AgentX support

[Frank Fock]

Guy.HINDELL____opentrade.co.uk wrote:
> 
> 
> However, I notice that the 'subagent2' binary works against the UCD agent
> (when
> I walk .1.3.6.1.2.1.37 I get three oids back), but the 'subagent' binary
> does not.
> They both work with the agentX++ master agent. I'll keep playing with it.
> 

The subagent example does not implement the ATM MIB, thus it is ok
that it does not return any oids below .1.3.6.1.2.1.37. Instead it
should return some strange OIDs after MIBII (strange because they are in 
the within IETF tree - which I know is illegal, but I wanted to test it 
anyhow).

> Also, I compared the behaviour of the subagent atm_mib with the original
> agent++ example running as a stand-alone agent, and that gives the same
> result,
> but needed a slight fix to compile - gcc complains about the use of an
> snmp_community_mib in agent.cpp which doesn't exist if SNMPv3 is not
> defined. Needs
> a #ifdef :-)
> 

These erratas in the examples are already fixed in AGENT++v3.4i which
will be available very soon...

Best regards,
Frank