On 2026-07-17 04:51 PM, Arthur Olson via tz wrote:
"Earth Rotation Records Spur October Vote to Avert Negative Leap Second"
https://www.techtimes.com/articles/320185/20260711/earth-rotation-records-spur-october-vote-avert-negative-leap-second.htm
Per the article, "Nobody has built software to survive that."
@dashdashado
This is an interesting and reasonably well researched article.
However, there is a badly misstated paragraph:
"A negative leap second would do the opposite: instead of the clock
pausing, it would skip forward. The sequence would run 23:59:57 →
00:00:00, with 23:59:58 and 23:59:59 simply not existing. The last
minute of the affected UTC day would contain only 59 seconds."
It says: " The sequence would run23:59:57 → 00:00:00, with 23:59:58 and
23:59:59 simply not existing" No. No! This is exactly incorrect! The
single label, 23:59:59 is omitted. The sequence would be:
YMD 23:59:57
YMD 23:59:58
YMD 00:00:00
YMD 00:00:01
It then says "The last minute of the affected UTC day would contain only
59 seconds.", that is, the day duration would be 86399s, which is
correct, but in contradiction to the previous sentence.
By the way, less important perhaps, it says: "A negative leap second
would do the opposite: instead of the clock pausing, it would skip
forward.". The clock does not 'pause' on a positive leap-second. The
proper UTC label, YMD 23:50:60 accounts for that. However, the NTP label
does 'pause', or 'freeze'. An NTP disciplined system would show a
sequence of:
YMD 23:59:58
YMD 23:59:59
YMD 23:59:59
YMD 00:00:00
YMD 00:00:01
A lot of the rest of the report is more or less good. Indeed,
preparation for a possible negative leap-second has not been accounted
for in most timekeeping applications as far as I can tell. Fear of
Negative Leap-Second (FOMLS) is a legitimate concern.
Speculation of when a negative leap-second might be required is entirely
guess-work. Many people like to try.. There was a "stall" in positive
leap-seconds between 1999 and 2006 when many of these same topics were
discussed. But it then reversed and continued positive leap-seconds.
"Prediction is difficult, especially about the future".
In any event I think it's not nearly as disruptive or dangerous as the
Sunshine Protection Act.
-Brooks
