On Mon, Oct 3, 2016 at 10:21 PM, Al Kossow <aek at bitsavers.org> wrote:
the will probably be 68000
unisoft kernels i've used weren't 010 with the 451
i'll have to dig around for what bits of the 451 kernel i still
have around. unisoft kept the mmu parts pretty well isolated since
the did so many hw ports
I'd be interested in looking over something along these lines. I have
some 80s hardware that we put a 68000 + 68451 on, but it would easily
take a 68010. It was going to be a massive wad of async ports and
some sort of WAN port (X.25 was one technique). It never made it out
the door as a product, but I have the prototypes in the attic. Dual
Z8530 serial cards (4 ports) with a Motorola DMA engine or 8-port PIO
serial cards on top of a pedestrian 68000+68451 design w/2MB of 41256
DRAM and an early Sony 3.5" floppy drive - designed 1983 to 1984 ISTR.
We used a Perkin-Elmer 7300 Uniplus box w/System III as our software
design workstation because it was an inexpensive way to get 4
developers cranking out C code to 68000 object code to link into our
own format for the hardware target. I still have the P-E 7300...
tried to sell it at Hamfests 20 years ago and nobody was remotely
interested.
We never finished the app for this thing, but I always wanted to see
about tossing UNIX on it - plenty of serial ports, and at least a few
hundred KB of removable storage... just have to figure out what to put
on it in the way of a mass storage device - perhaps hacking one of the
DMA serial cards to add a 5380 SCSI chip in place of a serial chip.
So... short form - I'd love to read over some 68K UNIX code to see
what it would take to make it run on orphan hardware from 30 years
ago.
-ethan