thanks, comment below...

--- Ki-Hyung Kim <[EMAIL PROTECTED]> wrote:
> > Q2: Did you use the handling rules as per the draft, or did you modify the
> > link-layer source address at every hop, as implied by your post a while
> back?
> I didn't change anything of the format document on the simulation. Also the
> link-layer source address is changed at every hop as everybody expected.

As I explained before, this departs from the handling rules in the format draft.
According to those rules (I won't send them again), the source address does not
change hop-by-hop. The assumption is that if an intermediate node wants to send 
an 
RERR to its previous hop, it would do a reverse forwarding decision. To be 
clear,
if multiple paths were kept, presumably an intermediat node would have more 
than one
reverse path to the originator of a message. In this case, it is not clear to 
which of
the possible predecessors should the RERR be sent. Perhaps to all of them, 
which costs
a bit more.

You say that the "link-layer source address is changed at every hop as everybody
expected".

One conclusion is that given that this contradicts the current handling rules, 
then it is

clear that they need to be clarified by explicitly mentioning that no such 
thing is done.

Another possible conclusion of your "as everybody expected" is that between 
these
choices:

a) link-layer source address is changed at every hop
b) link-layer source address is not changed at every hop

the draft has obvioiusly made the wrong choice in going with "b". You have 
verified that
having only one link-layer source address (the one in the 802.15.4 header)
works under handling rule "a". Have you also evaluated under handling rule "b" 
in order
to compare
both? Or is "b" obviously wrong for some reason?

-gabriel

_______________________________________________
6lowpan mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/6lowpan

Reply via email to