21 Dec
2016
21 Dec
'16
11:06 p.m.
On 21 Dec 2016, at 21:37 , Michael Kjörling wrote:
Just imagine how differently things would have turned
out had a major
software company, in the dawn of the 1980s, decided to seriously
pursue its UNIX variant rather than a clone of CP/M as it turned out.
That could refer to either IBM or MS. At the time IBM perhaps wasn't a software
company and MS wasn't major.
In any case, I don't think it would have made much difference:
[1] Even early Unix did not work well without a hard disk and/or ample RAM. Two years ago
I ported LSX and whilst it is amazing to see a minimal Unix work with only 48KB and a
floppy disk, it is much less usable than CP/M on 8088 class hardware. The constant
swapping to floppy disk kills performance and the 3 FIFO process design quickly becomes a
nuisance to the Unix experience.
[2] Hard disks and 256KB RAM did not become common on PC's until perhaps 1985, when
DOS had already entrenched itself. Also, by 1985 almost half of all computers in the world
were PC's.
[3] By 1980 CP/M had already an eco system with some 8,000 applications around it,
including full screen word processors, spreadsheets, etc. CBASIC and dBASE had already
spawned a cottage industry for line-of-business apps. These could be ported with modest
effort to DOS. It would have taken years for a similar Unix eco system to develop.
In short, my guess would be that if IBM and MS would have pursued PC/IX and Xenix for the
PC in the dawn of the 1980s, Gary Kildall would have become a billionaire pursuing CP/M-86
for the PC.
Paul