Re: [Ntp] CLOCK_TAI (was NTPv5: big picture)

[Martin Burnicki <[e-mail]>]

Kurt Roeckx wrote:
> On Wed, Jan 06, 2021 at 11:24:21AM +0100, Martin Burnicki wrote:
>>>
>>> Indeed. I am confident that the client can be very simple and provide
>>> the right time, even with leap seconds occurring.
>>
>> Of course. But the limitation is in the server. Right now, it is
>> sufficient to have a time source like DCF77. If you enforce using of
>> TAI, that's not sufficient anymore.
> 
> If you combine it with the leap second info, you can get TAI and
> UTC from DCF77.

Of course, I know.

My point is simply to avoid dependencies on other mechanisms if you only
need the current UTC or wallclock time.

This works perfectly fine with the current NTP version.

>> AFAIK, there's no current OS that doesn't support UTC.
> 
> I would argue that no OS really supports UTC. They support
> something like UNIX time, which is not the same as UTC since it
> can't represent the leap second.

Of course I know that, too.

However, for many applications this is sufficient, and for those with
higher level requirements it can easily be extended if the protocol
provides "UTC-like" time and a *consistent* TAI offset. See my example
in a previous message.

> If the OS has TAI support, you can calculate UTC from it if you have
> leap second information. But the OS or libraries don't support UTC,
> it's currently up to the application to make UTC from TAI.

Anyway, the current version of the protocol provides what these systems
need, and with the proposed way you can easily satisfy other requirements.

Martin
-- 
Martin Burnicki

Senior Software Engineer

MEINBERG Funkuhren GmbH & Co. KG
Email: [e-mail]
Phone: [phone]
Linkedin: https://www.linkedin.com/in/martinburnicki/

Lange Wand 9, 31812 Bad Pyrmont, Germany
Amtsgericht Hannover 17HRA 100322
Geschäftsführer/Managing Directors: Günter Meinberg, Werner Meinberg,
Andre Hartmann, Heiko Gerstung
Websites: https://www.meinberg.de  https://www.meinbergglobal.com
Training: https://www.meinberg.academy