Hi Daniel,
+> The paradox comes from the fact that the Mindcraft NT-Linux
+> benchmark measures situations that are almost never occurring in real
+> life, and may concern at most a few computers in the world.
+> Look at
+> http://cs.alfred.edu/~lansdoct/mstest.html
+> explaining how irrelevant the benchmark is.
Thank you for this link, it's great! After reading this, I feel much
better now! :-))))))
+> On the other hand a more realistic Web server benchmark suite
+> was done by the German C't magazin. As expected for more realistic
+> cases Linux looks much better:
+> http://www.heise.de/ct/english//99/13/186-1/
I have the c't and have read the article as well (in german ;-) ).
Both documents together make a pretty good picture, why my experiences
are so different from the Mindcrash re-test.
In other words: in the *real* world Linux is in most cases far better
performing than NT. So, whoever uses NT (after the PCweek lecture)
and says "yes" to a one on one reality check with a current Linux,
will fall flat. Face down.
So I can only recommend to do so! Ask your favourite NT-"enemy" for
a one on one reality check! - And have fun!
+> My guess is that NT pro people studied carefully Linux to find
+> a weak point on which NT is superior, and then invent a benchmark to
+> claim general superiority. They forgot to check whether the
+> effective superiority corresponds to a widespread need.
My current guess goes far more indepth. The guys from Microsoft are
not dumb. - Especially those who make the waves in the press. They know
how to bring an enemy down. And I wonder if the whole Mindcrash-stuff
wasn't in reality some sort of boobytrap.
Scenario: You have a system from which you really urgently need that it
has to win a one on one comparison. What can you do to assure this?
Simple trick of an illusionist: Set everything up in a way that you can
be sure that you are the one who wins!
So M$ went and benched a lot in their own labs until they found a
benchmark, where Linux did't perform too well, even if tweaked to the max.
Necessary for this is that this benchmark looks like it has any sense
in the real world. Since Webserver and Fileservers are quite important
stuff, it was nice to find a combination (setup) in which Linux lost
clearly. - Then ask your internal gurus what they think, if this problem
can be fixed within a week with heavy testing. If the answer is no, everything
is fine for the great Houdini-Trick.
Now you go and tell a third party to make a comparison with the exact
benchmark setup. Even play your role as "evil emperor" and tell them to
tweak Linux so badly that everybody with a brain will cry out loud for
a revenge under "fair" conditions. - And this is the trap.
Even a retest will (of course) show that YOUR system is superior. Thus
you wait for a couple of weeks to let the waves go high and then graciously
grant a retest. The enemy never ever has a chance to win this competition.
(Does this feel familiar to you, Alan? ;-) You never had any real chance.)
And best of all: Now you come out of this competition with a white collar.
You've been fair, and the enemy lost. - Isn't this an nice opportunity
to push this great news in the media?
Somehow I got the strange feeling that we've tapped into Microsoft's
illusionists trick.
You want to see another illusionist trick?
Try this one:
http://www3.mcps.k12.md.us/users//rsfay/magic/index.html
The trick is based on the same concept as the one from Microsoft. You
will be astonished *how* easy one can be fooled!
Best regards,
Herbert
"We have found solutions for all solvable problems. From now on there is
only the need for the NT-Hotline." A colleague about the closing of a Unix-HL
because too few users made calls (800 Users, 20 Calls a day).
-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
