Comments in line. On 3/21/06, Carsten Bormann <[EMAIL PROTECTED]> wrote: > Hi Ian, > > > I do not think the MANET protocols should be re-engineered in 6lowpan. > > I believe that a L2 implementation of minimal DYMO can fulfil 6lowpans > > requirements. > > > > I think packet optimizations are acceptable to explore, but protocol > > behavior should not be changed. > > > > I think that the charter should state exactly what type of > > "optimizations" are to be explored. > > Good point. "Optimization" is highly ambiguous. > No, we don't want to create another protocol; in other environments, > what we are trying to do would have been called "profiling", i.e., > defining a subset that is interoperable in the specific domain. > > >> Produce "Optimization of the Neighbor Discovery Protocol for > >> 6lowpans" > >> to define how to apply the existing Neighbor Discovery protocol in a > >> 6lowpan. > > > > Again, I think we should reuse IPv6 ND or use a MANET hello protocol > > (OLSRv2 or SMF neighborhood discovery seem suitable). Again, packet > > formats could be optimized, but the protocol behavior should not. > > Here I'm not so sure. I wouldn't want to rule out new protocol > development in this space. > What makes you confident that either ND as is or the MANET protocols > will fit the bill?
I am not an expert in ND or 6lowpan, but my intuition says that an existing algorithm can be harnessed to accomplish ND or neighborhood discovery in 6lowpan with little modification, except for the packet format of course. Ian Chakeres _______________________________________________ 6lowpan mailing list [email protected] https://www1.ietf.org/mailman/listinfo/6lowpan
