That would make sense and save a byte in all cases but is this
compliant with RFC 8025?"

"
Values of the Dispatch byte defined in [RFC4944] are considered as
   belonging to the Page 0 parsing context, which is the default and
   does not need to be signaled explicitly at the beginning of a 6LoWPAN
   packet.  This ensures backward compatibility with existing
   implementations of 6LoWPAN.
"

On Wed, Jul 5, 2017 at 12:50 PM, Carsten Bormann <[email protected]> wrote:
> You indeed don’t need page 1 for IPHC.
> You need to be on page 1 once you need 6LoRH.
>
> (My proposal was to simply define 6TiSCH to start in page 1.  No idea whether 
> that removes any ever-so-remote compatibility with 6LoWPAN or if there is any 
> other reason to start off in page 0.)
>
> Grüße, Carsten
>
>
>> On Jul 5, 2017, at 12:02, Simon Duquennoy <[email protected]> wrote:
>>
>> In Thomas' example
>> https://tools.ietf.org/html/draft-munoz-6tisch-examples-00#section-3.6.1
>> there is a page dispatch to page 1 (0xf1) followed by IPHC (no routing
>> header). In this case, couldn't one choose to elide the page dispatch
>> and directly include IPHC? Or is the IPHC different from a page 0
>> IPHC?
>> Just checking if we're on the same page (no pun intended ;))
>>
>> Thanks,
>> Simon
>>
>>
>> On Wed, Jun 28, 2017 at 1:41 AM, Michael Richardson
>> <[email protected]> wrote:
>>>
>>> Thomas Watteyne <[email protected]> wrote:
>>>> Simon, all, FYI, we agreed on Friday that using paging dispatch is the
>>>> right way forward. I propose we continue discussing on the plugtests
>>>> ML if that's going to create problems.
>>>
>>> Yes, but the point is that a non-6loRH node will not be able to decode other
>>> than "page 1", and we have no signaling mechanism to tell a 6loRH node to
>>> "fall back".
>>>
>>> That was intentional... We discussed having a flag in the RPL DIO to say if
>>> there were old nodes present, but decided it wasn't worth it.
>>>
>>>
>>> --
>>> Michael Richardson <[email protected]>, Sandelman Software Works
>>> -= IPv6 IoT consulting =-
>>>
>>>
>>>
>>
>> _______________________________________________
>> 6tisch mailing list
>> [email protected]
>> https://www.ietf.org/mailman/listinfo/6tisch
>>
>

_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch

Reply via email to