Re: [ubl-ndrsc] [Fwd: FW: [oagis-users] UserArea Extensions forVersion 8.0]

["Eve L. Maler" <eve.maler@sun.com>]

You *can* do that.  You can specify ##other (namespaces other than this 
one) and you to turn off validation with processContents.  But would we 
really want to put this between every element?  (I suppose this is one 
way to allow a context methodology that's closer to true XSD derivation 
-- you turn all customizations into restrictions!)

The usual practice with wildcards is that they're limited to places 
where you're quite sure that your own vocabulary is not responsible for 
the semantics of something that's nonetheless essential to provide.

	Eve

Eduardo Gutentag wrote:
> Wouldn't it be nice if you could specify "here you can put anything from 
> any namespace other
> than the current, or here you can put anything from namespace x, and it 
> will yes/not be validated"?
> 
> Eve L. Maler wrote:
> 
>> I would be concerned about putting wildcards everywhere.  It would 
>> mean that the vanilla UBL library would allow all kinds of junk, and 
>> XSD validation would mean little.  But there may be locations where we 
>> want to use it strategically.  (Eventually we'll need a rule about 
>> wildcard usage anyway...)
>>
>>     Eve
>>
>> Burcham, Bill wrote:
>>
>>> When I look at that BOD extension example (below) I see an "open 
>>> content" model (as described by Roger Costello in this ancient 
>>> correspondence 
>>> <http://lists.w3.org/Archives/Public/www-xml-schema-comments/2000JanMar/0159.html>and 
>>> in part three of his eggcelent XML Schema Tutorial 
>>> <http://www.xfront.com/xml-schema.html>).  I think UBL should define 
>>> an open content model, i.e. just about everywhere in a valid UBL 
>>> document instance, it should be easy to hang elements from non-UBL 
>>> namespaces.   What do you think?
>>>
>>> -Bill
>>
>>
>>
> 


-- 
Eve Maler                                    +1 781 442 3190
Sun Microsystems XML Technology Center   eve.maler @ sun.com