Re: comments on draft-ietf-sming-reqs-02.txt: Subject Categories

[Juergen Schoenwaelder <schoenw@ibr.cs.tu-bs.de>]

>>>>> David Harrington writes:

David> I have a question, though. I am of the impression that the RAP
David> WG found subject categories to be either undesirable or
David> questionable in their utility. I don't remember what led to
David> that impression. Can an SPPI expert comment on whether this was
David> a concern, and if so explain what the concern was? If it is an
David> undesirable feature, just as IMPLIED is in SMI, I would not
David> like to see it categorized as a requirement.

I do not consider myself an SPPI/COPS-PR expert. But anyway, the
COPS-PR protocol requires a Client-Type value to function properly.
The subject categories clause in the SPPI is used to define it.

In contrast to that: The IMPLIED keyword is not required to make SNMP
work (except for existing definitions that use IMPLIED). In other
words, you can "remove" IMPLIED and still write MIBs that work with
SNMP is it does today. If you remove the subject categories clauses
from the SPPI, then you have to invent another mechanism to define
the Client-Type value used in protocol exchanges.

I am saying that we need to address this issue. Whether we keep the
subject categories as they are defined in the SPPI or invent something
new or establish a registry is something we have to work out once we
are done with the requirements.

/js

-- 
Juergen Schoenwaelder      Technical University Braunschweig
<schoenw@ibr.cs.tu-bs.de>  Dept. Operating Systems & Computer Networks
Phone: +49 531 391 3289    Bueltenweg 74/75, 38106 Braunschweig, Germany
Fax:   +49 531 391 5936    <URL:http://www.ibr.cs.tu-bs.de/~schoenw/>