#musicbrainz

Grrreat.

honty half hour happy hug fest should be called happening, not fest!

petros: Ping?

Pawng

Freso: Pong.

"The Copper Family has/had member(s) Bob & Ron Copper" http://musicbrainz.org/show/artist/?artistid=32... I think this should be changed to two separate links; one each for Bob and Ron. Thoughts?

That AR should link to Bob Copper and Ron Copper imo

Bob & Ron Copper is a performance name and a group, not actually people entities

http://musicbrainz.org/show/artist/?artistid=51... and http://musicbrainz.org/show/artist/?artistid=52...

Cool. Also there is a group being the "parent" of another group, which I'm looking at changing as well.

How does the "Similar artists" thing work? Is it automatically generated, or can I add some similar artists?

mikemorr: they are generated from dmoz

gioele: Thanks. I had no idea about that.

the "word" smexy shoudl burn in hell forever and hever

iit makes me gag

Smexiful

Sexy mexican?

Esta muy fuego

loi

Dick Twang?

Sounds kinky

hi ruaok

Ahoy ruaok

ahoy!

thye've been kicking my ass since I've gotten back...

Hayfever?

yeah.

imagine the pollen that comes from living in the middle of this: http://flickr.com/photos/mayhem/2360079093

Sure looks nice though

yea. mixed bag for sure.

:(

bah phase mix

bh foods boling

you can't use %any_tag% in picard, only those of http://wiki.musicbrainz.org/PicardScripting?act... ?

Lin_ is now known as Lin

mclemo: but you can define new variables with $set

I just started using Picard to organize a rather large library of music; for one particular artist, it has the same album listen 3 times. The album is complete, but picard sees 1 album with only 1 song, 1 album with 7 songs, and 1 album with 2 songs (for example)... Why is this?

The "albums" all have the same name

although they do have diffferent times.

because picard is not so smart

it picks what looks best for each individual file

luks: ahh, dang :)

drag/drop them to the right one and save

So, in the musicbrainz database this particular album might have 3 listings (for different releases) and picard seems to think some files go with one, when some go with another?

yes

luks: okie; just want to understand what's happening :)

luks: btw, from time to time somebody requests/suggests a better algorithm to match to only one album. i've never had a closer look, but is there something promising?

outsidecontext: well, the only good way would be to do lookups for all files, and decide which albums to load at the end

some people suggested to pick one release and then stick to that when doing other lookups

but the problems start if the first release was wrong

(being new here, feel free to smack me down if stating the obvious or that which has already been discussed) perhaps having a score based on various information would work...

Vlet: how would that help?

for example, if all songs in album A are in the same folder as album B and all the songs from album A+B are the only files in the folder, then add 1 to the 'score' that they are the same...

let's say you have 13 files, and there are two releases - 12 tracks and one with a bonus track, it analyzes the first file and loads the 12-track one

if Album A has the same name as the folder it is in, +1

then it scans the 13th file and finally finds the right rleease

Vlet: folder based scoring is not a good idea

Vlet: if you already have files in the right folders, they are probably also already tagged

well, that's why I was thinking 'scoring', not using it as a basis for data... just hints

so you can just skip the scan step

i agree with luks, the only reliable way would be to decide after loading. picard is supposed to work on pretty screwed up tags, too

this is something that could be done in the tagging wizard

I have a pretty large 'communal' library, so there's a lot of screwy-ness to the tags, even though as far as file/folder structure, it's fairly sound

but for the main interface I'd prefer the more interactive version, ieven if it loads wrong albums

luks: probably better your way; once I initially clean up this library, it's not like I'm going to have hundreds of dups like I do now.

Vlet: I think you will be done much sooner working with small batches, cluster, and then cluster lookup

loading many files and using the 'scan' function will only slow you down

luks: yes, I'm going one letter at a time

A's, B's, etc

139,731 tracks :o

What does the (12/14; 12*) mean next to an album? 12 out of 14 tracks, but then why the second "; 12*"?

12 unsaved files

thanks :)

And thank you for the nice application; very pleased so far (assuming you're the developer)

luks: I don't think having stuff in folders means they're tagged...

Is there a tool that I can use to fingerprint files from the CL? (picard will look for the fingerprint in the tags, right?)

you can rip stuff and they're usually put into folders and you can download whole albums and they normally have folders

I can't really imagine a common situation where someone would have files from albums spread across multiple folders

the only other thing I can think of is people who just download random tracks and keep them all in the same folder

nikki: even itunes defaults to folders based on artist

you can change that option, but most people don't. it's not practical for a large collection

itunes default is an artist->album heirarchy

Hahahaha!

Oh my, that is pretty decent :D

yeah, it's real fun tracking server prgram erros without any kind of error logging or even error handling

error handling, we don't need no fucking error handling. our software is perfect!

I've to debug some brain dead RPC functions that just silently ignore all errors

Snicker.