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

    bored-art-40741

    09/08/2015, 9:20 PM
    Another 25% is the NGSession via like 4 levels of HashMap indirection
  • e

    enough-analyst-54434

    09/08/2015, 9:20 PM
    softValues uses identity == (eq) ... is that a problem for the zinc Cache usage? I see Option[(..., ...)] for example
  • w

    witty-crayon-22786

    09/08/2015, 9:20 PM
    this is a guava cache
  • e

    enough-analyst-54434

    09/08/2015, 9:20 PM
    ack
  • w

    witty-crayon-22786

    09/08/2015, 9:21 PM
    oh, i see.
  • w

    witty-crayon-22786

    09/08/2015, 9:21 PM
    um. maybe.
  • w

    witty-crayon-22786

    09/08/2015, 9:21 PM
    ...that would mean that we missed constantly though
  • w

    witty-crayon-22786

    09/08/2015, 9:21 PM
    sec.
  • w

    witty-crayon-22786

    09/08/2015, 9:23 PM
    ...oh. it uses identity for values? hm... i wonder what for...
  • w

    witty-crayon-22786

    09/08/2015, 9:23 PM
    that should be fine, because i don't know what else it would be compared to
  • w

    witty-crayon-22786

    09/08/2015, 9:23 PM
    i'm probably not understanding though
  • w

    witty-crayon-22786

    09/08/2015, 9:23 PM
    (reading)
  • w

    witty-crayon-22786

    09/08/2015, 9:24 PM
    ...perhaps to use a reference queue to delete the precise entry? hm
  • w

    witty-crayon-22786

    09/08/2015, 9:29 PM
    it looks like it uses 
    valueEquivalence
    in 
    containsValue
    , 
    replace
    , and 
    remove
  • b

    bored-art-40741

    09/08/2015, 9:29 PM
    btw YourKit considers 97% of this memory to be strongly reachable
  • b

    bored-art-40741

    09/08/2015, 9:30 PM
    Assuming that its notion of strong reachability corresponds to non-weak-reference reachability
  • b

    bored-art-40741

    09/08/2015, 9:30 PM
    And I don't know what else it would mean
  • w

    witty-crayon-22786

    09/08/2015, 9:30 PM
    (correction: 'soft'... i was using the wrong adjective)
  • w

    witty-crayon-22786

    09/08/2015, 9:31 PM
  • b

    bored-art-40741

    09/08/2015, 9:31 PM
    ah okay
  • b

    bored-art-40741

    09/08/2015, 9:32 PM
    Everything seems to be strongly reachable
  • b

    bored-art-40741

    09/08/2015, 9:32 PM
    So if this cache is in fact being populated, it doesn't seem to be populated using weak refs
  • w

    witty-crayon-22786

    09/08/2015, 9:32 PM
    soft
  • w

    witty-crayon-22786

    09/08/2015, 9:32 PM
    but no: not necessarily
  • b

    bored-art-40741

    09/08/2015, 9:33 PM
    Sure, they're synonymous in this situation
  • w

    witty-crayon-22786

    09/08/2015, 9:33 PM
    the other explanation is that the compile in progress strongly reached them
  • b

    bored-art-40741

    09/08/2015, 9:33 PM
    I don't think so
  • w

    witty-crayon-22786

    09/08/2015, 9:33 PM
    no, not synonymous
  • b

    bored-art-40741

    09/08/2015, 9:33 PM
    It's used interchangeably by the docs you just linked
  • b

    bored-art-40741

    09/08/2015, 9:33 PM
    Oh, or maybe it just means it can be either
1...316317318...2420Latest