Ancient 8086/80286 unixes?
dkelvey at hotmail.com
Sat Sep 8 09:36:16 CDT 2007
> From: cclist at sydex.com> To: cctalk at classiccmp.org> Date: Fri, 7 Sep 2007 10:07:40 -0700> Subject: Re: Ancient 8086/80286 unixes?> > On 7 Sep 2007 at 9:58, Bernd Kopriva wrote:> > > Steve Ciarcia's Trump Card (Z-8001)> > >From the article at http://www.dtweed.com/circuitcellar/trumpcrd.pdf> > It seems that the big advantage cited by Ciarcia was the ability to > run a compile-in-place languge called TBASIC to gain performance. > Given the prices on his cards, it hardly seems like a bargain.> > If one wanted to goof around with the Z8000 family as a coprocessor > card, I'd be sore tempted to wire up something with a Z8002 and 64K > of SRAM. You get the instructions without the expense and you don't > have to deal with the (awful) Z8000 segmented mode. And relatively > easy to do a lashup. Instead of Ciarcia's bucket, you could probably > do with a single Z8036. Fit the whole thing on a "short" card.
You can still use a Z8001 because it boots to non-segmented mode
and looks like a Z8002. The trick is to find the memory management
chip and add it to the Z8001. One can still use the Z8001 without
this and treat it as a Z8002. You can just ignore the segments.
Still, if you want to get the CP/M8000 running, you'll need to
embrace segmented code.
Once you are running segmented code, the only reason you need
to worry about it is that you can't run continuous code or data for
more than 64K.
The main disadvantage of running segemented is that many instructions
take more code space. One can still switch back and forth between
segmented code and non-segmented. This saves a lot of code space.
If one does their memory decoding correctly, one can have 64K code
and 64K data space in either segmented or non-segmented modes.
( the M20 does this for part of the memory map ).
> > IMOHO, only Motorola and NS of all of the "16 bit" chip producers > ever understood the importance of a continuous non-segmented memory > space. Given the time of introduction, I consider the 68K to be a > marvel of MPU design. Too bad IBM didn't adopt it for the PC.
That is true, it mades the 68K look real good for it's time. Still,
it had a memory limit, just bigger than 64K.
> > Cheers,> Chuck> > >
Kick back and relax with hot games and cool activities at the Messenger Café.
More information about the cctalk