Thanks, shirro. I had gotten the impression that Eben, JamesH, et al, were providing some sort of bundle of the ES APIs, including anything special to know about using the GPU blob (there are Easter eggs in the blob? Do tell! Of course, eggs as a Python pun is fully intended ). Fortunately, I'm already pretty well-versed at ES 2.0 from iOS and WebGL (although I don't know of anyone who can remember it all without looking stuff up), so, I've got a leg up that most probably don't.
One of the most fundamental goals of Pi-finity! is to provide a means for kids of all ages to create 3-D worlds and games (whatever they turn out to be, sophistication-wise) without having to become Jim Clark (SGI founder and GL/OpenGL inventor) level experts at geometry pipeline manipulation. Although some bemoaned the move from ES 1.1 to 2.0 to become more aligned with the workflow in OGL 3.x (and 4.0 this year?), including GLSL, etc., the flexibility that provides more than outweighs the increased effort needed to get even just a tetrahedron with different-colored sides spinning under moving light sources.
I have enough experience with OGL (and GL before it), ES (essentially now a subset of OGL), and effective user interface design that I at least have an inkling about what can be done and, I hope, better ways to do it, in terms of providing world/game-creation tools (much of which I'm hoping to leverage from lessons learned by others developing earlier game engines and tools). I am going to need help with some of the more gnarly details about how things are organized and implemented in the Pi. However, I've been fortunate enough to attract the attention of some developers here who also have decades of 3-D and game development experience (not all with GL-based technology, though). Between us, I hope we can all work with the Foundation developers to get some impressive 3-D Pi stuff off the launch pad sooner, rather than later.
It's been interesting doing my initial work on a Pi via ssh and VNC to the UK from SillyCon Valley. It's crashed/frozen an average of about once a day, but, that may be due to a flaky micro-USB power cable. It reminds me of my days remotely logged into an SGI Iris 2400 workstation from my house late at night, solely on the command line over a 1,200 bps modem. Back then, I couldn't see what was going on with the display, and I had to either wait until the last users had logged off for the night, or Unix-messaged them to coordinate me blasting something onto the display (which wasn't always quite as coordinated as some liked ).
Of course, all I could do was verify that my code didn't core dump, and that status messages echoed to the command line and log files showed that things appeared to be working. I wouldn't get to do visual debugging until the next day/evening when my hour, or so, of reserved display time in the lab was available. Thank God (and an IBM resarcher, IIRC) for VNC, along with the folks who have brought us international fiber-optic network speeds. I'm typically only seeing about 200 ~ 500 ms delays round-trip! That sure beats the pants off operating blind and rapidly earning frequent-flier miles on red-eyes between SFO and LHR
The young whippah-snappahs today have no idea how good they've got it, dad-gummit!
The best things in life aren't things ... but, a Pi comes pretty darned close!
"Education is not the filling of a pail, but the lighting of a fire." -- W.B. Yeats
In theory, theory & practice are the same - in practice, they aren't!!!