W3C home > Mailing lists > Public > xml-dist-app@w3.org > November 2000

RE: DR604

From: Williams, Stuart <skw@hplb.hpl.hp.com>
Date: Wed, 15 Nov 2000 17:51:00 -0000
Message-ID: <5E13A1874524D411A876006008CD059F191F90@0-mail-1.hpl.hp.com>
To: "'ohurley@iona.com'" <ohurley@iona.com>
Cc: xml-dist-app@w3.org
Hi Oisin,

I think the easy fix is to add the word 'content' to the end of DR604. Which
becomes:

"...This requirement implies it must be possible apply many transport or
application protocol bindings to the XP message without information loss
from the message (content). "

In terms of what intermediaries can touch and what transformations they can
perform on the semantics of an interaction... its not clear to me that we
have established any stated requirements - eg. do say request/response
interactions fully nest through a bunch of intermediaries or is it a
hop-by-hop chain of request/response pairs; what can an intermediary that
signs a message touch... but thats for a different section...

Regards

Stuart



> -----Original Message-----
> From: Oisín Hurley [mailto:ohurley@iona.com]
> Sent: 14 November 2000 17:15
> To: Williams, Stuart; xml-dist-app@w3.org
> Subject: RE: DR604
> 
> 
> Hi Stuart
> 
> > The semantics of an interaction are also bound up in the 
> events that arise
> > during the course of an interaction, not just the content 
> of the messages
> > exchanged.
> 
> Indeed, this is true. However I think it may be fair to say 
> that there is
> an 'expected interaction semantics' which is present prior to 
> the actual
> message interchange and that this is what is meant by this 
> requirement.
> 
> Coupled with the requirement for extension, it may  be possible for
> intermediaries to change then course of the envelope and thus depart
> from the expected interaction semantics. The provision of routing
> instruction makes this a viable means to tailor the exchange to your
> particular requirement or to resolve damage or queue issues.
> 
> > Delivery sequence may have semantics, single delivery may be
> > important, silent loss may be an issue, loss of a 'fault'
> > response may be an
> > issue. I'd like to fully understand what is scoped as 
> "information loss".
> 
> Hmm. We all know that information cannot be destroyed, merely 
> transformed :)
> I think what this means is that the information in the basic 
> message model
> (i.e. body of the message, all mandatory headers and 
> attachments) must reach
> the intended destination intact. The rule is that 
> intermediaries in the
> chain
> of delivery must not remove or mutate these data items.
> 
>  cheers
>    --oh
> 
> --
> ohurley at iona dot com
> +353 1 637 2639
> 
Received on Wednesday, 15 November 2000 12:54:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:57 GMT