Roy Schestowitz wrote in <1633672.6i4Jm8u41i@xxxxxxxxxxxxxxx> on Sat April
29 2006 00:29:
> I can think of at least 4 TCO studies (links on demand) which clearly
> indicated that Linux is cheaper. The one from IBM, for example, claims
> Linux to be 40% cheaper than Windows. I am yet to see a TCO study, other
> than that which was funded by Microsoft, which actually claims GNU/Linux
> is more expensive. Evidently, Microsoft invested many millions in
> disinformation that blinds some senior managers.
I was out for a little while earlier today and actively trying to think of
reasons that Linux TCO could become more then that of Windows TCO, just as
a "devil's advocate" type of thing. When you compare base systems to each
other, I can't think of how that could be. However, I did think about
something: Microsoft may be just bending the truth a little bit, as
opposed to spreading complete and total disinformation.
I'm not exactly sure how they're computing TCO, and since, as I mentioned
before, it's almost like black magic trying to find apples to compare to
apples, it probably doesn't matter all that much anyway. However, if you
considered the TCO to be everything from the operating system and
administration of it, down to the applications and servers running on the
operating system, and the administration of it, you might figure that Linux
based systems (or any other UNIX-like system) may incur slightly more in
the way of costs.
However, the only situations that I can think of at the immediate moment
where this would be true are those where the people involved aren't even
passingly familiar with the command line, the core system, or the way that
things work on it. There are many people who, given either environment,
can find their way around, usually with a set of notes that they've taken.
They're the ones that are comfortable using either one, but they're going
to be slowed down a bit by using their notes instead of just knowing how
each system works and what to do.
By that same logic, that means that somebody who's very proficient with
Windows, and not so much with Linux-based systems or other UNIX-like
systems, will find that they're going to be slower at working with those
systems until they learn them more. I wonder if Microsoft's TCO
measurements are comparing Microsoft system administrators with actual
Linux system administrators (and the appropriate application
administrators), or if they're using one type of person for both and
tainting their studies that way? Interesting thing to think about.
> By the way, diregard Erik's followup. He is a notorious FUDmeister who
> sees and seizes an opportunity in your relative uncertainty. Search the
> C.O.L.A. archive to find a handful of threads on this topic.
I didn't find his followup to be that horrid, at least in this round.