From owner-freebsd-advocacy Wed Jun 30 2:55:13 1999 Delivered-To: freebsd-advocacy@freebsd.org Received: from relay05.netaddress.usa.net (relay05.netaddress.usa.net [204.68.24.185]) by hub.freebsd.org (Postfix) with SMTP id 04F2315259 for ; Wed, 30 Jun 1999 02:55:10 -0700 (PDT) (envelope-from jesus.monroy@usa.net) Received: (qmail 4410 invoked from network); 30 Jun 1999 09:55:09 -0000 Received: from nw173.netaddress.usa.net (204.68.24.73) by outbound.netaddress.usa.net with SMTP; 30 Jun 1999 09:55:09 -0000 Received: (qmail 1599 invoked by uid 60001); 30 Jun 1999 09:55:05 -0000 Message-ID: <19990630095505.1598.qmail@nw173.netaddress.usa.net> Received: from 204.68.24.73 by nw173 via web-mailer() on Wed Jun 30 09:55:05 GMT 1999 Date: 30 Jun 99 02:55:05 PDT From: Jesus Monroy To: Gregory Sutter , Jordan K.Hubbard Subject: Re: [Re: [Linux vs. NT, take 2.]] Cc: advocacy@freebsd.org, krooger@debian.org X-Mailer: USANET web-mailer () Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable Sender: owner-freebsd-advocacy@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Gregory Sutter wrote: > On Tue, Jun 29, 1999 at 01:29:51PM -0700, Jordan K. Hubbard wrote: > > > Is there any validity to the discussion on -hackers that real-world= > > > application performance doesn't corroborate the poor benchmark = > > > results > > > (as far as FreeBSD is concerned)? I'm less concerned that = > > > benchmarks > > = > > Plenty. Netbench is notorious for not actually testing the load > > balancing abilities or performance degradation curve as the number of= > > users increases, both important factors in "real life" testing. > = > Another factor that is not taken into account in any benchmark is > general OS stability. NT may be able to kick our butts in some > application performance tests, but can they continue to deliver that > performance for a year without administrative intervention? I > doubt that most NT boxen are anywhere near that stable. Uptime > and platform stability is as important as performance. > = WRONG! This factor will not be taken into account by buying managers. I've worked with several companies on similar issues only = to draw a blank. One company for instance found a bug in a competitors unit. The bug manifested itself when a user sent 1024 bytes through the buffer, not 1023 nor 1025, "1024 bytes"! The bug would the lose the page in the buffer, never to be seen from again. The company then, after getting it's hands on the competitors mailing list sent everyone a one sheet page explaining the bug. The even went as far as send a diskette with a program to reproduce the bug. The results was less that 5% were interested in something different. Which amounted to hitting the red button on the front of the unit every day. --- "I'd rather pay for my freedom than live in a bitmapped, = pop-up-happy dungeon like NT." http://www.performancecomputing.com/features/9809of1.shtml ____________________________________________________________________ Get free e-mail and a permanent address at http://www.netaddress.com/?N=3D= 1 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message