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

    happy-kitchen-89482

    10/22/2019, 5:36 PM
    We mark all our tests as ITs
  • h

    happy-kitchen-89482

    10/22/2019, 5:36 PM
    and in fact I'm running tests in the toolchain repo, not even pants's own tests
    h
    • 2
    • 9
  • h

    hundreds-breakfast-49010

    10/22/2019, 6:48 PM
    right now we don't have a good way to send a python enum (or our wrapper around a python enum) to/from rust, right?
    a
    • 2
    • 1
  • h

    hundreds-breakfast-49010

    10/22/2019, 6:48 PM
    I'm not sure if this is worth building, but it is a limitation
  • h

    hundreds-breakfast-49010

    10/22/2019, 7:36 PM
    updated https://github.com/pantsbuild/pants/pull/8495 (the PR with my proposal for subprocess execution)
    🔥 1
  • a

    average-vr-56795

    10/23/2019, 12:06 AM
    Heads up, I moved rust tests around https://github.com/pantsbuild/pants/pull/8512 - slightly disruptive, but hopefully not too bad to rebase past. I also cherry-picked this onto 1.22.x to make future cherry-picks easier.
    👌 1
    e
    • 2
    • 5
  • h

    hundreds-breakfast-49010

    10/23/2019, 9:15 PM
    are the twitter folks still doing meetings? there's surprisingly little chatter today
  • a

    aloof-angle-91616

    10/23/2019, 9:19 PM
    lots of them!!! london folks only come here so often
  • a

    aloof-angle-91616

    10/23/2019, 9:20 PM
    we are talking about the NailgunProcessMap PR right now and i mentioned your interactive subprocess PR yesterday
  • h

    hundreds-breakfast-49010

    10/23/2019, 9:32 PM
    oh they're already in town, cool
  • h

    hundreds-breakfast-49010

    10/23/2019, 9:33 PM
    I just noticed some test failures on that PR that were my fault, and added a fixing commit to those
  • h

    hundreds-breakfast-49010

    10/23/2019, 9:34 PM
    but yeah it could definitely use some eyes on it when people have a chance to look
  • w

    witty-crayon-22786

    10/24/2019, 11:50 PM
    🎊 2
    💵 2
    🙏 2
  • a

    aloof-angle-91616

    10/25/2019, 12:20 AM
    it was so incredible to be able to have the discussions we did today in just one room, i'm so so glad to be part of this community
    🎉 2
  • a

    aloof-angle-91616

    10/25/2019, 6:12 AM
    https://github.com/pantsbuild/pants/pull/8542 allowing 
    @union
    base classes to define 
    @abstractmethod
    and 
    @abstractproperty
    methods, and allowing `@rule`s to use an 
    @union
    class as an input selector to generate a version of the rule for each 
    UnionRule
    variant
  • h

    hundreds-breakfast-49010

    10/25/2019, 7:39 PM
    do we have the ability to implmeent a subsystem as a plugin?
  • w

    witty-crayon-22786

    10/25/2019, 7:42 PM
    Yep
  • w

    witty-crayon-22786

    10/25/2019, 7:43 PM
    register.py has a place to declare them, but Tasks can also transitively depend on subsystems
  • w

    witty-crayon-22786

    10/25/2019, 7:44 PM
    And finally, 
    optionable_rule
    also registers the Subsystem
  • h

    hundreds-breakfast-49010

    10/25/2019, 7:56 PM
    there are files called 
    report.py
    , 
    reporting.py
    , 
    reporter.py
  • h

    hundreds-breakfast-49010

    10/25/2019, 7:56 PM
    all right next to each other in the 
    reporting
    dir
  • h

    hundreds-breakfast-49010

    10/25/2019, 7:59 PM
    anyway, is there any person who's particularly familiar with the reporting infrastructure I can ask questions of? particularly as it pertains to the v2 engine, I'm not sure if reporting is even currently happening in a v2 context
  • w

    witty-crayon-22786

    10/25/2019, 8:12 PM
    it isn't, afaik.
  • w

    witty-crayon-22786

    10/25/2019, 8:12 PM
    ...oh. wait.
  • w

    witty-crayon-22786

    10/25/2019, 8:12 PM
    we create workunits if that is enabled.
  • w

    witty-crayon-22786

    10/25/2019, 8:13 PM
    but it won't be streaming.
  • h

    happy-kitchen-89482

    10/25/2019, 9:44 PM
    That reporting is pretty v1-centric
  • h

    happy-kitchen-89482

    10/25/2019, 9:44 PM
    I think you can ignore it for now
  • h

    happy-kitchen-89482

    10/25/2019, 9:47 PM
    Can a rule declare that it returns an 
    Optional[Foo]
    ? Do we expect that to work? We have a rule that declared that it returned just 
    Foo
    , but we were in practice also returning None sometimes. It all seemed to work fine, but I figure we might have been skating on thin ice.
  • w

    witty-crayon-22786

    10/25/2019, 9:53 PM
    i believe that you would get an exception if it returned None.
1...667668669...906Latest