Hi Alex,

On Mon, Nov 29, 2010 at 10:22 AM, Alexander Irion
<[email protected]> wrote:
> another question to the IncrementalCompileOperations: I did some local 
> adaptions in it to ensure a flawless constant frame rate on my system.
>
> 1.
>
> First is to take the estimated compile time into account, whether an object 
> can be compiled in this frame or not. Now, I saw that you began to add 
> something similiar, but it's not yet used as a criteria for the loop. Will 
> this be added?

Yes, I'll be adding the estimated time to compile into account.  First
I need to get a reliable time estimation, in my own tests the results
have been rather surprising and not amenable to the straight forward
time estimation model I've used in initial testing.
> 2.
>
> Measuring the CPU time that is left for the current frame is good, but it's 
> useless if the GPU is behind the comands in the pipeline. To circumvent this, 
> I added a glFinish after each texture or drawable upload. Are you planning to 
> add something similiar for this?

glFinish breaks the asynchronous facility provided by the OpenGL fifo,
using it would lead to lower frame rates and increased likelihood of a
frame break.   I've thought about adding the option of doing a
glFinish() in different parts of the frame, but it won't be the
default a the OSG is primarily a tool for getting the best performance
out of the hardware.

Robert.
_______________________________________________
osg-submissions mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org

Reply via email to