Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
signals and events
4 views
Skip to first unread message
Nicholas Clark
Jul 13, 2005, 10:47:17 AM7/13/05
to poni...@perl.org, perl6-i...@perl.org
What is the state of signals and events support? As far as I can tell from
watching which perl 5 regression tests fail or hang under parrot, currently
parrot can't behave as a passive embedding target - it assumes that it needs
to take ownership of all signals.
watching which perl 5 regression tests fail or hang under parrot, currently
parrot can't behave as a passive embedding target - it assumes that it needs
to take ownership of all signals.
Is this likely to change soon? Currently ponie has had to add various skip
blocks to perl 5 regression tests that throw/catch signals, and this doesn't
seem ideal as it implies that code changes would also be needed by existing
CPAN code to work properly under ponie.
Nicholas Clark
Leopold Toetsch
Jul 14, 2005, 12:21:49 PM7/14/05
to Nicholas Clark, poni...@perl.org, perl6-i...@perl.org
Nicholas Clark wrote:
> What is the state of signals and events support? As far as I can tell from
> watching which perl 5 regression tests fail or hang under parrot, currently
> parrot can't behave as a passive embedding target - it assumes that it needs
> to take ownership of all signals.
> What is the state of signals and events support? As far as I can tell from
> watching which perl 5 regression tests fail or hang under parrot, currently
> parrot can't behave as a passive embedding target - it assumes that it needs
> to take ownership of all signals.
Parrot currently captures SIGHUP for testing (I don't remember if other
signals are blocked) And yes, there is no interface whatsoever to pass a
signal on, or some such.
> Is this likely to change soon?
We need an API.
.
> Nicholas Clark
leo
0 new messages