inefficient
Veteran
Nite_Hawk said:Shifty Geezer said:That doesn't sound to me like on the fly backwards compatibility. If it is, why are only a select number of games going to work with it? Also the statement ends with 'and will most likely use emulator software for those limited titles'. Is that MS talking, or IGN?
Still seems rather inconclusive either way to my reckoning. :? I't be nice to hear from Allard or Bach how BC actually works.
My guess is that they are using a combination of approaches if they are not just recompiling the games. They have the API on the new machine which theoretically is backwards compable with the old API, so once you have windows API calls they should theoretically work. (This is basically what happens in wine under linux). The other problem they have to solve is converting the x86 code into PPC. This is the tricky part. Do they do CPU emulation and try to make a virtual celeron core run on one of the PPC units, or do they try to make some kind of C++ dynamic recompiler? Emulating the entire core would probably take a lot of work, and it'd probably take a lot of time to get it bug free. Doing dynamic C++ recompilation may be easier, but what do you do about any asm code that was written?
Nite_Hawk
Dynamic C++ recompilation? As if the c++ source code was distributed with games...
And regarding wine. Go try play some win32 game on a non x86 linux box and come back to rethink your theories.