#musicbrainz-devel

/

      • Mineo
        the depreciation stuff is probably because I've got a newer version of some library, yeah, but that doesn't cause the tests to fail at least :)
      • 2014-07-01 18247, 2014

      • ijabz1
        ianmcorvidae Could you finish off the search-server git deployment please, i.e create repos under metabrainz or github synced with bitbucket
      • 2014-07-01 18223, 2014

      • misterswag joined the channel
      • 2014-07-01 18209, 2014

      • ijabz joined the channel
      • 2014-07-01 18219, 2014

      • Gentlecat joined the channel
      • 2014-07-01 18244, 2014

      • nikki joined the channel
      • 2014-07-01 18219, 2014

      • bitmap
        jesus2099: yeah, I'd like to support gids for the link_type field eventually, although it's low priority right now
      • 2014-07-01 18204, 2014

      • bitmap
        I'm not aware of any list for the ids other than in the HTML
      • 2014-07-01 18205, 2014

      • demonimin joined the channel
      • 2014-07-01 18206, 2014

      • ocharles
        Mineo: afaik that code is up to date
      • 2014-07-01 18249, 2014

      • ocharles
        also, hi all!
      • 2014-07-01 18221, 2014

      • ianmcorvidae
        hello! how goes it in haskell-land?
      • 2014-07-01 18238, 2014

      • ocharles
        very well :)
      • 2014-07-01 18249, 2014

      • ocharles
        though spending more time in clojurescript land these days
      • 2014-07-01 18210, 2014

      • ianmcorvidae
        also neat, though :)
      • 2014-07-01 18212, 2014

      • ocharles
        We're down to one frontend developer, so we've decided that everyone should be a bit more full-stack
      • 2014-07-01 18228, 2014

      • ocharles
        Completely in love with NixOS for deployments too
      • 2014-07-01 18235, 2014

      • ocharles
        lots of exciting technology to play with :)
      • 2014-07-01 18241, 2014

      • ianmcorvidae
        makes sense. sounds a bit familiar for MB too, as you might imagine
      • 2014-07-01 18248, 2014

      • ocharles
        indeed
      • 2014-07-01 18213, 2014

      • ocharles
        that's sort of why I pushed for it so much. "I could sit here and wait for Kris to have time to write this, or I could just roll up my sleeves and see if I can help"
      • 2014-07-01 18221, 2014

      • ocharles
        nikki informs me things are ticking along well over here
      • 2014-07-01 18239, 2014

      • ianmcorvidae
        as well as ever, better than sometimes, I'd say
      • 2014-07-01 18255, 2014

      • ruaok waves at ocharles
      • 2014-07-01 18204, 2014

      • ocharles
        hey ruaok
      • 2014-07-01 18208, 2014

      • ruaok
      • 2014-07-01 18210, 2014

      • ocharles
        ianmcorvidae: that's awesome!
      • 2014-07-01 18213, 2014

      • ocharles
        ruaok: yea, it's kick ass
      • 2014-07-01 18217, 2014

      • ruaok
        Gentlecat is knocking it out of the park!
      • 2014-07-01 18221, 2014

      • ocharles
        very impressed with the result of that
      • 2014-07-01 18225, 2014

      • warp agrees
      • 2014-07-01 18230, 2014

      • warp waves _o/
      • 2014-07-01 18231, 2014

      • ruaok
        (to use a shitty american metaphor)
      • 2014-07-01 18235, 2014

      • ruaok
        hi warp
      • 2014-07-01 18200, 2014

      • ocharles
        i get to hear shitty american metaphor's all day. our owner is very west coast :)
      • 2014-07-01 18211, 2014

      • ruaok
        ouch.
      • 2014-07-01 18213, 2014

      • ocharles
        "dude, this shit is fucking awesome. you're killing it man" is the usual pep talk
      • 2014-07-01 18219, 2014

      • ruaok
        I'm really trying to change.
      • 2014-07-01 18224, 2014

      • ocharles
        i don't mind it, it's a bit better than than a british "yea. that looks good"
      • 2014-07-01 18228, 2014

      • ruaok
        ha
      • 2014-07-01 18203, 2014

      • ianmcorvidae
        sounds decent enough as a pep talk, personally always found "killing it" to be a weirdly violent compliment, but :P
      • 2014-07-01 18213, 2014

      • ocharles
        hence my tweet :)
      • 2014-07-01 18210, 2014

      • ocharles
        but yea, we're coming along well. took a lot more time than everyone would have liked to find a process that fits us, and to make sure everyone is moving in the same direction
      • 2014-07-01 18211, 2014

      • ruaok has purposefully not looked at twitter today
      • 2014-07-01 18232, 2014

      • ocharles
        so while we would have hoped to launching *something* within 8 months of starting (!) we're at least very close to that now
      • 2014-07-01 18240, 2014

      • ocharles
        in fact, first customer wants to be live on jul 23
      • 2014-07-01 18251, 2014

      • ianmcorvidae
        heh, having trouble nailing down process, that sounds entirely unfamiliar </sarcasm> ;)
      • 2014-07-01 18256, 2014

      • ocharles
        :P
      • 2014-07-01 18215, 2014

      • ocharles
        for too long people wanted a digital process. I finally said it wasn't working, and that we should get a stack of cards and stick them on a whiteboard
      • 2014-07-01 18228, 2014

      • nikki
        stack of cards \o.
      • 2014-07-01 18230, 2014

      • nikki
        \o/ too
      • 2014-07-01 18242, 2014

      • ianmcorvidae
        (slightly kidding; was at a talk at open source bridge about codereview and I could honestly say we do a really good job there)
      • 2014-07-01 18249, 2014

      • ianmcorvidae
        heh, where are those pictures from our chicago meeting? :P
      • 2014-07-01 18254, 2014

      • nikki
        somewhere...
      • 2014-07-01 18205, 2014

      • nikki
        I never did find a way to make a more readable version
      • 2014-07-01 18213, 2014

      • nikki
        last time I tried, I crashed my image editor :P
      • 2014-07-01 18217, 2014

      • ianmcorvidae
        haha
      • 2014-07-01 18225, 2014

      • ocharles
        the code reviews and the two week release cycle actually work remarkably well here
      • 2014-07-01 18242, 2014

      • ocharles
        we're currently trying for a more continuous release process, with an emphasis on "if you start it, you finish it"
      • 2014-07-01 18254, 2014

      • ianmcorvidae
        yeah
      • 2014-07-01 18208, 2014

      • ocharles
        we had a big problem of people getting to 80% and then it would never quite get finished. so the whole system feels shitty and is full of "oh yea, don't click that. we didn't build that bit"
      • 2014-07-01 18208, 2014

      • ruaok
        I tooks some pictures too. should I look for them?
      • 2014-07-01 18211, 2014

      • ianmcorvidae
        I dunno how well it'd work for a newer project, but it does seem to work well for MBS that's so established (overall)
      • 2014-07-01 18251, 2014

      • ianmcorvidae
        ruaok: nah, I was just reminded of our room full of cards all over the floor, when ocharles mentioned using stacks of cards :)
      • 2014-07-01 18208, 2014

      • ocharles
        btw, it seems like selenium is half useful now
      • 2014-07-01 18213, 2014

      • ruaok
        k, I probably have readaonble ones.
      • 2014-07-01 18224, 2014

      • ocharles
        we're not doing much with it, but it doesn't seem to be quite as painful as it was before
      • 2014-07-01 18231, 2014

      • ianmcorvidae
        that's good to hear
      • 2014-07-01 18238, 2014

      • nikki
        mine are readable, it's just that the images are huge and don't fit into one image
      • 2014-07-01 18259, 2014

      • nikki
        (huge both in pixels and in filesize)
      • 2014-07-01 18218, 2014

      • ocharles
        we're not doing much though :)
      • 2014-07-01 18237, 2014

      • nikki
        I might dig them out in the next couple of days and try with the gimp instead of this shitty osx thing, since I guess I probably won't have much internet access on the train
      • 2014-07-01 18208, 2014

      • ianmcorvidae
        ah, parens, feels like home :)
      • 2014-07-01 18213, 2014

      • ocharles
        ah dynamic typing, feels like hell :P
      • 2014-07-01 18228, 2014

      • ianmcorvidae
        not bad, though. still a decent amount of overlap with things like mechanize and qunit tests, but it is nice to be able to glue them together a bit better than you can with separate backend/JS tests
      • 2014-07-01 18236, 2014

      • ianmcorvidae
        haha
      • 2014-07-01 18200, 2014

      • ocharles
        ah yea, mechanize is nice. i forgot about that
      • 2014-07-01 18221, 2014

      • ocharles
        ok, gotta run - good catching up!
      • 2014-07-01 18222, 2014

      • ianmcorvidae
        indeed
      • 2014-07-01 18225, 2014

      • ianmcorvidae
        later! nice to talk
      • 2014-07-01 18258, 2014

      • ruaok
        OMFG. AOL paid an invoice.
      • 2014-07-01 18224, 2014

      • ruaok
        this time the threat of a cake didn't do much. Not sure if anyone is listening @aol
      • 2014-07-01 18236, 2014

      • ianmcorvidae
        haha
      • 2014-07-01 18248, 2014

      • ruaok
        well, if they pay all the invoices, thats another $15k in the bank.
      • 2014-07-01 18207, 2014

      • Gentlecat
        ianmcorvidae: have time for sandbox stuff?
      • 2014-07-01 18238, 2014

      • ianmcorvidae
        yeah, let's look into that
      • 2014-07-01 18249, 2014

      • ianmcorvidae
        I guess you'd like one for your PRs, basically?
      • 2014-07-01 18202, 2014

      • Gentlecat
        right
      • 2014-07-01 18235, 2014

      • CallerNo6
        I don't think AOL can be shamed.
      • 2014-07-01 18212, 2014

      • ianmcorvidae
        Gentlecat: alright, gentlecat-geordi.mbsandbox.org should be set up with a fastcgi socket at /home/gentlecat/geordi/fcgi.sock
      • 2014-07-01 18226, 2014

      • Gentlecat
        oke
      • 2014-07-01 18228, 2014

      • ruaok
        CallerNo6: I hadn't really tried very hard.
      • 2014-07-01 18230, 2014

      • ianmcorvidae
        (so, just a geordi clone right in your home, and something like the CB/reports.mbsandbox fcgi script)
      • 2014-07-01 18250, 2014

      • ianmcorvidae
        you should be able to create a DB for yourself fine, unless you'd like an extra DB user/role to use
      • 2014-07-01 18209, 2014

      • navap
        ruaok: I forgot to mention yesterday during the ssl talk: other-158
      • 2014-07-01 18210, 2014

      • mb-chat-logger
      • 2014-07-01 18214, 2014

      • navap
      • 2014-07-01 18256, 2014

      • navap
        There's also mbs-6228 where https on subdomains displays the main site like https://forums.musicbrainz.org/
      • 2014-07-01 18257, 2014

      • mb-chat-logger
      • 2014-07-01 18221, 2014

      • ijabz1 joined the channel
      • 2014-07-01 18249, 2014

      • Gentlecat
        ianmcorvidae: I get same errors with SQL again :(
      • 2014-07-01 18252, 2014

      • Trohoky joined the channel
      • 2014-07-01 18225, 2014

      • Gentlecat
        ls
      • 2014-07-01 18241, 2014

      • Gentlecat
        ouch, wrong window
      • 2014-07-01 18253, 2014

      • Mineo recommends plugins for irc clients that turn 'ls' into '/names' :)
      • 2014-07-01 18240, 2014

      • ianmcorvidae
        Gentlecat: looks like you changed the search path, works even after that?
      • 2014-07-01 18253, 2014

      • Gentlecat
        no, same error
      • 2014-07-01 18255, 2014

      • ianmcorvidae
        hm
      • 2014-07-01 18206, 2014

      • ianmcorvidae
        sudo -u gentlecat /usr/local/postgres/bin//psql -U gentlecat gentlecat_geordi < /home/gentlecat/geordi/geordi/sql/tables.sql ran fine for me, so the DB should be there now
      • 2014-07-01 18211, 2014

      • kepstin_ joined the channel
      • 2014-07-01 18232, 2014

      • Gentlecat
        yep, thanks
      • 2014-07-01 18251, 2014

      • Gentlecat
      • 2014-07-01 18211, 2014

      • ianmcorvidae
        hm
      • 2014-07-01 18215, 2014

      • ianmcorvidae
        old python version, maybe
      • 2014-07-01 18242, 2014

      • Gentlecat
        most likely, works fine locally
      • 2014-07-01 18220, 2014

      • ianmcorvidae
        ah, yeah: "Changed in version 2.7: Support for multiple context expressions."
      • 2014-07-01 18220, 2014

      • ianmcorvidae
        I guess we can just convert it to nested with statements instead, which should work in 2.5+
      • 2014-07-01 18239, 2014

      • Gentlecat
        can we update python?
      • 2014-07-01 18255, 2014

      • Gentlecat
        because I have some issues with 2.6 in CB too
      • 2014-07-01 18214, 2014

      • ianmcorvidae
        rika's running an old version of ubuntu, but possibly there's a PPA
      • 2014-07-01 18234, 2014

      • ianmcorvidae is adding one, we can see how that goes
      • 2014-07-01 18251, 2014

      • ianmcorvidae
        can be added to virtualenvs on creation, presumably
      • 2014-07-01 18219, 2014

      • ianmcorvidae
        (once it's installed, obviously)
      • 2014-07-01 18251, 2014

      • Mineo
        yeah, virtualenv has a -p switch to tell it which python binary to use
      • 2014-07-01 18213, 2014

      • ianmcorvidae
        okay, /usr/bin/python2.7 exists now
      • 2014-07-01 18219, 2014

      • ianmcorvidae
        default /usr/bin/python is still 2.6
      • 2014-07-01 18255, 2014

      • Gentlecat
      • 2014-07-01 18215, 2014

      • ianmcorvidae
        hm
      • 2014-07-01 18249, 2014

      • ianmcorvidae
        hm, might be the version of virtualenv? not sure
      • 2014-07-01 18237, 2014

      • Gentlecat
      • 2014-07-01 18256, 2014

      • ianmcorvidae
        have you tried the workaround in the comment on the answer?
      • 2014-07-01 18234, 2014

      • ianmcorvidae
        I'm not sure where the virtualenv.pyc file they're talking about is hiding
      • 2014-07-01 18252, 2014

      • Mineo
        there's one in /usr/lib/pymodules/python2.6
      • 2014-07-01 18219, 2014

      • ianmcorvidae
        duly removed the .pyc then
      • 2014-07-01 18242, 2014

      • ianmcorvidae
        actually I'll just kill all the .pyc in that folder, heh
      • 2014-07-01 18249, 2014

      • Gentlecat
        oh, I was trying to apply that crazy fix
      • 2014-07-01 18216, 2014

      • ianmcorvidae
        hm
      • 2014-07-01 18223, 2014

      • ianmcorvidae
        still not working with just the .pyc removal, so
      • 2014-07-01 18223, 2014

      • Gentlecat
        nope, same error
      • 2014-07-01 18255, 2014

      • ianmcorvidae
        doing the virtualenv chain thing works, it seems
      • 2014-07-01 18215, 2014

      • ianmcorvidae
        (create a 2.6 virtualenv with --no-site-packages, install virtualenv, use that virtualenv to create the py2.7 one)
      • 2014-07-01 18220, 2014

      • Gentlecat
        hmm, do I need to create another venv inside the first one?
      • 2014-07-01 18234, 2014

      • Gentlecat
        I'm stuck at step 3
      • 2014-07-01 18239, 2014

      • ianmcorvidae
        you can create the venvs in whatever folder you want
      • 2014-07-01 18253, 2014

      • ianmcorvidae
        my process was
      • 2014-07-01 18240, 2014

      • ianmcorvidae
        mkdir py7-venv; mkdir py6-venv; virtualenv --no-site-packages py6-venv; source py6-venv/bin/activate; easy_install virtualenv; virtualenv -p /usr/bin/python2.7 py7-venv
      • 2014-07-01 18259, 2014

      • ianmcorvidae
        and then you can source py7-venv/bin/activate and it appears to work
      • 2014-07-01 18254, 2014

      • Gentlecat
        looks like psycopg fails to install https://gist.github.com/tsukanov/fccd9c63497d0a18…
      • 2014-07-01 18232, 2014

      • ianmcorvidae
        oh, that needs the -dev package, at least it did on rocko
      • 2014-07-01 18234, 2014

      • ianmcorvidae
        wonder if that's in the ppa