Re: [Ntp] Antw: [EXT] Re: New Version Notification for draft‑gruessing‑ntp‑ntpv5‑requirements‑03.txt

[Miroslav Lichvar <[e-mail]>]

On Tue, Oct 19, 2021 at 03:59:14AM -0700, Hal Murray wrote:
> 
> [e-mail] said:
> > One of the requirements for NTPv5 should be that it can be properly
> > implemented on current versions of all major systems, i.e. TAI cannot be the
> > only supported timescale. 
> 
> Depends on what you mean by "supported".
> 
> I think we should split things into two parts.
> 
> The time used in time-transfer should be TAI to keep things clean.  No leap.  
> No smear.

I'm sure that would look great on paper, but how exactly do you
implement it? You cannot convert POSIX time_t to TAI just from the
timestamp alone. When you receive a timestamp from the kernel (e.g. in
the SO_TIMESTAMP control message), you don't know if it was captured
before or after the leap.

The other important issue is that the TAI-UTC offset is not always
known.

NTP is a simple protocol for exchanging timestamps. It shouldn't
enforce what kind of timestamps are exchanged.

It would be nice if all ideas proposed for NTPv5 had to be provided
with a proof-of-concept implementation. We would have it ready in no
time.

-- 
Miroslav Lichvar