Hi,
On Tue, Mar 20, 2018 at 5:57 PM, dmccunney <dennis.mccun...@gmail.com> wrote:
> On Tue, Mar 20, 2018 at 3:05 PM, Dale E Sterner <sunbeam...@juno.com> wrote:
>>
>> HP's resources are so great that it would just be a tiny project for them.
>
> HP does things for *money*. There is no money in developing drivers
> for DOS, and the developers who are *capable* of doing it have others
> things to do for HP that *will* generate revenue. People who can
> write drivers do not grow on trees.
Indeed, many developers (e.g. Linux kernel) are paid contributors.
Even GNU tries to fundraise via paid support.
> It's the same reason you don't see drivers contributed as volunteer
> efforts. Programmers who *can* do that will spend their time doing
> stuff they get *paid* for.
No, there are actually still a lot of volunteers. Most software
projects are done on a voluntary basis. While some of them aren't very
important (but still fun), many good things would grind to a halt
without certain select volunteers. It's definitely more of a an
"amateur" (for the love of it) niche rather than commercial. Even Pat
Villani wrote his FreeDOS book as a labor of love (and took something
like two years of his free time to do it).
Things like GNU are meant to be open to everyone, commercial or
otherwise, as long as you're willing to do the work, and then continue
to share the fruits of that labor freely. Admittedly, proprietary
software still has a big stronghold, but it's not true that nothing
good comes without money. Hard work and dedication come for many
reasons.
> DOS was an OS created for 16bit bit CPUs with hardware imposed limits
> on the amount of RAM accessible. Early CPUs were a lot *slower*, too
> (Remember the 8088 used in the original IBM PC ran at 4.77
> *megahertz*, and could address 1MB RAM.)
The 8086 was meant to succeed the 8080, so one of the requirements was
to break the 64 kb [sic] barrier, going at least as good as 128 kb.
Even back then, despite memory scarcity and high cost, there was still
a growing need for more room to breathe. Actually, I think the 8086
was originally a stopgap measure to compete with the Z80. The iAPX 432
was meant to be the real cpu of the future. When those plans stalled,
they kinda went back and just extended the 8086 (286, 386, etc).
OS/2 1.x didn't appear until 1987 or so (requiring 286), and RAM
totals were still low. Windows 3.0 (1990) still supported real mode!
Even Win95 could (allegedly, slowly) run atop a 386 with only 4 MB of
RAM (thus probably why bloatier NT wasn't propagated to end users for
several more years).
> Current CPUs are all at least 32bit, and increasingly 64 bit, capable
> of accessing gigabytes of RAM, with multiple cores and *designed* for
> multi-tasking that DOS doesn't *do*.
One of the new features of OS/2 was breaking the 640 kb barrier, but
DOS extenders beat them to it!
DPMI was invented by Windows 3.0 as an improvement over VCPI. Thus,
you could better cooperate with DOS extenders under multitasking
systems.
Borland, IBM, Novell and many others used it. I think even MS' last
16-bit OMF linker was PharLap extended (circa 1994 for VC 1.52).
Most DPMI servers are either bound to the .EXE itself or run as
loadable TSR. So no, it's not really bundled inside the kernel itself,
but it still achieves the same goal. (While there is 16-bit DPMI, most
are 32-bit.)
You can access gigs of RAM with 32-bit DPMI. And DJGPP has been around
for ages (starting in 1989, still updated). Even other attempts were
made (EMX, MOSS, HX).
Novell DOS 7 / DR-DOS 7.03 supported multitasking (but only 64 MB per
task). Not sure of all the other obscure DOS clones, but there were
many attempts over the years (Desqview/X? that used DJGPP v1 for its
SDK).
Don't forget that Doom (1993) was DOS extended (32-bit) via Watcom.
Similarly, Quake 1 (1996) was built using DJGPP v2 (beta) and CWSDPMI
r1. No, not a port, but the original commercial versions!
> FreeDOS on hardware like that as the main OS is a pointless waste of time.
Not bad for running DOS stuff! That's the whole point. Granted, it
doesn't support billions of hardware drivers by default like the Linux
kernel, but it still works.
Also, it's good for benchmarking. But most people (or so I thought)
kept it around for the low-level access and easy way to reformat /
rescue / reinstall things.
But admittedly, the days of flashing the BIOS in DOS are almost gone.
Hey, you can also do programming for fun, word processing, or gaming
under modern DOS, but the big lacks are things like sound support
(which is relatively unimportant overall, IMHO).
> It can't make use of the hardware. It only becomes reasonable as a guest OS
> under something
> like VirtualBox, VMWare, or Xen, where it's one of an assortment of things
> the machine is doing.
It certainly can, potentially, but there aren't a lot of public
("free") drivers for many things. FreeDOS is a volunteer project, and
(despite difficulty) it has tried fairly hard to make things easier.
Too bad it's so underappreciated.
The idea that money will fix all ills is only weakly true in rare
cases. Microsoft has tons of money, so why haven't they done xyz?
(Because even they can't do literally everything.)
> The currently ballyhooed Internet of Things is occurring because stuff
> like 32bit ARM CPUs which can actually run a full TCP-IP stack and
> connect to the Internet are cheap enough to be embedded in places were
> you might have used an 8bit CPU at best in the past to keep the costs
> affordable. Hardware is *cheap* and getting cheaper.
Talk is cheap. (A poor carpenter blames his tools!)
And actually, I'm told that graphics cards doubled (or tripled?) in
value due to cryptomining. So scarcity can certainly drive prices up
again. Don't expect that even RAM prices will always be cheap.
DOS can do some networking (although no Javascript in the browser),
but you need suitable packet driver. That is indeed hard to find, so
Linux does have better support (but there's no obvious reason it
couldn't be ported to DOS somehow).
> But you really need to learn more about what has occurred in computing
> since you became proficient at DOS and DOS apps, whether or not you
> actually use it.
Give me a break. If we caved to every fad, we'd have nothing left. DOS
is maybe an unpopular example, but checking old mirrors of Simtel or
Garbo shows about twenty years worth of software. TWENTY YEARS! And
there's still tons more. Sure, you can discount anything if you try
hard enough, but it's unfair to say that it is all useless in the
modern age. Software productivity doesn't work like that. Some things
will never be rewritten for Linux, and some "old" things still work
fine. Hey, I like Linux (et al), but let's not pretend that it can do
literally everything.
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freedos-user mailing list
Freedos-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-user
