RE: Significant W3C Confusion over Namespace Meaning and Policy

I tend to agree with Patrick. Should there really be a single RDDL document for the XSLT namespace even though there are significant differences between v1 and v2 or identifiers and representations for each distinct version of XSLT? Given a choice I'd prefer the latter to the former. 
Perhaps versions shouldn't be indicated via opaque strings or numbers but instead be URIs which dispense RDDL documents when dereferenced. 
The road to to success is always under construction.   


From: on behalf of Elliotte Harold
Sent: Fri 2/18/2005 2:44 AM
To: Patrick Stickler
Cc: ext Norman Walsh;
Subject: Re: Significant W3C Confusion over Namespace Meaning and Policy

Patrick Stickler wrote:

> Exactly!
> And if some application encountered a term grounded in
> the Docbook namespace (sans any indication of a particular
> version of the Docbook document model) and tried to
> find out how to interpret that term by dereferencing
> the namespace name URI, (presumably) getting a namespace
> document, that namespace document would not be able to
> tell the application how to properly interpret that term.
> The application might thus be aware of various options,
> but ultimately, it would have to guess.

You assume that the interpretation of a term varies from one version of
a model to another. In practice, that's not unheard of but uncommon.
Most of the time, the same terms carry the same meanings from one
version to the next. It's only the new terms added in new versions that
take on new meanings.

Of course a RDDL document can contain content that's specifically
relevant to particular versions of a vocabulary. That's what the
xlink:arcrole attribute is used for.

Elliotte Rusty Harold
XML in a Nutshell 3rd Edition Just Published!

Received on Friday, 18 February 2005 14:25:24 UTC