[Tinyos Core WG] Meeting: 3/19

[Philip Levis]

On Mar 20, 2008, at 10:22 AM, Jan Hauer wrote:

> We missed the call (due to daylight-saving time), but then had a short
> internal discussion on TEP125: a 802.15.4 header can vary between 3
> and 23 byte in length (up to 37 with security), but in the current
> T-Frame/I-Frame CC2420 implementation we expect a 15.4 header with a
> specific length (9 byte) and a specific format. If we want to stick to
> static header sizes then the "802.15.4 Header" part of T-Frame/I-Frame
> should be exactly specified. The drawback is that we then lose general
> interoperability with 6lowpan networks. Alternatively if we allow
> dynamic 15.4 header sizes then, e.g. the cc2420_header_t would
> probably need to allocate the maximum size, which in turn means that
> message_t gets "bloated". Comments ?

How does TinyOS lose interoperability if it only sends and receives a  
specific 15.4 header format?

The I-Frame is intended so that TinyOS nodes can co-exist with  
6lowpan ones (that sense of interoperate). Clearly, a full 6lowpan  
stack would require more flexible header configuration than what the  
CC2420 and RF230 stacks currently support.

Phil