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

    agreeable-sugar-43925

    11/03/2016, 6:49 AM
    For bootstrapping pants, can I run it using Cygwin or something like that?
  • a

    agreeable-sugar-43925

    11/03/2016, 6:54 AM
    Didn't really work, I'm going to try using the windows bash command.
  • w

    wide-energy-11069

    11/03/2016, 6:57 AM
    @agreeable-sugar-43925 linux and osx are recommended. windows is a uncharted territory for pants.
  • l

    limited-country-20626

    11/03/2016, 7:01 AM
    I think unsupported, full stop.
  • l

    limited-country-20626

    11/03/2016, 7:01 AM
    but, you can use virtualbox
  • a

    agreeable-sugar-43925

    11/03/2016, 7:02 AM
    I'll just dual boot linux
  • a

    agreeable-sugar-43925

    11/03/2016, 7:02 AM
    I've been wanting to do it for a while
  • l

    limited-country-20626

    11/03/2016, 7:02 AM
    that would also work : )
  • l

    limited-country-20626

    11/03/2016, 7:05 AM
    if you have any trouble with it, keep virtualbox in mind. It's free, and a bit smaller of a commitment. There should be negligible perf penalty if you are mostly working on Pants itself.
  • l

    limited-country-20626

    11/03/2016, 7:06 AM
    but I dual boot my OSX box personally
  • w

    witty-crayon-22786

    11/03/2016, 5:32 PM
    @limited-country-20626 : saw tweets that might have indicated that you'd taken a crack at a survey...?
  • l

    limited-country-20626

    11/03/2016, 5:33 PM
    Yeah, I need to just go ahead and open it up to general contribution. I will do so today.
  • w

    witty-crayon-22786

    11/03/2016, 5:34 PM
    huzzah
  • w

    witty-crayon-22786

    11/03/2016, 5:34 PM
    thanks.
  • w

    witty-crayon-22786

    11/03/2016, 5:50 PM
    hackweek idea: port "list-owners" into the core, and have 
    ./pants test tests/python/pants_test/engine/test_isolated_process.py
    do the right thing
  • w

    witty-crayon-22786

    11/03/2016, 6:02 PM
    @fancy-queen-20734 : thinking of introducing zero-arg intrinsics
  • f

    fancy-queen-20734

    11/03/2016, 6:03 PM
    hm. ie like SelectLiteral, but a rule?
  • w

    witty-crayon-22786

    11/03/2016, 6:03 PM
    the difference with SelectLiteral is that it's exposed... the caller knows it exists
  • w

    witty-crayon-22786

    11/03/2016, 6:03 PM
    it would be a replacement for SelectLiteral in most/all cases
  • w

    witty-crayon-22786

    11/03/2016, 6:04 PM
    rather than 
    SelectLiteral(value, Type)
    , you'd just 
    Select(Type)
    , because the literal is available
  • w

    witty-crayon-22786

    11/03/2016, 6:04 PM
    "somewhere"
  • f

    fancy-queen-20734

    11/03/2016, 6:04 PM
    Yeah. that makes sense
  • w

    witty-crayon-22786

    11/03/2016, 6:04 PM
    (probably overloading "intrinsic" at this point)
  • f

    fancy-queen-20734

    11/03/2016, 6:04 PM
    then providing those becomes intrinsic to the engine
  • w

    witty-crayon-22786

    11/03/2016, 6:04 PM
    yea.
  • f

    fancy-queen-20734

    11/03/2016, 6:05 PM
    We’d probably need a extension mechanism for intrinsics then
  • f

    fancy-queen-20734

    11/03/2016, 6:05 PM
    or, maybe have a intrinsic provider that essentially provides services
  • w

    witty-crayon-22786

    11/03/2016, 6:08 PM
    @fancy-queen-20734 : so , in this branch, intrinsics are back to just being functions
  • w

    witty-crayon-22786

    11/03/2016, 6:09 PM
    and the key differentiator is that they're 1 arg, and statically resolvable
  • w

    witty-crayon-22786

    11/03/2016, 6:10 PM
    i've gotten rid of all of the intrinsic node types
1...153715381539...2420Latest