[TriLUG] Re: Why IExplorer is (sometimes) so fast?

[James Manning]

> [Chris Merrill]
> As described, that is at the TCP/IP level (in establishing a socket).
> This is all complete by the time the socket is considered "opened",
> right?  I would think this is well beneath the HTTP or socket layer
> and would, therefore, be unobservable from the socket or HTTP level.

I (well, we) had noticed this behavior (especially the FIN/ACK
nastiness - I hate half-closed sockets hurting my web server perf)
when doing SpecWeb and TPC-W benchmarking of Linux/Windows hybrid
environments at IBM.  I couldn't tell you the particular IE version,
but it was a good number of years ago (98/99) so I couldn't really say
what things are going on now, although it certainly wouldn't surprise
me for these to be going on.

Ethereal is such a great tool, btw.  Learn it, love it, live it.
Ethereal on Windows almost makes life bearable.
-- 
James Manning <http://www.sublogic.com/james/>
GPG Key fingerprint = B913 2FBD 14A9 CE18 B2B7  9C8E A0BF B026 EEBB F6E4