#general
- w
witty-crayon-22786
09/14/2015, 3:59 PMyea. will likely enable globally with a blacklist - f
freezing-motorcycle-14222
09/14/2015, 3:59 PMAlso, what do we do with Github pull requests after the review board has been merged in. Do we just close them? - w
witty-crayon-22786
09/14/2015, 3:59 PMyea - f
freezing-motorcycle-14222
09/14/2015, 3:59 PMThanks - a
ancient-stone-50795
09/14/2015, 4:11 PMOoh, this is nice. http://blog.rbcommons.com/2015/09/14/introducing-new-special-user-permissions/ Let me at it, there are lots of reviews I want to close. - m
mammoth-pharmacist-54215
09/14/2015, 4:42 PM@freezing-motorcycle-14222: the options are called pycheck not lint, that’s probably why you didn’t see them. ./pants --help-all |grep pycheck - m
mammoth-pharmacist-54215
09/14/2015, 4:42 PMIn fairness this work was based on work that one of the tweeps did internally before my time. Most of my work was porting it in a sane way to pants. - h
happy-kitchen-89482
09/14/2015, 5:47 PM@witty-crayon-22786, @bored-art-40741: While no one would be happier than I to get rid of global strategy, let's wait until we're actually fully onto isolated before pushing that... - w
witty-crayon-22786
09/14/2015, 5:47 PM@fancy-queen-20734: re https://rbcommons.com/s/twitter/r/2805/ : zinc/sbt don't recompile input files if they don't need to. but if the outputs for some files are missing, it will need to recompile - w
witty-crayon-22786
09/14/2015, 5:47 PM@happy-kitchen-89482: yea, was just toying with it. it's on hold until we figure out that last performance thing - f
fancy-queen-20734
09/14/2015, 8:24 PM@witty-crayon-22786: I had another thought about it. The potential performance regression caused by the patch is the same one we’d get if we used cache_target_dirs as it’s currently implemented with compile - w
witty-crayon-22786
09/14/2015, 8:33 PM@fancy-queen-20734: yea. when cody was implementing it we talked about having an additional 'stable location' type flag for cache_target_dirs - w
witty-crayon-22786
09/14/2015, 8:34 PM@fancy-queen-20734: but stable target directories aren't a great default... cf, this bug - f
fancy-queen-20734
09/14/2015, 8:34 PMI’d previously sketched out adding the hash to the compile directories - f
fancy-queen-20734
09/14/2015, 8:35 PMit’s tricky without adding an instance var to the strategy to hold onto a dict - f
fancy-queen-20734
09/14/2015, 8:35 PMso that the directory can be reproduced consistently - w
witty-crayon-22786
09/14/2015, 8:52 PMyep - w
witty-crayon-22786
09/14/2015, 8:52 PMin any case, we can't use complete wipes in the context of incremental compile, except when we've hit - f
fancy-queen-20734
09/14/2015, 9:00 PM@witty-crayon-22786: is that because of the intra-target incremental compile done by zinc? - f
fancy-queen-20734
09/14/2015, 9:04 PMI have mixed feelings about adding a post hit pre extract callback to the artifact cache interface. It ends up touching all the same things as my idea for doing a deletion in cache -> cache transitions, but requires the caller to know what they are doing - w
witty-crayon-22786
09/14/2015, 9:05 PM@fancy-queen-20734: there is only 1 caller who shouldn't be using cache_target_dirs - f
fancy-queen-20734
09/14/2015, 9:06 PMthat’s just compile, right? - w
witty-crayon-22786
09/14/2015, 9:06 PMexactly - f
fancy-queen-20734
09/14/2015, 9:06 PMfor performance reasons - w
witty-crayon-22786
09/14/2015, 9:06 PMbecause incremental compile - w
witty-crayon-22786
09/14/2015, 9:06 PMfull circle - a
ancient-stone-50795
09/14/2015, 9:18 PMHey tweeps, how long does it take you to do a
for most usersgit pull - a
ancient-stone-50795
09/14/2015, 9:18 PMSay if they are pulling in a day’s worth of changes. - b
billowy-alarm-80641
09/14/2015, 9:19 PMin our monorepo? - a
ancient-stone-50795
09/14/2015, 9:19 PMYes, in the monorepo