W3C home > Mailing lists > Public > www-tag@w3.org > February 2005

Re: [XMLVersioning-41] Comments and Suggestions on Draft Extensibility Finding

From: Karl Dubost <karl@w3.org>
Date: Tue, 22 Feb 2005 15:56:00 -0500
Message-Id: <e95b52866df21af4fe61d274c9944885@w3.org>
Cc: <www-tag@w3.org>, <noah_mendelsohn@us.ibm.com>
To: "David Orchard" <dorchard@bea.com>

Le 21 févr. 2005, à 14:46, David Orchard a écrit :
> I am interested in exploring language design beyond "simply" xml and 
> xml
> schema, but I retain the worry that the more abstract the discussion,
> the smaller the audience or the less useful particular audiences will
> find the material.  The finding already is almost too general for my
> tastes as I believe that XML Schema is the most popular choice of 
> schema
> language for xml design.

I do not completely agree with that.
The extensibility/versioning issue is completely related to many 
languages and without really broadening the topic. You may want in a 
finding gives some recommendations which are specifically related to 
XML, but definitely XML technologies are not the only ones at W3C.

History of CSS gives us a very good school case of a non XML technology 
dealing with Extensibility/Versioning problems (or more exactly not 
dealing with it).

CSS 1.0 is a specification. In the language, no mechanism has been 
designed to be able to identify, outside or inside, the version of the 
language. CSS becoming de facto a version-less technology.

CSS 2.0 extends the features and functionalities of CSS 1.0 by being a 
superset without being able to identify the language, except if a 
parser identify an property which is not part of CSS 1.0, then it means 
it's CSS 2.0

So the rationale of the CSS WG is somehow coherent. As long as you 
don't deprecate or obsolete or change the semantics of your features. A 
superset of the previous version doesn't absolutely require versioning.

But the mess might happen soon with CSS 2.1 and CSS 3.0 because it will 
not be anymore a superset and they might redefine some semantics, or 
drop some features.

What are the implications? For example, It's a huge difficulty to 
create a validator, because we don't know a priori which version of the 
language is given to the validator.



-- 
Karl Dubost - http://www.w3.org/People/karl/
W3C Conformance Manager
*** Be Strict To Be Cool ***

Received on Wednesday, 23 February 2005 00:02:30 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:32 GMT