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

[C Chauvenet <[e-mail]>]

Hi,

The last part of my mail you didn't mentioned in your reply talked about the 6loWPAN work which has done some previous work on compression.
I think the 6LoWPAN WG and people involved into it have answers to the point you are exposing.
I'm feeling that doing some compression in RPL is not so far from doing compression like 6LoWPAN.
Your point seems to be related to the general compression mechanism, not especially RPL.

What do you think ?

ps : What about saying "hi" and "regards" in mails, and assume your opinion through a real name and affiliation ?

respectfully,

Cédric.

Le 3 sept. 2011 à 08:38, Ietf Roll a écrit :


Cedric wrote:

That's said, I don't see any problem for trying to design some improvements for RPL.
I think this rule is the same for every protocols.

[rav] Yes and RPL certainly needs improvements. :-)  And compressing out a bunch of unused bits would be a good improvement.

As already mentioned in this discussion, it's all about *optional* improvements.
Designing new mechanism like compression doesn't mean that you HAVE to implement them.
So I don't  see why nodes won't interoperate if they all rely on the same RPL RFC.

[rav] Yes you are correct that nodes can interoperate if they all adhere to the same RFC (or in RPLs case - lots of RFCs already - and more if you want to do P2P and more if you want to to do non-storing mode, but I digress).
         We are talking about a change to the core specification that will cause interoperability problems from the outset.  If I rely on the uncompressed spec RFC (oh wait, it isn't an RFC yet) I can't speak to you if you rely on this new compression
         scheme.  If we decide to go forward this this, then we have to have some way to negotiate between nodes to say if you are or are not using compression (oops another I-D). OR everyone has to implement both which goes against the notion
         that we are designing constrained nodes.

         If this compression is a good idea, then let's make it part of the core spec, get rid of the extra wasted bits and try to make RPL useful on constrained nodes.  This draft, if it is adopted by the WG, should deprecate the RPL core spec and fix it.

rav