[Tinyos-help] Re: [Tinyos-devel] CTP + LPL
pal at cs.stanford.edu
Wed Feb 6 19:08:13 PST 2008
On Feb 6, 2008, at 2:10 PM, Matt Welsh wrote:
> While I like this idea, are we going through a TEP vetting process
> before adding new calls in the radio stack interface? I could think
> of a lot of things that one might want to add to the CC2420 radio
> stack but I thought the idea of the TEP process was to air these
> ideas through feature extension proposals before just adding things
> when they seem like a good idea. Then again maybe we need to draw a
> distinction between "public" interfaces and "private" ones; I
> suppose one could consider the internals of the CC2420 stack to be
> "private" -- that is until any upper level code starts relying on
> the existence of this interface and makes it impossible to port to
> other radio stacks.
Matt is correct; it would be problematic to add CC2420-specific event
handlers to multihop (HIL) protocols. One exception is MultihopLQI,
which is CC2420-specific.
If this event is going to become part of an HIL, then it should
probably make its way into an HIL interface, e.g., one of those
covered by TEP 105. It would be worthwhile to discuss -- in this
case, probably cursorily, as if it isn't, neither is LPL -- its
ability to be hardware independent.
More information about the Tinyos-help