[time-nuts] leapseconds, converting between GPS time (week, second) and UTC

[Martin Burnicki]

Fiorenzo Cattaneo wrote:
> Yes correct. NTP behaves as intended. I called NTP issue as that's
> what commonly called in the industry, ...

Agreed, but that's simply wrong. If you ran a PTP daemon instead of an
NTP daemon you had exactly the same problems when the PTP daemon passes
a leap second announcement to the kernel.

> ... but in reality it's applications
> which are unable to deal with time steppings.
>
> The LEAP second smearing that Google (and others too, now AWS as well)
> is as you say, just a hack to avoid application problems.
> 
> I am starting to realize I need to very precise (no pun intended) when
> writing on this mailing list.

;-)

> Please accept my apologies, this is my first thread I contribute to.

No need to apologize, IMO. I also post only rarely on this list because
most other folks here are much more familiar with the kind of stuff
discussed here. Anyway, very interesting reading. ;-)

Martin