#general
- a
ancient-stone-50795
11/04/2015, 6:29 PMOtherwise, you'd think they might have setup their java compilation under bazel? - e
enough-analyst-54434
11/04/2015, 6:36 PMThey may just be blocked by impedance mismatch. FWICT chromium is a real community project and Bazel is not yet. They may be waiting on that. - a
ancient-stone-50795
11/04/2015, 6:36 PMpants is a real community project ... - e
enough-analyst-54434
11/04/2015, 6:37 PMI'll take a look at their build system as it realtes to the bug he linked tomorrow am for coffee and see if I can say something informed - m
mammoth-pharmacist-54215
11/04/2015, 7:10 PM@happy-kitchen-89482: very cool stuff in https://rbcommons.com/s/twitter/r/3074/diff/1#index_header thanks for all the work on this. - f
fancy-queen-20734
11/04/2015, 7:21 PM+1, I’m pretty excited about that changeset - r
red-cpu-25099
11/04/2015, 7:32 PMIs it bad to call self.setUp twice in pants BaseTest (in one test)? - m
mammoth-pharmacist-54215
11/04/2015, 7:33 PMnot sure if I understand you… you shouldn’t manually have to call setUp pytest should do that for you before the test. - b
bored-art-40741
11/04/2015, 7:38 PMCorrect - r
red-cpu-25099
11/04/2015, 7:38 PMI'm trying to figure out what happens in antlr_gen test, as I see now it's the only place with manual setUp call. So I'm going to remove it. Objections? - m
mammoth-pharmacist-54215
11/04/2015, 7:39 PMmind providing a link to the code you are looking at so I can take a peak? - b
bored-art-40741
11/04/2015, 7:40 PMThat test is old old in several ways - b
bored-art-40741
11/04/2015, 7:40 PMWe have much better constructs for testing tasks now - m
mammoth-pharmacist-54215
11/04/2015, 7:40 PMyeah, I think thats safe to remove - m
mammoth-pharmacist-54215
11/04/2015, 7:40 PMthe setup I mean not the test - b
bored-art-40741
11/04/2015, 7:41 PMIt is not safe to remove - b
bored-art-40741
11/04/2015, 7:41 PMThough perhaps it should be replaced with a line that explicitly clears the graph since that's why it needs the call - m
mammoth-pharmacist-54215
11/04/2015, 7:41 PMoh, I see this isn’t a test method… sorry misread - m
mammoth-pharmacist-54215
11/04/2015, 7:42 PM@red-cpu-25099: looks like the test should be rewritten to not do this - r
red-cpu-25099
11/04/2015, 7:42 PMI will rewrite the test, thanks! - b
bored-art-40741
11/04/2015, 7:43 PMTake a look at some modern task tests - b
bored-art-40741
11/04/2015, 7:43 PMThis process is much easier and cleaner now - e
enough-analyst-54434
11/04/2015, 7:43 PMThat was some pretty crappy review by Stu and I - e
enough-analyst-54434
11/04/2015, 7:43 PMsorry about that - r
red-cpu-25099
11/04/2015, 7:44 PM@bored-art-40741: any test names as examples? - b
bored-art-40741
11/04/2015, 7:44 PMOne sec, let me poke - b
bored-art-40741
11/04/2015, 7:48 PMcontrib/go/tests/python/pants_test/contrib/go/tasks/test_go_buildgen.py - b
bored-art-40741
11/04/2015, 7:48 PMLooks pretty modern and uses the cleaner way of constructing a graph and then getting a task context for it - e
enough-analyst-54434
11/04/2015, 7:54 PMJust stay away from
- note the note there, buildgen is special enough to need it - you should notalias_groups