Re: [babel] Genart last call review of draft-ietf-babel-rtt-extension-04

[Juliusz Chroboczek <[e-mail]>]

Thanks for your review, Roni.

> The document was changed in 02 from experimental to standard track. There is
> text in the end of section 4 about the algorithm being experimental

Right.

> from the mailing list i noticed that the reason to make it standard
> track is to allow using the TLVs by other WGs.

My understanding is that the goal is to allow other WGs to depend on the
protocol without incurring the complications of a downref.

Here's some background.  On the one hand, the algorithm has been
extensively tested in simulation, and deployed in production on hundreds of
routers, with no ill effects.  On the other hand, we don't fully understand
why it works -- in fact, I'm surprised it works as well as it does, I'd
expect at least some instability.

Hence, we're quite confident that this protocol is both useful and safe to
deploy on the Internet, which makes it a candidate for a Standards Track
document.  At the same time, we don't want to preclude experimentation
with other algorithms, possibly easier to analyse from a theoretical
standpoint, and hence we prefer to stress that the algorithm is experimental.

> Of course you could split the document to two documents one standard
> track and the other experimental but as i said this is up to the group
> and i have no real objection to publish the document as is.

I'd rather not split the document, since the first part would lack
rationale if we did.  If there are strong objections, we could conceivably
move the algorithm into an appendix, but I'd rather not do that, as I feel
that the document reads well as it stands.

Thanks again,

-- Juliusz