Re: SOAP Encoding Schema broken

I've been thinking about this a bit more. In the array case, we're already
covered by[1], assuming people actually use itemType attributes. If they
don't use itemType then we're back to needing specific element QNames in
order for type assignment to work.

So what will people do in the real world? Use itemType? Use specific element
QNames? Both? Neither?

I'm not sure that validation against the encoding schema is all that useful,
even if I change it to remove the current ref problem. It will only provide
type info nodes in arrays or generics accessed by position. And it will only
do it for arrays if people haven't already used itemType. So that really
just leaves generics. Is it worth the effort to get the encoding schema
right and write rules for type assignment when the utility is so limited?

Gudge

[1] http://www.w3.org/2000/xp/Group/1/10/11/soap12-part2.xml#enctypename

----- Original Message -----
From: "Martin Gudgin" <martin.gudgin@btconnect.com>
To: "Jacek Kopecky" <jacek@systinet.com>
Cc: "XML Protocol Discussion" <xml-dist-app@w3.org>; "Yves Lafon"
<ylafon@w3.org>
Sent: Friday, April 26, 2002 9:49 AM
Subject: Re: SOAP Encoding Schema broken


>
> ----- Original Message -----
> From: "Jacek Kopecky" <jacek@systinet.com>
> To: "Martin Gudgin" <mgudgin@hotmail.com>
> Cc: "XML Protocol Discussion" <xml-dist-app@w3.org>; "Yves Lafon"
> <ylafon@w3.org>
> Sent: Friday, April 26, 2002 9:18 AM
> Subject: Re: SOAP Encoding Schema broken
>
>
> > Gudge,
> >  these elements are only used where the edge label doesn't
> > matter,
>
> Oh, I agree entirely.
>
> > and they are used so that xsi:type becomes unnecessary.
>
> xsi:type is *never* necessary, although it is sometimes useful.
>
> >  There are two scenarios - the edge is a reference or is not a
> > reference.
>
> Agreed.
>
> >  In case the edge is a reference its element information item
> > carries no relevant type information.
>
> Right, it's an edge and only an edge.
>
> > In case the edge is not a
> > reference there is no problem with the current schema.
>
> Agreed.
>
> >  Since the type information has no meaning on the reference, I
> > prefer your middle solution, element named "reference" which has
> > an attribute "ref" and that's all.
>
> OK, presumably we'd need to frame this so that such elements are only used
> in arrays and generics accessed by position.
>
> Gudge
>
>

Received on Friday, 26 April 2002 06:10:39 UTC