W3C home > Mailing lists > Public > www-ws-arch@w3.org > July 2002

Re: Reliable messaging; solution or requirement?

From: Mark Baker <distobj@acm.org>
Date: Wed, 10 Jul 2002 20:18:53 -0400
To: Joseph Hui <Joseph.Hui@exodus.net>
Cc: www-ws-arch@w3.org
Message-ID: <20020710201853.J5470@www.markbaker.ca>

On Wed, Jul 10, 2002 at 02:21:40PM -0700, Joseph Hui wrote:
> FWIW, RM, which addresses the reliable delivery message (or a
> set of messages, in stream or datagram fashion) from one source
> to one or more destinations, where more than two pieces of
> *application* (as opposed to system or network) software
> are likely involved in the process, goes way beyond the problem
> your described as "... two pieces of [application] software
> distributed on an unreliable network ..."  

Mea culpa, of course I meant "two or more pieces ...".  I was still
stuck in TCP mode.

For example, here's how you could implement the classic dining
philosophers coordination problem with REST/HTTP;

http://www.xent.com/pipermail/fork/2001-August/002923.html

To answer Suresh; a "goal" in this context is the problem that is trying
to be solved.  For example, ordering shoes, getting a stock quote,
setting the thermostat, etc..

And while I'm here, to David; 2PC is just another solution to some kinds
of reliability problems, but it isn't what I was talking about (I happen
to think that it is unsuitable for the Internet).  I was talking about
the general problem of reliable coordination that these reliability
technologies (2PC, RM) try to be a solution for (in part or in whole).

MB
-- 
Mark Baker, CTO, Idokorro Mobile (formerly Planetfred)
Ottawa, Ontario, CANADA.               distobj@acm.org
http://www.markbaker.ca        http://www.idokorro.com
Received on Wednesday, 10 July 2002 20:07:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:01 GMT