#musicbrainz

That's it ! 1AM I'm done with my edits :)

hmm, only 8pm here ;)

n1md4: British?

... but before I get started, too much other stuff around that has to be done first.

so not getting started at all.

owell.

n1md4: Also, I do appreciate fanart.tv - they actually rely on MBIDs for their music section (so basically, you can "only" add fanart for stuff which has been added to MB), and they have a limited scope that doesn't overlap with MB's.

Well, it does a tiny bit now with the CAA, I guess. But still.

Yeah, British.

I've had too many 4AMs this week!

Without other responsibilities I'd be awake at night and sleep all day :)

TheAudioDB would be good if they'd followed in fanart.tv's footstep's and relied on MB data, which TADB could then attach moods, genres, images, whatever to in addition to the data MB supplies/-d.

n1md4: A lot of us would. :)

hehe

* =>

from the theaudiodb site, it looks like their metadata is straight-up just musicbrainz imports

kepstin: Most of it is.

although they store a copy locally with possible user modifications :/

well, I said no more edits .. 2AM is a reasonable time to start thinking about bed

Being imports, it also means they don't get updates from the MB metadata.

yeah

and they track at our release group level, and ask folks to pick a representative tracklist for the album.

do you ever have a meeting of heads of these various projects?

it could really pay off!

unless they happen to meet at music tech meetups or hackfests, no

Well, we have some "collaborating" with Discogs.

But that's mostly just getting a data dump (which may be freely accessible anyway?) and being allowed to import that through Geordi.

yeah, the discogs data dumps are freely available, http://www.discogs.com/data/

We have a pretty good relationship with the BBC.

the foss community survives because (not only) there's a sea of ideas - but sometimes it's good to gather the waters, and do something intelligent, instead of swimming against the tide of one another ... or have I just had too much to drink :)

Spotify has been represented at at least the last two MusicBrainz Summits.

how'd you feel about spotify?

I'm not sure I like them, but don't know how to say why

They don't carry a lot of the music I listen to.

likely their licensing I suppose ...

admittedly, I'm not sure I'd really use spotify, they don't carry a lot of the music I listen to

hh

(Like I said, most MB's have a tendency to like more obscure music. :p)

*MB'ers

(I have yet to hear of a streaming service that carries japanese pop music, particularly doujin stuff like touhou arranges...)

too much exposure to mainstream, no doubt!

theres' too much out there to rely on 1 service alone

I listen to a fair amount of b-boy mixtapes and, well, generally hip hop, r&b, dancehall etc. mixtapes.

As well as underground hip hop artists.

That kind of non-commercial stuff only rarely makes it onto channels such as Spotify.

I'm tempted by stuff like google play which lets you upload to the cloud, but why would I do that when I could just carry around everything on a 160g ipod and save the cell data fees...

true

i couldn't be without my daap share ;) although "shamefully" my music tastes are reasonably mainstream ... I blame the radi.

radio*

Heh.

I never listen to the radio anymore. :p

the one radio station in the area that I almost care about, the country music station, my receiver can only pick up in mono... it sounds better as an internet stream :/

Freso should move to where the good radio stations are!

_mostly_ I don't like the idea that someone else controls what I can or can't listen to, so create my own library.

n1md4: +1

even if it's likely never to be taken offline (spotify etc)

I'm a collector too :)

I don't know. I wouldn't bet on Spotify being around in, say, 10 years.

But I can make sure my FLAC rips etc. are.

That's true.

Oh yes! I stream with compression, but keep a good backup of flac

I mean, I remember a time before Facebook, where MySpace was the lay of the land.

n1md4: Are you a Linux user by any chance?

You can tell!

:)

So's my wife and children .. maybe not by choice

n1md4: You might want to try and look into mopidy.

It's like mpd-on-speed.

interesting! I've been looking to upgrade my DAAP server, and really doesn't have to remain DAAP either .

It supports several "source"/"input" types, e.g. play from file, play online radio stream, ... play from Spotify.

And it allows several way to connect to - including mpd protocol (so all mpd clients can use it, of which there is a fair amount), a web client (actually, there are several to pick from), and some others as well, IIRC.

I must say (slightly off topic) I'm a big fan of bootstrap sites, they just click (no pun intended) with me.

:p

hmm. one fun interesting thing to do with an online streaming player might be to look into using webrtc transmission; you get to use opus codec then for using slightly less bits :)

kepstin: :p

Definitely going to look in to this some more

And fans of MusicBrainz. :p

I mean, for online streaming to headphones in a noisy place e.g. public transit, I'd probably be happy with 96kbit opus.

apt repo! love that

n1md4: Debian or 'buntu?

debian on servers and ubuntu for the desktop

:)

not a huge fan of my shuttleworth .. too much business like, but fair play to him, it's still foss

ooh. mopidy already stores musicbrainz ids in their data model

kepstin: As I said, they're MB fans. ;)

(of course the first thing I do is look for their data model in the docs)

Hehe.

okay, i know what I'm doing for the next 30 minutes :) brb

ooh. pluggably library support

pluggable*

... I was supposed to be off doing other stuff

Haha.

mopidy is a neat project. :)

guess I should package it for exherbo :)

I'm sure they'd love that.

Be sure to ping them about it in #mopidy of you do.

Also, they're usually happy to help plugin/extension authors and users as well.

if you get a lot of 'no matching tracks' where do you go? can you access a larger fingerprint database?

well, with the puid linkage gone from musicbrainz, no, there really isn't any bigger fingerprint database.

echonest has fingerprints, but they're optimized for a different purpose and tend to cover more mainstream stuff and less obscure stuff.

the best solution, of course, would be for you to match the tracks to musicbrainz using additional metadata and give us your fingerprints...

Thoughts on the table header and medium separation? http://i.imgur.com/oizD5J0.png

If you want it, here's the musicbrainz.org version http://i.imgur.com/rsnOMhv.png

match the tracks to musicbrainz using additional metadata and give us your fingerprints..., how would I do this sorry?

do you have any information about the artist/track title/album or anything like that? you could try doing lookups on musicbrainz via that info

nod will do after it finishes scanning

im seeing a few 'gold' cd icons amongst the silver ones, what does this tell me?

dna6android: for example use beets

dna6android: it has a workflow for submitting acoustids

in a semi-automated fashion

if your music has decent existing tags

navap: I like the lighter touch. If I were looking for a cue for medium separation, I'd probably look for indentation before weight/color.

(for that matter, has anybody ever needed to refer to the column headings "#" and "Title" to figure out what they're looking at?)

wow, the ws/2 api response for that release is wicked slow.

http://mbjs.kepstin.ca/?release=ba9f7a0c-03fe-4... :)

(handily pre-cached for your viewing pleasure)

certainly less... compact than navap's layout :)

...

http://mbjs.kepstin.ca/?release=d7861c53-af4e-4... and http://mbjs.kepstin.ca/?release=ba9f7a0c-03fe-4... should probably be merged.

is there a performance report of slowest releases?

If only we could get rid of the column headers. I think it looks weird with whitespace immediately below them for rating and length

The columns are obvious for a release, but on other pages you get weird things that need to be labelled

ah. I hadn't considered that.

Duration is probably clear enough without a header. Ratings, maybe?

I also like the lighter version. Could 'rating' and 'length' be put on the same line as the disc title?

Actually even rating might not need a header

"Score", on the other hand, will never make any sense to me, header or no header. Because: lucene.

(thinking of other pages, obviously)

kepstin: fyi, even if the puid database was still there, it would be a smaller one :)