…n the Collection
…flex) discovered that they weren't being received.
…as needed. Fixes agorman's test case from irc.perl.org #reflex.
Agorman (irc.perl.org #reflex) discovered that run_all() called more than once asplodes ugly-like. He provided a nice test case, which I've added to Reflex's tests. It uses Reflex::POE::Wheel::Run, and that might need to be removed to make the test more portable.
…l don't work, however.
"emits" is a shortcut for Reflex::Trait::EmitsOnChange. "observes" is a shortcut for Reflex::Trait::Observed. Names and locations may change.
…otes within this commit.
If this can be made to work, it represents an incremental migration path from POE to Reflex. Begin with a POE program, and gradually incorporate Reflex code. Eventually (har) port everything over.
Agorman in irc.perl.org #reflex discovered that multiple PID watchers receive notification of all exiting child processes. This commit limits notification to just those PIDs that are watched.
…. Thanks to "batman" in irc.perl.org #reflex.
…agorman in irc.perl.org #reflex.
…ts absence in freenode #perl.
It's up to the Writing role's consumer to decide how flushes are to be triggered. The Writing role is only to be concerned with writing.
… timers. About 97% done with the renaming.
…ed to change a lot of code and docs that used Reflex::Timer, which has bee deleted.
…thods. I still don't like the function names, but they're good placeholders for now.
…flex::Util::Methods because Reflex::Role will take that over.
It expands Reflex::Role::Parameterized with additional syntax for callback names, method names, and attribute parameters.
… too useful to be sequestered into a streaming role.