DNS IPv6 Transport Operational Guidelines
draft-ietf-dnsop-ipv6-transport-guidelines-02

[Ballot comment]
Reviewed by John Loughney, Gen-ART

Nits:

1) RFC 2026 boilerplate - needs updating
2) No headers / footers / page numbers / page breaks, etc.

[Ballot comment]
>    transport. The concern is that a resolver using only a particular
>    version of IP, querying information about another node using the same
>    version of IP can not do it because, somewhere in the chain of
>    servers accessed during the resolution process, one or more of them
>    will only be accessible with the other version of IP.

hard to read. suggested reword: 

  The concern is that a resolver using only a particular version of
  IP will be unable to resolve a name because somewhere in the chain
  of servers accessed during the resolution process, one or more of
  them will only be accessible only with the other version of IP.


>    through a "translator", i.e. they have to use a recursive name server
>    on a so-called "dual stack" host as a "forwarder" since they cannot
>    access the DNS data directly.

Is "forwarder" the best name for this? Sounds to me like what a
"proxy" does. Any reason not to call it that?

>    With all DNS data only available over IPv6 transport everything would
>    be equally simple, with the exception of IPv4 recursive name servers
>    having to switch to a forwarding configuration.

What does this mean? What is there to "switch to"? Recursive servers
just do this.

In the normative references section, I have a hard time believing that
most (if indeed any) of the references are actually normative. Not
that it matters a whole lot though.