#general
- w
witty-crayon-22786
04/07/2016, 4:25 PM@fast-megabyte-33579: the fact that it isn't idempotent is a serious bug that we discovered recently, and which i'm currently investigating - w
witty-crayon-22786
04/07/2016, 4:29 PM@silly-crowd-96129: welcome! let us know if you have any questions. - f
fast-megabyte-33579
04/07/2016, 4:32 PMthanks stu - f
fancy-queen-20734
04/07/2016, 5:55 PMHow would people feel about enabling test duration logging for ci to help identify tests that take a long time? - w
witty-crayon-22786
04/07/2016, 5:55 PM+1 - f
fast-megabyte-33579
04/07/2016, 5:57 PMI'm looking at https://github.com/pantsbuild/pants/blob/master/src/python/pants/backend/jvm/targets/jar_library.py#L43 which suggests I should be able to exclude a sub-dependency from a jar_library but it's erroring out with "unknown arguments" when I try - am I just misunderstanding? - w
witty-crayon-22786
04/07/2016, 5:58 PMshould be able to - w
witty-crayon-22786
04/07/2016, 5:58 PMhttp://pantsbuild.github.io/build_dictionary.html indicates that it accepts a list of exclude objects - w
witty-crayon-22786
04/07/2016, 5:59 PM@fast-megabyte-33579: oh, wait. no... jar_library does not support excludes. - w
witty-crayon-22786
04/07/2016, 5:59 PMat that level, you'd want to exclude on the jar itself - w
witty-crayon-22786
04/07/2016, 5:59 PM@fast-megabyte-33579: find in page for
on the build_dictionaryjar: - f
fast-megabyte-33579
04/07/2016, 6:00 PMah! - f
fast-megabyte-33579
04/07/2016, 6:01 PMthanks - f
fast-megabyte-33579
04/07/2016, 6:09 PMis there a way to get an ivy report for where a jar is coming from (in this case slf4j-log4j12) - w
witty-crayon-22786
04/07/2016, 6:10 PMyes..../pants resolve.ivy --open $target - f
fast-megabyte-33579
04/07/2016, 6:10 PMah yes - c
curved-manchester-66006
04/07/2016, 8:00 PMhttps://pantsbuild.github.io/3rdparty_jvm.html talks about having a tree of 3rd party dependencies per org instead of one big 3rd party BUILD file. Is that "BUILD file per point in the package oraganization (like
) or per team in my organization (teams X & Y each have their own 3rdparty build file)?org.apache.foo.sweetlib - w
witty-crayon-22786
04/07/2016, 8:01 PM@curved-manchester-66006: per point in the package organization - w
witty-crayon-22786
04/07/2016, 8:01 PMwe do - w
witty-crayon-22786
04/07/2016, 8:01 PM3rdparty/jvm/org/apache/foo/sweetlib:sweetlib - w
witty-crayon-22786
04/07/2016, 8:03 PMteams having their own independent 3rdparty definitions would mostly encourage unnecessary fragmentation - w
witty-crayon-22786
04/07/2016, 8:03 PM...as it does in polyrepo setups =P - c
curved-manchester-66006
04/07/2016, 8:03 PMWhen a team has 3 binaries that only work with special versions of guava (or other weird things) they define those in the
files for said binary?BUILD - w
witty-crayon-22786
04/07/2016, 8:03 PMno - w
witty-crayon-22786
04/07/2016, 8:03 PMthey still define them in 3rdparty, with TODOs - w
witty-crayon-22786
04/07/2016, 8:03 PMand the owners of 3rdparty track those as tech debt - w
witty-crayon-22786
04/07/2016, 8:04 PM(just spieling about how we do it, basically) - w
witty-crayon-22786
04/07/2016, 8:04 PMwhen you end up with multiple versions of something in 3rdparty, the discouraged version has a version, and the encouraged version doesn't - w
witty-crayon-22786
04/07/2016, 8:05 PM
vs3rdparty/jvm/org/apache/foo/sweetlib:sweetlib3rdparty/jvm/org/apache/foo/sweetlib:sweetlib-1.4 - c
curved-manchester-66006
04/07/2016, 8:07 PMGotcha
it is3rdparty/jvm/org/apache/foo/sweetlib-teamFooShamefulHack-1.4