>With that said, I guess what I kind of envision is
something very low level...
I've only done a little bit of VCS programming, but I've been lurking for a
long time. I don't want to discourage you from using C in the least, but
you reminded me of another idea I had a while ago, and haven't really
thought through.
A general-purpose, low-level language like C has a big disadvantage for a
specialized, resource-constrained platform like Stella. It's really easy to
clearly describe exactly what you want to do, as long as it's at the
"register level". And if you take it too far, you'll run out of resources.
The other extreme might be more interesting, from a computer science point
of view. In theory, you could define your own special-purpose, high-level
language, specific to Stella. Then you could directly express high-level
concepts like scanlines, cpu-cycles, TIA resources, etc. Ideally, the
compiler could figure out the best way to implement it on the 2600 - given
enough information about what you are trying to do. At first it would be
rather "dumb", but over time you could build-in the known Stella tricks
(register allocation, cycle counting, etc.). Of course, you would want the
ability to mix-in assembly routines easily, as well.
I'm not an expert in compiler theory or anything, but it seems like you
could make something really unique with that approach - truly leveraging the
power of your modern computer to make the most out of the minimal hardware
of the 2600.
-Frank
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://stella.biglist.com
