cem@intelca.UUCP (Chuck McManis) (10/31/85)
Consider the history of microcomputers for a moment (this is net.micro
after all) and you will see that binary compatibility is VERY important.
CP/M (the first recognized standard) was based entirely on 8080 binary
compatibility. Later the Apple II became a "standard" due to the shear
volume of Apple IIs and clones. Then the IBM PC came out and it is now
the "defacto" standard in microcomputers.
Then there was the 68K and it had a.) No major supporter (CP/M-68K ?
even DRI didn't seem to like that O/S) and b) No recognition until
Apple came out with the Mac. So someone gets the idea to port UNIX to it.
Great that make 15 incompatible operating systems for one architecture.
Binary compatiblity makes it worthwhile (read profitable) to develop
software for a given processor, if the kernel is sufficiently flexible
then it is even better since a wide variety of hardware is also supportable.
The only way I can see UNIX becoming the "standard" will be if someone
(preferably AT&T) can develop a p-code (u-code ?) that all compilers could
generate, then all "ports" of UNIX would include the p-code assembler
as part of ld, it would have to be a very clearly specified code. Only
then would software portability be viable, the next problem would be
magnetic media format but that is an entirely different can of worms.
--Chuck
--
- - - D I S C L A I M E R - - -
{ihnp4,fortune}!dual\ All opinions expressed herein are my
{qantel,idi}-> !intelca!cem own and not those of my employer, my
{ucbvax,hao}!hplabs/ friends, or my avocado plant. :-}peter@graffiti.UUCP (Peter da Silva) (11/04/85)
> The only way I can see UNIX becoming the "standard" will be if someone > (preferably AT&T) can develop a p-code (u-code ?) that all compilers could > generate, then all "ports" of UNIX would include the p-code assembler > as part of ld, it would have to be a very clearly specified code. How about pcc intermediate code? Since most 68000 and other non-PDP-11 UNIX 'C' compilers seem to be based on PCC, and since the interemediate code should be hardware independant (modulo byte ordering), this would allow 'C', Pascal, and Fortran-77 portability. Pcc mightn't be the fastest or the best compiler, but it's common enough... -- Name: Peter da Silva Graphic: `-_-' UUCP: ...!shell!{graffiti,baylor}!peter IAEF: ...!kitty!baylor!peter