W3C home > Mailing lists > Public > www-ws@w3.org > July 2001

RE: Reliable HTTP

From: Richard P. King <rpk@watson.ibm.com>
Date: Wed, 18 Jul 2001 15:09:50 -0400
Message-ID: <3B55DEFE.BF30F0FE@watson.ibm.com>
To: www-ws@w3.org
Nick,

I'd like to respond to your question just a little differently than John did.
The terms client and server are used because those are the roles that are found
at the HTTP level.  Two nodes could very easily regard each other as peers, in
the sense that either can initiate communication with the other, but who,
during any particular communication session, must each take just one of those
roles.  These two peers could, for example, exchange messages in both
directions by having two HTTPR channels.  One channel is only used by A to PUSH
messages to B (let's call them), while the other is only used by B to PUSH
messages to A.  A's role is the client role on the first channel and the server
role on the second.  No polling is done and there is, at a higher level,
complete symmetry using a pair of asymmetrical channels.

Richard.
Received on Wednesday, 18 July 2001 15:10:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:38 GMT