Hi Dirk,
many thanks for carefully looking through the draft. Please see
comments
inline.
On 27.03.2014 16:30, [email protected] wrote:
Sorry that I missed the preceding WGLC - I do think that this
document
is ready for publication. It has greatly improved since version 00
;-)
Though some (minor) nits came to my mind after re-reading:
p.1.
Updates: 5568 (if approved) => shouldn't be added 5949 since it does
also update PFMIPv6?
I don't think so. The update of 5568 is with the PrRtAdv-Messages.
5949
does not contain such things, as there is no explicit messaging
between
MAGs and the MN. Mobility Options are explicitly under the control of
IANA.
As mentioned by others for prior versions there is still mixed usage
of FBack, Hack, ... and FBACK, HACK, ...
Same for PMAG/NMAG and pMAG/nMAG.
Oh yes, that was in the figures ...
p.10ff
"Section 3.3. Protocol Operations Specific to PFMIPv6" and Figs.
4/5
do include "PMAG (PAR)" and "NMAG (NAR)" - isn't it all about PMIP -
so no relevance for AR? Otherwise I would expect a statement like
that
also a mixed scenario MIP/PMIP is in focus here ...
I tried to find out whether this was explained in prior posts but
didn't catch any ... sorry if I missed it!
Actually the terms PAR and NAR in parenthesis are used to indicate
the
correspondence with FMIP ... it does not consider mixed terms, but
should help the reader to see that this is all about the same
"abstract
entities" here.
p.15
sect. 4.1.3 is on NAR, so I guess:
of the PAR => of the NAR
Yes, thanks.
the NAR joins the groups subscribed
for forwarding on the tunnel link. < sounds puzzling to me
=> the NAR joins the groups the MN has subscribed
for (which are then forwarded by PAR) via the tunnel link. <
is it
that what is meant?
Yes, thanks. This is better.
p.21
number of muticast records => number of multicast records
Thanks, fixed.
or Section 4.2 of [RFC3376]) for the => or Section 4.2 of [RFC3376]
for the
Thanks, fixed.
p.23
5.5. Length Considerations: Number of Records and Addresses
I understand why the maximum number of multicast address records
is 72
or sources for MLDv2 is 89 (also given in
http://tools.ietf.org/html/rfc3810#section-5.1.10), but I miss a
consideration of specific limitation due to 8-bit Length format in
new
Mobility Header Multicast Option (Fig.7).
Have I misunderstood something or isn't there a much stricter limit
for multicast address records to (512-2-4)/(4+16) < 26 (w/o source
addresses) ??
I guess you hit a point: By bringing back length formatting to
standard
mobility options recently, we missed that this will not fill an
Ethernet
packet. I don't think this matters much, but we definitely should
adjust
the section on length considerations.
for that multicast address to their MLDv2 (IGMPv2) equivalents
=> for that multicast address to their MLDv2 (IGMPv3) equivalents
Thanks, fixed.
Hope this helps
Yes, it definitely does.
We will wait for the next days to pass the call deadline and resubmit
then.
