On Mar 31, 2010, at 8:59 AM, Carsten Bormann wrote:
It seems that there is good enough support for 'ff:fe'
(ff:fe00 due to the way this is put in the middle of two 24-bit
fields.)
Of course the fully expanded version of the IID would be
0000:00ff:fe00:xxxx.
As discussed before, 6lowpan-hc can no longer simply point to
Section 6 of RFC 4944 since the PAN ID is no longer used.
As far as I can see, HC does not have to change 4944 at all.
HC with the results of this discussion simply is more efficient if
the 4944 conditional "if no PAN ID is known, 16 zero bits may be
used" is true.
From prior discussions on the list, the result of 802.15.4-2003 is
that the PAN ID is always known so the case where no PAN ID is known
is never true.
So I see no need for either:
I see two approaches:
1) 6lowpan-hc updates Section 6 of RFC 4944 to indicate how an IID
is generated from a 802.15.4 short address.
2) 6lowpan-hc becomes self contained by saying how to expand a
802.15.4 short address into an IID (note the slightly different
choice of wording). An update to RFC 4944 would be left for
another draft (presumably the ND draft).
Instead, I would say:
Note: This form of compression only addresses the case where instead
of a PAN ID, 16 zero bits are used according to section 6 of RFC
4944. This case appears to be the common implementation practice
for RFC 4944. If, however, a PAN ID is used, only the case SAM=01/
DAM=01 can be used to obtain some compression.
Unclear whether setting the PAN ID to zero is the common
implementation practice. This issue came about due to
interoperability concerns. The conclusion from that thread was that
the PAN ID is always known and so you should never assume zero for
those bits. Of course, that spawned additional concerns about the use
of PAN ID in the first place, leading to where we are today.
At this point, I would prefer to include a short section within
6lowpan-hc that simply says how to reform the IID from an IEEE
802.15.4 short or extended address to make the HC draft self contained
and decouple it from RFC 4944.
--
Jonathan Hui
_______________________________________________
6lowpan mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lowpan
