Mobile IPv6 Fast Handovers for 802.11 Networks
draft-ietf-mipshop-80211fh-04

[Ballot discuss]
In my view, deployment of IAPP could be harmful.  I strongly
  encourage the removal of all references to IAPP.  Task Group r
  (which is working on 802.11r) is working on better layer 2 support
  for fast roaming.  Further, IAPP is not really needed to make the
  central points.  I suggest the following changes:

    (a) Remove the discussion of IAPP in section 2.

    (b) In section 4, remove step 6.  Likewise, please delete the
        reference to "step 6" elsewhere.  I found two references in
        the security considerations.

    (c) Remove reference [8].

  In section 4, the document says:
  >
  > Note that in some existing 802.11 implementations, steps 1-4 are
  > performed by firmware in rapid succession (note that even in these
  > implementations step 3 is sometimes performed in a host driver,
  > especially for newer WPA implementations). ...
  >
  Earlier, there is a discussion of 802.11i.  Here, WPA is discussed.
  The relationship of the two is going to be lost on most readers.  I
  think it is best to refer to IEEE 802 standards, and avoid the WiFi
  Alliance marketing terms.

[Ballot discuss]
In my view, deployment of IAPP could be harmful.  I strongly
  encourage the removal of all references to IAPP.  Task Group r
  (which is working on 802.11r) is working on better layer 2 support
  for fast roaming.  Further, IAPP is not really needed to make the
  central points.  I suggest the following changes:

    (a) Remove the discussion of IAPP in section 2.

    (b) In section 4, remove step 6.  Likewise, please delete the
        reference to "step 6" elsewhere.  I found two references in
        the security considerations.

    (c) Remove reference [8].

  In section 4, the document says:
  >
  > Note that in some existing 802.11 implementations, steps 1-4 are
  > performed by firmware in rapid succession (note that even in these
  > implementations step 3 is sometimes performed in a host driver,
  > especially for newer WPA implementations). ...
  >
  Earlier, there is a discussion of 802.11i.  Here, WAP is discussed.
  The relationship of the two is going to be lost on most readers.  I
  think it is best to refer to IEEE 802 standards, and avoid the WiFi
  Alliance marketing terms.

[Ballot comment]
>    otherwise be possible.  By tunneling data between the old and new
>    access routers, it is possible to provide IP connectivity in advance
>    of actual Mobile IP registration with the home agent or correspondent
>    node.  This removes such Mobile IP registration, which may require
>    time-consuming Internet round-trips, from the critical path before
>    real-time service is re-established.
   

Could be made more clear. Last sentence in particular is hard to
follow. (remember, orient this towards general reader).

>    The RtSolPr message is used to request information about the
>    router(s) connected to one or more APs.  The APs are specified by
>    link layer address in the RtSolPr and associated IP-layer information
>    is returned in the PrRtAdv.  In the case of an 802.11 link, the link
>    layer address is the BSSID of some AP.

Might be good here (and elsewhere in this section) to be more clear
about which fields are used to carry the info. E.g.,

      The APs are specified in the New Access Point Link-Layer Address
      option [rfcxx ] in the RtSolPr ...

and in paragraph 2:


>    This list would supply the necessary BSSIDs to fill into the
>    RtSolPr messages.

Actually, what field is used for this?

>    from Section 4 and the lettered events are from Section 5.  For
>    example, "1abcde234567fg" is the sequence where the MN performs a

where do these gobbledigook "1abcde234567fg" strings come from and why
do you use them?

Oh, I see. Another sentence or two explanation would be good. E.g.,
add something like:

For example, "1" denotes step 1, "abcde" denotes steps a-e), etc.

>    changing radio conditions may cause nAR to be suddenly unreachable. 

s/nAR/NAR/?