> Interesting and enjoyable post. I disagree with some of your
> conclusions. I think the U.S. is probably computer saturated
> so the explosive growth in PC sales of the late nineties is
> just not going to be there, whereas China and the rest, are
> yet to 'computer up' their vast populations .. hence they
> show growth.
The U.S. has millions of people who use computers, and those
computers have been pre-packaged with Windows. Ford made
millions of Model T's and only offered them in Black. It took
them years to adopt standards such as stick-shift, automatic
spark advance, mixture control, and even the clutch, break, and
accelerator pedals we know today.
Just as people who buy cars look for new innovations in style,
engineering, and functionality, which drives new sales, the same
can and should be true with computers. Part of the problem is
that there hasn't been a healthy competitive market.
Think about it, each time Microsoft has faced really intense
competition, the entire industry shifted paradigms. When Apple
came out with Basic in Rom, keyboard and motherboard on the same
chassis, and a built-in video driver, it push Microsoft to shift
from toggle switches, components, and S-100 cards to the
When Apple introduced the Lisa, Bill Gates scrambled to find
some way to beg, borrow, or steal Windows technology.
Eventually, he got it from Steve Jobs, and even borrowed many
ideas and concepts from X11R3. Even the look and feel of
Windows 3.1 came from Motif and the HP widgets.
When Sun began shifting the market with it's SLC and IPC UNIX
workstations, Microsoft scrambled to introduce Windows NT,
complete with true preemptive multitasking, protected processes,
and interprocess communication. Bill Gates announced that
Windows NT would be a "Better UNIX than UNIX". At the time, Sun
had 15% of the desktop market.
When Yddragasil Plug-n-Play Linux offered self-configuring
installation, Internet access, including e-mail, ftp, and web
browsers, Bill Gates delayed "Chicago" to make sure that it had
these features, even though it caused a delay right into the
last week of Microsoft's Fiscal Year 1995 (end of August).
When Linux began outperforming Windows NT 4.0 in both
performance and reliability, Microsoft made huge efforts to
improve the stability and performance of Windows NT 5.0 which it
reneamed Windows 2000 after numerous embarrasing "Blue Screen of
Death" demonstrations by Bill Gates.
When Linux users began to point out the advantages of a user
configurable interface, with a number of different layouts and
preference settings, Microsoft introduced numerous user
configurable "look and feel" settings, beyond those offered in
Windows 98 or Windows 2000. They also tried to be more backward
compatible with Windows 98.
Even today, Linux and Apple's OS/X UNIX based system have put
the fire to Microsoft. Microsoft will probably have to more
agressively persue a true UNIX variant. In addition, Microsoft
has seen that the Open Source development model is working
remarkably well, and has begun to try and capture some of the
key advantages without sacrificing control of the product.
Competition, real competition in which Microsoft holds less than
60% of the market, and has no exclusive control of the OEM
distribution channels, is good for the industry and the computer
using public as a whole. It keeps both Microsoft and the
competitiors on their toes.
If Linux were to capture 90% of the market, the distributors
would be less inclined to compete. They would have less
incentive or need to innovate, and they would be trying to milk
as much value out of the established technology before
introducing new technology innovations, much the same as DEC did
in 1983, as IBM did in 1992, or the way Microsoft did in 2002.
> The problem with Win959898SEMEWinNT3.142kXP2003 is that there
> are TOO many versions of Windows and people are balking ..
> heading to their old favourites because the new ones are not
> *that* compelling.
This is very true. Microsoft's biggest competiton is -
Microsoft. Microsoft had to convince people to stop using
Windows 3.1 and switch to Windows 95, without losing the market
to Linux, Solaris, or UnixWare. Microsoft then had to convince
corporate users to switch workstations to Windows NT 4.0 rather
than staying with Windows 95. At the same time they had to make
sure they didn't lose control to Solaris, SCO, FreeBSD, or Linux.
Even today, Microsoft has to try to convince corporate managers
and consumers to purchase new Windows XP machines rather than
stay with Windows 98 or Windows 2000, while at the same time not
losing control of the market to OS/X, or Linux.
What is significant is that Linux has been the most consistent
and persistent threat for almost exactly 10 years. Microsoft
hoped that Linux would starve itself to death, and instead,
Linux has captured the majority of the server market by unit
volume, and has consistently grown it's Client based market share.
> The solution, of course, would be for MS to slow down for a
> while and concentrate on new apps and programs. XP on the
> desktop and the new Server should be fine for quite some time
> and Microsoft might contribute more to society by focusing on
> skillfully patching them and creating innovatative apps to
> run on them.
The problem for Microsoft is that it's being *ed on all
sides. The Linux kernel has been radically improved several
times in the last few years, and performance numbers are
substantially better than NT or 2000. The reliability of Linux
has come to approach the reliability of UNIX, while Windows NT
and 2000 still have serious problems with 3rd party software,
especially in complex configurations.
In the GUI front, KDE 3.0 and GNOME 2.0 are creating whole new
levels of expectations. Windows will again have a hard time
keeping up, and as KDE and GNOME become more widely available,
especially preinstalled on OEM machines, Microsoft's claims that
Linux is "unfriendly" or "a command line oriented system" will
be completely ineffective.
On the Scalability front, Linux can now scale from PDAs and
"book-case" routers and wireless hubs, to Z-900 mainframes
capable of running thousands of virtual "Linux servers" within a
Linux has completely redefined clustering, and the business
version, Grids, has also become a whole new breakthrough in
performance, reliability, and security.
Linux has even redefined the standards of system administration,
with more autonomic systems, self-maintenance, and superior
remote management which provides GUI interfaces, even over
low-bandwidth dial-up or cellular wireless environments.
Microsoft can't stand still, but they have to do more than try
to use contract changes to the EULA to force corporate customers
and OEMs to accept ultimatums and extortion threats. At the
moment the EULA of XP, the *y traps, and the "Do or Die"
threats have probably done more for Linux than anything the
Linux user community could ever have done.
Microsoft might even find that Linux could be profitable. They
could sell the same type of library they sold for Solaris, AIX,
and HP_UX, to enable developers to compile windows applications
to Linux. They may find that they can get the benefits of
loosely coupled applications such as those available in Open
Source, while providing their own "shells", even such as .NET or
it's successor, to exploit Linux functionality.
Microsoft has figured out that it pays to play nice with UNIX,
as they illustrate in their "1 degree of separation"
commercials, where they show the advantages of having a nice
friendly Microsoft GUI interface to UNIX based CAD/CAM servers.
Bill Gates and Steve Ballmer are good businessmen. They will
have to make some radical adjustments, but if they can adjust,
they could find that Open Source/Linux could be as profitable
for Microsoft in the next 10 years as the Open Standards based
Internet was in the previous 10 years.