oh well

I hope this can be automated on our build servers

yvanzo i was bit confused the other day when you said about the tests,i dont see any other way that would be easier than what you said though.

Not sure whether I should have reported it here or on the #musicbrainz channel (where I did). It appears there wasn't a replication packet produces for 1500 UTC. I've just downloaded the 1600 UTC one.

No real drama for me, just reporting in case it points to a potentially wider issue.

bitmap ^

yvanzo: What's the replication packet produce?

This is what I got in the logs: http://paste.debian.net/1105171/

pristine__: changes from the main DB to be replicated by mirrors, see https://musicbrainz.org/doc/Replication

adhawkins: thanks for reporting; looks like our hourly cron script did die before producing that packet

No worries bitmap, happy to help.

I assume the next one would have picked up those change too?

yes, no data appears to have been lost, since it didn't clear the dbmirror pending tables

Cool.

still concerning though. I'll have to review these scripts to make sure they're robust in the event of DB connection failures

unclear why we hit our max connections limit on bowie

Ok.

yvanzo: Thanks :)

bitmap: nope it doesnt work 🤔

error?

icu-config not found

```icu-config: command not found

```

oh great https://bugs.debian.org/cgi-bin/bugreport.cgi?b...

fixing all the modules that depend on that is going to be fun :|

aahhh !

<BANG>

It’s World Habitat Monday!

People up for reviews: reosarevok, iliekcomputers, Freso, pristine__, yvanzo, zas, ruaok, bitmap, Mr_Monkey, CatQuest, aidanlw17 – anyone else, please let me know (explicitly) ASAP.

reosarevok: Go!

Boing !

Hi all

Freso: I am up for a review

Yay!

hi

boing! :D

No reosarevok?

:(

iliekcomputers: Ping?

reosarevoked

reosarevokky

Well, pristine__ then! Go!

Aww

:(

hi pristine__ !

Thanks :)

Hi all <3

Was a good weekend. Reached India on 2nd

Worked on improving develop.sh of LB

Worked on a few open PRs and merged a few

And testing CF algo is it is deterministic or not

It will be but just in case.

Read a lil spark stuff.

fin!

Umm....

Freso: go!

🙋

Now yes reosarevok

reosarevok:

(was cycling home, sorry)

reosarev-OK

:)

HAH

During the last two weeks I’ve been around and about. I was also at some summit last weekend and handling the live streaming there. Curious bunch of people there, but generally great craic going around. :)

Now I’m back home and exams are starting to show on the near horizon and I’ve also started to get ready to think about GCI and preparing to go to GSoC mentor summit next week…

I think that’s all I have to say for now. Mostly business as usual.

fin!

reosarevok: Go frealz.

:D

threso

(People still up: iliekcomputers, yvanzo, zas, ruaok, bitmap, Mr_Monkey, CatQuest, aidanlw17, amCap1712 – anyone else, please let me know (explicitly) ASAP.)