Re: SMF in Manet and MPR

That none of us even wrote an ID suggesting it is fairly indicative of that we all - independently I think - thought of this as a hack. Not even an experiment as IETF uses the word (though an experiment in the broader use of the word).

(An informational ID describing it could have been written, but clearly no one thought that worthwhile.)

> On 30 Jan 2024, at 09:43, Henning Rogge <[email protected]> wrote:
> 
> On Tue, Jan 30, 2024 at 9:51 AM Christopher Dearlove
> <[email protected]> wrote:
>> RFC 5444 is mandated for use on the MANET UDP port (and IP protocol) by RFC 5498 and repeated in RFC 8245. Any other use is optional, but there’s no good reason to use it on the data plane.
>> 
>> There was an (or possibly more than one) experimental - I don’t believe it even made it to an ID, but was implemented and reported in at least one paper - use of OLSR for an SMF-like approach to multicast, intercepting the data packets in the IP stack, encapsulating them as a new OLSR message type, flooded using OLSR, then converted back to IP data plane at destinations. This was rather a hack and the SMF approach was to supersede that, keeping data in the data plane where it belongs.
> 
> the OLSR (v1) implementation of olsr.org had a "BMF" plugin that did
> exactly that... encapsulating IP traffic into OLSR messages to forward
> them.
> 
> It wasn't a nice thing, good that we did not standardize this.
> 
> Henning Rogge

_______________________________________________
manet mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/manet