Hi Robert > I just read the paper. It's a interesting and should be flexible > approach to generating smooth roads by using coarse geometry and pixel > shaders to compute the tex coords to give a final smooth look. I > wouldn't have though it'd take too much work to implement either.
in case anyone is interested there is one more paper. http://www.j3l7h.de/talks/2005-02-01_Paving_Procedual_Roads_with_Pixel_Shaders.pdf > The OSG has very capable multi-threaded paging which is more general > purpose than this approach. What is missing is would the procedural > instancing and reuse of geometry for the trees/grasses. If you > implemented the later then you could use the standard PagedLOD for > handling all the multi-threaded paging to get the scalability. I'm long reader of osg list but my tasks are mostly in ai, physics, planning small team work. I read it to keep my understanding of graphics up to date . and now I'm temporarily even in video processing industry, so when I return to graphics world I think will do something like that > Intel's > job is selling CPU's not teaching one the best way to implement > something given the complete range of hardware available... yes still they implemented quite intresting way to separate and sync threads - which could be used to scale future demos based on osg ( not just OpenScenegraph will be split between threads, but physics, AI logic, etc - it is important in a light that there are seems a roadblock for Mghz for processors for a while and not only Intel will follow a road of many processors and cores. > With the integration of Present3D into OpenSceneGraph body it's my > plan to start putting presentations together that show off a few more > of the capabilities of the OSG in a more accessible way. this is a great news. Hope really there will be exciting demos in a near future. Best regards Sergey _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
