https://pantsbuild.org/ logo
Powered by
#general
  • w

    witty-crayon-22786

    04/04/2016, 8:35 PM
    @melodic-kite-14017: in 
    [compile.zinc] args
    . i gave an example last week
  • w

    witty-crayon-22786

    04/04/2016, 8:36 PM
  • m

    melodic-kite-14017

    04/04/2016, 8:38 PM
    must have missed that with the other discussion, sorry for the distraction
  • a

    adorable-glass-89507

    04/04/2016, 9:06 PM
    @gentle-potato-32845: yup it was me
  • a

    adorable-glass-89507

    04/04/2016, 9:07 PM
    what are you thinking about?
  • g

    gentle-potato-32845

    04/04/2016, 9:38 PM
    how to contain the pants
  • g

    gentle-potato-32845

    04/04/2016, 9:38 PM
    @adorable-glass-89507: are you building services inside their containers, or doing a build-then-copy-to-runtime-container thing
  • g

    gentle-potato-32845

    04/04/2016, 9:39 PM
    if building inside, how are you dealing with build speed and caches and stuff?
  • h

    happy-kitchen-89482

    04/04/2016, 9:42 PM
    @jolly-chef-92794: can you take a peek at https://rbcommons.com/s/twitter/r/3643/ when you get a chance? I think you’re the only person who knows that code well, and I fear unintended consequences of that change.
  • j

    jolly-chef-92794

    04/04/2016, 9:42 PM
    @happy-kitchen-89482: Sure, just give me a few minutes. I’m finishing up some internal oncall work
  • a

    adorable-glass-89507

    04/04/2016, 9:58 PM
    @gentle-potato-32845: we're using a docker container to run pants commands, both to test and build services -- we volume map 
    ~/.cache/pants
    , 
    ~/.cache/artifact_cache
    , ivy cache dir (not the ivy lock dir) and the monorepo source to the jenkins slave host
  • a

    adorable-glass-89507

    04/04/2016, 9:58 PM
    and yes, we're building services inside the container
  • a

    adorable-glass-89507

    04/04/2016, 9:58 PM
    so we share a bunch of folders
  • a

    adorable-glass-89507

    04/04/2016, 9:59 PM
    and we'll either map the artifact folder via nfs or configure a remote cache as well
  • a

    adorable-glass-89507

    04/04/2016, 9:59 PM
    I think @jolly-doctor-21349 is going to try the nfs solution first
  • a

    adorable-glass-89507

    04/04/2016, 9:59 PM
    it's working pretty well now, except for race conditions unrelated to pants (we use docker compose to run our integration tests and it's been flaky)
  • m

    mammoth-pharmacist-54215

    04/05/2016, 12:27 AM
    @billowy-beach-68498 started at Twitter today, he will be working on the build team here.
  • m

    mammoth-pharmacist-54215

    04/05/2016, 12:27 AM
    Welcome
    🎉 2
  • b

    bored-art-40741

    04/05/2016, 12:34 AM
    Welcome!
  • m

    mammoth-pharmacist-54215

    04/05/2016, 12:35 AM
    I know a lot of tweeps have had problems in the past with twitter.com addresses getting blocked, what does @billowy-beach-68498 need to do to avoid those issues?
  • m

    mammoth-pharmacist-54215

    04/05/2016, 12:42 AM
    the only thing I remember doing was making sure that git config wasn’t setup to use my twitter.com email… not sure if anything else is needed.
  • r

    rough-minister-58256

    04/05/2016, 12:44 AM
    fairly sure you need to use a non-twitter.com email for rbcommons - iirc, its their mailer that has issues sending email on behalf of the domain.
  • m

    mammoth-pharmacist-54215

    04/05/2016, 12:44 AM
    that sounds right
  • r

    rough-minister-58256

    04/05/2016, 12:44 AM
    I have a @twopensource.com addy that’s just fwded to my @twitter.com
  • e

    enough-analyst-54434

    04/05/2016, 2:21 AM
    Its more than their mailer I think. I do not understand the DNS records and their use, but a second datapoint is slack itself which bans adding the twitter.com domain wholesale
  • w

    wide-energy-11069

    04/05/2016, 8:14 AM
    https://github.com/pantsbuild/pants/issues/3142 may be causing some funkiness in ci. at least the intellij ci tries to invoke pants but it hangs after git checkout different version of pants
  • b

    bored-art-40741

    04/05/2016, 1:36 PM
    @wide-energy-11069 That's known
  • b

    bored-art-40741

    04/05/2016, 1:36 PM
    I'm going to put out a patch to optionally clean up that lock today.
  • b

    bored-art-40741

    04/05/2016, 2:27 PM
  • b

    bored-art-40741

    04/05/2016, 2:56 PM
    This test seems to fail sporadically on Jenkins, not deterministically: http://jenkins.pantsbuild.org/job/pants_ci.integration_tests/168/SHARD=5/console
1...112011211122...2420Latest