Re: spatial/temporal issues (was Re: putting time into snmpv2

[Keith McCloghrie <[e-mail]>]

Michael,

> In Regards to your letter <[e-mail]>:
> > The "brilliant insight" I tried to convey in the few minutes that were
> > left in the DC meeting was that issues of the spatial and temporal
> > semantics of MIB variables are completely unrelated to the issues of
> > security, access control, and proxy that are addressed in RFC 1351.
> > In the mail I've read so far, no one has contested this numbingly
> > obvious claim.
> 
> mtr also brought up a similar point in his introduction to the new party mib
> stuff.  He indicated that "Security" is a part of the "SNMP Administrative 
> Model" which also addresses other things (proxy, other protocols, etc.).
> 
> Do you think the "SNMP Administrative Model" is currently addressing too much?
 
The scope of the "Administrative Model" derives from the work done by
the SNMP Security WG over the 3 years leading up to RFCs 1351/1352/1353.
Those RFCs specify a framework which encompasses not only authentication
and privacy, but also proxy and access control.  The decision to have 
the SNMP Security WG continue, with a charter unchanged in its scope
and in parallel with the SNMPv2 WG, retains that scope within the 
Administrative Model.

All proposals are trying to build upon 1351 in which spatial semantics are
tightly bound to security, proxy and access control.  Specifically, a) spatial 
semantics are specified through 1351's definition of proxy, b) both security 
and proxy parameters are contained in the partyTable, and c) access control 
for proxied parties is specified through the combination of 1351's partyTable 
and aclTable.  In fact, the proposal presented by Steve loosened that 
binding from 1351.  Chuck's proposal moves further away from 1351.

> Like you said, you did not get much time at the end of the session.  

I would like to correct this misunderstanding.  While I regret that there 
was insufficient time at the SNMP Security meeting on the Wednesday 
evening for questions on Chuck's proposal, and I also regret that we were 
not able to continue the discussion on Friday (and Chuck might want to 
thank Jeff for defending him in that respect on Friday morning), I would 
not like anyone to get the impression that the process was anything but 
fair on the Wednesday evening.  

Specifically, even though Chuck's proposal was not mentioned on the 
agenda sent out on the mailing-list the prior week, Chuck was given the 
choice of whether he wanted to present his proposal before or after 
Jeff and Steve's presentations, and given that choice, Chuck chose to 
go last.  Even then, each presenter had between 30 and 45 minutes, and 
the meeting went past 10pm specifically to allow Chuck to finish.

If I am overreacting here, to an unintended implication, then I apologise.
But I feel it's important that the record be set straight.

> Are you going to post your points/proposal to the mailing list?  

I fully encourage Chuck to post his proposal to the working-group.

Keith.