🚀 go-pugleaf

Rachel Polanskis wrote:
> 
> In article <3409F706.794B@engr.sgi.com>,
>         David Chatterton <chatz@engr.sgi.com> writes:
> > Rachel Polanskis wrote:
> >>
> >> It's not Linux, but...
> >> http://www.standishgroup.com/syst.html
> >> http://www.sun.com/realitycheck/standish.html
> >>
> >
> > Personally I am unimpressed by these sort of articles, but I note that
> > M$ have equivalent articles proclaiming NT vs Solaris on their web
> > sites. Unfortunately, the industry takes Solaris (and therefore Sun
> > hardware) as the UNIX benchmark, which is not the best performer at a
> > lot of things.
> 
> But we don't have to guess who you're barracking for, do we?!
> ;)
> 
> From: <chatz@engr.sgi.com>
> ^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> ;)
> 

Sure, but the best benchmarks are spread amongst all the vendors - on
the M$ site they compare a dual pentium system against a Sun server for
web serving and say that NT beats UNIX for Web serving, especially on
price/performance. There are three things wrong with that statement, SGI
and DEC have in the past had better Web servers than Sun (although Sun
are doing a lot of work to catch up), and secondly, UNIX vendors have a
totally different price structure allowing you to get significant
discounts on the advertised price (ie 20-40% off). Most PC vendors are
on very very small margins so they can't do this. Thirdly, where is the
linux comparison on the same pentium system?
 
> s/Solaris/UNIX/g
> 

I prefer not to do that, but my bias is showing through again :)

David
 
-- 
David Chatterton                            (61-3) 9882 8211 (Tel)
R&D Software Engineer		            (61-3) 9882 8030 (Fax)
Performance Tools Group               http://reality.sgi.com/chatz
Silicon Graphics Pty.Ltd., 357 Camberwell Rd, Melbourne, Australia