Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: A voodoo masterpiece

Re: A voodoo masterpiece

From: Holger Baer <holger.baer_at_science-computing.de>
Date: Wed, 10 Nov 2004 01:29:36 +0100
Message-ID: <cmrndg$a12$1@news.BelWue.DE>


Jonathan Lewis wrote:
> Note in-line
>
"Holger Baer" <holger.baer_at_science-computing.de> wrote in message news:cmnnpf$eiq$1_at_news.BelWue.DE...

 >>
 >> But Don and Mike finally put the cards on the desk: They can't be bothered
 >> with
 >> proofing anything they gibber out, the so called "real-world phenomenom"
 >> (and
 >> in their cases that means throw hardware at it and the problem will go
 >> away)
 >> must suffice.
 >>


> No - read Mike's comment carefully. It is a little out of context, but
> he makes the perfectly valid statement that "proofs" undertaken on
> a PC can break down at high loads on 'real systems'. You only have
> to consider that PCs tend to be single-CPU systems to realise how
> true that can be. It's why I sometimes give away free consultancy
> time in exchange for temporary access to large-scale hardware.

I tried to reread Mike's comment. However, the page seems no longer to exist - or maybe it will reexist on the 18th? ;-)

I agree that Mike's statement is valid, nethertheless it seemed to justifiy a behaviour where one can postulate any cause/effect relation ship one wants to because one has seen it in such a complex, heavy loaded system that creating a testcase was beyond the reach of a test system, yet still get away with generalizations that this postulated (but never proofed) cause/effect will apply to other systems.

Now I have to go and find some really large-scale hardware to get your free consultancy time :-)

Cheers
Holger Received on Tue Nov 09 2004 - 18:29:36 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US