Re: Didn't I Start This? (Was: Autotopology)

["Karl Auerbach, Empirical Tools and Technologies, 408/427-5280" <[e-mail]>]

 > I appreciate the work that you as chair and the other members
 > of the SNMPv2 work group has accomplished (and are to be
 > congratulated). However, I believe this discussion should
 > remain cross-posted because it seems that SNMPv2, as defined,
 > will break most existing discovery/autotopology mechanisms.

I don't see how V2 is going to break something that didn't work in
the first place.

I've found SNMPv1 to be of only minor use in autodiscovery because:

1) it is still relatively rare in real networks

2) many implementations are broken or support miniscule, trivial mibs,

3) there is no completely ubiquitous community string, and

4) people don't put useful information into their system group data

5) The standards do not clearly say "SNMP agents ought to respond to queries
   received on a broadcast address."

			--karl--