Re: URN UUID question

[Joel Kalvesmaki <[e-mail]>]

Thanks, Dale,

> From: Joel Kalvesmaki <[e-mail]>

> >
> > I would like people who do not own or have access to a domain name to be
> > able to mint names.
>
> The current tag scheme doesn't allow that, but it's fairly simple to
> get around that problem.  (Although it's difficult to imagine anyone
> who would use this system but not have an e-mail address.)  We need
> some person or enterprise that has a legitimate authorityName to
> delegate parts of its URI space to owners who do not have
> authorityNames.  So we could separate:
>
>     tag:[e-mail],2014:delegations:person-A:[person A's string]
>     tag:[e-mail],2014:delegations:person-B:[person B's string]
>     tag:[e-mail],2014:delegations:person-C:[person C's string]
>     tag:[e-mail],2014:delegations:person-D:[person D's string]
>
> We could probably set up an automatic delegation system much
> resembling how tag URIs are delegated by allowing each possessor of a
> telephone number a set of strings:
>
>     tag:[e-mail],2014:delegations:17816479199,2014:[my string]
>
> would be automatically delegated to me because I possessed the use of
> that E.164 number on 2014-01-01.
>

My caveat on users' access to domain names wouldn't apply to email
addresses (I can safely assume users of this data model will have, at some
point, some email address), but your delegations idea is very interesting
to see.

> My concerns are akin to those that have motivated the architects of
> > Canonical Text Services[1][2][3] to develop the convention
> > "urn:cts:..." to provide names for ancient literary works, their
> > fragments, and their versions. their naming scheme stands
> > independent of server performance, etc. But it also can be easily
> > incorporated into registries that facilitate Semantic Web
> > applications.
>
> Good Lord, haven't they thought to properly register those schemes?  I
> don't see them listed in
> http://www.iana.org/assignments/urn-namespaces/urn-namespaces.xhtml .
>

Sorry I can't speak to this, since I'm not using any CTS urns, and I wanted
to avoid creating a new urn scheme. I think what the architects have
published so far suggests that they still have work to do, esp. on defining
validity and developing rules for a registry (or explaining why no registry
would be needed).

Best wishes,

jk
-- 
Joel Kalvesmaki
kalvesmaki.com