#metabrainz

anshuman73: thanks :)

The thing is LB-114 is currently a GCI task. I'm not sure if you can work on it as some student from gci might want to take that task up. However I'd leave the decision to armalcolite and Freso,

LB-114: Create a Profile Page with editable features https://tickets.metabrainz.org/browse/LB-114

zas, bitmap: did either one of you *create* version 1.3.2 for picard in jira? I think 1.4 should have been renamed (or just kept as the next version number) because we already have a lot of tickets with fix version 1.4

ferbncode: Although you can pretty much take it up once GCI is over and no one else has worked on it

zas, bitmap: scratch that, there are just a few tickets with the wrong fix version, I'll move them to 1.4...

ferbncode: P.S, I'm a mentor around here. Just a student myself :) You can pretty much ask your queries in general here. You'll probably get better responses

I'm not *

ferbncode: ^

stupid question: how do I run tests for listenbrainz?

anshuman73, ferbncode: It does not matter if the task is for GCI or not. If someone wants to contribute, then they are welcome. anshuman73, the GCI tag for the task was added probably because some student might want to do it during GCI. (also you showed some interest during GCI)

armalcolite: I still have that interest ;) I just wanted to point the query to you, as I don't know what one should do in this case :)

anshuman73: Its fine. If you want to do it you can ask someone to not take it.

with the ~12 days left of GCI, GCI students should have the priority of issues that they'd like to work on before opening it up to others :)

armalcolite: I'd rather not do that from what I've learnt in the past few days.

:P

ferbncode: You'll find a lot of tasks to work on at https://tickets.metabrainz.org

ferbncode: you are even free to suggest a issue/feature and work on it.

gcilou: yes, absolutely true. But i am suspecting no one might take it. Also since it is a design task, it probably won't be a good idea to take it with only ~12 days left.

armalcolite: true. But there is still the possibility of any student claiming it and working on it, so it'd be best to wait the 12 days before someone else trying to work on it just to be safe. If we think it's too difficult to finish as a gci task by the end of GCI, we could ask Freso to remove the task and remove the GCI label from the issue

gcilou: looks like a good idea. :D

Mineo: i'm prolly responsible of wrong fix version, wasn't certain which one to use... 1.4 is the correct one, right?

yes it is

it's also the one listed under "Unreleases versions" in jira

(which is how I remember how to chose the correct one)

Ok, sorry for the hassle, it wasn't clear for me

no problem :)

Mineo: do you by any chance still have GSoC proposal(s) related to the new search somewhere?

trying to get familiar with the whole thing

and what issues are still there

gcilou, anshuman73, armalcolite, ferbncode: FWIW, I'm with armalcolite: If someone wants to work on a ticket, they should just work on it. We can remove it from GCI no problem. GCI is also about learning open source, and sometimes someone will have tackled your pet bug/issue/feat. req. before you get around to it.

Freso: ahh, that sounds more convincing.

ferbncode: if you're working on, just make sure you assign the ticket to yourself and inform Freso so he can remove the task from gci I guess

No need to inform me before the ticket has actually been done.

Plenty of people assign themselves tickets and never actually do them...

Not that I'd ever do such a thing...

Freso: that sounds good to me

can someone have a look at my comment added on LB-114 , I'm pretty sure my "noobness" can be washed out of it and the ticket can be better explained. (or suggest me how to do so)

LB-114: Create a Profile Page with editable features https://tickets.metabrainz.org/browse/LB-114

zas Mineo updated comments on PR-525

MEB-64: Make MetaBrainz.org translatable https://tickets.metabrainz.org/browse/MEB-64

I'd say this is out of scope of this ticket

hmm, where will I get my translated strings from then ?

"This task is about making the source code for https://metabrainz.org/ translatable using Python's gettext implementation and integrating that with our translation platform."

as stated by https://codein.withgoogle.com/dashboard/task-in...

translation platform linking to transifex

well, only see what the ticket says

alright.

ruaok: I don't even know what a W-9 is so I guess that one's for you :)

you can assume that the transifex project is there

Gentlecat: Is there a way to run critiquebrainz server without using docker? If I run 'python manage.py runserver' I get redis.exceptions.ConnectionError: Error -2 connecting to critiquebrainz_redis:6379. Name or service not known.

anshuman73, Freso: got it. thanks :)

tigranl: there is, but it will be more complicated

in this case you need to also run a redis server locally and specify it in the custom config file

there might be other obstacles

[picard] samj1912 opened pull request #527: PICARD-864: Minor fixes to #525 (master...picard864) https://github.com/metabrainz/picard/pull/527

Gentlecat: Also, what should I do with less files? Because I don't see working styles on pages.

are you running CB with docker compose?

reosarevok: yep, I have a w9 at the ready

after you start the development version using the regular command, connect to the container: `docker-compose -f docker/docker-compose.dev.yml run critiquebrainz /bin/bash`

then `npm install` and `./node_modules/.bin/gulp`

zas: didnt get your latest comment on the PR-525?

Mineo: zas how should I proceed with PR-517? Should I move the ignore option to general page?

Yep, thanks

this is not very well documented currently

needs some improvement

Guys, I just had a question about musicbrainz that I have no idea how to answer: What's the recommended way of running the server in production? Is it via starman, plackup or starlet??

It seems that the old server user starman, documentation says plackup, and musicbrainz-docker uses starlet.

UmkaDK_: we use starlet in production, but plackup for development (it's simpler and already comes with Plack)

starman or any other PSGI compatible server would work just as well

but I would recommend starlet only because that's what we use :)

:)

Are there any disadvantages to using plackup in prod?

(Just in case I'm feeling lazy and don't want to update the start up scripts) :)

well, I think it's single-process, whereas starlet supports many workers. this might not have any consequence if traffic isn't high

I see ... never looked at starlet properly... so, is it just an upgrade on starman?

it's supposed to be a bit faster, but I think we only switched because it worked better with Server::Starter

which allows you to SIGHUP the server process and restart the workers without downtime

Cool! Thanks bitmap!!

nice, four peers downloading the mbs vm image from me atm

bitmap: zas added tests for id3 deletion

updated PR-523

excellent

samj1912: thanks, great, about PR-517 i think it is ok in the Fingerprints page, as it makes only sense when fingerprints are enabled (or i missed something?)

could we have a test for performer: too?

yeah, I thought so too, that's why I placed it in fingerprints page, I had earlier placed in general page along with the others

bitmap: tested that in general for all formats

where?

ah, it was added already

thanks

yeah, the reason I didn't catch the performer bug in id3 when I wrote that test earlier is because the test_format file uses id3v2.4

ill add 2.3 tests once I figure out all the problems with PR-522 and PICARD-748

PICARD-748: Picard repeats/duplicates field data https://tickets.metabrainz.org/browse/PICARD-748

https://github.com/metabrainz/picard/pull/525#d... raises something, should we continue to support python 2.6, it seems to me 2.7 is out since ages now, and many features are lacking from 2.6 compared to 2.7. Since we are about to move to py3 (for 2.0) not much an issue, but i wonder if Picard is really tested against 2.6.

we were using only 2.7 in our earlier travis build since ages so I think it seems logical to phase out 2.6

zas: I thought 2.7 was already minimum version.

no issue dropping 2.6 for me

(another) dumb AB question: with the cli, if I try to scan a file w/o/MBID, I get `[:( nombid ]`...

... but if I retry the file, I get `[:) done ]` , presumably because it passes the "already examined this file" test.

CallerNo6: Yep.

That seems misleading to me. Worth a ticket?

CallerNo6: Gopher it!

Maybe there's not really a better way to do it.

CallerNo6, I think "nombid" is actually "nom bid". Food auction.

oh, no wonder I was confused!

Until then, I usually do `sqlite3 ~/.abzsubmit/filelog.sqlite` and then run `DELETE FROM filelog WHERE reason NOT NULL;`.

Freso: that's my point, we are supposed to be 2.6 compatible, restricting the use of 2.7 functions, but at the same time, no test or build is made against 2.6 (which means it is unlikely to be 2.6 compatible in practice). I guess dropping 2.6 is a non-issue. We just need to update documents.

zas: My point was that I'd be fine with not support Py2.6 for the 1.4 release. :)

*supporting

Freso, cool trick, thanks

we last bumped the version in 2012 when 2.7 was already ~2 years old https://github.com/metabrainz/picard/commit/f3a...

Mineo: ^^

(I'm also for dropping Py2.x entirely, if we do a Py3+PyQt5 bump after the Picard 1.4 release.)

2012... omg. ._.

I think it's due time to let Py2.6 R.I.P.

yeah :P

Freso: we'll do, this release will be the last py2/qt4 (with perhaps only fix releases afterwards)

🙌

and I suppose we'll be doing minimal maintenance on the 1.x branch and don't want people filing bugs for 2.6

Mhm. That is a good point too.

👍

bitmap: exactly... but i'm not aware of any 2.6 only issue in JIRA for Picard anyway

CallerNo6: I really like SQLite. :)

zas: Some Debian user will come along in 5 years when they finally get Python 2.6 in stable and come upon some obscure bug...

>_>