rayon

2016.06.06


  • PRs
  • Should try_fold have its own “boolean” signal?
  • If you combine with try_reduce, then the reduce will already cancel work early, so there isn’t much advantage for it to do so.
  • Do we want to make any kind of guarantee?
  • We could say that, presuming you don’t combine with something that aborts early like try_reduce, try_fold never communicates across splits, and thus if you do try_fold(..).for_each(..) you are guaranteed to execute at least the items before the first None is found

2018.05.30


First inaugural rayon meeting :)

Issue Triage Notes


PR Notes

2018.02.07


Attending

  • cuviper
  • jake
  • nikomatsakis

Status

  • min rustc version: technically we are compatible with 1.12.0 though rand makes it hard
  • usize question

2018.01.24


Attending

  • cuviper
  • KodrAus
  • jake
  • nikomatsakis

Agenda

  • Management and decisions going forward (nikomatsakis)

Milestone List

  • Pending decisions
  • frustration that u64 ranges don’t fit into IndexedParallelIterator
  • this is because it uses usize for compatibility with std