While preparing for the IETF meeting, I re-read CCID-4 and have some comments on the current revision, which I would like the authors to consider. Best wishes, Gorry Comments: ---- Page 8 first para. Is it sufficient to assume the header size is 36 bytes, given the large space that could in future be allocated to DCCP options? - I think the text should mention options processing. - There are also options in IPv4 (and IPsec), and the important case of IPv6 which may need to be mentioned. - My suggestion is that at least these issues should be highlighted, and an IPv6 case given explicity. ---- Is the CCID-3 Loss Intervals option defined for other CCIDs beyond 3? - A quick look at the Registry suggests that some options that are desirable for CCID-4, were defined only in CCID-3 specific registries. - Most of the IANA considerations also applies to this new CCID. - Do you wish to redefine these here also for CCID-4? ---- Editorial work: --- Abstract: /This document contains the profile/ - Could this be changed to: /This document specifies an experimental profile/ --- Abstract: It's good to say also that this experimental, but I think this appears more than once at the moment, without actually explaining WHAT experiment the community needs to do to verify that is safe. --- Page 7: Point 2. - First sentence seems hard to parse. --- Page 8: Point 5. - I suggest that "[RFC828] MUST be considered as normative" --- References: I expected CCID-3 to be normative to this document. I also expected CCID-3-Drops to be normative to this document. - Do others agree? ---
