On 06/05/2012 20:58, jouni korhonen wrote:



Alexey,



Hi Jouni,



Thanks for the review. See some initial comments inline.


On May 5, 2012, at 8:58 PM, Alexey Melnikov wrote:




I am the assigned Gen-ART reviewer for this draft. For background on
Gen-ART, please see the FAQ at
<

http://wiki.tools.ietf.org/area/gen/trac/wiki/GenArtfaq>.

Please resolve these comments along with any other Last Call comments
you may receive.

Document: draft-ietf-netext-access-network-option-10.txt
Reviewer: Alexey Melnikov
Review Date: 5 May 2012
IETF LC End Date: 9 May 2012
IESG Telechat date: Unknown

Summary: This draft is ready as a Proposed Standard

Major issues: none

Minor issues:

Minor: [ANI] and [TS23003] seem to be Normative (as per their use in 3.1.1)



If it is OK from the RFC process point of view, we can put these
non-IETF references into the normative section.


I think that is fine in general. You should double check with the 


responsible AD whether the references are stable (as otherwise they 


might not be suitable for referencing normatively).



Nits/editorials:

1.  Introduction

    This document defines a new mobility option, the Access Network
    Identifier (ANI) option and its sub-options for Proxy Mobile IPv6,
    that can be used by the mobile access gateway to signal the access
    network information to the local mobility anchor.  The specific
    details on how the local mobility anchor uses this information are
    out-of-scope for this document.  These mobility options are optional
    and are not mandatory for the Proxy Mobile IPv6 protocol.

Nit: Last sentence: "optional" and "not mandatory" are the same thing on my book.



Proposal for new text:

    "These mobility options are optional for the Proxy Mobile IPv6 protocol."



Perfect :-).



Strictly speaking PEN numbers are not limited to 4 bytes. However you have a registry for types of identifiers, so a new value can be allocated for bigger-than-4-bytes PENs.



Right. So we could just remove the 4 octet length requirement and use a
"natural" length indicated octet coding for the PENs. For example

ANI Length = 1 ->  PENs 0-255
ANI Length = 2 ->  PENs 0-65535
ANI Length = 3 ->  PENs 0-16777216
...

That would be ok?


That would work, as long as the extra complexity (which is marginal in 


this case) is Ok with the WG.