Re: [xml-dev] Validate **against** a schema OR validate **with** a

[Rick Jelliffe <rjelliffe@allette.com.au>]

DavePawson wrote:
>  
> Is that the heart of the matter Rick?
> Thinking of todays list of 'applications' to take a simple SGML 
> validation and partition it into many parts for different reasons,
> isn't augmentation just another option along the way (that some
> people want)?
There is no way for an XSD schema or an instance to communicate to 
processing software that the full infoset of the document will only be 
obtained if default values specified in the schema are added. With DTDs, 
the problem does not arise: you have to add the defaults (if 
standalone!=yes etc). With RELAX NG, the problem does not arise: there 
is no notion of defaulting.  For XSD, you have to look at the particular 
schema (does it provide defaults?) and know something about the instance 
(does it assume defaults?) and both (do the defaults that the instance 
assumes match the defaults in the schema I will use?) and then work out 
how to configure your system to make sure that defaulting is done (if 
there is in fact a built-in way.) 

>
> If W3C won't address it...
>
XML and XSD and all the spaghetti in between are their standards: no-one 
else will take it on. And there is no groundswell for XML 2.0 which 
could consolidate and fix these: mostly when I see calls for XML 2.0 it 
relates to some syntactical piddliality.
>
> But clearly not so amongst the readers here.
But we are clearly the solution to any and every problem, and never the 
cause...

Cheers
Rick Jelliffe