Erik Hofman wrote:

Hi Josh,

Josh Babcock wrote:

OK, I thought I was cool, but I'm not. When I got Xorg working, I forgot that there's a bug in sawfish that resets LD_LIBRARY_PATH and that I had a hack somewhere that was pointing apps to the old XFree86 radeon driver. So now that's fixed, and glxgears is great (and faster) and blender is happy, but fgfs has different ideas. It's obviously not using direct rendering. I'm getting about 1/3 FPS about of it, the same I get with GL_ALWAYS_INDIRECT=1.


You have to make sure that most options like "Enhanced runway lighting", "Specular highlight" and "3d clouds" are turned off. These might trigger some functions that make the driver fall back to software rendering.

Erik

_______________________________________________
Flightgear-devel mailing list
Flightgear-devel@flightgear.org
http://mail.flightgear.org/mailman/listinfo/flightgear-devel
2f585eeea02e2c79d7b1d8c4963bae2d


Ok, in answer to some previous questions, yes, I did recompile after switching the LD_LIBRARY_PATH, and I also just recompiled with Melchior's LDFLAGS but to no avail. Here are my rendering settings as of now:


Contents of "/sim/rendering"
fog (fastest)
shading (true)
skyblend (false)
textures (true)
wireframe (false)
horizon-effect (false)
enhanced-lighting (false)
distance-attenuation (false)
specular-highlight (false)
bits-per-pixel (16)
debug (false)
realism (5)
static-lod/
random-objects (false)
draw-otw (true)
specular-lighting (false)

If I set draw-otw to false, I get a sky colored screen, with occasional flashes of torn view that look like a horizontal band of what I should be seeing taking up about 1/6 to 1/4 of the screen. The flashes probably only last one frame, and do not occur every frame. The framerate is way up though. I'm not sure where the BPP number comes from, I'm running my xserver at 24 BPP.

Josh

_______________________________________________
Flightgear-devel mailing list
Flightgear-devel@flightgear.org
http://mail.flightgear.org/mailman/listinfo/flightgear-devel
2f585eeea02e2c79d7b1d8c4963bae2d

Reply via email to