Re: Clarification on use of encodingStyle attribute

Sorry, I guess I wasn't clear.  The questions you raise are good ones, but 
neither is the one I intended.  My question boils down to, is the 
following reference legal given that it crosses two separate activations 
of our encoding style?


<!-- across headers -->
<SOAPENV:Header>
   <n:e1: SOAPENV:encodingStyle="http://www.w3.org/2001/12/soap-encoding">
        <n:target id="targ">somestring</n:target>
  </n:e1>
  ...maybe some other headers here...
   <m:e2: SOAPENV:encodingStyle="http://www.w3.org/2001/12/soap-encoding">
        <m:link ref="targ"/>
  </m:e2>

</SOAPENV:Header>

How about the following?

<!-- within a header -->
<SOAPENV:Header>
   <n:e1>
        <n:target id="targ" 
 SOAPENV:encodingStyle="http://www.w3.org/2001/12/soap-encoding">
                     somestring
       </n:target>
       ...maybe some other elements here 
       <n:link ref="targ"
 SOAPENV:encodingStyle="http://www.w3.org/2001/12/soap-encoding"/>
       ...maybe some other headers here...
    </n:e1>
</SOAPENV:Header>

Now we get to the real crux of the question:  If the above are legal, then 
presumably each results in a single graph.  Now, if we change the examples 
as follows, do we have two disjoint graphs?  There's nothing to connect 
them except that they live within the same SOAP envelope,

<!-- across headers -->
<SOAPENV:Header>
   <n:e1: SOAPENV:encodingStyle="http://www.w3.org/2001/12/soap-encoding">
        <n:target id="targ">somestring</n:target>
        <n:link ref="targ"/>
  </n:e1>
  ...maybe some other headers here...
   <m:e2: SOAPENV:encodingStyle="http://www.w3.org/2001/12/soap-encoding">
        <m:target id="targ2">somestring</n:target>
        <m:link ref="targ2"/>
  </m:e2>

</SOAPENV:Header>

and the following?

<!-- within a header -->
<SOAPENV:Header>
   <n:e1>
        <n:target 
SOAPENV:encodingStyle="http://www.w3.org/2001/12/soap-encoding">
                <n:inner1 id="targ" >
                     somestring
             </n:inner1>
                <n:link ref="targ" >
       </n:target>
       ...maybe some other elements here 
        <m:target 
SOAPENV:encodingStyle="http://www.w3.org/2001/12/soap-encoding">
                <m:inner1 id="targ2" >
                     somestring
             </m:inner1>
                <m:link ref="targ2" >
       </m:target>
    </n:e1>
</SOAPENV:Header>

The following is the same as the previous one, except that the encoding 
style is only activated once an apparent element.

<SOAPENV:Header>
   <n:e1 SOAPENV:encodingStyle="http://www.w3.org/2001/12/soap-encoding">
        <n:e2>
                <n:inner1 id="targ" >
                     somestring
             </n:inner1>
                <n:link1 ref="targ" >
       </n:e2>
       ...maybe some other elements here 
        <n:e3>
                <n:inner2 id="targ2" >
                     somestring
             </n:inner2>
                <n:link2 ref="targ2" >
       </n:e3>
    </n:e1>
</SOAPENV:Header>

I presume that this last example is different, in that we have now created 
a struct, which undoubtedly ties the graphs together.  Right?

I think that giving an interpretation to each of these examples would help 
to get everyone on the same page as to the rules for multiple graphs.  I 
would then make sure that the specification is clear on all of these 
points.  Thank you.

------------------------------------------------------------------
Noah Mendelsohn                              Voice: 1-617-693-4036
IBM Corporation                                Fax: 1-617-693-8676
One Rogers Street
Cambridge, MA 02142
------------------------------------------------------------------







Jacek Kopecky <jacek@systinet.com>
03/25/2002 01:37 PM

 
        To:     Noah Mendelsohn <noah_mendelsohn@us.ibm.com>
        cc:     Jean-Jacques Moreau <moreau@crf.canon.fr>, Henrik Frystyk Nielsen 
<henrikn@microsoft.com>, Martin Gudgin <marting@develop.com>, Herve 
Ruellan <ruellan@crf.canon.fr>, <soap@zaks.demon.co.uk>, 
<xml-dist-app@w3.org>
        Subject:        Re: Clarification on use of encodingStyle attribute


 Noah,

 I don't think your second statement is just an other way of 
saying the first.

 The attributes id and ref are only meaningful the known way when 
used on elements using the SOAP Encoding encodingStyle.

 Or at least I always thought so.

 Maybe this is the right time to 1) make the attributes namespace
qualified and say that SOAP Encoding references may only go to
SOAP Encoding IDs. Or to 2) say that the encodingStyle is applied
also on the targets of references. Or to 3) explain what happens
when a node in SOAP Encoding data has a different encodingStyle.

 In fact, I believe we have two slightly different issues here:

 A) can SOAP Encoding reference point to non-SOAP-Encoding ID?
 B) what's it mean if a node in SOAP Encoding data has a 
different encodingStyle?

 My positions are No on A (making the attributes ns-qualified 
would probably help us achieve this) and on B it's either 
forbidden or a terminal node.

 Best regards,


                   Jacek Kopecky

                   Senior Architect, Systinet (formerly Idoox)
                   http://www.systinet.com/



On Mon, 25 Mar 2002 noah_mendelsohn@us.ibm.com wrote:

 > So, what I think I'm hearing is: the placement of
 > SOAPENV:encodingStyle attributes has no direct bearing on the number
 > of graphs.  Stated another way: id/href can cross from one
 > encodingStyle scope to another?  I just think we have to be clear on
 > this.  Thanks.
 > 
 > ------------------------------------------------------------------
 > Noah Mendelsohn                              Voice: 1-617-693-4036
 > IBM Corporation                                Fax: 1-617-693-8676
 > One Rogers Street
 > Cambridge, MA 02142
 > ------------------------------------------------------------------
 > 
 > 
 >

Received on Monday, 25 March 2002 14:46:43 UTC