- From: Williams, Stuart <skw@hplb.hpl.hp.com>
- Date: Fri, 4 May 2001 10:34:47 +0100
- To: "'Mark Nottingham'" <mnot@mnot.net>, Martin Gudgin <marting@develop.com>
- Cc: XML Protocol Comments <xml-dist-app@w3.org>
Hi Mark, > -----Original Message----- > From: Mark Nottingham [mailto:mnot@mnot.net] > Sent: 03 May 2001 20:01 > To: Martin Gudgin > Cc: XML Protocol Comments > Subject: Re: Must understand mustUnderstand proposal > > > On Thu, May 03, 2001 at 03:49:21PM +0800, Martin Gudgin wrote: > > 3. Does an intermediary *always* remove the headers > targeted at it? If > > not then I think we need some way of annotating them as 'processed'. > > While SOAP assumes this, I'd note that there is very little > implementation of SOAP headers and intermediaries out there, and > arguably they comprise the most underspecified part of it. > > It seems like there are several headers which may be processed (ah, > that word again) by several intermediaries; for example, caching, > logging, etc. > > Then again, does mustUnderstand really make sense in that context? > These sorts of services are really advisory hints about what can be > done, not instructions about what must be done to provide the > service, so maybe the semantic of mustUnderstand *is* mustConsume in > this context. In the logging case it may be that the sender of the message wants the message logged by a 3rd for audit purposes. In such circumstances I think that mustUnderstand is more than advisory. > Thoughts? Can anyone think of modules that could be targetted at > multiple devices where mustUnderstand would make sense? A message path/routing module. Admittedly, we could discuss whether the path/routing header is the same header with modified content or a different header of the same 'type' at each intermediary along the path. > -- > Mark Nottingham > http://www.mnot.net/ Stuart
Received on Friday, 4 May 2001 05:35:07 UTC