Josh's guide to Operating Systems and the Processor Wars Or A Hard Look Into
the Crystal Ball by a Nearsighted Geek By Josh Turiel
Currently, users are being pummeled by the hype factories with all sorts of
propaganda on the different OS's (Operating Systems) and processors. IS
(Information Systems) workers are panicking, nasty letters are being written to
all the various magazines telling why X rules and Y sucks, and flame wars rage
on Usenet about these topics. What I'm going to try to do here, with a minimum
of bias and flaming, is briefly explain my views on the subject by listing all
the main contenders and what I think are the good points, bad points, and
prospects for each.
To explain where I'm coming from, I'm the IS manager for a small Boston ad
agency that does retail-oriented, graphic work. We use mostly Macs as clients,
but we have a few Intel boxes, too. These are networked with Intel-based
servers. I'm relatively platform-neutral myself, though I tend to prefer my
Mac a tad, and my own choice for my writing is an old PowerBook. But I'll use
anything that works, and the PowerBook was given to me by my company. I do all
my fun stuff on a UnixWare (since changed to Linux) box at home.
Well, here we go!
MS-DOS (5, 6, etc.): The upside of MS-DOS is that it runs on virtually every
Intel-standard PC made since 1981. This is also its downside. DOS is a
"lowest-common denominator" operating system, one that restricts machines to
running in real mode, wasting most of the horsepower of advanced systems. A
Pentium-100 machine running DOS is basically just a real fast XT. My favorite
description of DOS is "A boot sector virus."
PC-DOS: Now that IBM has split off their own DOS from Microsoft, and the Justice
Department has ended the traditional Microsoft bundling practice, it may gain
converts. Some decent features, and IBM's name, but it's still DOS.
DR-DOS: See above, but without IBM's clout, and built-in cheesy peer-to-peer
networking. Enough said.
Windows 3.X: Gives DOS some rudimentary multitasking ability and a prettier
interface. The biggest problems with the current generation of Windows are the
dissassociation of file management from the Program Manager, and the abundance
of incompatible driver versions. Despite all its flaws, Windows makes an
adequate program launcher; but with all its flaws, it often serves to give
users Mac envy. Since Microsoft managed to get Windows bundled with virtually
every PC sold over the last four years, it has a market critical mass that
ensures its dominance for some time to come. That does not necessarily
translate into eternal dominance, though, as students of history should
realize. DOS and Windows make up the lowest-common-denominator operating
system, an Ubersystem for the masses. At the moment, it has the favor of the
masses. But the masses don't ask for Windows. It's given to them with their
PC, like it or not. (Remember, once upon a time CP/M was the operating system
of the masses.) The forthcoming Windows 95 (previously known as Chicago or
Windows 4.0) will shed most of the DOS albatross from Windows. But it will be
a major transition, and as such it gives a window of opportunity to other
vendors (like IBM and Apple). DOOM, the bane of network managers everywhere,
was written using NeXTStep, to simplify development and to make it easily
portable to other platforms. Need I say more?
Apple System 7: Apple's current entry in the OS wars has one huge disadvantage:
it only runs on Macs. And though Apple has a massive market presence, rivalled
only be IBM and Compaq (over 30% of the total PC market between the three),
they are all alone. That said, System 7 is probably the most intuitive of the
major operating systems, and Apple's control of the Macintosh platform is
reflected by the crafting of the interface. The day Apple allows clonemakers
into the Mac market, this may become a problem. Microsoft has been hamstrung
by the lowest-common-denominator approach to operating systems that comes from
lacking control over the machines' design. If Microsoft designed the BIOS and
hardware for all the clone makers, Windows would be far better. Despite flaws,
Apple leads the way in things like transparent networking, advanced sound/video
support, and installation of expansion devices. But, without a way to expand
their market, Apple is doomed to life on the margins as PCs become even more of
a commodity.
IBM OS/2: Possible the best operating system of the bunch, technically, OS/2
does a tremendous job of multitasking DOS programs, and running Windows
programs. Unfortunately, that's pretty much all anybody uses it for. There are
few native OS/2 applications (but all sorts of development tools), probably
because of the bad reputation it got during its first incarnations (versions
1.X).
UNIX's best hope may, however, be in the project of a young Finnish computer
science student. Linus Torvalds wanted to learn more about operating system
design, so he wrote, with the aid of a loosely knit team of coders over the
Internet, his own UNIX work-alike operating system, dubbed "Linux". Linux
distributed free with full source code included, has virtually all the basic
set of UNIX tools available, and some companies have begun to port their
commercial applications to it. Ironically, since the source code is included,
it has been proving relatively simple to port Linux to other platforms. Linux
will not be likely to supplant any of the more mainstream operating systems
anytime soon (or ever), but the mere fact that a free UNIX with source code
exists has done more to ensure UNIX's survival into the next century than any
licensing move Novell, Sun, or any of the "regular" UNIX giants could have
pulled off without howls of protest from the shareholders. Novell tried to
sell "shrinkwrapped" UNIX to the masses in the form of UNIXware, but that
proved about as popular as NeXT's hardware.
There are also a lot of PDAs out there, touted by the cognecenti as the wave of
the future. Hogwash. Though I have a Newton myself (and like it), PDAs in
general are a tool for some small tasks only. I use mine to organize phone
numbers, appointments, and some faxing and e-mail. When I need to write for
real, I use either a good old notepad that weighs less and fits in my pocket
better, or my PowerBook, with it's full-size keyboard. PDAs are probably
best-suited for communications, as I use mine, or to be data-collection
devices. Of the ones that exist today, only Newton has some potential. I look
at Tandy's entry as a "mee-too" effort at best, and HP's excellent LX series,
while laudable, are more DOS computers, miniaturized to the nth degree. AT&T
had a facinating system with the EO communications devices, but they were too
expensive, and too ahead of the time, and have since been axed. As our
definition of what constitutes a computer evolves, so will the acceptance of
the PDA. Sony and Motorola's Magic Cap-based devices will probably make a dent
in the market over the next few years, as prices fall on hardware products, but
they're not real PDAs in my opinion. They're neat communications devices. In
summary, all of today's popular operating systems have a market presence of
some sort or degree. All have major shortcomings, and all have survival
prospects that, while not necessarily spectacular, are good. With operating
system technology at the point it's at, I think the major objective of a buyer
is to find the one that stinks the least. But remember, there already has been
an operating system that was flexible, easily portable, and (by the standard of
the day), very easy to use. It was called CP/M.
Opinions expressed in this article are those of the author and do not
necessarily represent those of the LANtern or of the BNUG. Trademarks used in
this article, generally identified by many upper-case letters (eg, UNIX, OS/2,
etc.) are the owners of their respective companies.