I have finally gone through the latest revision of the ds-lite draft and
overall I think it looks good. I only have two small comments/questions:

Section 4.2 and 4.4 defines an addressing scheme for the ds-lite tunnel and
also talks about the addressing of the home network but I have a problem
understanding the reasoning behind it and the conclusion. Why does the
tunnel addresses have to be well defined while the home network doesn't have
to, or does it? Since the NAT state always will be based on the IPv6 source
address the IPv4 addresses shouldn't really matter as long as it is a 1918
address. 

Section 6 seems very ambiguous and more about saying that in more complex
cases you might want to consider a different solution rather than defining
encapsulation methods. It needs more explanation and has to be a bit clearer
on what actually is expected when supporting other types encapsulation.
Right now I don't see what the benefit there would be in having multiple
encapsulation methods. 

A small nit is that the terminology changes throughout the document, which I
guess is due to the merger with the snat draft. This makes it look somewhat
like two different solutions. It is mainly chapter 5 that need to be changed
to be more in line with the rest of the document. 

-Mikael

_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires

Reply via email to