#musicbrainz

/

      • luks
        well, any tests are good
      • 2008-10-14 28836, 2008

      • luks
        just for making sure every field gets loaded is this fine
      • 2008-10-14 28813, 2008

      • luks
        LotR: thanks
      • 2008-10-14 28814, 2008

      • VxJasonxV
        I need to learn -D forwarding
      • 2008-10-14 28821, 2008

      • Muzzz
        "learn"?
      • 2008-10-14 28822, 2008

      • VxJasonxV
        I have Local and Remote forwarding down, but -D escapes me
      • 2008-10-14 28826, 2008

      • VxJasonxV
        learn and actually use
      • 2008-10-14 28836, 2008

      • Muzzz
        It's simply ssh -D <port> user [at host]
      • 2008-10-14 28847, 2008

      • VxJasonxV
        well, I know how to enter it
      • 2008-10-14 28848, 2008

      • Muzzz
        Then in whatever application whack in your SOCKS proxy as localhost:<port>
      • 2008-10-14 28851, 2008

      • Muzzz
        Simple \o/
      • 2008-10-14 28855, 2008

      • VxJasonxV
        figured.
      • 2008-10-14 28803, 2008

      • VxJasonxV
        Ok, so now I understand, thank you.
      • 2008-10-14 28809, 2008

      • VxJasonxV
        Now I just need to figure out if I have a need :P.
      • 2008-10-14 28819, 2008

      • Muzzz
        I find it useful for bypassing rapidshite IP download limiting
      • 2008-10-14 28823, 2008

      • VxJasonxV
        heh
      • 2008-10-14 28832, 2008

      • Muzzz
        Just keep tunnelling via a different host to get multipart RARs or whatever heh
      • 2008-10-14 28841, 2008

      • VxJasonxV sighs @ usenet for the web
      • 2008-10-14 28855, 2008

      • Muzzz
        I torrented the best thing ever the other day
      • 2008-10-14 28807, 2008

      • VxJasonxV
        I'm thankful that I don't deal with that nonsense (Rapidshare, MegaUpload, etc.) anymore.
      • 2008-10-14 28817, 2008

      • VxJasonxV
        I have my own webserver/server and do whatever I Want with it \o/
      • 2008-10-14 28822, 2008

      • VxJasonxV
        Define: The Best Thing Ever
      • 2008-10-14 28825, 2008

      • Muzzz
        40 folders or so, each containing 1-5 ZIP files. Each ZIP file containined a RAR part, which contained the actual VST software
      • 2008-10-14 28831, 2008

      • VxJasonxV
        O_O
      • 2008-10-14 28833, 2008

      • VxJasonxV
        FFS
      • 2008-10-14 28859, 2008

      • Muzzz
        This was a torrent, why you have multipart RARs in a torrent is stupid, why you have each RAR part zipped is stupid too
      • 2008-10-14 28801, 2008

      • VxJasonxV
        they really needed to compress those rar's 10K more didn't they?
      • 2008-10-14 28822, 2008

      • VxJasonxV
        actually, multipart (split files in general) are better for torrenting than one huge file
      • 2008-10-14 28829, 2008

      • VxJasonxV
        ok that's not true
      • 2008-10-14 28833, 2008

      • VxJasonxV
        they /can/ be better
      • 2008-10-14 28840, 2008

      • luks
        hm, why?
      • 2008-10-14 28841, 2008

      • Muzzz
        Not always when you consider the files the archives contained
      • 2008-10-14 28853, 2008

      • VxJasonxV
        because there's a greater chance of chunks being identical
      • 2008-10-14 28859, 2008

      • VxJasonxV
        maybe, I could be talking out of my ass
      • 2008-10-14 28806, 2008

      • Muzzz
        And even then I'm not convinced, torrenting is downloading in parts as it is so huge files are fine
      • 2008-10-14 28811, 2008

      • Muzzz
        O.o
      • 2008-10-14 28812, 2008

      • luks
        I don't think a hude compressed file will have identical chunks
      • 2008-10-14 28813, 2008

      • LotR
        how can it be better to uselessly stick the files into another file?
      • 2008-10-14 28815, 2008

      • luks
        huge
      • 2008-10-14 28827, 2008

      • VxJasonxV
        I don't mean the Zipped rars are a good idea
      • 2008-10-14 28846, 2008

      • VxJasonxV
        my train of thought is always that if you could torrent one iso, or a 30 rar part span of the iso, it's better to go with the rar parts
      • 2008-10-14 28849, 2008

      • LotR
        I think you *are* talking out of your arse
      • 2008-10-14 28851, 2008

      • VxJasonxV
        again, because of chunking and all that
      • 2008-10-14 28851, 2008

      • Muzzz
        VxJasonxV: woo, so you potentially save downloaded a tiny part, but then spend an age unarchiving it all
      • 2008-10-14 28853, 2008

      • VxJasonxV
        I probably am :D
      • 2008-10-14 28824, 2008

      • Muzzz
        People should man the fuck up and not multipart RAR files intended for torrens
      • 2008-10-14 28829, 2008

      • luks
        well, iso is a special case as it's completely uncompressed
      • 2008-10-14 28841, 2008

      • luks
        what most people download on torrent are heavily compressed files :)
      • 2008-10-14 28843, 2008

      • Muzzz
        luks, in which case a single archive would be fine
      • 2008-10-14 28848, 2008

      • luks
        right
      • 2008-10-14 28813, 2008

      • VxJasonxV shrugs
      • 2008-10-14 28815, 2008

      • VxJasonxV
        k, I'm just crazy
      • 2008-10-14 28822, 2008

      • Muzzz
        Yes, yes you are :P
      • 2008-10-14 28826, 2008

      • LotR
        so how much does an iso compress when you zip/rar it?
      • 2008-10-14 28833, 2008

      • luks
        depends on the contents
      • 2008-10-14 28839, 2008

      • LotR
        1% 5% 10%?
      • 2008-10-14 28851, 2008

      • luks
        an iso of zeroed files compresses pretty well :)
      • 2008-10-14 28842, 2008

      • LotR
        well, I would hope there'd be useful content in my ISO
      • 2008-10-14 28800, 2008

      • FauxFaux
        LotR: A full iso is basically a concatenation of the files, so, basically, not at all, assuming the contents are already compressed.
      • 2008-10-14 28843, 2008

      • LotR
        FauxFaux: right, and the usual content on an iso isn't *that* compressable, right?
      • 2008-10-14 28815, 2008

      • FauxFaux
        YOUR QUESTION MAKES NO SENSE
      • 2008-10-14 28818, 2008

      • FauxFaux
        Ahem.
      • 2008-10-14 28854, 2008

      • LotR
        what, like there isn't a most common type of content for torrented ISOs?
      • 2008-10-14 28811, 2008

      • LotR
        I mean linux distributions of course :)
      • 2008-10-14 28820, 2008

      • Muzzz shakes his head
      • 2008-10-14 28822, 2008

      • FauxFaux
        Well, software occasionally has hillariously poor ocmpression, unless it's a game or a linux distro.
      • 2008-10-14 28801, 2008

      • FauxFaux
        Having said that, NFSMW has about 900mb of 24bit 9billionhz wav on that compresses quite well, 'cos they're retards.
      • 2008-10-14 28826, 2008

      • LotR
        NFSMW?
      • 2008-10-14 28840, 2008

      • luks
        24bit audio in a game?
      • 2008-10-14 28843, 2008

      • FauxFaux
        It's a game.
      • 2008-10-14 28853, 2008

      • FauxFaux
        For the music, no less.
      • 2008-10-14 28809, 2008

      • Muzzz
        Need for Speed Most Wanted
      • 2008-10-14 28820, 2008

      • FauxFaux
        I'm pretty sure I mean that one, not at my machine right now.
      • 2008-10-14 28820, 2008

      • Muzzz
        Wiggatastic
      • 2008-10-14 28849, 2008

      • hawke_ joined the channel
      • 2008-10-14 28827, 2008

      • aCiD2
        "TODO This should be a trigger if we ever get a real DB" lol? I wonder how old that comment is
      • 2008-10-14 28835, 2008

      • FauxFaux
        Hee hee.
      • 2008-10-14 28843, 2008

      • FauxFaux
        svn praise it!
      • 2008-10-14 28846, 2008

      • aCiD2
        :P
      • 2008-10-14 28802, 2008

      • Muzzz
        *svn blame
      • 2008-10-14 28810, 2008

      • Muzzz
        Blaaaame
      • 2008-10-14 28834, 2008

      • LotR
        you mean git blame. aCiD2 wisely shuns svn :)
      • 2008-10-14 28840, 2008

      • MClemo
        svn is totally unusable!!!!!!11
      • 2008-10-14 28832, 2008

      • Muzzz has cravigns for HOT BEEF
      • 2008-10-14 28832, 2008

      • aCiD2
        :P
      • 2008-10-14 28839, 2008

      • aCiD2
        you can hot my beef
      • 2008-10-14 28847, 2008

      • Muzzz
        omnomnomnom
      • 2008-10-14 28812, 2008

      • Muzzz
        But if I log out of WoW to go cook beef I'll lose my Voidwalker :(
      • 2008-10-14 28817, 2008

      • luks
        aCiD2: I bet it's from the MB-on-MySQL time :)
      • 2008-10-14 28856, 2008

      • aCiD2
        :)
      • 2008-10-14 28801, 2008

      • aCiD2
        didn't even there was that!
      • 2008-10-14 28843, 2008

      • Tykling joined the channel
      • 2008-10-14 28803, 2008

      • ruaok joined the channel
      • 2008-10-14 28804, 2008

      • snarlydwarf joined the channel
      • 2008-10-14 28815, 2008

      • Sargun has quit
      • 2008-10-14 28838, 2008

      • aCiD2
        luks, ruaok: Is MB::S::Artist->Insert not meant to be called directly or something? It keeps throwing "Do called without BeginTransaction or AutoCommit" for me
      • 2008-10-14 28858, 2008

      • ruaok
        you need to open a transaction before calling it.
      • 2008-10-14 28802, 2008

      • FauxFaux
        Maybe you should begintransactionfirst. ¬_¬
      • 2008-10-14 28810, 2008

      • ruaok
        the idea is that you never mod the DB without being in a transaction.
      • 2008-10-14 28827, 2008

      • aCiD2
        ruaok: right, but why should I have to open the transaction before calling Insert, shouldn't Insert handle that?
      • 2008-10-14 28845, 2008

      • aCiD2
        Same with SearchEngine->AddRefs
      • 2008-10-14 28858, 2008

      • aCiD2
        FauxFaux: well duh, but I'm just wondering why Insert doesn't do that itself
      • 2008-10-14 28808, 2008

      • ruaok
        no. even insert has the same assumption.
      • 2008-10-14 28828, 2008

      • ruaok
        the modbot opens transactions and commits them if the insert/artist-insert go well.
      • 2008-10-14 28830, 2008

      • warp
        because the transaction might consist of multiple inserts or updates?
      • 2008-10-14 28838, 2008

      • aCiD2
        ok, I see
      • 2008-10-14 28839, 2008

      • warp is just guessing :)
      • 2008-10-14 28844, 2008

      • FauxFaux
        I expect insert simply doesn't error check for you?
      • 2008-10-14 28853, 2008

      • aCiD2
        so the client is responsible for beginning the transaction, gotcha
      • 2008-10-14 28856, 2008

      • FauxFaux
        Oh, or, maybe.
      • 2008-10-14 28820, 2008

      • aCiD2
        FauxFaux: error checking what? It checks it has enough info to do the commit, but doesn't check it actually commited
      • 2008-10-14 28828, 2008

      • ruaok
        insert is nothing but error checkls. :)
      • 2008-10-14 28834, 2008

      • warp
        haha
      • 2008-10-14 28800, 2008

      • ruaok
        but it assumes that a transaction is open
      • 2008-10-14 28802, 2008

      • aCiD2
        ruaok, I think I found a way to do testing as well with DBD::Mock
      • 2008-10-14 28809, 2008

      • aCiD2
        had a play with it while I was waiting for a lecture :)
      • 2008-10-14 28811, 2008

      • ruaok
        ohhh.
      • 2008-10-14 28820, 2008

      • aCiD2
        lemme show what I have
      • 2008-10-14 28828, 2008

      • luks
        aCiD2, ruaok: btw, speaking of DB code... can you please look at http://bugs.musicbrainz.org/browser/mb_server/bra…
      • 2008-10-14 28857, 2008

      • ruaok
        aCiD2, luks: I'm having to queue your requests for an hour.
      • 2008-10-14 28803, 2008

      • ruaok
        I'm prepping for the board meeting.
      • 2008-10-14 28812, 2008

      • aCiD2
        :)
      • 2008-10-14 28819, 2008

      • luks
        it's more or less a port of my python ngs code, but I think new code we write should be something like this
      • 2008-10-14 28827, 2008

      • aCiD2
        well when you're free, my attempt at testing is at: http://rafb.net/p/MfBPUJ45.html
      • 2008-10-14 28837, 2008

      • aCiD2
        (started testing methods)
      • 2008-10-14 28811, 2008

      • aCiD2
        luks: looking now
      • 2008-10-14 28817, 2008

      • aCiD2
        looks quite similar to what we do atm, what are the major differences?
      • 2008-10-14 28831, 2008

      • luks
        there is no artist->insert
      • 2008-10-14 28834, 2008

      • aCiD2
        my changes atm have taken stuff like "Insert" "SelectArtistsByName" out of the object level and into the package level
      • 2008-10-14 28844, 2008

      • luks
        the code for handling DB and for model are separate
      • 2008-10-14 28856, 2008

      • aCiD2
        ie, $artist->Insert becomes MB::S::Artist->Insert, in my branch
      • 2008-10-14 28804, 2008

      • aCiD2
        is that what you mean?
      • 2008-10-14 28810, 2008

      • luks
        yes
      • 2008-10-14 28821, 2008

      • aCiD2
        yea, that's def needed, only just started getting round to doing the changes
      • 2008-10-14 28825, 2008

      • luks
        well, except that I'd prefer it to be in a separate package
      • 2008-10-14 28836, 2008

      • aCiD2
        Like MB::S::Model::Artist? :P
      • 2008-10-14 28840, 2008

      • luks
        and $artist should not know about $dbh
      • 2008-10-14 28844, 2008

      • aCiD2
        right
      • 2008-10-14 28827, 2008

      • aCiD2
        well what I'm doing atm (only started this today, though it's been on my mind a while) is making all DB calls package level, and they accept $dbh... should I go the route of introducing a context?
      • 2008-10-14 28833, 2008

      • aCiD2
        or could we not just use MusicBrainz.pm for that
      • 2008-10-14 28803, 2008

      • luks
        they need more than $dbh
      • 2008-10-14 28803, 2008

      • aCiD2
        (doing $dbh atm because it lets me switch between DBD::Mock and DBD::Pg)
      • 2008-10-14 28811, 2008

      • aCiD2
        mmm? like what?
      • 2008-10-14 28820, 2008

      • luks
        they might need $dbh to the raw DB
      • 2008-10-14 28829, 2008

      • aCiD2
        ah, true
      • 2008-10-14 28833, 2008

      • luks
        cache shouldn't be a global variable, so alsosomething liek $cache
      • 2008-10-14 28851, 2008

      • luks
        which would also help testing a lot
      • 2008-10-14 28853, 2008

      • aCiD2
        ok
      • 2008-10-14 28830, 2008

      • aCiD2
        glad I'm thinking in the same direction as you (for once :P)
      • 2008-10-14 28842, 2008

      • luks
        btw, what is the roadmap for the TT branch you have set with ruaok?