Hi Maxim,
On Fri, Sep 17, 2010 at 3:59 PM, Maxim Gammer <[email protected]> wrote:
> Hi Robert,
> Yes, you variant works well. It will be included in the next version?
I have just checked in the following:
inline int RandomRateCounter::numParticlesToCreate(double dt) const
{
// compute the number of new particles, clamping it to 1
second of particles at the maximum rate
float numNewParticles = std::min(static_cast<float>(dt *
getRateRange().get_random()), getRateRange().maximum);
// add the number of new particles to value carried over from
the previous call
_np += numNewParticles;
// round down the number of particles.
int n = static_cast<int>(_np);
// take away the number of rounded number of particles leaving
the decimal place
// this is done so that two frames of 0.5's will results in
first frame 0 new particles, second frame 1
_np -= n;
// return the rounded number of particles to be created
return n;
}
I'm not entirely happy with it but at least it should avoid massive
numbers of particles when particle systems come back on screen.
The real problem that we are seeing still is that particle systems are
particle systems aren't being regularly updated whilst off-screen, or
at least what it appears to be the problem when testing the
osgparticle - if I move the particle off screen for a few seconds and
then pull them back I get an odd gap in the particle system. I've not
dug into the implementation closely enough to review exactly what is a
happening, perhaps this could be addressed as an optional behviour
i.e. keep particle systems running fully even when offscreen.
Robert.
_______________________________________________
osg-submissions mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
