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

[Frank Strauss <strauss@ibr.cs.tu-bs.de>]

Hi!

Durham, David writes:

DD> First, if there is no motivation for a requirement it is still
DD> fine for it to appear in the REJECTED REQUIREMENTS section. This
DD> section specified which proposed requirements were discussed but
DD> there was insufficient consensus to make them required to be part
DD> of sming. This will show that at least the requirement was brought
DD> up, even if insufficiently motivated. 

Just a reminder: 

  Chair> [...] Also by this date [April 15th 2001], motivation needs to be
  Chair> supplied for the existing requirements that don't have a clear
  Chair> motivation (i.e. use) or they will be removed from the requirements
  Chair> document (requirements without any motivation will not be
  Chair> considered).

So, I would like to see the proponents of those requirements that miss
a motivation (no matter in which section) to add a clear motivation at
least until friday.

DD> As for tables, if there is no motivation, it is clearly not an
DD> accepted requirement... Although I recall some motivation related
DD> to keeping the table nomenclature to remain consistent with SMIv2
DD> & SPPI terminology, and making translations between SMIv2 and
DD> SMIng easier. Afterall, many of the existing description clauses
DD> refer to tables.

From the perspective of the NMRG SMIng proposal, this is why the *SNMP
mapping* does talk about tables, while the core SMIng language follows
the general OO-motivated model of attribute grouping and inheritance
were appropriate. I don't think that it would be a good idea to talk
about tables in the core SMIng.

 -frank