#musicbrainz

/

      • kepstin-work
        since some names should be sorted by 'first name', not 'last name' :)
      • 2013-02-22 05335, 2013

      • kepstin-work
        e.g. many patronymic names
      • 2013-02-22 05351, 2013

      • tom_anderson
        ianmcorvidae: here's a better list of supported dbs: https://github.com/doctrine/dbal/tree/master/lib/…
      • 2013-02-22 05303, 2013

      • kepstin-work
        icelandic comes to mind there; the phone book in iceland is sorted by given name :)
      • 2013-02-22 05309, 2013

      • hawke_1
        kepstin-work: and then we do stupid stuff with the sort name. ;-)
      • 2013-02-22 05316, 2013

      • nikki
        tom_anderson: yes, that would work
      • 2013-02-22 05326, 2013

      • Leftmost
        Names are weird and very culture-specific.
      • 2013-02-22 05332, 2013

      • ianmcorvidae
        as long as postgresql's supported that's probably what I'd use if I do an install locally (since I have that installed and configured and such for MB already)
      • 2013-02-22 05346, 2013

      • nikki
        jesus2099: we don't have any guidelines right now, but that's what I normally use, yes
      • 2013-02-22 05353, 2013

      • CallerNo6
        hawke_1: stupid sortname tricks?
      • 2013-02-22 05323, 2013

      • jesus2099
        nikki: it avoids → this one looks cooler — no that one is best —  no — yes — no
      • 2013-02-22 05326, 2013

      • jesus2099
        :D
      • 2013-02-22 05353, 2013

      • Leftmost
        There are also issues of cultures with given name and family name reversed (so how do we know to display them in the right order?), cultures with multiple surnames, etc.
      • 2013-02-22 05309, 2013

      • tom_anderson
        ianmcorvidae I have not installed it on anything but mysql so far and look forward to hearing your experience.
      • 2013-02-22 05307, 2013

      • Leftmost
        I read a good article on the assumptions web apps tend to make in having people input names.
      • 2013-02-22 05314, 2013

      • jesus2099
        anyway who wants more than one field ? I didn’t get the start of discuss, probably
      • 2013-02-22 05316, 2013

      • tom_anderson
        Leftmost: By using last/first on the performer and just name as the alias I hope to capture that data on both sides.
      • 2013-02-22 05322, 2013

      • jesus2099
        (or are we all saying the same same)
      • 2013-02-22 05324, 2013

      • Leftmost
        Let me see if I can find it.
      • 2013-02-22 05326, 2013

      • ianmcorvidae
        tom_anderson: do you have any sort of data dump that can be imported, by the way? an empty DB is well and good, of course, but not quite the same as testing with real data
      • 2013-02-22 05347, 2013

      • ianmcorvidae
        jesus2099: not in MB, tom_anderson here is working on a different system -- which will link to MB, but is separate :)
      • 2013-02-22 05349, 2013

      • tom_anderson
        ianmcorvidae: that's a whole other can of worms I haven't started on
      • 2013-02-22 05355, 2013

      • ianmcorvidae
        ah :)
      • 2013-02-22 05313, 2013

      • hawke_1
        This sort if thing is another reason that having external auth is kind of nice (OAuth and the like) — lets you outsource all that naming crap that you don’t really care about to someone who has the time and energy to put into it.
      • 2013-02-22 05319, 2013

      • derwin
        tom_anderson: btree guy?
      • 2013-02-22 05325, 2013

      • tom_anderson
        db.etree
      • 2013-02-22 05330, 2013

      • ianmcorvidae
        hm, looks like your thing is also set up for apache
      • 2013-02-22 05331, 2013

      • hawke_1
        For usernames anyway, i.e. webapps — not for MB etc.
      • 2013-02-22 05339, 2013

      • derwin
        tom_anderson: oh hai ! I talked to you @ sfmusictech
      • 2013-02-22 05339, 2013

      • ianmcorvidae
        maybe I'll also see how your thing would work with nginx/php-fpm then :)
      • 2013-02-22 05353, 2013

      • tom_anderson
        derwin: my project: https://github.com/dbetree/db
      • 2013-02-22 05330, 2013

      • jesus2099
        ah thank ian
      • 2013-02-22 05332, 2013

      • jesus2099
        :)
      • 2013-02-22 05306, 2013

      • ianmcorvidae
        hawke_1: our oauth2 thing will ship to beta monday :)
      • 2013-02-22 05324, 2013

      • ianmcorvidae
        (only vaguely related -- but still :D)
      • 2013-02-22 05307, 2013

      • kepstin-work
        oauth2 to allow access/editing for stuff like tags, collections, etc? cool!
      • 2013-02-22 05334, 2013

      • ianmcorvidae
        for anything that currently requires auth, acoustid/geordi will probably also switch to it even though they don't need any editing access :)
      • 2013-02-22 05349, 2013

      • kepstin-work
        picard should of course switch to using an oauth flow instead of taking username+password
      • 2013-02-22 05310, 2013

      • ianmcorvidae
        well, given that luks is the one who implemented the oauth support it strikes me as likely that'll happen soon ;)
      • 2013-02-22 05327, 2013

      • Leftmost
      • 2013-02-22 05349, 2013

      • kepstin-work
        tom_anderson: ^^^ well worth a read :)
      • 2013-02-22 05301, 2013

      • tom_anderson
        I'm getting Bjorked now
      • 2013-02-22 05300, 2013

      • TheLastProject joined the channel
      • 2013-02-22 05304, 2013

      • kepstin-work
        icelandic names are strange tho; people have to choose from an approved list managed by the government or ask for a special exception that may or may not be granted.
      • 2013-02-22 05341, 2013

      • Leftmost
        I keep getting emails about attending IUC. I wanna go. :-P
      • 2013-02-22 05348, 2013

      • warp
        kepstin-work: over here the goverment may refuse to register certain names as well, though I am unclear what the exact rules are
      • 2013-02-22 05316, 2013

      • warp
        (though one of them is that the name should be written in latin script, which isn't that surprising)
      • 2013-02-22 05342, 2013

      • derwin
        kepstin-work: the same issue in other countries.. but I hear they're way ahead in freedom terms.. over in europe..
      • 2013-02-22 05354, 2013

      • kepstin-work can't find anything in canada on the subject
      • 2013-02-22 05356, 2013

      • derwin
        sure, you can't name your kid whatever you want, but...
      • 2013-02-22 05305, 2013

      • derwin
        they're WAY AHEAD OVER THERE, I HEAR
      • 2013-02-22 05311, 2013

      • derwin
        oh, and it's illegal to be a neo nazi
      • 2013-02-22 05316, 2013

      • derwin
        but other than that thoughtcrime
      • 2013-02-22 05320, 2013

      • derwin
        WAY AHEAD OVER THERE
      • 2013-02-22 05332, 2013

      • derwin ends parody of clueless leftist peer group, thank you, thank you
      • 2013-02-22 05341, 2013

      • warp
        haha
      • 2013-02-22 05317, 2013

      • kepstin-work
        i suspect you probably need your name to be in latin script here, but I can't find *any* documentation talking about this stuff.
      • 2013-02-22 05337, 2013

      • warp
        My given name is not part of the name I am registered under with my goverment. This is something I need to be aware of whenever someone will do a comparison with my passport (like when buying airplane tickets).
      • 2013-02-22 05343, 2013

      • kepstin-work
        hmm. for a birth certificate, you have to provide separate 'last name' and 'given name(s)'
      • 2013-02-22 05335, 2013

      • kepstin-work
        but as to what a name can actually contain, nothing.
      • 2013-02-22 05359, 2013

      • tom_anderson
        Björk says I can't have last/first names so I'll drop that from my schema and use name/alias(es)
      • 2013-02-22 05307, 2013

      • kepstin-work
        yay
      • 2013-02-22 05321, 2013

      • derwin
        tom_anderson: I told you that ppl in here would be able to help a lot :D
      • 2013-02-22 05350, 2013

      • Leftmost
        We're just huge nerds is all.
      • 2013-02-22 05307, 2013

      • kepstin-work
        walking edge case generators, all of us ;)
      • 2013-02-22 05333, 2013

      • kepstin-work
        the entire purpose of the style mailing list seems to be to let us smooth out our guidelines until we run out of edge cases ;)
      • 2013-02-22 05309, 2013

      • ianmcorvidae
        lol "run out of edge cases"
      • 2013-02-22 05339, 2013

      • nikki
        like there's such a thing
      • 2013-02-22 05348, 2013

      • Leftmost
        An optimal weighted minimum of edge cases and style weirdness.
      • 2013-02-22 05354, 2013

      • kepstin-work
        huh: for 'musicbrainz.org/doc/libdiscid', google is returning "A description for this result is not available because of this site's robots.txt"
      • 2013-02-22 05325, 2013

      • kepstin-work
        doesn't make sense, it should be allowed
      • 2013-02-22 05301, 2013

      • ianmcorvidae
        yes, it should :/
      • 2013-02-22 05316, 2013

      • ianmcorvidae
        unless we have a server with the wrong one in rotation
      • 2013-02-22 05320, 2013

      • ianmcorvidae
        yes!
      • 2013-02-22 05323, 2013

      • ianmcorvidae
        probably lolo
      • 2013-02-22 05325, 2013

      • drsaunde joined the channel
      • 2013-02-22 05354, 2013

      • ianmcorvidae
        fixed
      • 2013-02-22 05327, 2013

      • drsaunde joined the channel
      • 2013-02-22 05300, 2013

      • jesus2099 joined the channel
      • 2013-02-22 05336, 2013

      • puneet joined the channel
      • 2013-02-22 05303, 2013

      • JonnyJD
        kepstin-work: I think I know why you have problems with libmirage. They completely misunderstood library versioning. Or did they really have 7(!) not backwards compatible releases?
      • 2013-02-22 05325, 2013

      • kepstin-work
        JonnyJD: lots of not backwards compatible releases; it's very tightly tied to the cdemu project
      • 2013-02-22 05352, 2013

      • JonnyJD
        oh boy..
      • 2013-02-22 05307, 2013

      • kepstin-work
        hmm. libdiscid builds some executables 'discisrc' and 'discid' that it doesn't install, is that intentional?
      • 2013-02-22 05343, 2013

      • JonnyJD
        well, right now they are examples/tests. I was also thinking about including these, but they are not really "complete" command line utilities
      • 2013-02-22 05331, 2013

      • JonnyJD
        you know, with an actual interface, parameters, version information, man page
      • 2013-02-22 05333, 2013

      • kepstin-work
        other than that, everything looks good. builds fine, tests pass in sandbox environment, everything's installed to the right place :)
      • 2013-02-22 05327, 2013

      • JonnyJD
        Either way. Having something like discirsc using libdiscid would be nice. But I think just making discisrc "public" like it is now is not right.
      • 2013-02-22 05345, 2013

      • kepstin-work
        https://gist.github.com/anonymous/5016266/raw/ is all there is to the exherbo package :)
      • 2013-02-22 05335, 2013

      • JonnyJD
        I would actually like it if anybody takes that on and creates a discisrc project or something that combines discid and discisrc. In any way, it should be a different package in distributions.
      • 2013-02-22 05301, 2013

      • JonnyJD
        that is quite a short packaging file
      • 2013-02-22 05321, 2013

      • kepstin-work
        yeah, the 'require cmake' bit does everything for me.
      • 2013-02-22 05340, 2013

      • kepstin-work
        if there was some optional configuration it would be longer :)
      • 2013-02-22 05348, 2013

      • JonnyJD
        yes, looks like everything is done implicitely
      • 2013-02-22 05317, 2013

      • JonnyJD
        not sure if that makes it actually more difficult to understand, but learning exherbo is a bit too much right now
      • 2013-02-22 05322, 2013

      • kepstin-work
      • 2013-02-22 05327, 2013

      • JonnyJD
        I like it how arch linux does packaging
      • 2013-02-22 05301, 2013

      • JonnyJD
      • 2013-02-22 05328, 2013

      • JonnyJD
        it is a bit longer, but has the complete information and not much implicit stuff
      • 2013-02-22 05351, 2013

      • kepstin-work
        the exherbo system assumes that your package is autoconf based by default, so to do an autoconf build you basically leave the file empty.
      • 2013-02-22 05302, 2013

      • kepstin-work
        it's all well documented and easily overridden, however
      • 2013-02-22 05354, 2013

      • kepstin-work
      • 2013-02-22 05321, 2013

      • JonnyJD
        well, if it works. Everybody uses the distribution that works best for them.
      • 2013-02-22 05346, 2013

      • kepstin-work
        http://www.exherbo.org/docs/exheres-for-smarties.… shows the actual code that's run automatically
      • 2013-02-22 05303, 2013

      • Freso
        Very, very late to the discussion, but: My first name is "Frederik Sandberg" and my last name is "Olesen". My given name is "Frederik" and my family name is "Sandberg Olesen". Even for western cultures, splitting names up in multiple fields can be difficult.
      • 2013-02-22 05309, 2013

      • Freso
        tom_anderson: ^
      • 2013-02-22 05354, 2013

      • JonnyJD
        freso: that is weird
      • 2013-02-22 05310, 2013

      • Freso
        ianmcorvidae: OMGOMGOMGOMGOMG!! First URL lookups, and now OAuth on Monday? OMG!
      • 2013-02-22 05316, 2013

      • ianmcorvidae
        heh
      • 2013-02-22 05330, 2013

      • ianmcorvidae
        I guess you missed the meeting :P
      • 2013-02-22 05301, 2013

      • JonnyJD tries to make the versioning and so-versioning easier in libdiscid.
      • 2013-02-22 05333, 2013

      • kepstin-work
        JonnyJD: is 0.2 and 0.3 binary compatible?
      • 2013-02-22 05301, 2013

      • JonnyJD
        yes, you can link a 0.2 progam to libdiscid 0.3.x
      • 2013-02-22 05308, 2013

      • kepstin-work
        k, good ;)
      • 2013-02-22 05309, 2013

      • JonnyJD
        not the other way around, obviously
      • 2013-02-22 05315, 2013

      • kepstin-work
        well, yes :)
      • 2013-02-22 05321, 2013

      • JonnyJD
        we use the semver.org style
      • 2013-02-22 05330, 2013

      • JonnyJD
      • 2013-02-22 05314, 2013

      • JonnyJD
        that is in the end also what libtool is doing "somewhat", but semver.org is a good documentation and libtool is a mess (documentation wise)
      • 2013-02-22 05320, 2013

      • CallerNo6 agress, the oauth news is omg-worthy
      • 2013-02-22 05330, 2013

      • CallerNo6
        er, agrees even
      • 2013-02-22 05334, 2013

      • kepstin-work
        JonnyJD: soversioning properly is a bit painful
      • 2013-02-22 05316, 2013

      • derwin
      • 2013-02-22 05318, 2013

      • derwin
        last.fm playlisting for subsonic? yes pls
      • 2013-02-22 05323, 2013

      • JonnyJD
        the result is easy to understand, but the way with current, revision and age.. yuck
      • 2013-02-22 05308, 2013

      • JonnyJD
        I think I will drop all that stuff in the cmakefiles and I am looking for a way to make it easier in the autotools (which might be more difficult)
      • 2013-02-22 05343, 2013

      • Freso
        JonnyJD: If you think that's weird, you should ask me about my father's name sometime. :)
      • 2013-02-22 05345, 2013

      • kepstin-work
      • 2013-02-22 05355, 2013

      • Freso
        ianmcorvidae: I did.
      • 2013-02-22 05303, 2013

      • Freso
        ianmcorvidae: Did you see me send notes out from it? ;)
      • 2013-02-22 05316, 2013

      • ianmcorvidae
        haha
      • 2013-02-22 05322, 2013

      • JonnyJD
        kepstin-work: that is what I am reading (again). I understand the result, but the whole documentatio is a mess and setting numbers that are only used for calculations.. a mess
      • 2013-02-22 05307, 2013

      • JonnyJD is having dinner now
      • 2013-02-22 05331, 2013

      • kepstin-work
        JonnyJD: the key point i think is that it sets up versioning symbols in the object that make it easy to tell if given binary x will work with given library version y.
      • 2013-02-22 05342, 2013

      • JonnyJD
        I really like http://semver.org . That is what many people use since ages, but a good documentation for that.
      • 2013-02-22 05358, 2013

      • kepstin-work
        but yes, semver is nice, for human-readable versions :)
      • 2013-02-22 05301, 2013

      • JonnyJD
        kepstin-work: the key is, that should already be encoded in the version number directly, I think
      • 2013-02-22 05327, 2013

      • kepstin-work runs the automake build of 0.2.2 to see how the soversion was set there
      • 2013-02-22 05303, 2013

      • kepstin-work
        huh, it installs libdiscid.so.0.2.1
      • 2013-02-22 05313, 2013

      • JonnyJD
        yes, that broke in between
      • 2013-02-22 05321, 2013

      • JonnyJD
        we even had libdiscid.so.1 already..
      • 2013-02-22 05336, 2013

      • JonnyJD
        and that is, because libtools versions are a mess to understand
      • 2013-02-22 05346, 2013

      • JonnyJD
        I guess. I wasn't involved back then.
      • 2013-02-22 05326, 2013

      • kepstin-work
        the usual way to do backwards incompatible api revisions now is to change the library name, like 'libdiscid-1.so.0' (along with libdiscid-1.pc)
      • 2013-02-22 05338, 2013

      • kepstin-work
        makes parallel installation possible
      • 2013-02-22 05317, 2013

      • pierreghz joined the channel
      • 2013-02-22 05319, 2013

      • kepstin-work
        and the soversion is mostly useful so that a binary will give a link error when run against a too-old library
      • 2013-02-22 05349, 2013

      • hawke_1
        kepstin-work: How is it otherwise impossible? Is there some reason that libfoo.so.1 and libfoo.so.2 cannot be installed in parallel?
      • 2013-02-22 05307, 2013

      • JonnyJD
        they can bei installed in parallel
      • 2013-02-22 05311, 2013

      • JonnyJD
        *be