RE: extension attributes in WSDL 1.2

I'd just like to point out that putting a "required" attribute on attributes
is a very very bad idea.  XML does not do attributes on attributes, hence
the whole xlink/hlink mess.

This is one of the reasons why many people like to make elements out of
things, rather than attributes.  Maybe the spec should say the behaviour is
undefined, and that extension attributes that require processing SHOULD not
be modeled as attributes?

Cheers,
Dave

> -----Original Message-----
> From: www-ws-desc-request@w3.org [mailto:www-ws-desc-request@w3.org]On
> Behalf Of jbbkfri@netzero.com
> Sent: Tuesday, October 15, 2002 11:59 AM
> To: www-ws-desc@w3.org
> Subject: extension attributes in WSDL 1.2
>
>
>
> In section 4 (Language Extensibility and Binding) of the base
> 1.2 WSDL specification, the optional wsdl:required attribute
> is described as being used with a value of "true" to indicate
> that a WSDL processor must recognize and be able to process
> the extension element that wsdl:required is used on.
> Omitting wsdl:required is equivalent to stating it with a
> "false" value.
>
> For extension attributes, there is no method prescribed for
> determining if it is required or not, so I'm assuming that a
> WSDL processor is not required to understand or process
> extension attributes.
>
> In either case, nothing is said regarding the required or
> recommended behavior when a WSDL processor can not make sense
> of an attribute extension or element extension that is not
> required.  I would like to be able to make use of an
> extension and be assured that processors that did not
> recognize it would ignore it and continue processing.  It
> seems that at a minimum, the spec should indicate that a WSDL
> processor SHOULD ignore unknown extensions.
>
> Can anyone tell me what the intended behavior is when
> processing non-required extensions?
>
> Tnanks,
> Barry F
>
>
>
>
>

Received on Tuesday, 15 October 2002 19:25:52 UTC