#general
- e
enough-analyst-54434
09/21/2015, 5:22 PMYou can run locally if you're curious - e
enough-analyst-54434
09/21/2015, 5:22 PM./pants test.py --coverage=modules:[list,of,module,names] [targets] - e
enough-analyst-54434
09/21/2015, 5:22 PMmaybe drop the s on modules, I always forget - f
full-kilobyte-93429
09/21/2015, 5:23 PMthanks john! my previous run dropped 1.x% so i like to rerun - f
full-kilobyte-93429
09/21/2015, 5:23 PMdo the numbers vary that much? - e
enough-analyst-54434
09/21/2015, 5:23 PMNo that is very large - e
enough-analyst-54434
09/21/2015, 5:23 PMMore likely 1 unit test shard recorded, one did not - e
enough-analyst-54434
09/21/2015, 5:23 PMThe sum of the 2 is what's needed for a good measure - f
full-kilobyte-93429
09/21/2015, 5:24 PMgot it - f
full-kilobyte-93429
09/21/2015, 5:24 PMdoes coveralls automatically do this? - e
enough-analyst-54434
09/21/2015, 5:27 PMMerge the 2 - yes - e
enough-analyst-54434
09/21/2015, 5:27 PMWe need to send though: https://github.com/pantsbuild/pants/blob/master/.travis.yml#L76-L77 - e
enough-analyst-54434
09/21/2015, 5:28 PMAnd that gates on shard success, so you can have 1 of the 2 unit-test shards fail and get a low coverage number - b
bored-art-40741
09/21/2015, 5:36 PM@witty-crayon-22786 or anyone else: has there been any modification to the artifact cache downloading code in the past couple months? - b
bored-art-40741
09/21/2015, 5:37 PMIt was very carefully tuned to actually pipeline and decompress in parallel - b
bored-art-40741
09/21/2015, 5:37 PMAnd we're noticing with the upgrade that cache downloads are taking way too long - w
witty-crayon-22786
09/21/2015, 5:37 PMit's definitely still decompressing in parallel - w
witty-crayon-22786
09/21/2015, 5:37 PMbut no, i don't think that's changed - w
witty-crayon-22786
09/21/2015, 5:38 PMexcept that isolated runs in its own threads already - b
bored-art-40741
09/21/2015, 5:38 PMOh hm - b
bored-art-40741
09/21/2015, 5:38 PMIn parallel, I mean in a process pool - b
bored-art-40741
09/21/2015, 5:52 PM@witty-crayon-22786: Do we think average scala artifact size would get quite a bit larger? - b
bored-art-40741
09/21/2015, 5:52 PMAre we bundling both classfiles and jars now? - w
witty-crayon-22786
09/21/2015, 5:56 PM@bored-art-40741: well, you know we're double caching those at the moment - b
bored-art-40741
09/21/2015, 5:56 PMI did not - b
bored-art-40741
09/21/2015, 5:56 PMWell, I knew that we were in .pants.d - w
witty-crayon-22786
09/21/2015, 5:56 PMer, it's the same thing as the disk space - b
bored-art-40741
09/21/2015, 5:56 PMNot necessarily in the artifact cache - b
bored-art-40741
09/21/2015, 5:56 PMGot it - w
witty-crayon-22786
09/21/2015, 5:56 PMwas the simplest impl =/