#general
- w
witty-crayon-22786
09/05/2015, 12:15 AMbecause the lhs changes the type of the rhs? - b
bored-art-40741
09/05/2015, 12:15 AMI know why, too, Benjy and I labored over this code in zincutils for a while - w
witty-crayon-22786
09/05/2015, 12:15 AMick - w
witty-crayon-22786
09/05/2015, 12:15 AMthank you for the fix though 😃 - r
rough-minister-58256
09/05/2015, 12:15 AMheh - b
bored-art-40741
09/05/2015, 12:16 AMyeah, python still tries to do conversions for you when you do string operations like bytes.replace(unicode, more_unicode) - w
witty-crayon-22786
09/05/2015, 12:16 AM@bored-art-40741: would it be cleaner for zincutils to treat all inputs as binary? - b
bored-art-40741
09/05/2015, 12:16 AMIt does treat all inputs as binary, but maybe it should actually type check - w
witty-crayon-22786
09/05/2015, 12:16 AMi see - b
bored-art-40741
09/05/2015, 12:17 AMIf we want to be loosey goosey we could even assume the user wants unicode to be treated as utf-8, and default parameterize that - b
bored-art-40741
09/05/2015, 12:17 AMI'm personally a bit of a purist when it comes to bytes vs code points - b
bored-art-40741
09/05/2015, 12:18 AMI'm fine with asserting, and I'd prefer not to make assumptions about the user's string encoding life choices - b
bored-art-40741
09/05/2015, 12:23 AMDoes anyone have objections to making name_hashing=True in the isolated pants.ini? - b
bored-art-40741
09/05/2015, 12:23 AMAssuming all tests pass (including the one I'm going to add for this) it would be nice to backstop - w
witty-crayon-22786
09/05/2015, 12:26 AM@bored-art-40741: nope, that would be great. - w
witty-crayon-22786
09/05/2015, 12:26 AMwe'll flip it on internally as soon as we can. - b
bored-art-40741
09/05/2015, 1:33 AMIterating on scala integration tests: definitely not energy ⭐ recommended - w
witty-crayon-22786
09/05/2015, 1:42 AMheh - b
bored-art-40741
09/05/2015, 1:43 AMOkay, have a backstop integration test + fix - b
bored-art-40741
09/05/2015, 1:43 AMPutting together an RB now - w
witty-crayon-22786
09/05/2015, 1:47 AM...ah, poop. just noticed that the format version moved into zincutils - b
bored-art-40741
09/05/2015, 1:57 AMRB out - b
bored-art-40741
09/05/2015, 6:49 PMDo Twitter/other shops run pants within Jenkins? - b
bored-art-40741
09/05/2015, 6:50 PMFor as long as I can remember our output has been too aggressively buffered when running pants in jenkins, and I'm curious if anyone else has also had this problem - b
bored-art-40741
09/05/2015, 6:52 PMI'm not sure if it's python buffering differently because of the types of the output ttys, or if it's a jenkins problem - e
enough-analyst-54434
09/05/2015, 7:14 PMTwitter does and the only oddity in buffering I remember were single lines that mutated - which was only exhibited by the junit test runner ......... - e
enough-analyst-54434
09/05/2015, 7:15 PMThose lines looked like a stall, but they were not if you scrolled really far right - e
enough-analyst-54434
09/05/2015, 7:15 PMThat was pants output a year ago though - b
bored-art-40741
09/05/2015, 7:20 PMhm - b
bored-art-40741
09/05/2015, 7:20 PMThat seems to suggest it's jenkins then