#musicbrainz

aCiD2: no, I mean that mb_server has inherited a lot of stupid things over the time

aCiD2: I can think of 4 ways to convert an album to VA

The 2007 mixtape is, unfortunately, not as well documented with regards to artists as the 2008 was/is. :/

aCiD2: just implementing one that covers most cases (the release editor) would be my priority

Oh, I see what you mean

maybe I'll take a step away from the edit types tomorrow, and do more refactoring work

Refactoring == <3

also need to work out how to test this framework

monkeys

lots f them

80 hours a week

10 dorrar

Hey folks, I'm playing around with the xml feed, and despite having read TracksWithMultipleArtists, I just want to verify that a <track> will never have more than one <artist> .. yes?

for now that's correct

but never say never :)

<3 to whomever makes it possible for a track to have more than one artist! (So we can clear up in the "Foo & Bar" artists...)

That sounds like a hint towards future MB schema having more than one artists/track ;)

luks: on a more general level, I know SoC finished and you probably haven't followed much of my work... but do you think I'm still going in the right direction?

ie, copying and rewriting features from the MusicBrainz site into my site

Vlet: it would involve more than adding <artist> entries, so yes, you can be sure that with the current XML schema there will be only one

Thanks; for now, my parser will rely on that, but I think I may wipe my artist field and create a track_artists table for my application ;)

aCiD2: I don't agree with that completely

feel free to expand :)

aCiD2: as I said, there is a lot of things wrong, duplicated, etc.

i know

I've been changing some of that (like returning objects instead of ids and stuff)

so I don't see the point in implementing those

I mean more on the UI level

oh, at the user level, right

porting things at the code level is probably the best way to go

but I'd reduce the user side to the required minimum

mostly what I'm doing, but I do need a usable interface by january ;)

ok

which is exactly the reason why you shouldn't implement all that

yup

i think the only important ones I need now are "Edit All Tracks" and "Edit Release Events", then I'm pretty much done on this section

[17:10] Freso<3 to whomever makes it possible for a track to have more than one artist! (So we can clear up in the "Foo & Bar" artists...)

totally. dude

what type is http://www.discogs.com/release/358832? ep? album?

vinly 7"

zeno_: I'd say so yes

if the release title doesn't match the tracks it's usually ep or album

Wizzcat: which one to use though?

ep for sure

k, thx. also what if discogs lists the english translation (or phonetic spelling) and you have the orig language? use the origional?

always use what's on the sleeve

barcode == ASSIN

?

No.

http://en.wikipedia.org/wiki/Barcode vs http://en.wikipedia.org/wiki/Amazon_Standard_Id... :)

ah sorry

you can get barcodes from amazon, but it's somewhat more convoluted

ok one last q: what script is this? 森 オッサン チョイチョイ キリキリまい

Foriegn.

Kanji, right?

yep

yeah

so han (kanji) or kanja/kana?

kanji/kana

cool thx

jcazevedo has joined #MusicBrainz

what sort of stats?

you know, edits over a certain period, what kind of edits, etc.

just thinking about it while looking at cord, would save me having to randomly parse his edit log :p

plus it'd be fun for my own sake

^^remili1^^ is now known as ^^remilia^^

you'd require the moderations tables (which are large), but it'd be fairly easy to produce a report

one more thing I don't have any time to actually code, oh well :|

Stats are pretty.

we'll see if I ever manage to install mbz locally, tried it the other day but the instructinos were cryptic to say the least

perl deps all over the floor

inhouseuk: Iirc the moderation data isn't replecated, either.

Replicated.

Wizzcat: I managed it, can't be that hard.

Or you could download the vm. :)

Wizzcat: just don't install the website and it's easy :)

That's true.

FauxFaux: correct, they arn't replicated

so it's plain impossible? duh

or I guess I could make fake edits by myself :p

Well, you could rape bandwidth and re-import every night. :p

Or you could write the code and get it committed.

Wizzcat: you can still code against the slightly-out-of-date snapshot

oh you mean I can download it just not update?

yup

cause I don't care about that bit

FauxFaux: heh, the b/w raping briefly crossed my mind ;)

Well, you do, if you want current stats. I mean, you can't host it yourself and use serious languages.

I still think the moderation data should be split into smaller files... there's no point redownloading data from several years ago because it never changes

Sharding.

nikki: Heh, can you resume the download?

also, I don't think it's that hard to fake part of the moderation data by analyzing the replication packets. not votes and stuff, but what gets changed when at least

FauxFaux: no idea

True, LotR, I have http://faux.no-ip.biz/mbreports/neweditsof.php ¬_¬

FauxFaux: I doubt you can resume

With a true stream format you should be able to?

Well, you may need to drop the last meg or so to get the synch right.

FauxFaux: but it's a tarball with replication packet type files in it, no? the individual files grow, and you could resume that, but when they're in a tarball, not everything gets added to the end

Ah, I assumed it was just one big file single table.

the replication packets at least have two files that both have part of the data for each update

Bah.

FauxFaux: can you import the moderation tables seperately from the other tables?

I never understood the database import, I'm just so fucking happy when it's worked and I never want to see it again.

hehe

did you have that much trouble with it?

I think you can...

Yes, yes I did. :p

I believe the db snapshots are very similar to sql COPY output

I think I need to write to the authors of ExpressRip. They're sending a large number of useless requests to the freedb gateway

cmd=cddb+read+reggae+00000000&hello=anonymous+localhost+ExpressRip+1.53&proto=6

the 00000000 is the freedb id !

FauxFaux: you're weak

how is the traffic on the gateway these days?

I've grabbed the log files and I'm going to do some stats this week

sounds like quite the change from last time.

people have started to realise it is available

good

speaking of the gateway, what's with utf8 support? is that an eac issue?

yes

I'm forced to use freedb every time I rip something non-english, sux :|