Re: [Curdle] sntrup761x25519-sha512

[Ron Frederick <ronf%timeheart.net@localhost>]

On May 16, 2023, at 10:10 AM, Mouse <mouse%Rodents-Montreal.ORG@localhost> wrote:
> 
>>> (Work on how to use those curves can start earlier, but publication
>>> should wait until algorithms are official.
>> Can't do interop experiments without putting entries in the IANA
>> registries [...]
> 
> Hmm?  I can't see why not.  I could (putatively) implement something
> today and call it FredBloggsKex; if (say) PuTTY does likewise, we can
> interop test with one another even if we never even so much as tell
> anyone about it, much less convince the IETF to register the name.
> 
> Can't do *open to the world* interop testing, maybe.  But even then,
> you can do _almost_ all interop testing with a name@FQDN extension name
> rather than the final name.  _Most_ interop problems won't care what
> name you advertise it under, after all.

Agreed - I see no reason not to do interoperability testing with the qualified version of the name. I’ve been doing interop testing between AsyncSSH and OpenSSH for the sntrup761x25519-sha512%openssh.com@localhost key exchange since adding support in AsyncSSH in August of 2022, and I see from https://ssh-comparison.quendi.de/comparison/kex.html that SmartFTP and TinySSH also provide support for this already.
-- 
Ron Frederick
ronf%timeheart.net@localhost