[WireGuard] [noise] [ANNOUNCE] WireGuard Launched!

Jason A. Donenfeld Jason at zx2c4.com
Wed Jun 29 03:45:02 CEST 2016


Hey Trevor!

On Wed, Jun 29, 2016 at 1:12 AM, Trevor Perrin <trevp at trevp.net> wrote:
>
> Very cool, that's a lot of good work, and nice docs.

Hooray! Glad you like it. Perhaps there's a place for it on the Noise homepage?

>
> Are you finding it easy(er) to justify HKDF, yet, instead of something
> more custom? :-)

Alas alas the seductive siren that is Mr Krawczyk... :P

>
>
> A few editorial things (https://www.wireguard.io/protocol/)

Did you see the paper by the way? It's a much bigger overview than the
/protocol/ page.
https://www.wireguard.io/papers/wireguard.pdf
Everyplace where I include Noise crypto or reference Noise, I cite
you. This time I got the URL correct :)

>
>  * You should link the Noise website (which has nicer PDF/HTML specs)
> instead of the Github source (the markdown was written for Pandoc
> instead of Github flavor, so has some glitches in Github view).

Sorry; this was left over from before the site existed. Fixed now.

>
>  * You could omit the first "initiator.key = HMAC" call in processing
> the initial message with PSK, as that key isn't used.
>
>  * You could omit the first two "responder.key = HMAC" calls in
> processing the response message, as that key isn't used.

Yea, that's true. But I wanted to keep the general flow consistent. It
helps reflect more clearly back to what's really happening within the
Noise-level mechanics.


More information about the WireGuard mailing list