Forwarded by permission of the author.
Juliusz
--- Begin Message ---
On Dienstag 23 Dezember 2008 11:53:49 you wrote:
> Hi Henning,
>
> Henning Rogge <[email protected]> writes:
> > Did you ever looked at the PacketBB, TimeTLV and NHDP
>
> Yes, I have carefully read all three drafts. I have also carefully looked
> at AODV/DYMO and at DSR.
>
> PacketBB has a lot of good ideas, but it is too complex. In Babel 2,
> I took what I believe are the good ideas, and avoided what I believe is
> useless complexity. I gratefully acknowledge the influence of PacketBB,
> thanks to which a route announcement is on average 15 bytes in Babel 2 (as
> opposed to over 30 in Babel 1).
Maybe you should have a lock into the recent drafts... one of the "you still
have work to do" arguments of the IETF review team was that they had to reduce
the complexity... I think they have done a good work in draft 17 (which will
become the final RFC). They got rid of most special fields (which can just be
expressed as additional TLVs) and the strange "ordered" flag.
> I believe that Babel's neighbour discovery mechanism is smarter than the
> one in NHDP. If you disagree, please let me know.
In which way ? Maybe you can give me some pointers to look into the babel
documentation.
> As to TimeTLV, I don't understand what it's trying to solve. Babel 2
> represents times as simply a 16-bit number of centiseconds. This gives
> a range of 10 minutes with a precision of 10ms, and I fail to see why the
> complexity in TimeTLV would be needed.
PacketBB was made for all kind of mesh networks, including sensor networks.
And sensor networks might have timer values of minutes to hours. So they
designed a 8-bit transport format for small and long time values.
In addition to this TimeTLV solves the problem if different timeouts depending
on the hopcount (very important for fisheye implementations).
> Finally, my opinion of OLSRv2 is that it solves none of the fundamental
> problems with OLSR.
I think we can agree to disagree here... ;)
> All of this is my understanding. I'm very much open to new ideas, and
> I have no problem with making incompatible changes to Babel, so please let
> me know if there's anything I've missed.
I just wanted to start some idea exchange...
signature.asc
Description: This is a digitally signed message part.
--- End Message ---
_______________________________________________
Babel-users mailing list
[email protected]
http://lists.alioth.debian.org/mailman/listinfo/babel-users
