Re: Evaluation: draft-ietf-tsvwg-tcp-nonce - Robust ECN Signaling with Nonces to Experimental

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

I'm nervous about making this one Proposed without some experience.  I 
can't quite convince myself that it's robust against various kinds of 
packet loss, duplication, reordering, etc.  (Yes, I know and respect 
the authors, but...)

Beyond that, our experience with firewalls and the original ECN 
mechanism has me concerned.  Are there firewalls out there that are 
going to get upset at seeing a new TCP bit set to 1?  I don't know, but 
I don't particularly want to find out the hard way that an 
implementation conforming to a new Proposed standard can't talk through 
some firewalls.

I'd be delighted to approve this as Experimental.

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