Re: IMPORTANT -- Proposed change to Entity Sensor MIB
Margaret Wasserman <[email protected]>
| Newsgroups | gmane.ietf.entmib |
|---|---|
| Message-ID | <[email protected]> |
>
>Tools that rely on globally unique names are broken. Unless there is
>evidence that the majority of people use broken tools, I think we
>should not start to change names just because we have clashes with
>some proprietary MIBs.
The Wind River code generator (AKA the former Epilogue Envoy
product) is broken in this way. If you use it to generate
code for two MIBs will the same object descriptors, it will
generate conflicting function names.
Some of the tools that I have heard are "broken" include:
SNMP Research's Emanate code generator
HP OpenView
..but, I can't verify this from my own experience.
Also, using non-globally-unique names will make it harder to
access this MIB from any of the MIB browser products I've used,
if the conflicting Cisco-proprietary MIB is also loaded.
I'm still not sure, though, how big of a problem this is or
isn't. How many of the target devices for this MIB will
implement both this MIB and the Cisco-proprietary version?
Margaret