On Wed, Jan 9, 2013 at 2:32 AM, Michael Robinson
<plu...@robinson-west.com> wrote:

> A protected mode dos like the one under Windows 9x and Windows ME
> could be interesting and would justifiably deserve a different name
> like Freedos-32.  The problem with a dos environment is that there
> isn't an operating system taking care of all the hardware and
> providing standard calls to use it.  Most sound card support
> involved adding to your program in most likely a spaghetti fashion
> calls to a third party driver, closed source of course.  Windows 98
> may have had multitasking, but if that is true, it was more than
> just a single thread dos system.

I wouldn't call that a "protected mode DOS".  Win98/ME used DOS as a
real mode loader for Windows.  The protected mode portions were in the
Windows code, and once Windows was loaded, DOS was out of the loop.

> Gates made some very bad
> assumptions that crippled dos back in the day.  Assumption one,
> nobody will ever need more than 640k of memory for executable
> programs and drivers...  I imagine that other bad assumptions
> were made as well.

That wasn't a Gates decision, it was an IBM decision.  The 8088 CPU
used by the original PC had a one megabyte address space.  IBM chose
to reserve the memory above 640K for system level functions, and
confine user programs to the area below that.  Gates was writing to
IBM specs.

Given that you have a megabyte available, total, and some *will* need
to be reserved for the system, where *do* you draw the line?

> Actually, there is OS/2 which was supposed to be the competitor to
> Windows 9x and I'll bet that IBM is willing to release source code
> to it.  Maybe the freedos community should get it's hands on OS/2
> and develop it further.

If IBM is willing to release code, that's news to me.  The folks at
Stardock, who make the Window Blinds skinning utility, began under
OS/2, and tried to acquire the rights to it when IBM ceased
development.  IBM wasn't selling.  You can still get OS/2 from an
outfit called eComstation:

> What I'd like to see at this point is a focus on debugging and a focus
> on deploying Freedos via a rom chip.  It should be possible to get write
> once 1 meg+ memory chips now.  Why not install the freedos kernel,
> command.com, etcetera on such a chip?  If you can't overwrite the
> operating system executable, security is enormously improved.  For low
> power embedded processors that are say only 8 bit, freedos may be very
> useful.

You would do better to start with DR-DOS, which originated to serve
requests from Digital Research customers who wanted a ROMmable version
of DOS.  At the time, MS-DOS was not architected to allow the
separation of code and data putting DOS into ROM would require.
Selling DR-DOS over the counter was a later move.

> A hypervisor that can run dosbox and make modern hardware work
> with old dos programs anyone?  How about dosbox running on a Pentium 133
> or a Pentium 166 machine with 16 megs of ram?

Insufficient demand to justify the effort.

Master Java SE, Java EE, Eclipse, Spring, Hibernate, JavaScript, jQuery
and much more. Keep your Java skills current with LearnJavaNow -
200+ hours of step-by-step video tutorials by Java experts.
SALE $49.99 this month only -- learn more at:
Freedos-user mailing list

Reply via email to