Re: [6tisch] #41 (minimal): intended status for draft minimal (was: internded status for draft minimal)

[Tero Kivinen <[e-mail]>]

Xavier Vilajosana writes:
> Given that context can you please provide solid and objective arguments on why
> the specification of a the profile of IPv6 on BTLE is an standard and what we
> define in minimal (the 15.4e profile) it is not? 

Does minimal change anything in 802.15.4? I.e. is there something
there that is not already defined in 802.15.4 when suitable options of
the 802.15.4 is selected?

I.e. can someone take 802.15.4 and implement it, picking parameters
until he get the same parameters specified in the minimal and then
implement minimal without ever reading it.

My understanding is that yes, 802.15.4 complient implementation which
implements just what is already defined in the 802.15.4 can also be
complient with minimal, if it selects suitable options.

This would make it profile instead of new protocol.

When you start to implement things on top of that, for example the key
management or joining processes, then those parts are outside the
scope of 802.15.4 and you have to implement something not mentioned
there, then you are making new protocol.

On the other hand I do not think there is real difference between
those two, i.e. it does not matter whether it is profile or whether it
is new protocol for the IETF process point of view.

Currently the problem is that minimal is not clear what it tries to
be. The actual content of it is more like profile, i.e. it is just
telling which options of 802.15.4 to enable, but some of the text is
written in a way which would say it is new protocol as it copies
things from the 802.15.4.

It is not clear for me what should implementor do if he spots mismatch
between minimal and 802.15.4? Which one of them is authorative? If
this is profile, then 802.15.4 should be followed unless the mismatch
was explicitly spelled out. If this is new standard then this document
should be followed.

Also what does it mean when we have text saying for example "source
and destination address fields MUST be filled with an extended address
(64 bit)"? Does that mean that no minimal node can ever use any other
addressing formats?

I would assume it is supposing to mean that when using minimal profile
for the packets, we use those extended addresses, but anything that is
valid in the 802.15.4 network can also be used for other traffic in
the same network. I.e. this minimal just gives minimal
interoperability profile for the network so devices can use those
settings until they get something better.
-- 
[e-mail]