> I see no good reason why the same index mistake should be repeated for
> IPv6.
>
OK.
If I understand you, the attached proposal should fit your
requirements...
> Simple, but is a slight
> problem if one removes peers in the middle.
Note the "tricky" comment solving that problem... Just not allowing it !
I think this can be a realistic assumption. As Amos said, it is not
expected to change peer configuration very oftenly. In fact, this mib is
inspired on the way IPv6 Interfaces tables is done on rfc 2465, your
very first idea.
Check it. If sound, I will commit.
This archive was generated by hypermail pre-2.1.9 : Sat Dec 01 2007 - 12:00:05 MST