Hi Paul. On 8/14/07, Paul Martz <[EMAIL PROTECTED]> wrote: > My understanding from the OpenGL BOF is that current hardware is already > OpenGL 3-like, so current drivers are essentially OpenGL 2 layered on top of > OpenGL 3 hardware.
So how far back in hardware generations will OpenGL 3.0 support? NVidia 6/7 series? I presume Mt Events is just 8+ series? Did you hear any gossip about the likes of Intel w.r.t supporting OpenGL 3.0. Also any mention form the likes of Mesa w.r. OpenGL3.0? >I believe the plan for most vendors is to allow apps to > create either an OpenGL 2 or OpenGL 3 context at startup, then use the > appropriate API for that context. Not sure whether mixing heterogeneous > contexts in a single app would be allowed, but it seems theoretically > possible based on what I heard. Mixing GL2 and GL3 in one app is probably not necessary. > On a related topic, I understand there will be a convenience library of > shaders to mimic old fixed function features. I didn't spot this in the published docs. Do you have a reference? > > Did the vendors give any hint as the timeline they expect? > > Nothing specific. We were told "vendors have OpenGL 3 drivers in > development, of course the spec needs to be finalized first". The latency > for public OpenGL 2.0 drivers was ~6 months IIRC and I imagine this might be > comparable. It doesn't sound like this will require radically new hardware > beyond what already exists. > > > Will 2.x still be supported in 5 years? I can imagine that a > > large number of apps would still not be ported by then. > > If the PEX/PHIGS/GL ISV shift to OpenGL 1.x is any indication, 5-6 years > sounds about right. OpenGL 1.0 came out in 1992 and by 1998 most all ISVs > had abandoned PEX, PHIGS, Starbase, et al. Perhaps this shift will happen > quicker, as OpenGL 3 hardware is already out, there are fewer IHVs nowadays, > and OpenGL is already widely accepted and has a proven track record. Personally I'd like to see a quick transition, and see projects like the OSG helping this. However, starting with a clean slate scene graph would not necessarily help as it would require users to migration to a new scene graph to get to GL3, something that they would put off as it could be a lot of work to port across. Users working at a high level would probably be able to port across more easily though - for instance if users just load a scene graph and then navigate round it using an osgViewer based app then the majority of the functions and classes will be identical. Paradoxically it'll be the advanced users that get their hands dirty creating state themselves that are likely to require the most retooling... Robert. _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
