#musicbrainz-devel

/

      • alastairp
        hmm
      • when did mb start using PUID?
      • ruaok
        2006
      • alastairp
        for that matter, when did we start TRM?
      • kepstin-work postdates trms, but predates puids.
      • Freso
        I don't remember no-PUIDs, but I do remember dropping TRMs.
      • But I also joined up in '06, so...
      • ruaok
        hmmm. thats gonna take some digging
      • jesus2099 joined the channel
      • ijabz joined the channel
      • jesus2099
        warp bitmap : btw i just found it’s MB.Control.EntityAutocomplete();
      • héhé
      • hawke_1
      • jesus2099 has left the channel
      • (@alastairp)
      • ruaok
        woah. I just found a job offer from relatable.
      • I think it was sometime in late 2000, early 2001 when we started using TRM
      • hawke_1
        ruaok: The wiki page says 2000
      • “A version of the TRM client was first released on August 18, 2000 to help create MusicBrainz” says relatable.com
      • ruaok
        which is odd, given that MB wasn't started until sept 2000.
      • ah. that makes more sense.
      • alastairp
        right. thanks
      • ruaok
        that's right, they contacted my about cdindex and partially pushed me over the edge to go for mb.
      • alastairp
        well, it's in my thesis now. so it's official
      • ruaok
        its amazing how many orgs mb has outlived.
      • Freso
        :)
      • (@ ruaok and alastairp both.)
      • alastairp
        ruaok: you wouldn't know what tech trm was based on? I found a paper that was talking about "our company relatable"
      • so assume it's that, but i'm looking for independent verification
      • ruaok
        yes, trm was from a company called relatable.
      • but I dont know anything about the algorithms they used on the server.
      • alastairp
        rigth
      • Freso
        Go proprietary, closed-source software and algorithms! :D
      • ruaok
        hey, it helped us get started.
      • Freso
        Yep.
      • ruaok
        closing the metadata loop was important
      • Freso
        I'm glad we've finally arrived at AcoustIDs though.
      • alastairp
        ruaok: and you don't know about puid either?
      • Freso
        Well, arrived for now, anyway.
      • ruaok
        did you guys see this?
      • reosarevok
        Looks nice
      • ruaok
        I dare not ask if we should put flattr buttons on mb.org
      • Freso: maybe on musicbottle?
      • reosarevok looks at http://grooveshark.com/#!/artist/Tego+Calder+n/58489/albums , shakes head
      • hawke_1
        I like the idea of flattr buttons on mb.org
      • kepstin-work
        for the artists or for the mb developers? :)
      • hawke_1
        Heh
      • probably both
      • For the mb editors, a flattr link on every edit! :-D
      • ruaok
        all of the above.
      • I've asked if flattr can resolve all artist MBIDs
      • kepstin-work
        huh, then you have some way of thanking that guy who keeps adding all those naxos releases, for example ;)
      • ruaok
        would someone take this to the style list to ask if we should do something with this?
      • reosarevok
        ruaok: I will, once they reply :)
      • ruaok
        ok
      • hawke_1
        The only thing I find weird/unattractive about flattr is how the amount spent is dependent on how much you use it. (i.e. flattr one things in a month and they get a relatively big amount, one hundred things and they get a small amount)
      • reosarevok
        Well, I find that bit to make perfect sense
      • You have X money to share amongst all the stuff you like
      • the less you like, the more there is to share
      • hawke_1
        yeah, but the amount shared/paid isn’t proportional to how much you like something
      • just how many cool/liked things you found this month as opposed to last month
      • reosarevok
        Hmm
      • I could see it working with a few tiers
      • Like "I like this alot!" (with the picture of an alot)
      • And "I like this fine"
      • Or whatever
      • And one "alot" being worth twice a fine
      • dunno
      • hawke_1
        Yeah
      • That seems reasonable to me.
      • ijabz joined the channel
      • Prophet5 joined the channel
      • LordSputnik joined the channel
      • LordSputnik
      • nikki: should work on files in a subdirectory now
      • nikki: but it'll still move them to the top directory if you have more than one folder with music
      • nikki gets more discnumber errors
      • nikki
        yeah, it still moves them
      • LordSputnik
        even when they're all in a single subfolder?
      • nikki
        yes
      • LordSputnik
        ... that's weird
      • works for me :P
      • hmm, what happens if you add "print library_folder" at the end of the DetectLibraryFolder function in MassTagger.py?
      • (line 75-ish)
      • nikki
        nothing
      • LordSputnik
        that's extremely weird
      • nikki
        I even added a print "library_folder:" line and that doesn't appear either
      • LordSputnik
        hmm
      • can you run "ls -R" from the terminal in that directory and put it on gist?
      • (if you're on linux)
      • Mineo liked the PrintHeader function of masstagger.py - "MusicBrainz Warp - 0.1 created ben ockmore" :)
      • Freso
        ruaok: Do you have any idea how I would go about resolving MBID to Flattr URL?
      • Mineo likes, even
      • Oh, he's not on anymore.
      • Bugger.
      • kepstin-work
        Freso: presumably by adding a new AR type to artists, "flattr url" :/
      • Freso
        :|
      • LordSputnik
        Mineo: why do you like it? :P
      • oh
      • lol
      • ...
      • kepstin-work
        it's an awesome theme name, and warp.
      • Mineo
        and the app created its author!
      • kepstin-work didn't notice that.
      • kepstin-work should go home, then... :)
      • Freso
        I wish there was a way in JIRA to make "collections" of some sort of tickets.
      • kepstin-work
        ticket groups? ;)
      • Freso
        E.g., there are some tickets that I would potentially take on, but I don't want to submit myself enough to it at this point as to assign myself to them.
      • nikki
        you guys are obsessed with grouping
      • Freso
        I need Ticket Group Group Groups! D:
      • LordSputnik
        nikki: that's really weird...
      • nikki
        yeah :/
      • kepstin-work
        man, encoding opus in ogg streams would be so much easier if they had an opus wrapper library that generated ogg_packets like libvorbis does...
      • kepstin-work is manually wrapping the raw opus encoder output into packets for now.
      • LordSputnik
      • nikki: can you change the detect function again, to: http://pastebin.com/VAV5Nzvq
      • nikki: then we'll see where it's tripping up
      • kepstin-work
        hmm. that certainly wouldn't work if I ran it over my top-level music folder :)
      • kepstin-work would have to run it separately for each two-levels-deep subdirectory
      • LordSputnik
        kepstin-work: me too :P
      • kepstin-work
        whee, ~/Music/Opus/brainz-opus/Primary/Rotate/01\ -\ Primary\ -\ Fix.opus
      • my first subdirectory is by filetype, second one is by manual sorting.
      • ideally i'd adjust it so that the tagger could determine the second level automatically by tags on musicbrainz or something, and the first level should be easy enough to do programatically.
      • kepstin-work quite likes release 80ad4261-957e-4a74-847f-ddde2e10f4b5, so he usually uses it when trying new things :)
      • Leftmost
        opus, eh? What quality?
      • I've wondered about using it.
      • LordSputnik
        kepstin-work: i may do it in a smarter way, by going from the audio file upwards instead of the library directory downwards
      • ruaok joined the channel
      • kepstin-work
        Leftmost: I've been using 128kbit
      • which I can't tell apart from my old 160kbit vorbis or 192kbit mp3 files, so yeah.
      • most people seem to find it fairly close to a good aac encoder.
      • (in hydrogenaudio blind listening tests done at 64kbit, opus beat out a couple of good aac encoders)
      • nikki wonders how big a quality difference would have to be for her to notice
      • Freso
        1 bit, I'm sure.
      • kepstin-work
        the difference in that hydrogen audio test was barely above the margin of error, tho.
      • nikki
        Freso: 1 bit? in my dreams :P
      • Freso
        nikki: :)
      • nikki
        I did once try it with various mp3 bitrates, but I don't remember what the results were
      • probably that I can't tell 128 from 320, at least :P
      • kepstin-work
        some people can tell 128kbit mp3 apart, but usally not that much higher
      • Freso
        FLAC is definitely overkill, but I'm too much of a hoarder to lose anything, and I don't want to both store as FLAC and also store things as something else for playing them.
      • kepstin-work
        mp3 has some really annoying ringing/metallic artifacts if you're listening to something that pushes the quality mp3 can do at 128kbit
      • nikki
        I like flac.
      • kepstin-work uses flac for local playback in his house, but makes lossy copies of everything for portable use.
      • kepstin-work has ridiculous amounts of disk space.
      • Leftmost
        kepstin-work, that's what I do.
      • nikki too
      • nikki
        oggenc is great :D