[ogsa-wg] Teleconference minutes - 9 November 2005

[Karl Czajkowski]

On Nov 14, Mark McKeown modulated:
...
> Is this technical issue not addressed by IETF RFC 4122,
> "A Universally Unique IDentifier (UUID) URN Namespace"?
> 
> http://www.ietf.org/rfc/rfc4122.txt
> 

Thanks, Mark, for giving me the nudge to beat this dead horse...

I think it is an issue that the WS-Naming proposal _could_ easily
address by mandating UUID syntax in the AbstractName (or defining a
separate UUID-containing element QName that can be used in lieu of
AbstractName).  AbstractName has been left open to undisciplined and
ambiguous use.

As it stands, I fear the notion of an AbstractName without any
normative or runtime-encoded reference to a naming convention,
namespace, and/or naming authority is not actually useful. The UUID
syntax essentially defines a namespace based on Internet names, and
other approaches could also serve, such as using an "assertion" model
where the name and the one issuing the name are both identified (the
"EPR minter" as Frank Siebenlist has been putting it).

Much to my surprise, this does not seem to be a universally held
position within GGF!  These basic concerns were brought up in the BoF
in Seol last year, if not earlier in all of the related OGSI naming
discussions.


karl

-- 
Karl Czajkowski
karlcz at univa.com