#general
- e
enough-analyst-54434
08/28/2015, 9:00 PMbus using another user on github for the sync push to - w
witty-crayon-22786
08/28/2015, 9:00 PMsure - e
enough-analyst-54434
08/28/2015, 9:00 PMthe new user gets a new queue from Travis - e
enough-analyst-54434
08/28/2015, 9:00 PMIn fact - maybe I should be doing this for the OSX stuff! - e
enough-analyst-54434
08/28/2015, 9:00 PMI'm going to go landgrab some users - r
rough-minister-58256
08/28/2015, 9:01 PMheh - e
enough-analyst-54434
08/28/2015, 9:05 PMdamn - should have thought of that earlier!: https://github.com/pantsbuild-osx/ - e
enough-analyst-54434
08/28/2015, 9:05 PMI'll do one for pypy - r
rough-minister-58256
08/28/2015, 9:06 PMnice - a
ancient-stone-50795
08/28/2015, 9:36 PMIf only Travis-CI would let us pay them we wouldn’t have to be so clever! - e
enough-analyst-54434
08/28/2015, 9:37 PMSince they won't my guilt at doing this is assuaged - e
enough-analyst-54434
08/28/2015, 9:37 PM - a
ancient-stone-50795
08/28/2015, 9:38 PMLGTM - e
enough-analyst-54434
08/28/2015, 9:38 PMIf that works then I may hand off to a tweep to setup the pypy sync - e
enough-analyst-54434
08/28/2015, 9:38 PMI already got the travis-ci wired for the pantsbuild-pypy/pants repo, just needs script work - e
enough-analyst-54434
08/28/2015, 9:42 PMOK - cool that worked - e
enough-analyst-54434
08/28/2015, 9:46 PMAnd the pantsbuild/pants-for-travis-osx-ci repo is deleted now in favor of pantsbuild-osx/pants - you may need to adjust your build orb configs - b
bored-art-40741
08/28/2015, 9:52 PM++ - w
witty-crayon-22786
08/28/2015, 11:29 PMwoot! - r
rough-minister-58256
08/28/2015, 11:33 PMorb? - w
witty-crayon-22786
08/28/2015, 11:43 PM(ps: clean, uncached build and test of a large jvm target worked fine with pypy) - m
mammoth-pharmacist-54215
08/28/2015, 11:43 PMhow was the timing compared to CPython? - w
witty-crayon-22786
08/28/2015, 11:44 PMit's all compile time, and too variable to be interesting probably - w
witty-crayon-22786
08/28/2015, 11:44 PM63 minutes 😃 - m
mammoth-pharmacist-54215
08/28/2015, 11:44 PMfair enough - b
bored-art-40741
08/30/2015, 5:52 PM@witty-crayon-22786 @adventurous-twilight-44110 I think we need to revert https://rbcommons.com/s/twitter/r/2668 unless we can come up with a workaround - b
bored-art-40741
08/30/2015, 5:53 PMFatal missing deps becoming non-fatal on a second run because of artifact caching is a major regression - b
bored-art-40741
08/30/2015, 5:53 PMAfter thinking on this for a while I can't see any way to actually pull this out of the jvm compilation tasks - b
bored-art-40741
08/30/2015, 5:53 PMThe failure has to happen before they write out their artifacts - b
bored-art-40741
08/30/2015, 6:02 PMUnless the product that powers missing dep detection is stored in the artifact and rechecked even on an artifact cache hit or valid target? But I don't think that's the case