#musicbrainz-devel

alastairp: did you ever setup and use greylog?

yep

it's pretty good

I just saw that its made in Hamburg.

I should really try it. :)

ruaok: so. postgres 9.4

time to rip out ES ;)

really? that good?

well

no idea yet

I should try it

methinks that PG9.4 can't replace ES's search

well, I'm only using it for autocomplete atm, which I think pg does

so the only thing that was missing was indexing into the json

which works now

true that.

but I'm thinking about the flexibility of search that ES gives, which PG doesn't.

but that may not be so important to us.

not sure

well, sure

I'll take a look

Project critiquebrainz build #75: SUCCESS in 1 min 31 sec: http://ci.musicbrainz.org/job/critiquebrainz/75/

* roman: Added an option to use Sentry for logging.

Freso: thanks for the bug report. I hope this doesn't mean you're trying to submit with new versions of the extractor ;)

alastairp: http://acousticbrainz.org/9587391b-78bd-467c-8d... metadata->version - it's not like it's not sort-out-able. Anyway, I didn't know I wasn't "allowed" to submit using the -git version - it's not exactly neither news nor a secret that I am doing that, esp. considering that I'm the creator and maintainer of the -git package in the AUR:

One of the reasons I'm doing the -git package thing is exactly so that I can catch such bugs before they get released. I've caught and reported a number of e.g. Picard bugs as well due to using the -git package.

Picard devs: please check the 1.3.1 branch

It's. All. Wrong!! D:

(Not really.)

Anything in particular to look at/for?

if i forgot something

Freso: tbh I'm a little surprised with your comment. as a counter comment, we've been pretty open about needing to sort out versioning, and now with new changes going into essentia trunk that there's a possibility we could be changing output

as far as I'm aware, you're one of the few people not using our static builds, as we've been asking people to do since day 1

I guess our immediate todo now is to update the server to block unknown extractor hashes until we can come to an agreement on versioning

alastairp: Why? The hashes can be easily filtered out for any given data set as required as it is.

we're not really fans of bumping problems to our future selves. it's complicated enough as-is

this has already had hours of discussions gone into finding the best way of doing it. we have so many moving parts that we want to keep it as simple as possible

and having everyone reanalysing tracks every single time a commit is made to essentia is not the simplest way forward

I was just letting you know so you could hold off until we've come to some agreement on how to go forward with our versioning

Sure. I've already replaced e-a-git with e-a. I just don't get the fuss about it.

obviously you don't understand the problem as well as we do then

Exactly. So enlighten me. :)

I really can't be bothered to do that now

we can talk next year

huh. cranky alastair is cranky.

zas, i think https://github.com/musicbrainz/picard/pull/384 is missing

that's PICARD-663

fixed

Freso: how do you generate the changelog from JIRA for 1.3.1 ?

zas: Click on the "Changelog" link?

no 1.3.1 in Change Log yet

i found a way

zas: ...

Via http://tickets.musicbrainz.org/browse/PICARD/fi...

yes, found it

is there a way to remove http://tickets.musicbrainz.org/browse/PICARD-666 (my mess) ?

1.3.1 branch is ready, can someone build mac/win packages from it ?

Mineo_ JonnyJD_: Ping?

I'm about to AUR-package some beets plugins. Should I do beets-plugin or python2-beets-plugin?