I've read somewhere that the big enterprises are
more worried about applying patches than working
on they own business. Let's think about the time
you have to apply a patch on windows, around at
least 20 minutes, and you have to bottleneck
your network even if you have a server just for
keeping the updates do be done.

But, this is still the good side, what to say
when a big enterprise, something like one of the
biggest banks find a problem in windows, and MS
just tells you, wait until the next release?

We were working with NT 4, and wondering why our
servers could only deal with about 50-60 users
at a time!! Very bad, neither the MS people
could tell us why. Then we finally discovered
the problem was with the Thread Local Storage
Area (TLS), NT4 could have just 256 TLS per
process, well you make use of threads in an
environment like MS Transaction Server to slow
down tasks communication times, avoiding use of
inter process calls, but what if a thread calls
one component (COM), and then calls another, and
each time you make a call a new thread is
launched (possibly from a pool)? and with a 4
component path you in a only call you fire with
4 threads, which a mean of 4 threads per call,
256/4 = 64, here's the magic number!!

They told us then the problem would be corrected
next version of windows (NT5 or w2k), but, in
w2k we just have 1024 TLS's, well, something
better, now we may deal with 256 customers, but
what if we decide to fire parallel
transactions?, lets say,get information from
account in the mainframe, contact a partner
through a message queue, retrieve information
from database, and call some other transaction
all same time, then again you have (1024/4)/4=
64, just because you tried to make independent
things in parallel, and we even have a way to
tune it. We can do nothing about!! Well, for
sure, we may create independent processes for
all of that, and accept the overhead of
interprocess calls, since the multi threading
environment doesn't work.

And David (see his comments) is very right about
that you won't see
these advertising researches about Free
Software,
just because they don't have money to buy these
bespeak researches.

Besides, M$ is only the leader because of
piracy, and we cannot really on piracy for
building the picture of usability of a tool. And
we must remember this is one of the reasons
these software is expensive, the enterprises pay
for the copies people doesn't paid, this they
didn't see yet. They are paying for this
training people are having for use the software
they will be forced to use because is supposed
be cheaper since everyone is able to use it. On
the other side then, they wouldn't pay for this
piracy copies, they would pay for the service
from people who works with the software.

Is it really surprising to learn that a software upgrade is less expensive than switching to another technology?

What about after doing 2, or 3 upgrades? Sure, the initial cutover has to be more expensive, like switching from a Mac to a PC or the other way around, but that tell us nothing about what to do when making a new purchase, or what will the cost be after 5 years of patch and upgrade hell.

I think swapping systems is just a bad idea for many people. However, when new systems come in, they can take advantage of the new low cost.

After all, fixing your car is nearly always MASSIVELY CHEAPER than buying a new car, but we buy new cars all the time. New cars have new features, no capabilities, and often cost less to maintain going forward.

These studies are all created to justify MSFT's monopoly and keep these firms in power. I mean, you NEVER read from one of these research firms about anything totally new becoming hot (like the web browser) because they don't take risks, they just tell promote their best customer's best interests. They don't track small companies or think outside the box because that's not how they get paid.

We know this first hand because we have a fast growing client base, but they don't even know we exist. And our customers include hundreds of small businesses as well as a handfull of Fortune 500 customers. But they'll tell you that our competitors have the right solutions for "big business" even though those solutions have failed big businesses for years.

Think about the spam and virus problem of email today. They can only imagine blocking these things rather than taking a new approach.

It's like trying to fix MS-DOS rather than saying it's time for a completely new OS. They rarely see the future, but they make good press for free magazines that rely on selling stories about their advertisers.