https://pantsbuild.org/ logo
Powered by
#announce
  • h

    happy-kitchen-89482

    03/06/2017, 4:05 AM
    It happened locally, yes
  • h

    happy-kitchen-89482

    03/06/2017, 4:05 AM
    nuking ivy cache appeared to resolve it once, and then it recurred
  • r

    rough-minister-58256

    03/11/2017, 12:59 AM
  • r

    rough-minister-58256

    03/11/2017, 12:59 AM
    wonder if something like this would be useful for landing PRs to master?
  • r

    rough-minister-58256

    03/11/2017, 1:00 AM
    looks pretty easy to setup
  • r

    rough-minister-58256

    03/11/2017, 1:02 AM
    could potentially modify and run our own instance of it that could do other customizations like copying + pasting the PR body as the commit body, etc
  • w

    wide-energy-11069

    03/11/2017, 1:06 AM
    don’t we already ci for PR before merging to master? maybe I’m failing to see what more it offers
  • w

    wide-energy-11069

    03/11/2017, 1:07 AM
    also sometimes its cache issue, not sure if homu can help there either
  • f

    full-kilobyte-93429

    03/11/2017, 1:19 AM
    that's a SQ that takes care of testing (retry) && commit (messages, approvals)
  • f

    full-kilobyte-93429

    03/11/2017, 1:19 AM
    things we are doing manually
  • w

    wide-energy-11069

    03/11/2017, 1:22 AM
    ahh
  • w

    wide-energy-11069

    03/11/2017, 1:23 AM
    Note that Homu can only test one pull request at a time
    deal breaker?
  • f

    full-kilobyte-93429

    03/11/2017, 1:23 AM
    for huge monorepo yes
  • f

    full-kilobyte-93429

    03/11/2017, 1:24 AM
    or someone needs to takes care of operational issues
  • w

    wide-energy-11069

    03/11/2017, 1:28 AM
    given pants’ ci size, it may hurt as well. It means it will be bounded by the longest shard (50 min). Given 16 hrs of operation per day, it can only test 19 PR commits...
  • f

    full-kilobyte-93429

    03/11/2017, 1:29 AM
    50min ouch
  • w

    wide-energy-11069

    03/11/2017, 1:30 AM
    yep contrib shard sometimes times out due to travis limit
  • h

    happy-kitchen-89482

    03/12/2017, 4:59 PM
    The contrib tests run with 
    —fast
    mode turned off, which will not be the case in the new pipeline
  • h

    happy-kitchen-89482

    03/12/2017, 4:59 PM
    So that should speed things up
  • h

    happy-kitchen-89482

    03/12/2017, 5:00 PM
    But also the integration tests in general are slow for reasons I’m not clear on, they seem to take a lot of time doing work that should be cached.
  • w

    wide-energy-11069

    03/17/2017, 5:39 PM
    i’m releasing today. anything in particular to pay attention? afaict, it would be 1.3.0.dev14?
  • r

    rough-minister-58256

    03/17/2017, 5:53 PM
    did you have a target timeframe for the release today? I’m hoping to get the v2-by-default patch in today, but may be a race depending on how things go.
  • w

    wide-energy-11069

    03/17/2017, 5:57 PM
    @rough-minister-58256 i can do it in the evening as well if needed
  • w

    witty-crayon-22786

    03/17/2017, 5:57 PM
    not sure whether that goes in/out of dev14 matters
  • w

    witty-crayon-22786

    03/17/2017, 5:58 PM
    ...er. yea, i guess it matters
  • w

    witty-crayon-22786

    03/17/2017, 5:58 PM
    in a "spirit of the 3-4 week stability period" sense
  • r

    rough-minister-58256

    03/17/2017, 5:58 PM
    yeh
  • r

    rough-minister-58256

    03/17/2017, 5:59 PM
    feels like once we eek into the >dev10’s we’re approaching the end of current dev release cycle
  • r

    rough-minister-58256

    03/17/2017, 6:00 PM
    lets see how it goes - don’t want to hold things up, but hoping it should be relatively quick.
  • w

    witty-crayon-22786

    03/17/2017, 6:00 PM
    yea, would be good to get it in. not sure whether anyone actually uses the dev releases, but maybe
1...363738...206Latest