#musicbrainz

k

will do.

ok, i really need to go to bed.

ruaok: good luck! :D

thanks and sleep well!

latersz folks

There's not any client-side handling of multiple matches of stuff, which means just adding it to the web-service won't be easy.

Oh, except the cdids..

svn diff > patchname.patch to create patches, right?

Yeah. Unless mb has a specific format they're wanted in.

svn diff is a good default

ok

should I mark it as fixed when I attach the patch? (it does fix the problem)

Faux: It'd be a pain, but it parses the release, then each track - you could theoretically output the release id on each new release, then times to a file as it hits each, then have someone re-load the same files, and import the times from that temp file on reloaded matching release id

Would (either of) you two consider a patch adding http://faux.no-ip.biz/mbreports/cueread2.php? It's mostly sql this time. :)

BrianFreud: I don't even follow, never mind. :P

lol

a patch in what way, FauxFaux ?

same thing I said when I last looked at handling releases, instead of tracks, in picard

To get mb_server to do it via web-service or something, acid2_.

I don't under stand, how is it doing it now then?

and what needs to be patched?

PHP! :D

Oh, you mean actually add cueread INTO musicbrainz?

Yeah.

I don't see why not, dunno where you'd add it though

try to match unknown releases by overall times list, rather than just puid

I'd add it to the release-search web-service, and maybe the "other searches" page.

(probably)

FauxFaux: Give me the code then :)

Might as well have a play around in Perl :)

luks: Does PicardQT have any support for global variables, addressable from say taggerscript?

acid2_ is now known as aCiD2

There's really nothing more than a perl regex and some string concatenation, http://faux.no-ip.biz/svn/mbreports/cueread2.php. I'll have to go dig out the sql. :)

BrianFreud: no

Oh hi luks.

Cool, looking now FauxFaux

I'm imagining creating a matrix on file load, [releaseid, tracknumber, tracktime]. If I ever can get the tracktimes thing patched for the server, it could be used for then submitted a release's worth of times, but also would be good for FauxFaux's lookup function

Yeah!

ohhh a patch.

haha

thank you rob :P

hmmmm

:-)

I deleted one line, not really hug worthy lol

FauxFaux: Actually, you may be able to do it anyhow

(but learn more about how the web service works, which is a good thing :)

I guess it'd be piss-easy in the actual picard code, if I could get the damn thing to build. :P

hah

FauxFaux: That script has millisecond precision?

the add cluster as release code already pulls all the info from a cluster - you'd just want to do something similar, just pulling the track times and track numbers, then submitting it to a lookup...

Cue files have the microseconds in, and musicbrainz is stored to millisecond, aCiD2. Version 1 of that was trying to guess the discids, whcih is a lot faster without the CACHE. :p

:)

Man that's some hardcore SQL (to me)... I should learn more SQL :)

I never learnt joins... I just kinda stopped there and let Django handle it for me :P

The two problems you could run into: duplicated track #'s (causes issues now too with the ACaR script) and non-extant track # tags

The sql in cueread2 is just a select and grabbing the artist/release name from the id, the nasty stuff is just coming. :p

aCiD2: go ahead and commit that patch yourself. go get the credit you deserve. :)

the patch looks good, thanks for catching that.

haha, ok :)

I subscribed to mb-bugs, so I'm trying to catch the easy stuff as it comes

cause I know you lot the hard stuff ;)

ruaok: should I commit to trunk, or the RELEASE branch?

trunk

ok

its almost time to start thinking about another release.

lol

ruoak: you'd be the one to ask... internally, for the pages that can accept mbids in the url, how do those pages convert the mbids to rowids?

BrianFreud: GIve me an example page, and I'll take a look?

not that it stops rob answering, but it's one more thing for me to learn :)

BrianFreud: It's just in the database..

album { id integer, gid 36characterstring, name string, etc. }

aCiD2: Refresh for s(low)ql.

(don't look now aCiD2 )

ruaok: ha

BrianFreud: we load the object in question from the DB.

FauxFaux: 20 minutes!?

then we have teh rowid at hand

aCiD2: Heh heh heh.

actually, go ahead and change the script to use that... I'll let my friends on digg know >:)

Only needs to be once every so often. :p

ok, under "Changes mainly of interest to MusicBrainz Programmers" at http://blog.musicbrainz.org/archives/2003/09/se...

any of those pages already support mbid in the url

To pull off a "add times from release" plugin, as I see it, I need to patch editall.html to support time (and I guess trackname) edits in the url, but also add mbid -> release rowid lookup support, as the mbid is avail to picard, but the rowid is not

aCiD2: about the patch/bug... please only close tickets once is the patch in svn, otherwise they will get lost

Oh ok

thanks luks, won't make that mistake again :)

I've also created http://bugs.musicbrainz.org/report/33 a couple days ago

BrianFreud: Imo, things shouldn't be using rowids

I'm guessing I can copy the track/name edit in url code from edit.html, and the if/elseif/else from http://bugs.musicbrainz.org/browser/mb_server/t... for the mbid -> rowid lookup

I mean, that's what global unique id's are for

yes, well, editall has no support for either, but edit.html has support for edits in urls - but only if addressed via rowid, not mbid

Ah, I see

well http://bugs.musicbrainz.org/browser/mb_server/t... is what you want, I think then

and make sure you add $mbid to the %args in editall

that would give me $mbid=GUID support in the url?

I think so

haha

BrianFreud: I recommend you get a copy of Learning Perl btw

I'm reading it atm, while it's a bit light for me having coded before, it's a good intro

though I'm guessing this might be better if I want to look up a release, and not a track: http://bugs.musicbrainz.org/browser/mb_server/t...

luks: how does work the new report (http://bugs.musicbrainz.org/report/33)? does it checks "patch" in keyword and/or summary?

exactly

(t.summary LIKE '%[patch]%' OR t.summary LIKE '%patch %' OR t.keywords like '%patch%')

k. (I've just found the sql query export ;) )

Lies, learning perl is for girls, real men write php then transliterate it. :p

I believe you're allowed to use a language that isn't php, but I wouldn't recommend it.

errr PHP?!

hell no

D:

python <3

luks: http://musicbrainz.homeip.net/show/artist/tags....

I'm getting the cannot load tags message there.

last time it was JSON.pm, but I fixed that.

how can I debug what the issue is?

there are no accounts on that machine and the email verification doesnt' work.

sorry. :)

but I get no messages in the error log

that might be the problem

but this probably shouldn't depend on the email

- $m->comp("/comp/checkloggedin", 1, 1)

+ $m->comp("/comp/checkloggedin", 1, 1, undef, 1)

in http://bugs.musicbrainz.org/browser/mb_server/b...

I think that should fix it

logins are currently plain broken. :-(

SerialNumber has joined #MusicBrainz

on all servers, or just yours?

my dev server