Re: RPC issue: multiple body blocks

On Fri, Aug 17, 2001 at 03:03:40PM -0700, Frank DeRose wrote:
> The RPC task force (sounds official, doesn't it?) is currently looking at
> revisions to the spec that would disallow multiple "serialization roots" in
> RPC. This would not preclude the use of multi-referenced elements (data
> items pointed to from more than one place inside the RPC request/reply)
> inside RPC. But, boxcarring with multiple serialization roots would be
> prohibited.

You say serialization root, I say block ;)


> I should mention that, AFAIK, there is no way to prohibit someone from doing
> boxcarring inside a single serialization root. That is, I don't think there
> is any way to prevent anyone nor should we prevent anyone from having an RPC
> request/response whose one parameter is an array of requests/responses. So,
> there is an out.

That's fine; as long as there's one root, it's good for me.

Cheers



> > -----Original Message-----
> > From: xml-dist-app-request@w3.org [mailto:xml-dist-app-request@w3.org]On
> > Behalf Of Mark Nottingham
> > Sent: Friday, August 17, 2001 2:27 PM
> > To: Doug Davis
> > Cc: XML Distributed Applications List
> > Subject: Re: RPC issue: multiple body blocks
> >
> >
> >
> > Of course, it could also be generalised out to a more generic
> > boxcarring mechanism for SOAP messages in general, rather than a
> > RPC-specific one (preferred approach, I think)
> >
> > My motivation is that module authors need to be able to make certain
> > assumptions about RPC messages; this is a pretty basic one.
> >
> > Cheers,
> >
> >
> >
> > On Fri, Aug 17, 2001 at 02:16:13PM -0700, Mark Nottingham wrote:
> > >
> > > Yeah, boxcarring is nifty, but out of scope for us, IIRC. Another RPC
> > > effort can tackle it, IMHO. Otherwise, we need to explicitly support
> > > it, and deal with all of the headaches it incurs...
> > >
> > > Cheers,
> > >
> > >
> > >
> > > On Fri, Aug 17, 2001 at 05:07:29PM -0400, Doug Davis wrote:
> > > > So you want to disallow boxcarring if sec. 7 is used.  It'll
> > still be ok
> > > > to do boxcarring if some other RPC style is defined, right?
> > > > -Dug
> > > >
> > > >
> > > > Mark Nottingham <mnot@mnot.net>@w3.org on 08/17/2001 05:02:44 PM
> > > >
> > > > Sent by:  xml-dist-app-request@w3.org
> > > >
> > > >
> > > > To:   XML Distributed Applications List <xml-dist-app@w3.org>
> > > > cc:
> > > > Subject:  RPC issue: multiple body blocks
> > > >
> > > >
> > > >
> > > > Reading section 7.1, it's hinted that RPC messages are modeled as a
> > > > single struct in the message (note the use of 'single').
> > > >
> > > > However, I don't see anything explicitly prohibiting multiple body
> > > > blocks in a RPC message.
> > > >
> > > > While common sense dictates that RPC with multiple body blocks isn't
> > > > too useful, SOAP does allow them in the definition of a body, and RPC
> > > > doesn't give any solid guidance.
> > > >
> > > > I'd be more comfortable if we ruled out more than one
> > > > child of the body when the RPC convention is in use, except when a
> > > > Fault is present, of course.
> > > >
> > > > Thoughts?
> > > >
> > > > --
> > > > Mark Nottingham
> > > > http://www.mnot.net/
> > > >
> > > >
> > > >
> > > >
> > >
> > > --
> > > Mark Nottingham
> > > http://www.mnot.net/
> > >
> > >
> >
> > --
> > Mark Nottingham
> > http://www.mnot.net/
> >
> >
> 

-- 
Mark Nottingham
http://www.mnot.net/

Received on Friday, 17 August 2001 18:17:26 UTC