Hi Remi, Here are some comments. Tom P. Section 4.3: "o The filtering behavior for DCCP MAY be independent of the filtering behavior for UDP." Shouldn't that be "for UDP or any other protocol" or something similar. Certainly it's acceptable for the DCCP filtering behavior to be independent of TCP... "REQ-4: A NAT MUST NOT respond to an unsolicited inbound DCCP-Listen packet for at least 6 seconds after the packet is received." What does "not respond" mean? Not send the ICMP message? Not forward the DCCP-Listen packet? Both? Maybe this could be clearer if it's reworded in a positive vein -- "Upon receiving a DCCP-Listen packet a NAT MUST..." Section 5, REQ-5: Is this saying that the "established connection idle timeout" cannot be configured to be less than 2 hours 40 minutes? I see this language is stolen directly from behave-tcp, but I find it confusing. Section 6: The first paragraph seems inaccurate to me. It's actually much easier to modify the payload in DCCP than it is in TCP. As long as the modified packet remains within the PMTU, you don't have to do anything to the DCCP header except adjust the checksum -- you certainly don't need to modify the sequence numbers. You would only need to modify the sequence numbers if you added a packet. Of course changing the packet size *might* foul up the application, but the purpose of an ALG is to be application-aware, and so it would be highly unlikely to do something that would foul up the application. I agree with the intent here -- no ALGS -- but I don't think your argument justifies it. Maybe something more along the line of no DCCP applications currently need ALGs, so let's keep it that way. Section 10: Somehow I think an empty security considerations section isn't going to fly :-). I suppose you could start by referencing the behave-tcp security issues and describe what's different here. General: Your use of RFC 4787 terms isn't identical to RFC 4787. For example, you use "endpoint independent mapping" and RFC 4787 uses "endpoint-independent mapping" -- it makes it difficult to look up the meaning of the terms. > -----Original Message----- > From: dccp-bounces@xxxxxxxx [mailto:dccp-bounces@xxxxxxxx] On Behalf Of > Dan Wing > Sent: Thursday, August 07, 2008 9:24 PM > To: Behave WG > Cc: dccp@xxxxxxxx; rem@xxxxxxxxxxxx > Subject: WGLC: draft-ietf-behave-dccp-01 > > Behave is commencing a 2-week WGLC for "Network Address Translation (NAT) > Behavioral Requirements for DCCP", > <http://tools.ietf.org/html/draft-ietf-behave-dccp-01>. The WGLC will end > on > August 21. > > Please send comments to the author and to Behave. > > -d
