Re: [Roll] Adoption of draft-goyal-roll-rpl-compression as a new ROLL WG document

["Angelo P. Castellani" <[e-mail]>]

On Mon, Sep 5, 2011 at 16:01, Pascal Thubert (pthubert)
<[e-mail]> wrote:
> Hello Angelo
>
>> - do we really need two different RPL dialects to exist?
>
> I, for one, do not. I think that once decompression has occurred, we
> should get RPL packets as defined in the spec.
>
> If more than one compression method exist over a same MAC in a same
> network, or if nodes use compression and others do not, then we are in a
> situation that is equivalent to different MACs over the same band, or
> different cryptos. DIOs have to be sent twice, etc.
>
> So people will have to be careful as to when and how they deploy new
> stuff. It is probably preferable to limit the number of gateways that
> have to repeat their multicast DIOs over multiple networks. But that's a
> deployment issue.

Why shouldn't we avoid this complexity defining, using a set of
guidelines (e.g. if MAX_FRAME_SIZE<X use compressed-format-N), which
format MUST be used on any L2?

A.