Messages in this thread | | | Date | Wed, 30 Jun 1999 00:58:38 +0100 | From | "Manuel J. Galan" <> | Subject | benchmarks and baits |
| |
It seems that we are MS is striking hard against Linux and they have a lot of money and people to spread FUD, etc. over Linux.
Maybe their best strike so far has been the Mindcraft libel (and sequels...)
Lies, statistics and benchmarks:
Why are all these benckmarks essentially flawed compared to other kind of benchmarks: because they are specially artificious.
Benchmarks are normally artificious but regarding FPU, disk IO etc, I can say that I, myself, have more than 10 computers doing FULL FPU calculations full time or people using full disk IO for video streaming, etc... but who on earth has a single web server with four fast ethernets serving static web pages at the rate that benchmarks use?
( http://cs.alfred.edu/~lansdoct/mstest.html )
NOBODY.
Should Linux take MS bait and involve itself into a frentic benchmarkitis forgetting its main strengths: stability, reliability and low end hardware usability?
NO.
(Unfortunately it seems that this bait is very easy to take for free software developpers :-/)
On the other side, can we forget medium size hardware (high end servers, SMP, fast IO disk systems, etc)
AGAIN, NO.
The point of balance between these two trends is very difficult to find but I feel that at this moment it is the survival test of Linux vs. MS.begin:vcard n:Galan Moreno;Manuel x-mozilla-html:TRUE org:E.T.S. de Arquitectura (ULPGC);Matematicas adr:;;;;;; version:2.1 email;internet:manolow@step.es x-mozilla-cpt:;1056 fn:Manuel Galan Moreno end:vcard
| |