[Tinyos-devel] [Tinyos Core WG] TEP105: LowPowerListening DutyCycles

[Vlado Handziski]

On Mon, May 19, 2008 at 11:10 AM, Andreas Koepke <koepke at tkn.tu-berlin.de>
wrote:

>
>
> Philip Levis:
> >
> > On May 16, 2008, at 1:04 PM, Andreas Koepke wrote:
> >>>>>>>
> >>>
> >>> No. If there has been at least one packet that went out, cancel() will
> >>> return FAIL, since there is a non-zero probability of the message
> having
> >>> been delivered. But, as a side effect of this failed cancel, the mote
> >>> stops transmitting subsequent packets and signals a sendDone() with
> >>> SUCCESS.
> >>
> >> Which implies that an ACK has been received for  a unicast transmission.
> >
> > SUCCESS on sendDone() means the packet was transmitted successfully, not
> > that it was acknowledged.
> >
> > Phil
>
> I would not be so sure. The Data-LL traditionally solves the task of
> reliable transmission -- so SUCCESS does imply an ACK. At least for a
> "traditionally" educated network engineer.


The problem here is that the AM abstraction has traditionally been
understood as a "best-effort" or unreliable single-hop service. Thats why
the semantics of the result values are like they are. If we want to push a
traditional LL/MAC stack organization below AM, these assumptions would need
to be revisited, but this is the source of some of the confusion about the
LPL control also.

Vlado

Vlado
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.millennium.berkeley.edu/pipermail/tinyos-devel/attachments/20080519/179094f6/attachment-0001.htm