#musicbrainz-devel

looks like java's already there

you could put the shell script in /usr/local/bin I guess. though ~/bin seems just as suitable.

which I should have known because people run search servers

bitmap: there are errors in ./relationship-editor/Relationship.js and a few more places.

* ianmcorvidae: MBS-5200: load language for error pages

* ianmcorvidae: MBS-5200: introduce with_translations, migrate finalize_error/dispatch to use it

* ianmcorvidae: MBS-5485: escape artist.comment in rdfa_macros

* ianmcorvidae: MBS-4180: show preview and AC-being-changed separately

meh, internet is so slow right now

unzip: not found :/

hah

one sec

should be installed now

heh, of course it was the trailing commas again

heh

I need to break that habit now :/

Anyone advise on git workflow regarding this new vagrant stuff of ocharles'?

It's a repo with lots of submodules.

git pull --recurse-submodules

Do I just clone the top repro? Will that effectively clone the submodules too?

Use --recurse-submodule.

Or -submodules

Or something like that.

I want to make some modifications with a view to submitting a pull request later.

For the clone command as well.

They're on github. Do I fork the repro there or what?

adhawkins: Just fork ocharles' repo.

And that'll fork the submodules too?

adhawkins: The submodules point outside his repo.

No.

The submodules aren't forked.

Well, one of them is his. It's likely I'll want to modify that too.

They're stored separately.

You can fork that then, but you'll have to edit the submodule then, to point to your fork.

Erk...not really looked at submodules to be honest.

What about if I create a branch in ocharles' repo? Does that branch submodules too?

adhawkins: :)

Try to fork and clone his repo first.

You'll then notice a .gitmodules in the root of the repository.

Yep, seen that.

This is what points to the remote git repositories.

Yep

Each submodule is then additionally stored as a file or something in the git repository, only containing the remote git commit hash id.

And that's it.

Ok, better do some reading then I think!

No code of the submodule is stored in the git repository.

All that is stored is the hash id and the repository url.

So the hash points at a specific commit? It doesn't track a particular branch or master?

Which, OTOH, is also all you should need to replicate the code. :)

yeah, it tracks a specific sha1

Yep. A specific commit.

which can cause problems sometimes, but generally means more control

So how are you supposed to update one of the submodules and have the 'parent' pick up the changes?

You have to edit the sha1 every time?

you cd into the directory and pull, generally, and then there's some sort of incantation you do in the parent repository to tell it to commit the changes

adhawkins: The SHA1 in the submodule "file" is updated.

You `git commit` that like any other file change.

(you don't manually edit the submodule file, though)

No. :)(

* :)

Oh ok, so when you pull the submodule, the submodule file updates its sha1 automatically?

And there's no incantation to do in the parent directory.

ianmcorvidae: ^

git picks up the changed sha1 itself.

okay

ianmcorvidae: the github site isn't loading for me right now but I pushed a commit to my beta branch that removes offending commas

adhawkins: Yep.

bitmap: deploying now

I think I need to do some reading :)

adhawkins: You still need to `git add`/`git commit` the updated submodule "file" though.

Ok

adhawkins: It's really not as complicated as it sounds.

adhawkins: Just try and clone the repos and play around with it a bit.

(Well, I don't think it is anyway. :D)

* ianmcorvidae: i18n: N_ functions didn't get re-added to the Makefile, evidently

* ianmcorvidae: Update .pot files

Freso: ianmcorvidae: cheers.

adhawkins: You're welcome. :)

ianmcorvidae: it works now :) thanks

cool

I'm working with a database dump of Musicbrainz data and am just wondering what the purpose of the release_coverart table is, as it is empty

it stores coverart URLs?

er

that wasn't supposed to have the ?

but anyway

0 of them? :)

yes, initially

it's updated by a script

and on production, that script uses some AWS keys that we can't make public

ah...

so yeah

Isn't licensing bullshit lovely :)

you can run the script without it, and it'll update URLs for things other than amazon coverart

or you can ignore it entirely and CAA coverart will still work

is that script included in the VM? i.e. if I had that and just substitued my own keys would it work?

it should be there somewhere, and yes

Freso: Should have known it'd be covered in Pro Git :)

yeah, it's in the codebase

yeah but CAA coverart doesn't seem to have too much info at this point sadly (at least based on the things I've searched)

adhawkins: :)

Bookle: admin/RebuildCoverArtUrls.pl

awesome, thanks ianmcorvidae and nikki :)

adhawkins: I mostly just read the git man pages to figure things out. :p

so without amazon keys you're not going to get much no matter what :)

adhawkins: I find that they're surprisingly well written, for being man pages.

Freso: Yeah, I generally use them to refer back to when I don't quite know how to do something. I used Pro Git to get the basic concepts down, so expect it'll cover submodules well too

:)

I should refer CallerNo6 to Pro Git, actually.

But he's logged off again. Ah well.

Bookle: working on it ;) But yes

Not yet :)

yup, I'm sure someday it will be populated enough to use, the problem is I'm working on something that needs it today ;)

you could use the asins to guess the url

another thing that I noticed that will be very cool down the road is the Bookbrainz stuff... definitely interested in that :)

in the same way the picard cover art stuff does

(but I said "guess" because amazon doesn't always store the image in predictable locations)

nikki: yeah, I figured I would have to resort to playing around with Amazon urls in order to grab the cover art

In the same vein as Bookbrainz... is anyone here interested in doing a Moviebrainz?

adhawkins: in case you didn't see: http://beta.musicbrainz.org/ws/2/label/c80d3a32...

Yep, all done at my end too now I think.

Cheers

Bookle: As another half-formed media database? I hope we would get bookbrainz usable before moving on to Moviebrainz.

there's no reason why someone couldn't start moviebrainz if that's what they're interested in

hawke_1: I didn't mean at the same time, just the general idea

Bookle: Then, yes — there is interest in having such a thing.

Or maybe “VideoBrainz”

reosarevok would just say everythingbrainz :P

(since it would presumably cover TV shows and such as well)

I would!

I'd say AllBrainz but don't want to get sued by Kellogs

AllBrainz: The desire of zombies everywhere.

Bookle: skd5aner and me are tooting the tvbrainz horn every now and then over at mythtv's

Project musicbrainz-server_beta build #207: SUCCESS in 24 min: http://ci.musicbrainz.org/job/musicbrainz-serve...

mwtuea: Silence complaints about trailing commas

dekarl: tvbrainz?

hawke_1: moviebrainz, tvbrainz, <thing that gets aired on a tv/radio station>brainz

(I mean, how can you toot the horn of something that doesn’t exist yet?)

Probably in an attempt to get someone to change that? ;)

usualy its ranting about how bad the metadata sites in that domain are and that we really need someone to extend musicbrainz towards it ;)

dekarl, wait, wait, you don't like themoviedb and the like? :p

(and their nice extension towards music! :p)

I mash up themoviedb, thetvdb and tv/radio stations data to produce a guide feed and its not fun to work around all their limitations. (these being the best sites I could find)

reosarevok: well, its the best (as in amount of data, data model, license and editing process) site around. But its still so super primitive

dekarl: they thought MB was too complicated, so they started their own music site

(that uses MB data but keeps it "simple")

(well, it uses "their own" data, but imported from MB)