As far as Graphics go, Graphics context creation, and that’s it please.
If someone wants to write a cross platform game, they’ll use OpenGL.
If they want to appease Microsoft, they’ll use Direct3D.
And if they want performant code on all such systems, they need to write
both OpenGL and Direct3D code that basically do the same thing, but are
more optimized to their specific platforms. So one of the main reasons for
using SDL is lost.
I don’t want or care to know how SDL does its thing behind the scenes.
I just want it to be fast in both Windows and Linux/OSX. And that means
Direct3D for Windows, and OpenGL for the others.
Having everyone write both Direct3D and OpenGL code for their projects is
a complete waste of developer time. Sure, doing it in SDL is difficult, but
it’s a one-time job. Better to do it in one place, and have it optimized
and looked over by many people than to have every separate project do it.
whatever new API comes along. I want a simple layer (aka, SDL) over these
APIs that gives fast access to a framebuffer, allowing it to be stretched
and rendered in 2D. I agree that if you’re going full-on 3D mode then you
probably need to write API-specific code. But don’t force the average 2D
game developer that just wants speed to code multiple rendering backends.
Thanks,
Steve A.
Stella maintainerOn January 21, 2012 2:52:58 PM PoV wrote:
From my POV, I don’t want to write Direct3D code, or OpenGL code, or