>Yes, backward compatibility is a sham. x86 chips are not compatible
>with Pentium are not compatible with MMX, etc. Macs have the same
>problem. The 86xxx motorola chips are not compatible with the PowerPC
>chips. MacOS won't be compatible with Rhapsody.
Untrue. I have never encountered, nor have I ever heard of anyone who has
encountered, a valid 680x0 (not 86xxx, btw) instruction that did not run
correctly on a PPC chip (albeit in emulation).
As to your assertions re Rhapsody - the "Blue Box" layer of Rhapsody is a
fully-implemented MacOS layer. "Yellow Box" provides a similar
implementation layer for Windows 95 and NT. Note that these are fully
native implementations, not emulation layers. That means that a machine
running Rhapsody OS can run MacOS native apps, Windows NT apps and Rhapsody
apps, all at the same time.
Rhapsody is being developed in versions that run on both PPC and Intel
chipsets (and perhaps some others as well...). Rhapsody coded apps should
be fully portable between processors - everything I've seen so far leads me
to believe that Apple can accomplish this and will do so ontime or darn
near to it.
I'd go on but a) this topic isn't really salient to this list, and b) I'm
under NDA and really shouldn't get too enthusiastic about sharing what I've
been working with til it's released.
Anyone interested in Rhaposody should check out
http://www.apple.com/macos/rhapsody/ for starters.
>If something is "99.999%" compatible and executes thousands of
>instructions per second, how long will it run before you hit that "rare"
>incompatibility issue?
Depends. If an instruction that can cause an error is rarely called, then
it may be years or never at all. It's not like a computer executes every
possible instruction just for yucks.
! Jeffrey Fabijanic, Designer | "The Future exists,
! Primordial Software | First in Imagination,
! "Software of the First Order" | Then in Will,
! Boston, MA * (617) 983-1369 | And finally in Reality."