Fixed Function Pipeline
Removing the fixed function pipeline is simply a bad idea. Lets face it, as an indie-developer, we are targeting 3 year old intel on-board chipsets... maybe the fixed pipeline is available via effects files?
I have a nice little terrain engine in C++ - using DX8 - most users don't even have the big nine installed - that I would like to port over to XNA... it uses the fixed function pipeline and can blend 2 textures per patch via a coverage factor and modulate this with a shadowmap (if there are 4 blend stages)... and runs with high framerates even on intel on-board. For older hardware with only two stages, it simply do two passes.
So for indie-developers, wo loose 95% of our target audience due to the removal of the fixed function pipeline... grrr, not nice.
Now we could blame intel for doing their crappy hardware that has a wider installation base than all other "big players" have together... or just support the damn fixed function pipeline.
Another big deal is deployment, as others mentioned in this forum. In the past I've done some experiments with Java and a thin OpenGL wrapper to do casual game development - there were no problems to have a program local VM (with stripped down libraries) that just bloats the install around 5MB, is this possible with the .NET Framework too? An How-To would be nice here.
Thanks in advance,
Sascha.
P.S. I'm typing this message on a computer with Intel(R) 82845G Graphics Controller :-D
Confusing new programmers? That should not be the reason to remove the fixed function pipeline. It's like removing PL/SQL from Oracle as this can confuse SQL Server people... Let's remove Visual C++ from the studio line, it's confusing for Visual Basic users... I don't see a problem in exposing the fixed function pipeline.
Loosing fixed function is not a good idea just for that reason. It is
always nice to have fallback paths for older hardware. What I can see
in the bright XNA future are totally crappy Pong clones that need at
least shader 2.0 and cards with 256MB... come on.
:-D
I'm a big fan of C# and do all my tools with it via native interop. So it seems like a natural decision to use XNA and port my engine over. And it would be nice to have the chance and run a game on XBox360 and
maybe sell it sometimes in the future via XBox Live...
Agent Mulder: "The intel hardware is somewhere out there..."
Agent Scully: "So we have to support it!"
Personally I think dropping the FFP is a good move.
The varying capabilities of 3D cards has been one of the big pains of PC game development. There is no fixed platform, so you either write several versions of your renderer code, shaders, and so on, or only support the lowest common denominator (FFP).
Now we will have a forced fixed platform (Shader Model 2.0+), which will bully out inadequate hardware like Intel graphics chips (which sadly are in this laptop I'm using too).
