Re: [urn] example URN NIS

[Alfred Hönes <[e-mail]>]

Julian Reschke wrote:

> Hi there,
>
> is there a recommended NIS for URN examples? RFC 3406 doesn't seem to
> cover this...
>
> Best regards, Julian

<speaking as an individual>

Julian,
I assume you are looking for a dedicated URN Namespace for examples
(i.e. a formally assigned "NID").

There isn't one, and due to the dichotomy between the properties
of URNs (see RFC 1737 -- in particular the first 3 bullets in
Section 2 there: global scope, global uniqueness, persistency, etc.)
and the nature of 'hypothetical' examples (for educational purposes
only), the last time this has been considered informally, it has been
agreed that such 'example' URN Namespace would be a poor idea and it
should not be assigned by rfc3406bis.  (The discussion around the
URNbis BOF and in March 2011 has focussed on a namespace for
"experimental" URNs, but most of the arguments raised there against
these are similarly valid against "example" URNs.)

On the other hand, given the plethora of publicly assigned URNs that
can be used persistently without negative side effects, AFAICS the RFCs
so far in need of using concrete examples of URNs in general (and not
in the context of the discussion of some particular URN Namespace)
have used examples of existing bibliographic URNs, and doing so seems
to have well served the educational purpose, because everybody likely
has an idea of what kind of resource a urn:isbn might refer to.
Other well known URN the use of which should not incur any harm might
be considered in the IETF context; in particular URNs like
urn:ietf:rfc:3406  or urn:ietf:mtg:82  immediately come to my mind.

But maybe we need to reconsider this topic, if you see an urgent demand.

If (and only if) the use of something that looks like a URN (but isn't
permanently assigned and stable) for educational purposes is deemed
worth defining a form of URNs that can be easily identified as such,
a possible alternative to an entire URN Namespace dedicated to
example usage would be the creation of a branch in the 'IETF'
URN namespace. Beyond the urn:ietf branches defined in RFC 2648
(rfc, std, bcp, fyi, id, mtg), so far only the 'params' and 'xml'
branches have been defined (in RFCs 3553 and 3688, respectively),
and one might consider defining an "exceptional" 'example' branch
as well.


Kind regards,
  Alfred.