Re: [MIB-DOCTORS] [OPS-DIR] FW: Recharter: ISMS

["David B Harrington" <dbharrington at comcast.net>]

Hi,

I separated my responses, since one is about charter text and the rest
is an architecture and design discussion.

> Pedantically, you're correct.  The RADIUS client provides the 
> VACM extension
> with a policy name that's bound to a username and the VACM 
> extension uses
> that information to populate the MIB table.

Yes. My point is that the charter does not say this. The text is
wrong.
The charter says 
    a RADIUS-provisioned
    username-to-groupname dynamic mapping, that would provide a
binding
    between a user and preconfigured VACM policies via dynamic
additions
    to the securityToGroupname table.
What it should say is what you said:
	a RADIUS-provisioned policy name bound to a username, which
the VACM 
	extension will use to dynamically populate the
securityToGroupname table.

This may seem pedantic to you, but it is more than that to me. It took
the SNMPv2/SNMPv3 community ten years, including some very acrimonious
times, to get an architecture we could agree to, that was flexible
enough to meet the demands of multiple segments of the community. If
we want to change the architecture, then we should do so deliberately,
not as a side-effect of sloppy design. 

I strongly object to a WG goal based on the current wording, because
it is counter to the intentions of the SNMPv3 WG, and is not
compatible with the RFC3411 architecture.

I strongly support a WG goal based on the corrected wording.

dbh