Re: Web Services and Distributed Object Systems: a Difference in State.

Hi Robert,

On Thu, Dec 30, 2004 at 01:47:57AM +1100, Robert Mark Bram wrote:
> 
> Hi All!
> 
> I recently read with much interest the thread on this list entitled 
> "Stateful Web Services... " that began with this message:
> http://lists.w3.org/Archives/Public/www-ws/2004Oct/0015.html
> 
> I found this discussion enlightening because I am studying Web Services 
> and Grid Web Services (as a distributed object system) right now.

You might be interested in studying the Web as a distributed object
system too.  Check out;

http://www.w3.org/OOP/interfaces/

> Thus it is due to the above thread and much other material already studied 
> that I present my own discussion on the issue of state to do with Web 
> Services and Distributed Object Systems:
> http://phd.netcomp.monash.edu.au/RobertMarkBram/blog/detailPage.asp?id=41
> 
> All comments and criticisms are most welcome. This is naturally a work in 
> progress and represents nothing more than my understanding of the wisdom 
> of those who have designed systems before me...

I think a point of confusion remains in this article.  I agree that
Roy's terminology in his dissertation is a little bit confusing, but
where he refers to "client-stateless-server" (that you reference),
the "stateless" part there isn't an adjective for the noun "server",
it refers to the connector.  It's true that stateless servers, by
definition, can only participate in stateless exchanges, but the
converse isn't true; a server with state can still participate in
stateless message exchanges.  It just can't have *session* state, aka
state which can be referenced by messages for the purpose of
affecting the semantics of those messages.

Regarding "stateless clients", I don't think that's what you mean to
be discussing.  AFAIK, the only "stateless client" architectural styles
are "remote session" styles like telnet or VNC where session state is
*entirely* on the server.  What you appear to be describing in those
two sections are styles where the state of any session is split between
client and server.

I disagree with your summation of the tradeoffs of stateless vs.
stateful when you say "a stateful server [...] is more amenable to
message sequences i.e. application level protocols.".

I'll refrain from commenting on the rest, since it's so subjective.

Mark.
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca

Received on Wednesday, 29 December 2004 20:32:02 UTC