Re: [Ntp] NTPv5 draft

[Warner Losh <[e-mail]>]

On Mon, Dec 7, 2020 at 9:38 AM Salz, Rich <rsalz=[e-mail]>
wrote:

> My view is that it is no longer acceptable to design a protocol for
> deployment on the open Internet that has no authentication or message
> integrity and that people who disagree are out of consensus.
>

Time is harder, though, because most 'easy' authentication like https has a
time-element. If you don't yet know what time it is, it's hard to trust the
answers you get back if certificates are valid for only a given time and
some expired certificates have been compromised... But who doesn't like a
challenge... :)

BTW, I don't disagree. I'd love for there to be a good, authenticated time
infrastructure that included robust distribution of things like full leap
second history (and future) as a first class citizen, not a bolt-on like it
is today... But much of that info need not be at the lowest-level time
authentication bootstrapping since there's overlap with other areas and
services (the tz database, for example).

Warner