W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 1998

Re: Why gateways?

From: Roy T. Fielding <fielding@kiwi.ics.uci.edu>
Date: Wed, 25 Feb 1998 12:26:14 -0800
To: Mike_Spreitzer.PARC@xerox.com
Cc: http-wg@cuckoo.hpl.hp.com
Message-Id: <9802251236.aa01719@paris.ics.uci.edu>
X-Mailing-List: <http-wg@cuckoo.hpl.hp.com> archive/latest/5407
>I'm asking why the thing that initiates communication has to do so over HTTP,
>whereas the thing that ultimately serves it does not have to do so over HTTP.

Neither "has to do" anything.  The introduction introduces a bunch of
terms that are necessary for common communication about the HTTP
environment and its participants.  There are many things that can
be abstracted together to form the "user agent", just as there are
many things that can be abstracted together to form an origin server.
I included "gateway" specifically because people kept confusing
such things with a proxy.

I have a ton of ideas on how to rewrite the introduction to make the
abstractions a little more clear, and another ton of ideas on how to
reorganize the entire specification to make it readable, but have neither
the time nor the patience to restart the standards process.

BTW, the chain of communication is the only form described because
all other forms can be reduced to a chain for the purpose of discussing
the HTTP semantics.

Received on Wednesday, 25 February 1998 13:17:55 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 2 February 2023 18:43:04 UTC