Re: [TCLUG:17392] [OT] Some computing history?

[Philip C Mendelsohn <mend0070@tc.umn.edu>]

On Tue, 9 May 2000, Mike Hicks wrote:

> Anyway I was mostly just curious about the multitasking and possible
> protected memory functions and when they got in there..  Was CP/M a
> multitasking OS way back when the PC was being developed?

No, CP/M wasn't true multitasking, though maybe CP/M-86 was.  I'll ask my
Dad, 'cause he'd remember better than I.

> From what
> I've read, it seems that stuff came up in the mid-1980's (probably as
> the 286 was gaining popularity).

Oh, geez, no.  That stuff is at least 15 years older than that.  When my
Dad was in central engineering at DEC, they were using all that stuff in
RSX-11M in 1973.  And it wasn't really brand-spanking new
then.  Multitasking probably goes back to the late 50's, but I'm just
guessing and would have to look it up.

>  It also sounds like the system was
> designed much more intelligently than DOS -- the kernel actually knew
> how to handle serial ports and apparently had a terminal subsystem
> that quite similar to the way Unix handles things.  Of course, it
> appears that CP/M had a head start of several years on (Q)DOS.  
> Incidentally, the version of QDOS that Microsoft bought was apparently
> named 0.10...

Well, CP/M was better.  The fatal flaw with the PC is reliance on the
BIOS, which is not an interrupt driven BIOS.  That means that when you
call the BIOS, you get to wait for it to finish -- ick!  Linux salvages
things because it doesn't use it.  I have books on CP/M BIOS, but I don't
recall if it was interrupt driven, 'cause I didn't spend that much time
doing that level of mucking around.

Man, you're making me want to go write some Z80 assembly for old times
sake!

Phil M

-- 
Lottery:    a tax on people who are bad at math