Posted by Intenisfy on 9, 2001 at 3:37 PM:
In Reply to: Re: Naive question: can emulators go beyond emulation? posted by Mark Lord on 9, 2001 at 12:27 PM:
I understand the question.... and think it's valid, but, it misses the point of emulation completely.
Emulation, not only in MAME but in other emulators as well, tries to accurately reproduce the software running on the system it was programmed for.... slowdowns and all. (Remember, especially with MAME, the point is preservation and documentation NOT gameplay)
Of course, there are exceptions, like EMUDX and such which try to improve the game past it's original programming... Or the 1541 speedups in ccs64 (speeds up hard drive emulation WAY past the original specs, since it took like 10 minutes to load a game on the damn c64 :) --- )I'm sure it could be done, but I'd consider that more a simulation than emulation.
--Intensify
: Well the method of dynamic recompilation could be used (roughly converting the game's code to native x86 code). This would probably solve the slow down problems, but would probably bring new problems as well since the conversion operation is quite complex (especially if the programmers used a custom SDK or engine). bleem! uses dynamic recompilation if that tells you anything.
: : I mean, can emulators, for instance, not only emulate exactly a game, but also improve it by solving the original problems of it, specifically problems of speed. With MAME, you can unleash the power of your 800 mhz pentium 3 by deactivating the fps limiter, so the game runs faster, even too fast! But that's not the exactly what I'm thinking of. I'm thinking of solving the problem some modern 2D arcade games have, specially shooters: SLOWDOWNS. Do you think that in the future it will be possible to surpass the hardware limitations of the emulated hardware and, not only obtatin constant 60 fps, but also constant speed, no matter how many hundreds of sprites are on screen at the same time?