W3C home > Mailing lists > Public > public-html@w3.org > December 2008

Re: metadata content

From: Julian Reschke <julian.reschke@gmx.de>
Date: Thu, 25 Dec 2008 11:46:26 +0100
Message-ID: <49536482.5020705@gmx.de>
To: Ian Hickson <ian@hixie.ch>
CC: public-html@w3.org

Ian Hickson wrote:
> On Thu, 25 Dec 2008, Julian Reschke wrote:
>> Ian Hickson wrote:
>>> That does seem to a spec for it, indeed. How about "adobe:ns:meta/"?
>>>
>>> (Another reason to have a centralised registry rather than URIs or 
>>> java-like identifiers is to make it possible for conformance checkers 
>>> to catch typos.)
>> So are you saying that a single central Wiki-shaped registry for XML 
>> namespaces would actually work in practice?
> 
> XML Namespaces on the Web, or XML Namespaces in general? The use cases are 
> very different. For XML in general, there are multiple completely 
> unrelated markets, within which different rules will be needed. It doesn't 
> matter if a car manufacturer doesn't know about flower namespaces, because 
> they won't be interacting with flower shops.

Indeed.

> For HTML, where interoperability matters (i.e. "on the Web"), there is a 
> single, very large, market. It's a very different situation, and warrants 
> different solutions. Web sites for car manufacturers will be accessed with 
> tools that are also used for flower shops. There are no silos.

That is correct, but I don't see why that means a central registry is 
required. There are many case where one wants to put something on the 
web, but still use extensions that only some of the recipients will 
understand.

A good compromise is a mix of both, as currently practiced for Atom link 
relations.

BR, Julian
Received on Thursday, 25 December 2008 10:47:07 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:39:00 UTC