Re: [Ntp] Antw: Re: Antw: [EXT] Timescales, leapseconds and smearing

[Martin Burnicki <[e-mail]>]

Warner Losh wrote:
>     Can you be more specific? I only see:
> 
>                struct tm {
>                    int tm_sec;    /* Seconds (0-60) */
>                    int tm_min;    /* Minutes (0-59) */
>                    int tm_hour;   /* Hours (0-23) */
>                    int tm_mday;   /* Day of the month (1-31) */
>                    int tm_mon;    /* Month (0-11) */
>                    int tm_year;   /* Year - 1900 */
>                    int tm_wday;   /* Day of the week (0-6, Sunday = 0) */
>                    int tm_yday;   /* Day in the year (0-365, 1 Jan = 0) */
>                    int tm_isdst;  /* Daylight saving time */
>                };
> 
> 
> time_t encodes 23:59:60 Dec 31 and 0:0:0 Jan 1 with the same number
> (when converted from struct tm, POSIX is actually silent on how to
> encode the leap second because it doesn't acknowledge the leap second
> exists). POSIX ignores leap seconds. This fiction of 60 here is woefully
> incomplete. struct tm isn't really used for anything but presentation...
> These issues have been extensively litigated, discussed and explained in
> the LEAPSECONDS list.

Right. If "someone" happens to put a "60" into the seconds field, the
user can see that this specifies an *inserted* leap second.

I wonder how that structure could be used to indicate a *deleted* leap
second, of which I hope it will never occur. ;-)

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