Re: To meet or not to meet?

On Tue, Jul 20, 2004 at 12:42:33PM +0300, Romascanu, Dan (Dan) wrote:
 
> 1. Should the language about using the CLEI codes be more decisive? 
> SHOULD instead of MAY? For example: 'If CLEI codes are supported in 
> the operational environment where this agent is implemented, this 
> object SHOULD include a URI containing a Common Language Equipment 
> Identifier (CLEI) URI [XXX] for the managed physical entity.'

I guess we should leave that to the decision of the implementator
(or whoever configures this object). If we add more decisive text,
we might again have a debate whether the IETF features proprietary
identification mechanisms.

> 2.' If no additional identification information is known or supported 
> about the physical entity the object is not instantiated' - This object 
> has a MAX-ACCESS read-write. How would the creation or deletion of the 
> first URI in the list work? Would not it be simpler to use a zero-length 
> string for the case when no identification information is known?

I just kept this sentence from the original text. I don't mind allowing a
zero-length string instead.

> 3. 'Note that [XXX] would refer to a most likely informational RFC which 
> registers a URN namespace for CLEI codes.' - Does [XXX] really need to 
> be an Informational RFC? I am afraid that this would hang this MIB until 
> that RFC goes through our wonderful IETF process. Maybe the Telcordia 
> folks already have a document to be referenced as an Informative 
> Reference, or can create one on a shorter path.  

This MIB has already been hanging around for quite some time. While
it is popular to point to the slow IETF process, I believe that in 
many cases the speed in which a reasonable document is written does
contribute quite a bit to the overall speed. I do have a strong 
preference to have a URN namespace formally registered and all this
documented by an RFC. I do not think it is a complicated thing to
do - especially if someone from Telcoridia who knows how these CLEI
codes work writes this up. There are enough template RFCs out there.
There is no reason why such a document should not be ready to go to
the RFC editor as an individual submission directly after the next
WG meeting.

/js

-- 
Juergen Schoenwaelder		    International University Bremen
<http://www.eecs.iu-bremen.de/>	    P.O. Box 750 561, 28725 Bremen, Germany