Demo_Grass

Hello,

I had a chance to play with the latest release of python-ogre 1.1. It looks great!

I tried running the demo_grass but the grass doesn't seem to move.

Older versions of the code, the grass move quite nicely.

Is this demo broken?

Best regards,

I know this doesn't help hehe but
I've experienced the same problem

Having a quick look at the codes from both 1.1b and 1.0
I've came to notice they are exactly the same.
This is most confusing, since the change, then, must be inside PythonOgre.

Regards,
Danaugrs

Had a quick look and the issue is in the waveGrass function in Demo_grasss.py

At about line 105 you should have
offset.x = math.sin(self.xpos) * 5
offset.z = math.sin(self.zpos) * 5
For some reason the demo has *.05 which is wrong

Regards

Andy

Yay!
Andy has eagle eyes!

Cheers,
Danaugrs

thanks,
that fixed it!

On linux the Grass Demo with the code below currently runs at about 40FPS. About a sixth of the speed of the C++ version.


for reg in mStaticGeom.getRegionIterator():
## a little randomness
self.xpos += reg.getCentre().x * 0.001
self.zpos += reg.getCentre().z * 0.001
offset.x = math.sin(self.xpos) * 5
offset.z = math.sin(self.zpos) * 5
for lod in reg.getLODIterator():
for mat in lod.getMaterialIterator():
for geom in mat.getGeometryIterator():
geom.setCustomParameter(OFFSET_PARAM, offset)


If I replace the above code with the following the demo runs at about 160FPS. Has anyone else encountered this on Linux?

rit = mStaticGeom.getRegionIterator()
while rit.hasMoreElements():
reg = rit.getNext()
self.xpos += reg.getCentre().x * 0.001
self.zpos += reg.getCentre().z * 0.001
offset.x = math.sin(self.xpos) * 5
offset.z = math.sin(self.zpos) * 5

lodit = reg.getLODIterator()
while lodit.hasMoreElements():
lod = lodit.getNext()
matit = lod.getMaterialIterator()
while matit.hasMoreElements():
mat = matit.getNext()
geomit = mat.getGeometryIterator()
while geomit.hasMoreElements():
geom = geomit.getNext()
geom.setCustomParameter(OFFSET_PARAM, offset)



Yay!
Andy has eagle eyes!

Cheers,
Danaugrs

There's something called kdiff...
and it's free too.

@dermont

I guess I'll have to take a look at the generator code as it's certainly slower -- I did a quick benchmark and using the nice python 'iterator' interface is about 3 times slower than using the direct calls to the Ogre iterator

Cheers
Andy

Some of the OgreNewt demos appeared to be running slower as well, but I will have to get the C++ versions running to compare the difference.

@andy, thanks for the feedback. Wasn't sure if it was a problem with my build.