W3C home > Mailing lists > Public > xmlschema-dev@w3.org > January 2001

Re: [schema] Re: Forcing a certain document element

From: Eric van der Vlist <vdv@dyomedea.com>
Date: Sat, 13 Jan 2001 08:03:48 +0100
Message-ID: <3A5FFDD4.A90A14E3@dyomedea.com>
To: xmlschema-dev@w3.org
Robert Braddock wrote:
> 
> On Fri, Jan 12, 2001 at 11:54:28PM +0100, Eric van der Vlist wrote:
> > IMHO, since namespaces are aimed to disambiguate element names, ns1:elt
> > and ns2:elt should be considered as having a different semantic...
> 
> I agree.
> 
> > And if you import the definitions using unqualified types, you hide this
> > since you get similar element names from different namespaces that are
> > expected to have the same meaning !
> 
> I disagree :). While the tags don't carry the namespaces on them, they still
> are from that namespace, and any processing that needs to care about
> meanings will know that. (It could mechanically find out from the schema, if
> necessary.)
> 

Then you accept to be to be locked with W3C XML Schema and it's a still
deeper difference of perspective :) ...

<rant>
<disclaimer>W3C XML Schema is great</disclaimer>, but I tend to see it
(and any schema language) as an add-on spec that can help to validate
documents and build application when you seem to see it as a one the
basis of your architecture.

Seeing W3C XML Schema as an "add-on" I'd be as reluctant to lock the XML
applications I am developing as I would be to tie them to a specific
programming language or operating system.

I believe that XML would loose a lot from being locked with any schema
language. These schemata are complex technologies and AFAIK none of the
current ones can describe all the possible XML vocabularies and this
would lead to restricting the scope of possibilities. 

Furthermore, they are adding complexity to the parsing as well and are
probably not a good fit for applications to be run on smaller
appliances.

<conclusion>For all these reasons, I'd be very reluctant to build any
application based on the infoset modified by a schema
language...</conclusion>

</rant>

That's probably a very personal view, though.

Eric

> > Wouldn't it be simpler to use a single namespace then ?
> 
> Then you have name collision issues instead of nice encapsulation.
> 
> Granted that there is a rather serious disadvantage in going without global
> elements, since there are schema features that require global elements :(
> Having unqualified documents be incompatible with those features seems not
> quite right to me though. ?
> 
> Robert Braddock

-- 
------------------------------------------------------------------------
Eric van der Vlist       Dyomedea                    http://dyomedea.com
http://xmlfr.org         http://4xt.org              http://ducotede.com
------------------------------------------------------------------------
Received on Saturday, 13 January 2001 02:01:37 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 11 January 2011 00:14:19 GMT