draft-ietf-magma-mld-source-05.txt

[Thomas Narten <narten@us.ibm.com>]

nit in security considerations (and I also agree with Russ):

>    The ability to send MLD messages with the unspecified address can
>    lead to on-link abuse that is harder to trace.  However, due to the
>    expected behavior of RFC 2710 implementations to drop these messages
>    and the additional rule that Query messages with the unspecified
>    address be dropped, the impact of such messages will not affect
>    multicast capabilities on the link.

Middle sentence could be better. E.g., per RFC 2710, implementations
are required to do X. Not "expected behavior of 2710 implementations",
which leads one to wonder if this is required, or that implementations
just happen to do the Right Thing by chance, which wouldn't be
sufficiently reassuring.

Thomas