Re: [xri] XRD element order

[Will Norris <will@willnorris.com>]

On Jun 24, 2009, at 12:52 PM, Scott Cantor wrote:

> Will Norris wrote on 2009-06-24:
>> I think its worth noting that SAML, as an example, provides a pretty
>> healthy amount of flexibility with regards to element order in cases
>> where order really doesn't matter.
>
> Actually, I'd say it's the opposite, SAML has a fairly sequence- 
> oriented
> schema. And yes, that does cause problems, but it can be complicated  
> in XSD
> to produce schemas that don't have ordering but also need to  
> constrain the
> number of particular child elements that can appear.
>
> I guess I haven't seen a lot of sign that it causes problems for  
> people. The
> errors I see with SAML don't tend to be element order, more from  
> misusing
> things.

ahh, fair enough.  I guess I got a different impression (perhaps  
incorrectly?) from Chad when I talked with him.  I still like moving  
expires to be an attribute of XRD though, which I think should be a  
safe bet (and actually makes more sense anyway).

-will