httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jason A. Dour" <>
Subject Draft #2 of PCWeek rebuttal...
Date Thu, 12 Jun 1997 17:48:20 GMT

Better, Ken!  How's this, everyone?

It's always pleasing to see your product praised in print.  That's why the
Apache Group was at first happy with the positive press it received in
Michael Moeller's cover story, "FORT APACHE" (PCWeek, June 9, 1997)  and
Jim Rapoza's related article, "IT'S NOT EASY.  SO WHAT?"  (PCWeek, pg18,
June 9, 1997).  We were dismayed, however, by a statement made by Mr.
Moeller and by what we feel to be a misleading account of Apache's
performance in the benchmark graph from PC Week Labs. 

Moeller's article did a good job of explaining Apache's strengths and
popularity -- until the last column.  Moeller writes, "Anyone unfamiliar
with hard-core Unix programming will not be able to get Apache running,
said users."  To this statement, we must take strong exception.  Of the
hundreds of thousands of installations of Apache, how many does Moeller
believe to be administered by "hard-core Unix programmers?"  Apache
end-users are a diverse bunch: from computer novices, to system
administrators of varying ability and station, to Moeller's "hard-core" 
example, to everyday people who need good solid web server software. 
Moeller's "quote" from undisclosed "users" is a disappointing low-point
for an otherwise well crafted article.

As for the benchmark on page 18, we must wonder about the validity of the
claims it made.  Microsoft's IIS is shown overall to triple the
performance of two versions of Apache.  These figures are hard to believe. 
Internal testing as well as repeated user feedback indicates that Apache's
performance meets or far exceeds any of the competition's.  We believe the
reported numbers would be much different if the Apache configuration were
even minimally optimized to meet the benchmark's performance metric of

The Apache Group would welcome and appreciate having an opportunity
to work with PC Week Labs to verify and possibly correct these
benchmark results.  We would like to provide a configuration that
would be more appropriate to the platform and benchmark conditions
while remaining relatively unbiased toward the specific tests.  If
it can be demonstrated to our satisfaction that the test environment
was fair across the various servers, we will confirm the fact.

# Jason A. Dour <>                            1101
# Programmer Analyst II; Department of Radiation Oncology; Univ. of Lou.
# Finger for URLs, PGP public key, geek code, PJ Harvey info, et cetera.

Version: 2.6.2


View raw message