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

    rough-minister-58256

    01/29/2016, 1:05 AM
    👍
  • r

    rough-minister-58256

    01/29/2016, 1:05 AM
    @bored-art-40741: btw I have a PR up for that bug now, will get an RB cut later today/early tomorrow.
  • b

    bored-art-40741

    01/29/2016, 1:06 AM
    Excellent, thanks!
  • a

    ancient-stone-50795

    01/29/2016, 1:56 AM
    @mammoth-pharmacist-54215 see the pants 1.0 release plan doc https://docs.google.com/document/d/1HDIalsQRv2Ds_uylPznX_VPmMdJib-1PdaGNVvF7fu4/edit which has a strategy that we hashed out.
  • a

    ancient-stone-50795

    01/29/2016, 1:56 AM
    Pants1.0ReleaseRoadmap.txt
    Pants1.0ReleaseRoadmap.txt
  • w

    witty-crayon-22786

    01/29/2016, 1:58 AM
    made a truth table that someone might find interesting:
  • w

    witty-crayon-22786

    01/29/2016, 1:58 AM
    Copy code
    [compile.zinc]
# Do not use zinc incremental builds in CI. This has the effect of executing only clean or
# cached builds for each target, which maximizes correctness and the number of artifacts in the
# remote cache.
#
#   ----------------------------------------------------------------------------------------------
#   | incremental | incremental_caching | effect                                                 |
#   ----------------------------------------------------------------------------------------------
#   | False       | *                   | Clean but slightly slower builds; everything cached    |
#   | True        | True                | Faster but potentially dirty builds; everything cached |
#   | True        | False               | Clean builds; only some things cached                  |
#   ----------------------------------------------------------------------------------------------
#
incremental: False
  • b

    bored-art-40741

    01/29/2016, 2:02 AM
    Oh nice
  • b

    bored-art-40741

    01/29/2016, 2:02 AM
    One nit: 
    This has the effect ...
  • b

    bored-art-40741

    01/29/2016, 2:03 AM
    The "this" there is reversed from my initial expectation, I had to read that sentence a few times
  • b

    bored-art-40741

    01/29/2016, 2:03 AM
    (The this can easily refer to the thing after "do not" as opposed to just the whole sentence beforehand)
  • b

    bored-art-40741

    01/29/2016, 2:04 AM
    How does this interact with growth of .pants.d?
  • b

    bored-art-40741

    01/29/2016, 2:04 AM
    And the overhead of doing big incremental compiles?
  • w

    witty-crayon-22786

    01/29/2016, 2:04 AM
    .pants.d
    growth should be the same, because during incremental builds we copy from the old version to the new version
  • b

    bored-art-40741

    01/29/2016, 2:04 AM
    I've certainly noticed that incremental compiles have a BIG lag where something is happening before zinc gets going
  • m

    mammoth-pharmacist-54215

    01/29/2016, 2:18 AM
    @ancient-stone-50795: thanks, I didn’t realize the plan here was finalized.
  • a

    ancient-stone-50795

    01/29/2016, 2:10 PM
    This silly prompt patch is proving difficult to test. It turns off the prompt when there is no TTY. I guess that makes sense. I think I will have to jam in yet another flag to force prompting anyway during tests.
  • j

    jolly-chef-92794

    01/29/2016, 3:17 PM
    Heh, that’s awkward.
  • f

    fancy-queen-20734

    01/29/2016, 4:18 PM
    it sounds like one of those things where if you write the tests last, you may not be testing anything.
  • h

    happy-kitchen-89482

    01/29/2016, 4:48 PM
    What would the test be testing in that case? That test would have to be really useful to justify adding a flag to production code for it. And even if it is useful, maybe a backdoor env var, outside the options system? In fact we could formalize the concept of PANTS_SELFTEST_ env vars if we end up needing more than one.
  • a

    ancient-stone-50795

    01/29/2016, 4:48 PM
    Yes, I am debating the utility of adding a flag to test for another flag. The code is very simple, I'm going to float it without a test.
  • h

    happy-kitchen-89482

    01/29/2016, 5:27 PM
    I’m very comfortable with that.
  • h

    happy-kitchen-89482

    01/29/2016, 5:30 PM
    Can everyone else also not see the comments here:
  • h

    happy-kitchen-89482

    01/29/2016, 5:30 PM
  • h

    happy-kitchen-89482

    01/29/2016, 5:31 PM
    I just see a weird face-pile down the side
  • h

    happy-kitchen-89482

    01/29/2016, 5:31 PM
  • a

    ancient-stone-50795

    01/29/2016, 5:59 PM
    @fancy-queen-20734: could you take another look at https://rbcommons.com/s/twitter/r/3387/
  • a

    ancient-stone-50795

    01/29/2016, 5:59 PM
    @happy-kitchen-89482: it looks normal to me.
  • h

    happy-kitchen-89482

    01/29/2016, 6:00 PM
    harumph
  • h

    happy-kitchen-89482

    01/29/2016, 6:00 PM
    on Chrome?
1...937938939...2420Latest