Re: Evaluation: draft-ietf-pilc-link-design - Advice for Internet Subnetwork Designers to BCP

["Steven M. Bellovin" <smb@research.att.com>]

In message <200302281700.MAA04789@ietf.org>, IESG Secretary writes:
>
>Last Call to expire on: 2003-3-7
>
>	Please return the full line with your position.
>
>                    Yes    No-Objection  Discuss *  Abstain  
>
>
>Steve Bellovin      [   ]     [ X ]       [   ]      [   ]

I'm saying 'no-ob', but I think that my very first and very last points 
should be addressed by an RFC editor's note.


Section 6: Should there be mention of switches that snoop on IGMP?  I
suspect so.

Section 14: There's an equation I once derived that may be useful.  For 
a compression algorithm that resyncs every n packets (i.e., one that 
tries to send a new compression dictionary every n packets, so that it 
can get better interpacket compression), and with a packet loss 
probability of p, the effective packet loss probability will be 
multiplied -- if a packet is dropped, all subsequent packets up to the 
resync point are dropped.  The effective packet drop probability is
P = 1 - (\sum_{i=1}^n (1-p)^i) / n.  (I have no idea if they would want 
to incorporate that at this point.)

18: Another way to view the need for WEP is that it protects a link 
with a much-higher threat level.  In the case of, say, consumers with 
credit card numbers, the easy places to intercept them are at the 
edges.  The consumer edge (i.e., the home wireless LAN) can be 
protected with WEP; at the business edge, an attacker who can sniff 
them can more easily steal the whole database from the host.

The list at the end should include replay attacks.



		--Steve Bellovin, http://www.research.att.com/~smb (me)
		http://www.wilyhacker.com (2nd edition of "Firewalls" book)