On Thu, Dec 18, 2003 at 06:29:31PM -0800, Mr O wrote:
> Have these tests all been run at various times during the day
> and in various order to  prove your results? :) I know that if
> Mozilla is already open and I bring up a new tab I can have
> euglug.org loaded before I get to blink. Of course if you're
> counting loading time for the browser too then I'm at a loss.

Indeed, it makes most sense if you download the page to disk and then dump
the copy from disk in a loop about 1000 times and report the time of the
result.  Benchmarks should be as free of randomness as possible, which
means a network connection is out unless that is what you're benchmarking.
The loops guarantee that the disc caches have as little effect as
possible.  Take the resulting times and / 1000 them for the effective
performance differences.  You could do it only 10 times, but I suspect
that would give you inconcievably small numbers to work with.

_______________________________________________
EuG-LUG mailing list
[EMAIL PROTECTED]
http://mailman.efn.org/cgi-bin/listinfo/eug-lug

Reply via email to