#general
- b
bored-art-40741
04/01/2016, 3:29 PM@ancient-stone-50795: I'm trying to think of a way to test the blocking feature of the lock that isn't prone to some kind of extreme CI slowdown, and I'm not sure if it's possible - b
bored-art-40741
04/01/2016, 3:29 PMHowever I do think I came up with a way that's less likely to be affected by that - b
bored-art-40741
04/01/2016, 3:30 PMLet me know what you think, about to post it - b
bored-art-40741
04/01/2016, 3:41 PM@ancient-stone-50795: Posted https://rbcommons.com/s/twitter/r/3633/ - a
ancient-stone-50795
04/01/2016, 3:45 PM@bored-art-40741: LGTM - b
bored-art-40741
04/01/2016, 3:51 PMAha, I figured out an even cleverer way to do this! - b
bored-art-40741
04/01/2016, 3:52 PMWe can know the main process is blocking on the lock because of the message function - b
bored-art-40741
04/01/2016, 3:52 PMSo I can make this test both faster (no unnecessary waiting) and immune to slow CI - b
bored-art-40741
04/01/2016, 3:53 PMGoing to push that momentarily - b
bored-art-40741
04/01/2016, 4:01 PMPushed - b
bored-art-40741
04/01/2016, 4:03 PMGiven that the only followup changes were to comments and tests that I confirmed still work locally, I'm inclined to push this without waiting for another round of CI - b
bored-art-40741
04/01/2016, 4:20 PM@enough-analyst-54434: I'm about to land https://rbcommons.com/s/twitter/r/3633/ and I'd appreciate it if we could get it into this release - e
enough-analyst-54434
04/01/2016, 4:21 PMWill do. I'll be posting the CHANGELOG RB soon but will then be headed out and no back to actually do the release until this evening so plenty of time to get more stuff in. - b
bored-art-40741
04/01/2016, 4:21 PMCool, thanks - e
enough-analyst-54434
04/01/2016, 4:37 PMOK - this anticipates your lock fix landing and has it in the changelog: https://rbcommons.com/s/twitter/r/3641/ - b
bored-art-40741
04/01/2016, 4:38 PMJust landed - b
bored-art-40741
04/01/2016, 5:00 PMAnother global concurrency issue: https://github.com/pantsbuild/pants/issues/3130 - b
bored-art-40741
04/01/2016, 5:30 PMJust had a fully green run on Jenkins - b
bored-art-40741
04/01/2016, 5:31 PMOn the PR I'm working on that marks a couple of known-bad tests as xfail - b
bored-art-40741
04/01/2016, 5:31 PMIt seems like ivy file locking + ditching haskell has gotten us out of the non-deterministic nightmare - w
witty-crayon-22786
04/01/2016, 5:55 PMwoot - b
bored-art-40741
04/01/2016, 6:31 PM@ancient-stone-50795: Updated https://rbcommons.com/s/twitter/r/3632/ - b
bored-art-40741
04/01/2016, 6:33 PMThe xfails you were concerned about have been fixed/not xfailed. There's only one left, and it's definitely a real bug in a jdk environment that should be fixed. - a
ancient-stone-50795
04/01/2016, 6:36 PM@bored-art-40741: I approved, just maybe an issue with the comments - b
bored-art-40741
04/01/2016, 6:36 PMResponded - b
bored-art-40741
04/01/2016, 6:36 PMCould you drop/respond? - b
bored-art-40741
04/01/2016, 6:54 PM@ancient-stone-50795: ^^ - b
bored-art-40741
04/01/2016, 6:56 PMRandom ivy failure :\ http://jenkins.pantsbuild.org/job/pants_ci.integration_tests/164/SHARD=4/consoleFull - b
bored-art-40741
04/01/2016, 6:56 PMBut I don't think that's a locking issue, I think that's a legit "sometimes Ivy just fails" issue - a
ancient-stone-50795
04/01/2016, 7:00 PM@bored-art-40741: I don't have an issue if you just drop them yourself.