Andy Ross wrote: > Erik Hofman wrote: > > While I don't see a direct improvement in framerate I notice a real > > effect on the screen update. The old behaviour had a small bump in the > > update every second or so, while the new code elliminates that. > > This doesn't make much sense. All of the changes in that patch were > inside the per-frame loop. They certainly couldn't cause or fix > stutter over many frames, nor do I think have they been claimed to. > Perhaps something happened in the tile loader to do this? > > This is rapidly getting on towards voodoo coding, and I think perhaps > we should step back a bit. What, exactly, are the changes in this > patch that make it worthwhile? What does it eliminate? What is the > evidence for speedup? > > > BTW. This is probably only noticable on slower machines with a "slow" > > OpenGL implementation like the O2 I work on. > > The patch didn't touch any OpenGL code so far as I can tell, so I'm at > a loss to explain this one, too.
Well, for both remarks, the part of OpenGL that isn't hardware supported must be emulated by the CPU. Streamlining code might gain some speed for OpenGL which otherwise might have been waisted by something else. So that *could* be an explanation. Erik _______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
