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

Advanced Web Services & client/server

From: Mark Baker <distobj@acm.org>
Date: Sun, 8 Sep 2002 22:54:23 -0400
To: www-ws-arch@w3.org
Message-ID: <20020908225423.J9972@www.markbaker.ca>

Hi all,

I just wanted to toss out a comment about the "Advanced Web Services"[1]
paper that several members have authored.  It's certainly possible that
I'm misinterpreting the intent - and if I am I apologize - but it seems
to me as though the paper is suggesting that the Web (and/or REST) is not
already peer-to-peer.  If that is what it's suggesting, then I have to
disagree.

Specifically, the paper seems to assume that "client/server", which REST
builds upon, refers to software rather than roles.  i.e. that there is
a piece of software which shall forever play the client role (e.g.
browser), and a piece of software which shall forever play the server
(e.g. Web server) role.  This is not the case.  REST only depends on
roles; i.e. that in any given transaction, there is an initiator and
a terminator.  Software like that provided by KnowNow or my company,
which embeds a Web server into a Web browser (or vice versa 8-), has
already demonstrated that REST works well for peer-to-peer.

I'm certainly in agreement that we need more data flowing from Web
servers to Web clients, I just don't think we need any new architectural
work to enable it.

Thanks.

Perhaps we can talk about this at the f2f - but I wanted to get the
comment out on the public list too.

 [1] http://users.rcn.com/geoff2/aws1.0.pdf

MB
-- 
Mark Baker, CTO, Idokorro Mobile (formerly Planetfred)
Ottawa, Ontario, CANADA.               distobj@acm.org
http://www.markbaker.ca        http://www.idokorro.com
Received on Sunday, 8 September 2002 22:55:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:05 GMT