'hit rates' are not defined by HTTP; proxies, caching and so on are.
I'd characterize trying to derive hit rates as 'goofing around' more
than the use of mechanisms designed into the protocol...
See:
http://www.goldmark.org/netrants/webstats/
(in other words, you haven't chosen a very sympathetic audience)
Cheers,
On Mon, Jul 16, 2001 at 02:28:34PM +0200, Anthony Atkielski wrote:
> Does anyone know if there have been any changes in AOL's network
> these past two months or so that would have an impact on hit rates
> logged for Web sites outside their network? I know I've had to
> tweak journal analysis programs and sites in the past in order to
> get more accurate hit rates from AOL, but I've seen an abrupt drop
> in rates recently that I cannot immediately explain.
>
> Also, if there is a resource somewhere that summarizes all the ways
> that various ISPs goof around in ways that can cause hit rates to
> be less than representative, I'd appreciate a pointer. I've mostly
> made modifications empirically up to thise point (to account for
> proxies, caching, and so on, where possible). AOL seems to be the
> single largest offender.
>
>
>
--
Mark Nottingham, Research Scientist
Akamai Technologies (San Mateo, CA USA)
