> I'll note that while a theoretical explanation/study sounds interesting, > it's not required for publication of an IETF document.
In my opinion, without a theoretical understanding, it is difficult to provide adequate guidance to implementers. We have a deep theoretical understanding of what makes Babel work. This is why we can say with confidence that implementations MUST evaluate the feasibility condition and that the metric MUST be stricly monotonic, but we can safely leave the details of the route selection algorithm to the implementation. Not so with Babel-RTT. We've got an implementation that has been shown to work, but we have no precise ide which parts of the implementation are important and which are coincidental. In this situation, I don't feel confident to write up a protocol description. -- Juliusz _______________________________________________ Babel-users mailing list [email protected] https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/babel-users
