Re: minutes of RIPv2 meeting in Amsterdam

[Phil Budne <[e-mail]>]

	From: Gary Scott Malkin <[e-mail]>
	Date: Wed, 4 Aug 1993 13:17:16 -0400

	Just because OSPF did it one way, doesn't mean we can't do it another
	way.  Especially if our way is advantageous to us.  Unless someone
	can show that the OSPF way is significantly superior, I think we
	should stick with the current proposal.

The current propsal makes too strong a connection between "unnumbered"
and "not seperately numbered" (which the current RREQ draft blesses as
the preferred method).  And while the OSPF wording does the same (by
defining "lessIF" to be zero if the interface "has an address" (not a
unique address mind you)) the contruct is still clearly superior, and
would allow use of A.B.C.D.N if lessIf were replaced with a plain
ifIndex.

The TRUE problem here is the ip Address Table, but the current
proposal is still just a crock (in that it introduces a
discontinuity). A consistant use of the current INDEXing would be to
ALWAYS use 0.0.0.I (particularly if ipAddrTable could show multiple
interfaces with the same local address), but you won't catch me
suggesting it!