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

    bored-art-40741

    10/07/2015, 6:11 PM
    Though I'm not sure if it's considered a bug
  • b

    bored-art-40741

    10/07/2015, 6:12 PM
    Certainly the inability to debug it warrants calling it bug
  • e

    enough-analyst-54434

    10/07/2015, 6:12 PM
    How do you know this? I reviewed and wrote some of that code and I'm not sure this is true!
  • e

    enough-analyst-54434

    10/07/2015, 6:12 PM
    It very well could be in pants use of the resolver!
  • b

    bored-art-40741

    10/07/2015, 6:13 PM
    I don't see how that's possible, but at the same time I'm a bit too swamped to get into anything pex related right now
  • e

    enough-analyst-54434

    10/07/2015, 6:13 PM
    Basically we pass in a cache
  • e

    enough-analyst-54434

    10/07/2015, 6:13 PM
    We may be misuing/miscoping the cache
  • s

    strong-bird-83478

    10/07/2015, 6:14 PM
    @enough-analyst-54434: so would it be worked-around by building the pex with "--disable-cache"?
  • e

    enough-analyst-54434

    10/07/2015, 6:14 PM
    No
  • e

    enough-analyst-54434

    10/07/2015, 6:14 PM
    So, again - pants python handling - as opposed to go or jvm or ...
  • e

    enough-analyst-54434

    10/07/2015, 6:14 PM
    Is confusing - don't draw coclusions too quickly
  • e

    enough-analyst-54434

    10/07/2015, 6:14 PM
    pants.pex - which you folks use - has 0 issues here
  • e

    enough-analyst-54434

    10/07/2015, 6:15 PM
    The backtrace is in the test.pytest command - which also happens to build an unzipped pex
  • e

    enough-analyst-54434

    10/07/2015, 6:15 PM
    its the latter that fails
  • e

    enough-analyst-54434

    10/07/2015, 6:16 PM
    you can't pass --disable-cache to the latter
  • e

    enough-analyst-54434

    10/07/2015, 6:16 PM
    Basically I'll request anyone with opions on all this actually work the issue
  • e

    enough-analyst-54434

    10/07/2015, 6:17 PM
    code/tests - no guesses
  • s

    strong-bird-83478

    10/07/2015, 6:17 PM
    got it
  • a

    ancient-stone-50795

    10/07/2015, 6:19 PM
    @enough-analyst-54434: That did not seem to fix the problem on my end. Details in a sec
  • a

    ancient-stone-50795

    10/07/2015, 6:22 PM
    Copy code
    14:16:56 01:00         [chroot]
               Waiting for background workers to finish.
14:16:59 01:03   [complete]
               FAILURE

Exception message: Could not satisfy all requirements for six>=1.4:
    six>=1.4(from: unittest2), six==1.9.0(from: pantsbuild.pants.testinfra==0.0.51->pantsbuild.pants==0.0.51)
  • e

    enough-analyst-54434

    10/07/2015, 6:22 PM
    That just means you didn't actually use the new sdists
  • e

    enough-analyst-54434

    10/07/2015, 6:22 PM
    Using a clearly different rev is best here
  • a

    ancient-stone-50795

    10/07/2015, 6:23 PM
    Ok, wait I see… let me see...
  • e

    enough-analyst-54434

    10/07/2015, 6:23 PM
    Last tie this came up I think some form of sha suffix
  • e

    enough-analyst-54434

    10/07/2015, 6:23 PM
    ==...-abc123 or ==...abc123
  • a

    ancient-stone-50795

    10/07/2015, 6:24 PM
    I’m going to change the requirements to:
    Copy code
    +-f /Users/zundel/Src/Pants/dist
+pantsbuild.pants.testinfra==0.0.51-square-20151002-02
  • e

    enough-analyst-54434

    10/07/2015, 6:24 PM
    ty
  • a

    ancient-stone-50795

    10/07/2015, 6:25 PM
    Yeah, that worked!
  • a

    ancient-stone-50795

    10/07/2015, 6:25 PM
    So we do an emergency 0.0.52 release I take it?
  • e

    enough-analyst-54434

    10/07/2015, 6:26 PM
    Yes, as soon as CI greens for https://rbcommons.com/s/twitter/r/2942/ I'll send out a version bump/CHANGELOG RB
1...510511512...2420Latest