Re: Proposal: 'tag' URIs

From: Tim Kindberg <mailto:timothy@hpl.hp.com>:-

> We'd need an authoritatively managed persistent database that
> recorded all assignment and unassignment events, to a resolution
> of a day. Then 'immediately after' would be well-defined. I'm
> getting the point that we don't have such a thing currently.

While this is true, I think that there is very little need for such a
mechnism, given that these are conceptual resource identifiers rather
than entity locators. As *all* URIs are only worth as much as the
context within which they are used, the only real requirement
practically speaking is that they can be used in processors without
causing to many conflicts. If I create a namespace that no one uses,
then there isn't much of a problem if someone else starts using it and
neither person can prove who owns it. If on the other hand a
particular namespace is widely implemented, then there won't be too
much of a problem in deciding who owns the namespace - the proof will
be in the implementations.

Notwithstanding all of this, I don't feel that one should provoke such
debates by saying that a certain time period which is not easily
definable as "belonging" to a certain organization or person is indeed
defined for a particular URI scheme. What Iif I invented a URI scheme
that uses DNS and dates, but that enforces a "immediately before"
policy? Then you get conflicts. I serously think that if you don't own
a domain name for a certina period of time, then you must not use it.

--
Kindest Regards,
Sean B. Palmer
@prefix : <http://webns.net/roughterms/> .
:Sean :hasHomepage <http://purl.org/net/sbp/> .

Received on Friday, 27 April 2001 15:08:26 UTC