[net.micro.pc] Future of PC-DOS ??

ward@chinet.UUCP (Ward Christensen) (02/21/86)

I'm not looking for any 'product disclosures', but rather just trying to
open a discussion about the 'possibilities' of the future of PC-DOS.  For
example, we all complain about not having enough memory.  I have a 640K PC,
couple external hard disks, etc.  But I have RAM disk, print buffer, etc,
and have for example only 10K after Reflex loads!
  SO, what's the future got to offer?  Lets say dos 4.0 (or whatever) will
somehow efficiently use more memory.  Does that mean it MUST be built upon
say an 80286, i.e. something with more addressing?  If so, can ANY programs
be able to run in 'protected' mode?  Or might the future of PC-DOS be based
upon expansion into multi-tasking, but using, say, low memory bank switching
to swap various users in and out?  (For example, leave the first 128 or 256K,
but swap out the next 256 or 384K?) (make above ..any EXISTING programs...)
  I suppose another possibility (especially with Microsoft's involvement
in LIM mem std.) that programs will be divided up into little chunks, and
will do the appropriate LIM memory swaps?
  Thanks in advance for any 'creative' ideas.  Maybe IBM/M'soft will be
listening!  (P.S. does the '386 offer anything to help - compatibility, etc?)

mdf@osu-eddie.UUCP (Mark D. Freeman) (02/24/86)

Summary:

In <354@chinet.UUCP> ward@chinet.UUCP (Ward Christensen) writes:
>I'm not looking for any 'product disclosures', but rather just trying to
>open a discussion about the 'possibilities' of the future of PC-DOS.  For
>example, we all complain about not having enough memory.  I have a 640K PC,
>couple external hard disks, etc.  But I have RAM disk, print buffer, etc,
>and have for example only 10K after Reflex loads!
>  SO, what's the future got to offer?

The March 25, 1986 issue of PC Magazine has several small articles/references
concerning a possible new standard in the works for memory resident programs.
The purpose of this would be to allow TSR (terminate and stay resident) prog-
rams to find out what other TSR software is currently installed and internally
compensate (so that three programs aren't trying to grab the same interrupts
at the same time, etc.), to ACCURATELY determine available memory, and to de-
install wherever order it was loaded in relative to other TSRs presently in-
stalled.

I also read that the new versions of most TSRs will know how to use Lotus-Intel
expanded memory, so that they don't consume all of your 640K before you can
even load in a 'primary' program like Lotus 1-2-3.

Note:  this will all become obsolete when Windows take over the world and/or
when DOS can address >640K, but I thought it might prove interesting in the
meanwhile.
-- 
< < < < < < < < < < < < < < < < < < < <> > > > > > > > > > > > > > > > > > > >
Mark D. Freeman						    mdf@osu-eddie.uucp
StrongPoint Systems, Inc.				    mdf@osu-eddie.arpa
Guest account at The Ohio State University		 !cbosgd!osu-eddie!mdf

I speak, therefore I disclaim everything I say.
< < < < < < < < < < < < < < < < < < < <> > > > > > > > > > > > > > > > > > > >

chinn@butler.UUCP (David Chinn) (03/03/86)

> 
> Note:  this will all become obsolete when Windows take over the world and/or
> when DOS can address >640K, but I thought it might prove interesting in the
       ^^^^^^^^^^^^^^^^^^^^^

       As I understand it, DOS itself can address >640k, it is the
       way IBM mapped out the memory that imposes the 640k RAM 
       limitation.

    ... uw-beaver                        david m chinn
	   !{tikal,teltone}                      box 639
	       !dataio!butler!chinn    		 redmond,  wash 98073