On Thu, Apr 22, 2004 at 08:36:07AM +0100, Zefram wrote:
> And a final thought before I head off on holiday: consider the event
> sequence
>
> 0. start a signal watcher
> 1. receive a signal
> 2. start a second watcher for the same signal
> 3. send another signal
> 4. process events
>
> At the mom
On Thu, Apr 22, 2004 at 08:16:43AM +0100, Zefram wrote:
> Joshua N Pritikin wrote:
> >> 0. (bug) ->pending() should do an asynccheck
> >
> >That changes the documented behavior too much.
>
> How so?
Because merely calling asynccheck never eliminates the race condition.
It just makes it "shorter."
And a final thought before I head off on holiday: consider the event
sequence
0. start a signal watcher
1. receive a signal
2. start a second watcher for the same signal
3. send another signal
4. process events
At the moment, both watchers get events with hits=2. The correct
behaviour would be f
Joshua N Pritikin wrote:
>Ick. That's probably a bug. What I'm going to do is just drop any
>pending signals when you can stop(). Will that work for you?
Interesting. ->stop is actually documented to cancel pending events;
I hadn't noticed that before. So that change does seem to be necessary
