Removing the function call overhead is not what I call an *OpenGL*
optimisation. This should be left to the programmer to cache what he
needs in order to make less OpenGL calls.
What we need really are better render targets and fast/functionnal VBOs.
Francois Hamel
-----Original Message-----
From: mac-opengl-bounces+fhamel=email@hidden
[mailto:mac-opengl-bounces+fhamel=email@hidden] On Behalf
Of Jose Commins
Sent: Tuesday, March 29, 2005 1:45 PM
To: email@hidden
Subject: Xcode and OpenGL optimisations
I've been reading up on the future plans for MacOS and
especially OpenGL optimisations. One thing which has always struck me
about OpenGL is the function call overhead. I've always thought that
Xcode could be tweaked to replace GL calls with code that writes
straight into the card's display list (something we did a long, long
time ago with one of our graphics chips - and before anyone pips up with
the question, yes, we did use a hardware mechanism to make the list
thread/multi-cpu safe :).
Of course one can use OpenGL macros in Xcode to make things a
little less painful (highlighted by a discussion on this list in regard
to avoiding a context-check on every call) and I am curious as to what
could be on the horizon about further speeding up OpenGL on the Mac.
Regards,
Jose.
--
Death, taxes and Microsoft. If you put it that way, the first two don't
seem so bad.
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Mac-opengl mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/mac-opengl/email@hidden
This email sent to email@hidden
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Mac-opengl mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/mac-opengl/email@hidden
This email sent to email@hidden
