me> Maybe Apache is causing all the load, and Mozilla is just
me> getting the left overs....  But then how would that Mac
me> and P3 laptop display the same page coming from the same
me> machine in less time than I can count?

Chris> Apache is not the problem!  My classic quite happily
Chris> handles the apache manual pages like lightning with browser
Chris> caching turned off.  It even stands a full 10Mbit load
Chris> serving average 5k pages.

Yea, I meant that in jest, and forgot a smiley or two.  They only
time Apache slows to a crawl is when I run a ksh CGI script that
passes 5 or 6 SQL queries off to PostgreSQL then mucks up the output
with a pipe of about 5 sed's in a row.  (Oh, did I mention that
the largest query is about 20 selects unioned together, and each
one has around 3 joins and a subquery to assist in the joins.)

And I can run that with Lynx about 3 times a minute.  (Yes, it was
real ugly, and I knew better, but it was easy to cut and paste,
and it wasn't for real use, and no I don't have or want a copy of
that corpse anymore either. It was a schema test, so the tables
only had about 20 entries in them, and now the schema is all
different too.)

Somehow, I don't think the Mozilla "USPS priority delivery" of
index.html, foo.jpeg, and bar.jpeg ("Usually delivered in 2 to 3
days") is caused by my Apache/PostgreSQL setup.  I specifically
used a static page so I could easily see the size and not have
other stuff involved.  And as I stated above, I'd like new hardware
to make it a nanosecond world instead of a microsecond world, but
I'd rather keep the wife, kids, and house right now, so if it's
not free, I probably won't get it. 8^)

Someone (or group) needs to pull their head out of the sand and
realize that it does not take a 400MHz G4 or a 900MHz P3 to display
5KB of text and 100K of JPEGs in less than 30 seconds.  Or better,
2 minutes 50 seconds to get a window frame (with background still in
the content region!) up after I start a browser in an Xterm.

Jon