Re: Protocol Benchmarking (with Accept examples - long)
hoesel@chem.rug.nl (frans van hoesel)
Errors-To: secret@www0.cern.ch
Date: Thu, 3 Feb 1994 21:09:45 --100
Message-id: <9402032006.AA01385@Xtreme>
Errors-To: secret@www0.cern.ch
Reply-To: www-talk@www0.cern.ch
Originator: www-talk@info.cern.ch
Sender: www-talk@www0.cern.ch
Precedence: bulk
From: hoesel@chem.rug.nl (frans van hoesel)
To: Multiple recipients of list <www-talk@www0.cern.ch>
Subject: Re: Protocol Benchmarking (with Accept examples - long)
X-Listprocessor-Version: 6.0c -- ListProcessor by Anastasios Kotsikonas
Content-Length: 1403
> >
> True, but 99% of the time the client will simply accept the first
> type given by the server. The negotiation will only take place
> less than 1% of the time, and in the case of negotiation the server
> will most likely have to make some on the fly conversion that will
> probably be slower than the latency.
but the client has to tell the server it accepts that type isn't it
(at least that is what learned from your text)
now the cpu time is irrelevant when it comes to saying yes I do accept
that type, give me the file... but I can tell from my modem ligths that any
roundtrip is very bad and surely much more then 1%. Often its true that once
the request is answered and the data is coming the connection is pretty fast
(compared to my 19k2 connection) but in the initial phase the modem is just waiting,
- frans
>
> :lou
> --
> **************************************************************************
> * T H E U N I V E R S I T Y O F K A N S A S *
> * Lou MONTULLI @ Ukanaix.cc.ukans.edu *
> * Kuhub.cc.ukans.edu ACS Computing Services *
> * 913/864-0436 Ukanvax.bitnet Lawrence, KS 66044 *
> * UNIX! Cool! I know that! Jurassic Park - The Movie *
> **************************************************************************
>