W3C home > Mailing lists > Public > xml-dist-app@w3.org > January 2001

RE: [DS8] Conversational message exchange - revised

From: Anderson, William L <WAnderson@crt.xerox.com>
Date: Thu, 18 Jan 2001 12:46:35 -0500
Message-ID: <BF85F885643CD21188230008C756293502E1E9B6@crte128-2.wrc.xerox.com>
To: "'Mark Needleman - DRA'" <mneedlem@dra.com>, Ray Denenberg <rden@loc.gov>
Cc: john_ibbotson@uk.ibm.com, xml-dist-app@w3.org
Mark, I like your suggestion. The more detailed, specific scenarios should
be provided as examples.

Bill Anderson

> -----Original Message-----
> From: Mark Needleman - DRA [mailto:mneedlem@dra.com]
> Sent: Thursday, January 18, 2001 10:42 AM
> To: Ray Denenberg
> Cc: john_ibbotson@uk.ibm.com; xml-dist-app@w3.org
> Subject: Re: [DS8] Conversational message exchange - revised
> 
> 
> Ray John
> 
> maybe the way to go is to generalize this and say exchanged in a long
> running process .... Examples of such processes include  business
> communications, information retrieval (and perhaps a couple of more to
> indicate the types of long running exchanges we have in mind)
> 
> mark
> 
> 
> On Thu, 18 Jan 2001, Ray Denenberg wrote:
> 
> > john_ibbotson@uk.ibm.com wrote:
> > 
> > > DS8  Two trading partners are engaged in a long-running 
> business process
> > > which involves multiple  message exchanges. ....
> > 
> > John -- as we discussed yesterday, I would like to see this 
> case accomodate the
> > information retrieval scenario.  Could we expand it to include this:
> > 
> > 
> > Two partners are engaged in an information retrieval 
> session which involves
> > multiple message exchanges, and multiple message patterns.
> > 
> > This information retrieval scenario is modeled in terms of 
> a client/server
> > protocol; one partner is the client who wants to retrieve 
> information from the
> > other partner, the server.
> > 
> > Inititially, the partners may exchange messages to set 
> and/or negotiate
> > parameters that will be in effect for the remainder of the 
> process. The client
> > may then send a query and the server responds with a count 
> of documents
> > resulting from processing the query.  The client may then 
> request transmission
> > of the first N documents (full text of the documents, 
> specified portions only,
> > or just metadata) then the next N documents, and so on. The 
> client may
> > subsequently send another query (thus a second result set 
> would be created),
> > request resulting documents, and might subsequently request 
> additional documents
> > from the first result set.
> > 
> > Message exchange is patterned mostly in terms of 
> request/response messages (e.g.
> > for negotiation, query, and document retrieval) where the 
> client sends a request
> > and the server sends a response. However, there are some 
> messages that the
> > client may send that require no response. There are also 
> messages that the
> > server may send asynchronously; for example, the client may 
> send a request (a
> > query) and the server might send a message requesting 
> clarification about the
> > query; the client then sends the clarification, and 
> subsequently the server
> > sends a response to the query.
> > 
> > 
> > 
> > --
> > Ray Denenberg
> > Library of Congress
> > rden@loc.gov
> > 202-707-5795
> > 
> > 
> 
Received on Thursday, 18 January 2001 12:46:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:58 GMT