Re: entity mib support for tcif

On Thu, Oct 02, 2003 at 09:59:23AM -0400, Kaj Tesink wrote:
>
> >I agree here. We should call it what it is. Either we support the idea
> >that we need such an object or we don't. But hiding it in some
> >obscurely named object does not seem to make sense to me.
> 
> 
> that is fine with me too; that was my original proposal. but dave's proposal
> allows for accommodating other things than CLEIs as well; see this snippet:

While it is true that you can encode almost everything which is small
enough into an OID value, I doubt that this is a really practical and
useful approach.

Even though I understand Dave's concerns, I think that we should make
practical decisions. If CLEI codes play a significant role, then why 
not put something into the MIB that supports this market? It should
be possible to spell out clearly in the description clause of the object
under which conditions this object applies and that ENTITY-MIB
compliance does not require to implement this object or to buy any official
CLEI codes. In fact, it might be possible to turn this object into an
opaque string object which holds an arbitrary identification code which
in some environments might happen to be the CLEI code. With my limited 
knowledge about CLEI codes, I think a string type is more appropriate 
than an OID type.

/js

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