https://pantsbuild.org/ logo
Powered by
#general
  • b

    bored-art-40741

    10/01/2015, 4:54 PM
    yeah
  • b

    bored-art-40741

    10/01/2015, 4:55 PM
    btw we're probably going to upstream this to contrib now that 
    compile_classpath
    is the only interface it needs
  • w

    witty-crayon-22786

    10/01/2015, 4:55 PM
    that would be lovely.
  • w

    witty-crayon-22786

    10/01/2015, 4:55 PM
    this patch has actually had a very nice effect on our ivy usecase though
  • w

    witty-crayon-22786

    10/01/2015, 4:55 PM
    we gain some neat and safe flexibility
  • w

    witty-crayon-22786

    10/01/2015, 4:56 PM
    sources must be source compatible with all versions they're built against
  • w

    witty-crayon-22786

    10/01/2015, 4:57 PM
    but com.twitter.common, for example, is validated as compiling safely with guava 16 through 19
  • w

    witty-crayon-22786

    10/01/2015, 4:57 PM
    (we don't run its tests with all of those though... oy)
  • w

    witty-crayon-22786

    10/01/2015, 6:32 PM
    @bored-art-40741, @ancient-stone-50795 : i think you might also be able to adjust the zinc log level... currently we pass exactly pants own log level through
  • w

    witty-crayon-22786

    10/01/2015, 6:33 PM
    and that means that stuff printed at INFO is dumped out
  • w

    witty-crayon-22786

    10/01/2015, 6:33 PM
    could disconnect that, or just have a conditional where 
    -ldebug
    passes though DEBUG but otherwise always passes 
    -log-level=WARN
    to zinc
  • w

    witty-crayon-22786

    10/01/2015, 6:38 PM
    ...oh wow. passing warn is a big improvement
  • w

    witty-crayon-22786

    10/01/2015, 6:39 PM
    we should probably do that by default in pants.
  • w

    witty-crayon-22786

    10/01/2015, 7:05 PM
  • b

    bored-art-40741

    10/01/2015, 7:06 PM
    ++
  • a

    ancient-stone-50795

    10/01/2015, 7:23 PM
    Doesn’t matter much to us. With the reporting console settings I have set, I never see messages from zinc at all unless there is a compile failure.
  • w

    witty-crayon-22786

    10/01/2015, 7:31 PM
    we didn't want to turn it off entirely, because we have plenty of warnings that we want to nag folks with
  • w

    witty-crayon-22786

    10/01/2015, 9:12 PM
    i wonder how, in the long run, we can begin to hit the cache for the "nearest" artifact for an incremental compile.
  • w

    witty-crayon-22786

    10/01/2015, 9:13 PM
    ...actually, i take that back. don't want to think about the wacky bugs that could introduce =x
  • a

    ancient-stone-50795

    10/01/2015, 9:44 PM
    Are there any pressing RB’s out there that need to make it into the release tomorrow? I’m doing the release and I have meetings in the afternoon, I’d like to get it done early if possible
  • a

    ancient-stone-50795

    10/01/2015, 9:45 PM
    Maybe cut the release around noon Eastern time.
  • l

    limited-terabyte-34878

    10/01/2015, 10:01 PM
    @happy-kitchen-89482: can you take a look at my test task mixin RB when you get a chance? https://rbcommons.com/s/twitter/r/2902/
  • l

    limited-terabyte-34878

    10/01/2015, 10:01 PM
    thanks
  • b

    bored-art-40741

    10/01/2015, 10:01 PM
    Heads up, Benjy is on vacation atm
  • l

    limited-terabyte-34878

    10/01/2015, 10:02 PM
    ah
  • l

    limited-terabyte-34878

    10/01/2015, 10:02 PM
    thanks
  • w

    witty-crayon-22786

    10/02/2015, 12:14 AM
    is the cache fetch timeout configurable? looks like 4 seconds at the moment
  • w

    witty-crayon-22786

    10/02/2015, 12:18 AM
    Copy code
    Error while reading from remote artifact cache: HTTPSConnectionPool(host='<http://build.twitter.biz|build.twitter.biz>', port=443): Read timed out. (read timeout=4.0)
  • w

    witty-crayon-22786

    10/02/2015, 12:18 AM
    from the UK =(
  • b

    bored-art-40741

    10/02/2015, 12:19 AM
    Tansy has a patch out for this now!
1...486487488...2420Latest