Re: [xml-dev] The purpose of a namespace URI is ...

[Liam R E Quin <liam@w3.org>]

On Sat, 2012-12-01 at 12:44 -0800, Dan Vint wrote:
> Coming from the SGML world, I prefer to think of them as a Public 
> Identifier in a different format.

This is not a good way to think of a namespace URI.

There are some things in common - a namespace URI is a name - but some
important differences:
. Public identifiers in SGML are dereferenced indirectly (officially by
sending a fax to ISO last time I checked but that was a long time ago);
there's no expectation that anyone will ever try to dereference a
namespace URI (people do, because it looks like you ought to be able to,
and some people like to put human-readable explanations of the namespace
at the namespace URI, but there's no guarantee and software should not
do this automatically)
. Public identifiers explicitly contain versioning in their syntax. Two
public identifiers differing only in version or in language are known by
SGML software to be related. Two namespace URIs differing by even as
little as a single bit are considered 100% unrelated by namespace
processing software, and are not expected to refer to similar things at
all.

>  I process documents and I don't 
> typically use JAVA or other programming language to deal with them. 
> As such, I prefer to have a unique identifier for versions changes. 
> If not for minor changes, anything that introduces a change that 
> would break backward compatibility should be identified.

It's usually better to do this with a version attribute rather than to
use a new namespace URI and thereby claim you're making a new and
entirely unrelated language.

Liam

-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
Ankh: irc.sorcery.net irc.gnome.org freenode/#xml