Hello,
I've just look at the VAO branch, I have not tested -not enough HD space:/ so I 
can't give a full feedback
I can only compare with my own impl with this reduced perspective, I don't 
understand all so don't be regardant if i misunderstood the code:
-I like the idea of deporting vao in state..It seams a good place for them for 
state sorting.
-But I haven't spot the way you intend to share VAO among Geometries:
*you don't seam to use a singleton for bo affectation to geoms (so you're 
assuming user is in charge of sharing bo correctly among geoms?) 
*it seams you haven't changed primset so I'm asking myself how could you manage 
basevertex when  geoms arrays share same bos 

In my impl, I use a singleton to manage bo sets involved in vaos, it 
concatenate bufferdata in these bo and so bos and vao are shared among 
geometries
Further I also use drawElementBaseVertex to draw them in order to take into 
account offset induced by concatenation of bds in bo 
I did a test in debug build (on a GTX640) with a triangle indexed city and 
default useVBO 
it originally frames at                             40FPS 
+adding bo sharing among geometries :  55FPS
+adding useVAO                                    70FPS(The Draw charge drops 
from 12 to 6 even if i unbind vao after each geom draw)

https://raw.githubusercontent.com/mp3butcher/OpenSceneGraph

Again, these are just observations of gl features that accelerate drawing and 
that i would be glad to find in osg and not guidelines on what/how to achieve 
their impl.

I'll give you further feedback when i'll have actually test the branch


robertosfield wrote:
> On 8 July 2016 at 13:49, Mathieu MARACHE <> wrote:
> 
> > Great news ;-)
> > 
> 
> Was hoping to finish it today, but have a birthday family walk and
> dinner planned for the rest of today so heading off line.
> 
> I have a prototype of VAO functionality kinda working.  Performance is
> on big city models better than without VAO but still not as good as
> with display lists...
> 
> I'm currently refactoring the internals of how vertex data is passed
> to try and avoid some of the CPU overhead.  I expected there is couple
> more days work getting it all working smoothly.
> 
> Robert.
> _______________________________________________
> osg-submissions mailing list
> 
> http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
> 
>  ------------------
> Post generated by Mail2Forum


------------------
Read this topic online here:
http://forum.openscenegraph.org/viewtopic.php?p=68365#68365





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

Reply via email to