On 16/10/2025 13:38, Jeffrey Haas wrote:
Gorry,
On Oct 16, 2025, at 6:45 AM, Gorry Fairhurst <[email protected]>
wrote:
On 14/10/2025 06:47, Ketan Talaulikar wrote:
Hi Gorry,
I would like to follow up on your DISCUSS and comments on this
document. The authors have posted v35 of the document that clarifies
what is meant by "optimizing". Can you check if that is now
clarified and ok with you?
If not, then I would like to offer some suggestions (to you and the
authors) to consider:
Lightweight BFD Authentication
CSPRNG Based BFD Authentication
Alternating BFD Authentication
Alternating Between Regular and Lightweight BFD Authentication
Thanks,
Ketan
Thanks Ketan, I see the improved abstract, and I do think this is
helpful.
I am still unsure the present title is appropiate. I would still ask
if a change of title us possible. From my understanding, which is
limited to what I saw in the final document, I think any of these
would be good.
I'd suggest either: Lightweight BFD Authentication or Alternating BFD
Authentication
Really, let's not. Neither of these is particularly better.
For "lightweight", one would think the rather long security AD review
about the use of words like "strong" would tell us to try to avoid
"light". Light vs. what? It's not intended to be light - the
mechanism uses a pairing of a stronger mode and a less strong mode -
which is what made the security AD reviewers "happier". And, in the
future, a "stronger lightweight' mode is possible - or likely. If you
pair SHA-3 and CHACHA, is it still "light"? Probably not.
And this isn't even particularly alternating. The authentication mode
gets a sub-state with this optimization profile, not alternating
between between one auth type and another. (We did evaluate that as an
implementation choice and had to discard that option.)
Instead, I'd ask if the text discussing what optimization is being
referred to is clear in the document, particularly the abstract.
... but, I think others may understand better any sensititivies
surrounding the title, so I will take advice from you, let's discuss
briefly. I am now eager to close this DISCUSS and see the document
published!
Gorry
(WIT AD)
-- Jeff
Thanks for the clarifications
I still think the title could be better (and if, in future, I am
presented with a smilar title I expect to still query that wording).
However, for this document, I agree the latest abstract does explain the
title, thank you for the update.
I will now clear my DISCUSS, I wish good fortune in deploying this
protocol change!
Gorry
(WIT AD)
