[time-nuts] Leap second? Yay or nay?

[Poul-Henning Kamp]

In message <6A49BBA9110943DEA397D3F7929F32E3 at garadm>, "Jean-Louis Noel" writes:
>Hi,
>
>From: "Poul-Henning Kamp" <phk at phk.freebsd.dk>
>
>> Testing software for correct handling of leap-seconds is a major undertaking
>> which very few people have the kit and skill to do.
>
>A software that crashes/behaves badly when the time is updated is
>a bad software.

"A timescale that is not predictable is a bad timescale."

Yes, platitudes like that make for good sound-bites, but it doesn't
bring us any closer to a solution.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.