Re: [babel] Intdir telechat review of draft-ietf-babel-rtt-extension-05

[Antoine FRESSANCOURT <[e-mail]>]

Hello Juliusz, 

Your answer gives me the opportunity to give a minor remark I forgot to mention in my review: In the acknowledgements, I think you want to thank Jean-Paul Smets (instead of Smetz). 

Regarding my remark on the algorithms, given that the description of said algorithms in the references is 2 or 3 lines of text, I was thinking giving this text directly is not harmful to the whole document. But to be honest, I am quite nit-picking here. If you don't agree I can live with the current text given that I think the draft is already really well written. 

Best regards,

Antoine

-----Original Message-----
From: Juliusz Chroboczek <[e-mail]> 
Sent: mardi 13 février 2024 00:19
To: Antoine Fressancourt <[e-mail]>
Cc: [e-mail]; [e-mail]; [e-mail]; [e-mail]
Subject: Re: Intdir telechat review of draft-ietf-babel-rtt-extension-05

Thanks, Antoine.

> * In some parts of the draft, the authors personalize their text using 
> "we" or "our" (for instance in section 3.3 or in the introduction of 
> section 4). The authors might want to rephrase those sentences using a more neutral form.

Agreed.

> * In section 4.1, I was frustrated by the use of vague terms like 
> "fairly" or "somewhat" which gives the impression that the phenomenon 
> described is difficult to observe, and may not require the use of a 
> proper smoothing algorithm.

I'll try to find some more serious-looking terms.

> * In section 4.1 and 4.3, the text mentions external references to 
> describe algorithms that the implementer could be using to avoid 
> strong oscillations in the route selection algorithm. I would have 
> appreciated to have a rough intuition of the behavior of those 
> algorithms in the text directly, while keeping the reference to point to more detailed descriptions.

I'm not sure how to do that.  Both "exponential average" and "hysteresis algorithm" are familiar terms, I'm not sure how I can make them more accurate without spelling out the algorithms in detail.

Thanks again,

-- Juliusz