https://pantsbuild.org/ logo
Powered by
#development
  • w

    witty-crayon-22786

    11/01/2016, 11:49 PM
    good call @billowy-beach-68498 ... was sortof blindingly obvious once it was in place
  • w

    witty-crayon-22786

    11/01/2016, 11:58 PM
    unit tests:
    Copy code
    tests/python/pants_test/engine/test_rules.py FF..FF.F.......F..FFF.FFFFFFF....FF
tests/python/pants_test/engine/test_graph.py FFFFFFFFFFF
tests/python/pants_test/engine/test_build_files.py FFFFFFFFFFFFF
tests/python/pants_test/engine/test_selectors.py ......
tests/python/pants_test/engine/test_engine.py FFFFFFF
tests/python/pants_test/engine/legacy/test_structs.py ..
tests/python/pants_test/engine/test_struct.py ........
tests/python/pants_test/engine/test_storage.py .....
tests/python/pants_test/engine/legacy/test_address_mapper.py .FFFFFFFFFFF
tests/python/pants_test/engine/test_isolated_process.py .FFFFF
tests/python/pants_test/engine/test_path_globs.py ........
tests/python/pants_test/engine/test_scheduler.py FFFFFFFFFFFFF
tests/python/pants_test/engine/test_fs.py .FFFFFFFFFFFFFFFF.sFFFFFsFFssFFsFF
tests/python/pants_test/engine/test_parsers.py ..............
tests/python/pants_test/engine/test_mapper.py .............FFFFFF
tests/python/pants_test/engine/legacy/test_graph.py FFFF
tests/python/pants_test/engine/test_legacy_engine.py ...
tests/python/pants_test/engine/test_round_engine.py ................
tests/python/pants_test/engine/test_addressable.py ......................................
  • w

    witty-crayon-22786

    11/01/2016, 11:59 PM
    so maybe not tonight
  • f

    fancy-queen-20734

    11/02/2016, 12:14 AM
    lol
  • w

    witty-crayon-22786

    11/02/2016, 10:31 PM
    aha, so that's why it slowed down
  • w

    witty-crayon-22786

    11/02/2016, 10:31 PM
    Copy code
    -rw-rw----+  1 stuhood  admin  923235591 Nov  2 15:30 viz.0.dot
  • r

    rough-minister-58256

    11/02/2016, 10:32 PM
    hah
  • r

    rough-minister-58256

    11/02/2016, 10:32 PM
    thats alottta text
  • w

    witty-crayon-22786

    11/02/2016, 10:32 PM
    36 MBs of dot file a second... yay rust
  • w

    witty-crayon-22786

    11/02/2016, 10:33 PM
    it will finally be a flag. i swear it.
  • w

    witty-crayon-22786

    11/02/2016, 10:33 PM
    by the old gods and the new
    🙂 1
  • w

    witty-crayon-22786

    11/02/2016, 10:35 PM
    so, variants are a pretty big punt, but they're out over here: https://github.com/pantsbuild/pants/issues/4020
  • w

    witty-crayon-22786

    11/02/2016, 10:36 PM
    not used anywhere in 
    engine.legacy
    , so primarily breaks the examples.
  • w

    witty-crayon-22786

    11/03/2016, 8:09 PM
  • w

    witty-crayon-22786

    11/03/2016, 8:26 PM
    @fancy-queen-20734 : and, not 100% sure intrinsics need to exist either... have been adding thoughts to https://github.com/pantsbuild/pants/issues/4022
  • w

    witty-crayon-22786

    11/03/2016, 8:27 PM
    in particular, i think that if you injected the ProjectTree as a singleton, and then banned caching of tasks with direct dependencies on singletons, they could just be normal tasks
  • w

    witty-crayon-22786

    11/03/2016, 8:28 PM
    ...there is more there.
  • f

    fancy-queen-20734

    11/03/2016, 8:28 PM
    So, you’d only be able to cache tasks if they had a transitive dep on singletons?
  • w

    witty-crayon-22786

    11/03/2016, 8:29 PM
    ...it's not a fully formed thought now that i examine it a bit more
  • w

    witty-crayon-22786

    11/03/2016, 8:30 PM
    if you assume that the output of the task is something stable (ie: i looked at the filesystem and grabbed this stuff and/or computed this snapshot), then it's fine
  • w

    witty-crayon-22786

    11/03/2016, 8:30 PM
    but if the task just aliases the thing you give it, then it's leaky
  • w

    witty-crayon-22786

    11/03/2016, 8:31 PM
    so you really do need to hide references to the filesystem singleton somehow.
  • w

    witty-crayon-22786

    11/03/2016, 8:31 PM
    (which is what we were doing with the "context")
  • w

    witty-crayon-22786

    11/04/2016, 1:15 AM
    all tests passing: https://rbcommons.com/s/twitter/r/4270/ (...or disabled)
  • r

    rough-minister-58256

    11/04/2016, 1:17 AM
    noice
  • w

    witty-crayon-22786

    11/04/2016, 5:33 PM
    @enough-analyst-54434 : finished commenting on https://rbcommons.com/s/twitter/r/4345/ ... sorry for the false start
  • w

    witty-crayon-22786

    11/04/2016, 5:34 PM
    one substantive edit, but other than that looks great
  • e

    enough-analyst-54434

    11/04/2016, 5:35 PM
    Cool - thanks for kicking the tires
  • w

    witty-crayon-22786

    11/04/2016, 5:35 PM
    @enough-analyst-54434 : ah, one more thing though: patch now contains some other edits
  • w

    witty-crayon-22786

    11/04/2016, 5:35 PM
    and/or is the wrong patch
1...188189190...906Latest