entPhysicalUris object wrap-up

Hi,

I'd like to finish this up so I can submit an I-D update.
I have some issues to clarify, which require WG consensus.
I would rather remove this object than hold up progress
any longer, but if issues get resolved quickly then it
would be better to finish it now.

1) object purpose

There is strong opposition to a generic URI(s) placeholder
object by a couple people.  This is an object that contains
any semantics related to a physical entity, for which the
syntax happens to be encoded as a URI string. (very low
cohesion). 

I propose that this object be changed back to entPhysicalClei.
It should contain the empty string (for none) or a URI
string conforming to the CLEI URN specification (from Kaj).
If we think of other required semantics in the future, 
we can add new objects then.

2) references

As per Bert's comments, the citation to RFC 2396 will be 
removed from the DESCRIPTION clause, since it is redundant.
Should this be changed to point to the CLEI URN specification, 
or just add another document to the REFERENCE clause?
Where else should this spec be cited in the document?
There are normative references, so how long will the 
Entity MIB be held up waiting for the CLEI URN RFC?

3) conformance

Currently the new OBJECT-GROUP (entityPhysical3Group)
is in the MANDATORY-GROUPS clause.  If both objects
allow for null values, then this should not be a problem.

4) examples

I don't really have time (or enough info) to fill in the
examples in section 4, but I know somebody will eventually
ask about this before RFC publication.  I would really
appreciate some help updating the examples.  (I need to
add an example for entPhysicalClass==cpu(12) as well.)
Any volunteers? (4.1 and 4.2 can be done independently)
I would like to get this over with now ;-)


thanks,
Andy