The publish event cannot return a boolean b/c the event just gets
queued at that point so there still is no guarantee that the event
ever gets executed. Java Executors solve this problem by providing a
rejected execution interface. This allows the creator of a thread pool
to decide what to do with rejected events. One way doesn't work for
everyone, but the current approach to ignore events on disposal works
for the majority of cases I've encountered.
Is your app shutting down? Why was the Fiber disposed?
Mike
> --
> You received this message because you are subscribed to the Google Groups "jetlang-dev" group.
> To post to this group, send email to jetla...@googlegroups.com.
> To unsubscribe from this group, send email to jetlang-dev...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/jetlang-dev?hl=en.
>
>
Why is another thread disposing the fiber? I would handle this by
having the same fiber/thread dispose the thread that publishes to it.
If the same fiber is used, then it can update state accordingly to
know that a fiber is disposed prior to the publish.
The key to using jetlang (and most threading problems) is to try to
make all problems into isolated single-threaded, sequential units of
work. All state changes must be processed by the same fiber/thread.
The disposal of the fiber is a state change that the publisher must
know about prior to the publish.