Re: [manet] DYMO and other routing protocols

[Alexandru Petrescu <alexandru.petrescu at gmail.com>]

Charles E. Perkins wrote:
> Hello Alexandru,
>
> ext Alexandru Petrescu wrote:
> > I think the classification on-demand vs proactive is conceptual 
> > only, helping frame one's mind around two ways of obtaining routes.
>
> Well, leaving aside the notion that all work in the IETF is 
> conceptual only, I am not sure just what you mean.  In my view, 
> on-demand routing is one of the main contributions of [manet] to 
> routing technology.

Yes, that notion not purposefully stressed, I mean that it does not make
sense to push two different protocols just because the conceptual 
distinction makes sense.

On-demand routing (create routing data at most or all intermediary nodes
onyl when end nodes need to communicate) seems to me indeed to be a main
contribution of IETF manet to routing technology.  It is different than
other IETF routing in that it can accommodate potentially very variable
topologies.

> > That classification can't really be used to pit a protocol vs 
> > another.
>
> It has proved to be a very useful conceptual tool, helping a lot of 
> people organize their thoughts about the way protocols work.

I think yes.

> > Similary, DYMO says it's on-demand but it should advertise RREP (or
> >  RREQs) periodically if it needs to maintain up to date information
> >  at all times.  Making it reactive.
>
> I know you meant "proactive".  This is not in the protocol 
> specification, but for some particular cases I quite agree. A better 
> example would be "HELLO" messaging.

I think we talk here about a proactive need.  I take your mentioning of
HELLO just an example of semantics.  I see very well either OLSR's
HELLO, or periodic RREP, or ND's RAs used for proactive needs.

> Generally speaking, just because two classifications are not 
> mathematically contradictory, does not make them useless -- right??

Right.  They are useful so we can characterize behaviours.  But this
classification shouldn't entrench one into thinking a certain protocol
can't be both proactive and reactive (some times so other times so).

> We use blue and green to describe things as needed, and no one 
> complains based on the existence of turquoise or aqua.  If you would 
> rephrase your point to instead say that we should avoid being 
> constrained by the way we have classified things in the past, I would
>  very much agree with you.

I agree, I will rephrase my point to say a protocol can be acqua some
times blue other times green.

Alex



______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email 
______________________________________________________________________


_______________________________________________
manet mailing list
manet at ietf.org
https://www1.ietf.org/mailman/listinfo/manet