RE: [xml-dev] Re: determining ID-ness in XML

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

Going by his last reply, they want something 
that no one has to support and is unenforceable 
so no more reliable than what we have now.

A PI will certainly work for that. A namespace will 
work for that.  No one is a bit better off than 
they were before.  It becomes like RDDL: sort of 
a nice suggestion that doesn't have to be implemented 
so isn't a reliable means.

len

-----Original Message-----
From: Champion, Mike [mailto:Mike.Champion@SoftwareAG-USA.com]

> From: Bullard, Claude L (Len) [mailto:clbullar@ingr.com]

> Independent parties should not be extending 
> the rules for the system vocabulary.  That 
> is precisely what so many here beat MS up for. 

I thought Elliotte Harold was suggesting that both the namespace mechanism
as well as PIs provide a pre-approved way of extending the rules to
accomodate ID-ness. Unless I'm missing something, declaring an attribute to
be in the "id namespace" could work as well or better as the PI suggestion
... *if* we want to propose some interim solution that doesn't require a W3C
Recommendation.