Re: RDF & fragment ids; what breaks?

On Thu, Jan 23, 2003 at 11:32:26PM -0500, Tim Berners-Lee wrote:
> This (fragid hiding) is a form of end to end property.
> The significance of the the mime message is known only
> to sender/publisher and to receiver/browser.
> It is only in that context that the fragment ID makes sense.

I understand what you're getting at, but I don't believe this is
particularly valuable.  You argue that "RDF uses this hook to introduce
identifiers for arbitrary concepts", yet Roy and I and others are
saying that you can already do that without this hook.

FWIW, I have a hunch that there's nothing of importance that can be
accomplished with this hook, that couldn't be accomplished without it
(and therefore without the cost of using the hook); REST's constraints
are constraints on form, not function.  But, as it's just a hunch, I
won't bother trying to defend it, for now at least.

Plus I'm not sure that we're headed in the direction of proposed text
for webarch with this line of discussion, so I'm happy to drop it in the
hope that your discussion with Roy will do so.

And Jonathan wrote;
> Theoretical arguments aside it just wouldn;t be convenient to package each
> and every thing we talk about in a different document.

Sure, but I never intended to suggest that *everything* be identified by
a URI, only that there is a substantial cost in identifying things by
URI+frag (hence my avoidance of the word "break").  Judging by the RDF
I've seen out in the wild, which has lots of URIs with fragment ids, many
Semantic Web folks either;

- don't believe there's a cost
- aren't aware of the issue
- aren't given a choice, as many things they want to refer to are
identified only by URIs with fragment ids.

MB
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
Web architecture consulting, technical reports, evaluation & analysis

Received on Friday, 24 January 2003 01:08:02 UTC