Re: Proposal: 'tag' URIs

Tim and Sandro,

I think your proposal is a thing of beauty.

Namespace management is a very complex socio-technical issue.
By necessity, allocation of names within "URI scheme space"
and the top-level URN namespace must be tightly controlled, to
avoid problems of cybersquatting.  Furthermore, these top-level
spaces must be considered as public and subject to redefinition
through public participatory processes involving W3C, IETF and
other relevant groups.

What your proposal has the potential to achieve, it seems to me,
is the creation of a standard mechanism for auto-allocated
private namespaces.  The namespaces so created will each have
reasonably short and humanly understandable names.   The namespace
names will be globally unique by construction.  Very nice.

For example tag:cameron@sfu.ca/1 is an automatically allocated
namespace that I can use for whatever purpose I want, with the
confidence that, as a private namespace, I can retain control
of it.  I also know that no one else who follows the tag: 
specification will accidentally create conflicting tags.
Furthermore, if someone else deliberately tries to use this
namespace, I will be able to prove my ownership in a court of law.

The human readability of the tag specification will be an important
aid to future debugging.  If someone comes across a tag: URI
and is trying to figure out how it works in their application,
they will at least have a good clue about where to go to get
information on the namespace and the identifiers minted thereunder.

There is also the future potential of creation of some standard
for, perhaps, metaresolution services.   That is, a tag:
metaresolver might be defined to produce information about a
namespace automatically without necessarily being able to
carry out full resolution of tag: URIs.

Received on Saturday, 28 April 2001 11:47:47 UTC