gcanaday at gmail.com
Tue Jan 6 10:55:05 PST 2009
Yep. Touches are just worthy of mention because a lot of people skim over
that. Multiple scripts in the same object do what I described (you probably
already knew that tho) but multiple touches in the same prim don't trigger
touches in an easily predictable manner. It's not technically concurrency at
all, but the unknown touch queue can kinda wreck the effect.
The basic idea I was going for was that I would make a new script (even
duplicating code) wherever I would normally call a pthread_create() in C, and
give the child scripts no events at all that require outside interaction.
Looks like threads, though it isn't.
On Tuesday 06 January 2009 1:33:23 pm Dale Innis wrote:
> That's not actually an exception in any way, is it? All of the
> scripts still have a single thread and a single event queue. The
> passing of touch events just determines which queue(s) the touch event
> gets put onto. And note that touches aren't unique; the same applies
> to collisions, modulo bugs. (See the Wikis on llPassTouches() and
> llPassCollisions(), as well as "
> https://jira.secondlife.com/browse/SVC-3306 ".)
> Dale Innis
> On Tue, Jan 6, 2009 at 1:07 PM, Glen Canaday <gcanaday at gmail.com> wrote:
> > Each script is a single thread, with one exceptions. Touch events go to
> > the prim the touch event is in, unless there isn't one, in which case the
> > event goes to the root prim. Other than that, each script gets its own
> > even queue and they're stacked like you said.
> > I *often* multithread by using multiple scripts in the same prim.
> > --GC
> Click here to unsubscribe or manage your list subscription:
More information about the secondlifescripters