> I've been interested in writing my own emulators as well. The hardest and most
> important part ofcourse, is actually emulating the cpu. If you aren't using
> someone else's cpu core, you have to make your own. This can definitely take
> awhile, and will require accurate documents.
I'm just really curious about this. I had heard that too, and in fact, it was tricky, but I wrote a Hu6280 core a long time ago, for a planned TG16/PCE emu, and after I had it sort of minimally working (stepping bootcode in a debugger), I hit a point of "now what???"
It's weird, I had thought that was the hardest part, and yet, I failed (at the time) to be able to see the big picture, and make everything come together as an "emulator", not just a raw CPU core.
Hrmph. Maybe I just don't "get" emulation, somehow. Maybe I don't want to understand it fully, perhaps it would remove the "magic" of it all. I seem to be very good at low-level technical specifics (my background is asm coding, preferably directly to hardware interfaces), but can't seem to grasp the overall picture well enough, or something.
I admit that I'm rambling a bit, it's kind of late here, but.. do you have any suggestions? There have been so few very high-level (aka "theory of emulation") design documents/papers on the subject.
I think my life is just an asymptotic graph, never *quite* able to cross over the "finish line". :|
I think I'm slightly losing my mind too. Hrmm. Sorry for my rambling everyone.