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

    happy-kitchen-89482

    08/20/2015, 1:59 AM
    Will put you on the code review
  • r

    rough-minister-58256

    08/20/2015, 1:59 AM
    thx
  • h

    happy-kitchen-89482

    08/20/2015, 2:13 AM
    tomorrow though
  • r

    rough-minister-58256

    08/20/2015, 2:13 AM
    all good
  • h

    happy-kitchen-89482

    08/20/2015, 2:14 AM
    right now the test is failing because after my change we interpret 
    --jvm-options=
    as 
    --options
    in scope 
    jvm
    (as we should), but we also have options on various tasks called 
    --jvm-options
    .
  • r

    rough-minister-58256

    08/20/2015, 2:14 AM
    heh
  • h

    happy-kitchen-89482

    08/20/2015, 2:14 AM
    I'm very unclear on why this wasn't always failing
  • h

    happy-kitchen-89482

    08/20/2015, 2:14 AM
    my change shouldn't have made a difference
  • h

    happy-kitchen-89482

    08/20/2015, 2:15 AM
    I assume we only got away with this by luck or a bug masking the problem
  • r

    rough-minister-58256

    08/20/2015, 2:15 AM
    ah the dark arts of pants option management
  • h

    happy-kitchen-89482

    08/20/2015, 2:15 AM
    In this case we can always specify the fully-qualified option 
    --goal-task-jvm-options
  • h

    happy-kitchen-89482

    08/20/2015, 2:15 AM
    so this only affects users savvy enough to know about the shortcut options
  • h

    happy-kitchen-89482

    08/20/2015, 2:16 AM
    but still
  • r

    rough-minister-58256

    08/20/2015, 2:16 AM
    heh
  • r

    rough-minister-58256

    08/20/2015, 2:24 AM
    hmm, does 
    BackendConfigurationError: Failed to load the internal_backend.sitegen.register backend: No module named bs4
    ring a bell for anyone?
  • r

    rough-minister-58256

    08/20/2015, 3:02 AM
    ah, was an intentional breakage that was supposed to be caught by tests but a change moved it outside of the assertRaises scope; no wonder it seemed unrelated
  • m

    mammoth-pharmacist-54215

    08/20/2015, 5:48 PM
    couple of quick questions for people
  • m

    mammoth-pharmacist-54215

    08/20/2015, 5:49 PM
    1. Is there any reason someone would object to changing the default repl for python to ipython?
  • m

    mammoth-pharmacist-54215

    08/20/2015, 5:50 PM
    2. Does anyone have a problem with adding a —interpreter flag to repl that takes the interpreter to use?
  • r

    rough-minister-58256

    08/20/2015, 5:53 PM
    I would strongly object to #1 - vastly prefer the simple old python repl.
  • r

    rough-minister-58256

    08/20/2015, 5:53 PM
    also wouldn’t that mean bundling ipython with pants?
  • b

    bored-art-40741

    08/20/2015, 6:00 PM
    I almost long for the days when I could read every pants CR. We are moving at breakneck development speed right now
  • r

    rough-minister-58256

    08/20/2015, 6:01 PM
    heh true
  • r

    rough-minister-58256

    08/20/2015, 6:01 PM
    I remember when pants was still in our (previous) monorepo
  • r

    rough-minister-58256

    08/20/2015, 6:01 PM
    how far it has come
  • r

    rough-minister-58256

    08/20/2015, 6:02 PM
    pretty soon we’ll be hosting PantsCon
  • b

    bored-art-40741

    08/20/2015, 6:03 PM
    Given the growth over the last several summits we almost are already
  • b

    bored-art-40741

    08/20/2015, 6:08 PM
    Hey @ancient-stone-50795 has anything ever come of the third party jvm product refactor?
  • h

    happy-kitchen-89482

    08/20/2015, 6:25 PM
    @mammoth-pharmacist-54215: I'd also object to the repl being ipython by default
  • h

    happy-kitchen-89482

    08/20/2015, 6:29 PM
    Not sure I understand your --interpreter flag proposal
1...197198199...2420Latest