#musicbrainz

damn

alastairp, https://github.com/mderezynski/rimokon

I slaughtered the topic

alastairp, but it needs a modified libmpd

which supports MusicBrainz ID information

OK, cool

so it's kind of mpd-backed?

oph, awesome

yeah it's an MPD client

but MPD is terrible, terrible

I need so many workarounds

or maybe it's libmpd

I'm been after a nice distributed player for a /long/ time

well it's usable and it works

yeah, this is why I don't use mpd :)

I wish DBus supported TCP/IP

I'd start writing my own daemon instantly

I've /just/ started using upnp + xbmc + xbmc remote for android

ah

mpd is written as if this were 1995 or so

it needs a more modular architecture

Clint, MPD?

they should improve their protocol because it really limits what a client can do

for all I can tell, it sends next-song events whenever the f* it wants

yes, instead of continuing to bolt things on sideways

sometimes 5 in a row

it's really horrible

I have a friend somewhat interested in a better mpd

I should try and get him in

the problem with MPD is that it enforces a particular way in which clients can be written

and therefore most clients are as horrible as the MPD socket protocol allows

I think that's a general problem with remote players

xmms2 has the same problem, IMO

xmms2 does some things I don't really like

e.g. the database is implicit

you can play any song anywhere and it automatically adds it to the database

there must be a better way than using a plain socket, or at least they should improve their socket protocol

joe wreschnig (author of quod libet) started writing http://code.google.com/p/antiphon/

but it seems kind of dead

quod libet is quite a good player

he put some real thought into some of the features

one example for a horrible MPD client is Ario

but I guess people will use what's available

also people do stuff that is beyond my understanding as for why

e.g. Banshee can parse advanced queries only like this:

"date<1994"

it doesn't allow for whitespace

there's no reason why it shouldn't

Am I mad for expecting NGS to work under Perl 5.12 and Postgres 9.0.2?

with the number of dependencies, I wouldn't be too sure about Perl 5.12

PostgreSQL 9 should be fine

Man up and build the dependencies!

They are!

And have been since I woke up.

2 hours ago. ¬_¬

Hm, well that's all the dependancies installed and supposedly working.

Herp a derp. muz [at yggdrasil]:~/musicbrainz-server/postgresql-unaccent$ make -j15

Makefile:9: /usr/lib/postgresql/8.4/lib/pgxs/src/makefiles/pgxs.mk: No such file or directory

Looks hardcoded to me. /o\

its not.

Oh penises. It's because apt shat all over my system

Yeah, just started prodding about. I've got multiple instances of postgres on this machine. Am I mad? WHY YES.

man, I really should get on drafting a Release→URL "Official Homepage" AR. Pages as cool as http://thetrois.net/ deserve to be linked :)

Yeah, it's a really awesome site. Went on earlier to check the release names

hmm. I think I’m going to add track artists for those TROIS PROJECT releases.

they pretty clearly label the vocalists as track artists on the album art, and it’s a mix of multiple groups on each disc

Muz: yes, it should work with 5.12 - Perl has pretty strong back compat

well, insanely strong

Milosz: why not write a client for xmms2, which has a considerably more sane api?

well, we had a lot of problems with getting the old code running on 5.10 :)

ish, that was one very localised point

and that was doing some crazy glob referencing stuff so I'm not surprised that was brittle :P

with the number of modules the new code uses, it would be surprised if there wasn't some crazy code in there

the crazy code is mostly factored into a few important modules though, rather than in house

ocharles, I would just need to write a different source plugin

like the aptly named Variable::Magic

:)

which I probably am going to because I want to see how well xmms2 works

to do*

i've been happy with it

I think anything would be better than MPD

I first used existing clients and MPoD2 on my iTouch

but when I started _writing_ a client I really became upset

ocharles: yeah, I've found the most of my troubles have been with apt and Ubuntu more than anything.

Lesson learnt: don't assume that Ubuntu's assumptions are sane, and just do it yourself.

update-manager got worse in 10.10

for whatever reason

update-manager? Doesn't sound like a CLI tool to me. ¬_¬

I'm not being a heretic and X forwarding apps to my Windows machine, like FauxFaux is. ¬_¬

The pro way to do it. Also, Muz, 9.0.2 is totally the newest version of postregsql.

Yeah, I realise that now. It's what I'm running. ¬_¬

I read that as "compiling drunk"

sudo apt-get install -t unstable postgresql-9.0 # seems pretty easy, y u no debian Muz?

FauxFaux: because Ubuntu is user friendly.

KingJ, I ended up setting track artists on that THIRD ENSEMBLE: NOIR album; can you poke through the 10 edits starting at http://musicbrainz.org/show/edit/?editid=13803682 (use the Next Edit button at the bottom) and let me know if that seems OK?

kepstin: I need to check IRC more often. Checking through it now, yay for track artists

I'm adding an Audiobook to the database. There are 39 discs. Can I clone an existing entry, and edit the disc #?

KurtB: Not yet.

KurtB: Somebody told me that it will be implemented after NGS.

petteyg, Ah.... well, at least some one is thinking about it in the future.

kepstin: around?

yep

I'll PM you in a moment if that's okay, just got dragged into something

sure

(but if it's not that private, it might be better to just talk in here, so it can be logged :)