#musicbrainz

so you want self-criticising AI?

lol

I'm in a discussion about [untitled] tracks...

I want scripts that do "if a=b and b=c then a=c", not "a~~b, and b~~c, and c~~d, ... then a=z"

...in edit http://musicbrainz.org/show/edit/?editid=7688419

I wonder: when is a track name considered "widely known"?

brianfreud: but if a 'smart' script gets it right most of the time, it'll save you a lot of work, won't it? as for it screwing up, a good script would let you check the things it wants to do anyway..

symphonick: I think for sure when it's with that title on another official release

Ah, thanks.

LotR, yes, but not when part of that script's functionality is to rewrite every track title on a set of releases

hidden tracks suck, though. Why not print a title on the cover ;)

This case is 3+ weeks from official release, but it got a name on discogs

if it's only a fan-named track then it should be [Terry's Song]

the most annoying thing is 20 minutes silence before there's the hidden track

brianfreud: why not? it'd give you a list of proposed changes, and you could say which one were wrong, and do those by hand

lol, ok, starting to dumb this theoretical script down now :D

MClemo, yeah... who want's to wait for 20 minutes for a 1 min crap song anyway?

wants :-(

people who like surprises? :)

But you KNOW it's crap?

If it was a good song it would have made it to an official track.

brianfreud: i'd rather go for a fuzzy logic script and run it on a release at once and take a glance at the results before committing

symphonick: well... some hidden tracks are just jokes really. I like some, but I still hate that they're not a cd track of their own

xlotu, yes, but that's a very different thing than "run a script on all classical and just let it automagically fix them"

LotR, is 90 [silence] tracks between them better? :)

yes

90 [silence] tracks compresses better than 20 minutes of silence :D

I prefer 90 short [silence] before a 20 minute "[silence] / [untitled]" ;)

and most importantly, they also get removed easier :)

that's where 320 kbps cbr mp3 files really rock, on silence tracks!

Go mono!

Yes!!

lol

I'd like to think that mp3 had a "skip" frame, but probably not. :p

that would make far too much sense

luks: you can skip those, so yes.

you can enhance the format and call it supermp3

and do stick the superior id3 metadata format in the file!

Also, explicitly ban id3 tags, just to drive luks insane.

JINX

:)

just write vorbis tags instead :D

I'd choose a combination of vorbis comments, apev2 and id3v1

Or add a real tagging format to ogg.

btw, i recalled another format than mp3 that use id3 tags last night: wavpack

Ugh, yeah. Flaaaaaaaaaaaaaaaaaac for evaaaaaaar.

wavpack uses apev2

id3 or ape :)

granted, id3v1, but still idd3 :)

s/dd3/d3/

yes, id3v1 and apev2

lol, that sucks

the id3v1 part, I mean

id3v2 would make it not a wavpack file

"Spazpack".

wavpack do create slightly smaller files than flac... then again, flac got a sane metadata format and is less expencive to decode

almost every format that uses apev2 uses also id3v1

luks: what do they need id3v1 for?

nostalgia

Leeeeeeeeeeeeegacy.

to help developers incapable of writing apev2 parser :p

I yesterday found out that that stupid mp3gain program from sourceforge attaches the replaygain tags as apev2 tags to mp3 files

yey! death to mp3!

TRhat's what fooooooooooooooobar is for.

drop table trm;

drop table onmuzface;

Hawt

I left join'ed your mother

"Windows media ASF files (WMA, WMV) have their own tagging formats but also supports ID3 Tags embedded as attributes."... wuahahahaha

that's a lie

from wikipedia?

yeh

they define standard mapping of WM attributes to ID3v2 frames

but you can't store id3v2 directly there

yup: http://en.wikipedia.org/wiki/ID3

I hate the evil empire

*shrug* i dunno, they got some references to some microsoft sources

i just stay the hell away from those formats anyhoo

I had to learn all of them :/

ah, so your hospital stay recently clearly was some mental institute? =)

ASF baby! :D

heh

which ones?

I was surprised to find picard supporting tta, for example

ASF is actually one of the nicer ones

why not rdf with dublin core? :P

tta = raw stream + id3v2

didn't even have a tta codec installed at the time, but picard could read it

it's easy to support for a format without coding :)

oh, that was 20 minutes ago

hmmm, this is rather interesting... http://www.digitalpreservation.gov/formats/fdd/...

interesting to see a government archival site referencing info from wiki...

hmmm, ID3v2 supports up to 16 16 MB frames of data....

you can embed a mp3 in a mp3 :)

hmmm

or a player :)

would the internal mp3 play?

nope

Patent claims [...] in 1998 by the Fraunhofer Institute [...]. The practical impact of these claims is not clear to the compiler of this document. <-- funny statement on a .gov site :)

brianfreud: the easiest way to do that would be MPEG-4. an MP4 file is a just a container with multiple "slots" that can hold anything. audio, video, text, animation, etc.

and each of those slots can be synchonized separately

or together

brianfreud: vorbis tags support up to 4gb of data in 1 tag, iirc :)

eg. ARTIST=<4gb worth of text>

insanity

dump the entire MB database into a tag

and store it on fat32

hi all

I have a problem with libmusicbrainz - if no connection to the internet is available it fails with SIGABRT, is that a wanted behaviour?

hmm, nobody awake?

I am

foolip: are you familiar with libmusicbrainz?

in python a little

in C, not at all

well I use it in C - it works so far ok, but the problem is: if no internet connection exists mb_query_get_releases produces SIGABRT and crashes my app

I should think that is a bug

unless someone though it was nice to do error reporting via unix signals

yeah, that's what I think too...

do you know who maintains libmusicbrainz?

perhaps file a bug report?

no, I just file bug reports and then someone fixes stuff

(actually just once)

ok...

ping luks

sysrq_ is now known as ssyrq