[Make-wifi-fast] Fwd: [babel] Re: RFC 9616 on Delay-Based Metric Extension for the Babel Routing Protocol

[Dave Taht]

---------- Forwarded message ---------
From: Juliusz Chroboczek <jch at irif.fr>
Date: Tue, Sep 17, 2024 at 5:47 AM
Subject: [babel] Re: RFC 9616 on Delay-Based Metric Extension for the Babel
Routing Protocol
To: <babel at ietf.org>
Cc: <manet at ietf.org>, Baptiste Jonglez <baptiste at bitsofnetworks.org>


> A new Request for Comments is now available in online RFC libraries.
>
>
>         RFC 9616
>
>         Title:      Delay-Based Metric Extension for the
>                     Babel Routing Protocol
>         Author:     B. Jonglez,
>                     J. Chroboczek
>         Status:     Standards Track

For those who weren't following: the reviewers forced me to rework this
document quite a bit, and with hindsight they were right.

The most notable change is that the second part (oscillation avoidance)
has been rewritten in normative language.  This makes it slightly more
difficult to read (due to all the SHOULDs and MUSTs), but it makes it
clear that the algorithm is ready and safe for deployment.  I've done my
best to imply that it's not the only possible algorithm, and that the
first part (the packet format and RTT sampling) can be reused with other
algorithms.

Pascal Thubert (to whom thanks) coerced me into writing an applicability
section, with me screaming and kicking.  With hindsight, that was a good
idea.  (While I remain unconvinced by pure applicability RFCs, I now
understand that having an applicability section in an RFC in an otherwise
technical RFC is useful even when the applicability can easily be deduced
from the rest of the document.)

The one negative change is due to one AD insisting (in oh so many words!)
that I'm not allowed to use the word "blackhole".  I suppose that
gravitational singularities are taboo in their culture.

In case you're not a Babel fan (why?), this subprotocol is not specific to
Babel, and could easily be adapted to any other distance-vector protocol,
and with slightly more effort to a link-state protocol.  I'm open to
collaboration.

-- Juliusz

_______________________________________________
babel mailing list -- babel at ietf.org
To unsubscribe send an email to babel-leave at ietf.org


-- 
Dave Täht CSO, LibreQos
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.bufferbloat.net/pipermail/make-wifi-fast/attachments/20240917/81813792/attachment.html>