RE: Issue with encodingStyle

The question you ask is one that puzzled me for a long time, too. I
think the answer to it comes down to a conflict between two assumptions
of how applications will interoperate using SOAP:

1.	Applications define or agree on an XML wire-format
representation. That is, they agree on the syntax and grammar to appear
on the wire. The native data model on either side of an exchange is
immaterial provided it can process the XML structure.  

2.	Applications agree on a data structure in a non-XML.  The
application part on one end of the wire serializes this into some wire
representation, which could be almost anything, sends it, and the
application part on the other end then reconstitutes the wire
representation into a copy of the original, non-XML data structure.
When the wire representation is lexically XML, the message contains an
encodingStyle attribute revealing what serialization technique was used
so that the receiver can select from among the several that it is able
to decode.

-----Original Message-----
From: Tim Ewald [mailto:tjewald@develop.com] 
Sent: Thursday, February 07, 2002 7:49 AM
To: 'XMLDISTAPP'
Subject: RE: Issue with encodingStyle

So why not do away with the encodingStyle attribute altogether?

If the assumption is that a SOAP node should be able to process any
legal SOAP message, whether it has artifacts of an encoding or not, why
indicate that they exist. In short, what's the point of having a hint to
say that a message was encoded using a particular encoding style if the
SOAP node is expected to process the message in any case?

Tim-

> -----Original Message-----
> From: xml-dist-app-request@w3.org 
> [mailto:xml-dist-app-request@w3.org] On Behalf Of Andrew Layman
> Sent: Wednesday, February 06, 2002 8:12 PM
> To: noah_mendelsohn@us.ibm.com; cunnings@us.ibm.com
> Cc: xml-dist-app@w3.org
> Subject: RE: Issue with encodingStyle
> 
> 
> I agree with Noah's observation that we cannot and should not 
> try to forbid people from having structures without the 
> encodingStyle attribute but otherwise matching the pattern 
> used by an encoding.
> 
> Regarding the suggestion "For (a portion of a) SOAP block or 
> blocks to be recognizeable as encoded, the application MUST 
> indicate the encoding style using the encodingStyle 
> attribute....", I think this would be misleading, since what 
> it literally says is equivalent to "without the encodingStyle 
> attribute, an otherwise-clueless processor will not recognize 
> that a block is encoded in a certain way."  This is true, but 
> by using the capital-letters word MUST, which connotes a firm 
> requirement, it implies that encoding is mandatory, which is 
> not the literal meaning.
> 
> -----Original Message-----
> From: noah_mendelsohn@us.ibm.com [mailto:noah_mendelsohn@us.ibm.com] 
> Sent: Wednesday, February 06, 2002 4:44 PM
> To: cunnings@us.ibm.com
> Cc: xml-dist-app@w3.org
> Subject: RE: Issue with encodingStyle
> 
> Bob Cunnings asks:
> 
> >> What is your opinion of the proposal to change:
> 
> >> ..."SHOULD indicate their encoding style using
> >> the encodingStyle attribute"
> 
> >> to
> 
> >> ..."MUST indicate their encoding style using the
> >> encodingStyle attribute"
> 
> I think you're asking me, as you're replying to my note.  Honestly, I 
> don't think that particular proposal changes things very 
> much.  Either 
> way, if the encoding is named in the attribute, generalized SOAP 
> processors know about it.  Either way, if you leave it out, 
> you have a 
> perfectly legal SOAP message and the processor will take it to be 
> unencoded.  How we can prohibit someone from sending something that 
> appears to be encoded but really isn't, I'm not sure. 
> 
> What would make more sense to me would be:
> 
> "For (a portion of a) SOAP block or blocks to be 
> recognizeable as encoded, 
> the application MUSTindicate the encoding style using the 
> encodingStyle 
> attribute...."  That would have some teeth and would be OK 
> with me.  On 
> the other hand, I don't have much trouble with the status quo either. 
> Thanks very much.
> ------------------------------------------------------------------
> Noah Mendelsohn                              Voice: 1-617-693-4036
> IBM Corporation                                Fax: 1-617-693-8676
> One Rogers Street
> Cambridge, MA 02142
> ------------------------------------------------------------------
> 
> 
> 
>

Received on Thursday, 7 February 2002 13:13:48 UTC