#metabrainz

ruaok: which dump did you use?

[listenbrainz-labs] paramsingh opened pull request #53 (master…param/upgrade-python): Install Python 3.6 ourselves instead of using old jessie python https://github.com/metabrainz/listenbrainz-labs...

😂

iliekcomputers: https://www.nzherald.co.nz/world/news/article.c...

mooooin

if you're ever had the (dis)pleasure of interacting with RMS...

no idea who he is, and no idea what he said.. but great!?

yes, great.

well, that's quite a morning.

Good !

metabrainz/picard#4968 (master - 449b8f3 : Christian Clauss): The build passed.

Change view : https://github.com/metabrainz/picard/compare/24...

Build details : https://travis-ci.org/metabrainz/picard/builds/...

"I am resigning […} over a series of misunderstandings." Uh-huh…

he's not going to learn his lessons, nor go quietly. but he'll be invited to talk a lot less...

ruaok: wanna apply for president? :p

You know, for the lulz

me? I'd last all of about 15 seconds!

I mean, you can certainly take more than 15 seconds to defend underage prostitution, right

So, better than RMS?

Although you'd probably turn the FSF into the Flying Spaghetti Foundation, so maybe not

oh, now you're talking.

It's sad to see that RMS doesn't even attempt a critical look at himself but even when resigning just blames others

seems to be a trend these days.

iliekcomputers: listenbrainz-listens-dump-82-20190912-000002-full

CatQuest: pretty big name in nerdy-old-dude computer circles

And kind of a dick, like most nerdy old dudes in computer circles

(not sure why tbh but I struggle to find of anyone old and famous who isn't also kind of a dick, other than hopefully Brewster Kahle, who I hope hasn't said anything weird recently now :p )

CatQuest: https://en.wikipedia.org/wiki/Richard_Stallman for more info anyway

> 10:01 PM <iliekcomputers> Yeah, our python image doesn't support that for some reason

the changes I made to the python images earlier in the year _should_ include all optional dependencies

does it include a python module for xz? (are we still talking about xz?)

as I understand, if the xz libs are installed when you compile python, then tarfile will read xz files (same with gzip, bzip2)

actually, to be fair I'm not sure now what xz-utils is used for, because the python archive that we download is in xz. does this package have just the extractor or the build libs too?

> liblzma-dev

that one. perfect

Thanks for the ticket, alastairp :)

de nada

as I commented, if people use the output of the api correctly, it _should_ be backwards compatible, but I don't trust people :)

If they don't check the value, that's too bad :p

alastairp: heh, so JSON already sets forward anyway

lol, interesting

yeah, we deal with xml in the python bindings

which is why I noticed it

Guess because otherwise it would have an empty direction, while xml just would have no element

yeah, makes sense. xml does that "no element" thing better

mm, does that mean that the schema says that this element is optional? I wonder if it should also be changed to "required" if you're always going to add it?

It should, probably

I'll check

So this patch will be like five lines of code changes and a hundred of test changes, I assume :p

as it should be

Hmm

(it is optional, yes)

lolol

I'm definitely looking at an api output which doesn't show forward

reosarevok: I was actually readin that article and didn't understand why the "old nerd dick" thing came in

.. then i read the talk page and boy

I don't think we have "both" anywhere

alastairp: sure, it's not shown at all. It's weird that it defaults to both, because we never use both :p

(I guess that'd apply for what, sibling?)

alastairp: lol. So, we have Readonly our $DIRECTION_FORWARD => 1; Readonly our $DIRECTION_BACKWARD => 2;

Where the fuck did "both" in MMD come from then

down the rabbit hole he goes

It's been there from the beginning, it seems

Yup, literally there in musicbrainz_mmd-1.0.rng

The fuck :D

bitmap, yvanzo: I assume I should be removing "<value>both</value>" from https://github.com/metabrainz/mmd-schema/blob/f...

Should I also make https://github.com/metabrainz/mmd-schema/blob/f... non-optional or is it better to keep it optional for some sort of backwards compatibility even if we expect to output it everywhere after https://tickets.metabrainz.org/browse/MBS-10370?

MBS-10370: Add `direction: forward` to all forward relationships in WS

how does versioning of that document work?

is it possible that people have stored xml of the output of the ws pointing to a schema file which might change

so then if they perform validation, the file becomes "invalid"?

I guess keeping it optional is OK. perhaps a comment saying that it's left as optional for backwards compatibility but is now always output/

?

AFAICT we just update 2.0 nowadays

no, not the postgres autovacuum

oh man. a roomba is like an autovacuum

put a pg sticker on it

Nerds

if I had one, I'd do it. only metabrainz stickers so far.

and bookbrainz. was to hand.

he found the mural supplies though. that was a close call.

meh, not going to be a fun productive day today. nothing but english hacking, no time for code, methinks.

[musicbrainz-server] reosarevok opened pull request #1212 (master…MBS-10370): MBS-10370: Add "direction: forward" to XML WS rels https://github.com/metabrainz/musicbrainz-serve...

MBS-10370: Add `direction: forward` to all forward relationships in WS https://tickets.metabrainz.org/browse/MBS-10370