Re: --> Re: Turbo C++ & PCT

[dempson@swell.actrix.gen.nz (David Empson)]

In article <1994Feb6.234105.29718@iitmax.iit.edu>
CRAFJAY@minna.acc.iit.edu (Jay Craft) writes:
> 
> Ahhh. The package says that it requires 1 meg of memory and a 286.
> Did the other versions also say that, but worked anyway?  

Turbo C 2.0 requires 448k and an XT.  I don't have the TC++ 1.0
manuals handy, so I can't check that.

A few programs which claim to require a 286 will work on the PCT.  The
PCT's processor is an NEC V30, which has the instruction set of the
80186.  Several "286" programs are just using new instructions which
were introduced on the 186, so they will also work on the PCT.  An
example is 4DOS (replacement command processor), which improves its
performance by using 186 instructions where possible.

In other cases, they expect an AT-compatible BIOS and/or extended
memory, neither of which are present on the PC Transporter.

> Isn't there some kind of primitive virtual memory that works on an XT
> that might work?  Hmmm..

I doubt it.  If BC++ needs a 286, then it is probably expecting to use
extended memory, which cannot be simulated on an XT.  You might be
able to simulate XMS and/or EMS with some BIOS extensions or a TSR, and
help from the Apple, but it would be slow.

I've heard that you can get an EMS driver which uses the hard drive as
if it was expanded memory.  This is _very_ slow (especially with the
PCT's snail-like hard drive access speeds).

> I'll bet a company could make a few $$$ if they designed a, say,
> 386SX-on-a-board that was similar to a PCT... Or if AE would've 
> upgraded the PCT hardware.

... or the software, for that matter.
-- 
David Empson                                                               
dempson@swell.actrix.gen.nz                                                
Snail mail: P.O. Box 27-103, Wellington, New Zealand