Try using glDrawElements with an appropriate indexlist.
Try increasing the complexity of the fragment shader, (e.g by multiplying with one more uniform vec4) to see if the performance will continue...

If you don't use nSight, I'd suggest you use a fragment shader on a fullscreen triangle. (to avoid high primitive-setup costs, and <= 4 primitives per cycle setup, and transform-feedback...

The Xilinx ISE WebPack is available for linux, yes.

I am happy with: http://www.amazon.co.uk/SainSmart-Cyclone-EP2C8Q208C8N-Development-Blaster/dp/B006CHFE7Y/ref=sr_1_1?ie=UTF8&qid=1366972772&sr=8-1&keywords=altera+fpga
Most people prefer Xilinx, so:...

Thanks for this tool!
I have an immediate use for it :D.

Things to improve, but can live without them:
- Keyboard shortcut to autosave+compile;
- mark compile-error lines

I'd try something like:
1) remove all duplicate vertices
2) manually choose several triangles from the start of the intestine, which form a ring
3) repeat until no triangles left:
3.1) pick the...

Try
vec2f=fract(texcoord.st*vec2(width,height) - 0.5f);

Pixel center is {0,0} , so {-0.5,-0.5} is its bottom-left coord.

The modeler already gives you the normals per face-vertex exactly the way you want them. No need to parse "s" yourself.....

You don't need to handle smoothing-groups.
The modelling app initially uses them to generate normals per face-vertex; later to subdivide (tesselate) just before raytracing.

You only need to...

The multisampling one seems invalid: to do any proper custom blending, you will have to do per-sample processing: fetch each individual texel sample, calculate, write one sample. Exactly the same...

I don't see issues in adding this to core. In the worst case it's a glFlush() + added flushing of texture caches.
But for most things, I think it will be slower than image_load_store. (could you...

Could you test with this in the fragment shader:


layout(early_fragment_tests) in;

How do you check the value of y, what gpu and drivers are you using, can you reproduce it on another platform?
(it looks like a driver bug)

The compiler will create a temporary automatically, if it needs one.

Distance between current mouse position and previous mouse position. sqrt(dx*dx + dy*dy)

I bet you called glGetString(), though

Easiest way: if the distance between A and B (start and end position) is 50.3 pixels , draw 50+1 points. Each being slightly offset, the last one matching the end position.

Harder way, supporting...

I think I'd go with a solution like:


#ifdef GL_EXT_shadow_samplers
#define SHADOW2DPROJ(tex,coord) shadow2DProjEXT( tex, coord );
#define SHADOWSAMPLER(tex) uniform sampler2DShadow tex
#else...

Do a depth-only render with the objects being solid. You might need to add some depth-bias to the wireframe pass afterwards.

What about manual comparison of Z ? I think it should work, and furthermore not require that extension.

Things to try:
Make sure filtering is GL_NEAREST, and set before the compare mode;
Try doing the ref comparison manually.
glTexParameterf instead of glTexParameteri, the enum handling might...

glDrawPixels(Width,Height, GL_LUMINANCE, GL_UNSIGNED_BYTE, Data);

glDrawPixels(Width,Height, GL_RGB, GL_UNSIGNED_SHORT_5_6_5, Data);


http://www.opengl.org/sdk/docs/man/xhtml/glDrawPixels.xml

Could you test agnuep's demo, that does exactly the same:

http://rastergrid.com/blog/2010/02/instance-culling-using-geometry-shaders/

On startup (without navigating through mouse/kb), it gets...

Try
glEnable(GL_RASTERIZER_DISCARD);
during the transform-feedback.

Your gpu isn't old, so cannot be plagued by GF8800-type geometry-shader slowness.

The third option is fastest:
attach everything to the same fbo. (provided you have only one depth-buffer). Use glDrawBuffers() to switch around stuff. I think in most implementations the...