I have to admit that sometimes I feel sorry for CNN. Sometimes. I was looking at the headlines and I saw a story specuguessing about the next Leap second adjustment which is scheduled for the end of June 30, 2015 at 23:59:60 UTC. They equated it to "The next Y2K Bug" in their story, which is unusually accurate for CNN given what happened at 00:00:00UTC on 1 Jan, 2000. I tried to find the story, but to my bemusement I couldn't. There were others though..... June's leap second will be a kind of deliberately-induced Millennium Bug No, the Linux leap second bug WON'T crash the web • The Register Adding Leap Seconds is Far More Dangerous For The Internet I remember when I was punching holes in the Atlantic in the 80's and 90's I had to enter leap second adjustments at least twice. They were a PITA, but the procedure was pretty straight forward. I remember the last one specifically because our Weapons Officer was extremely concerned about whether or not I had entered the extra second correctly. (I did) The thing is......it didn't matter then, and it doesn't matter now. Submitted for your approval.........
I think The Register has it right: "If this all smacks of a manufactured, Daily Mail style panic of the sky-is-falling proportions to send you running for central Montana with a high-power telescopic rifle and a large tin of baked beans, then you’re right – it is."
I remember listening to the leap second on time station WWV several times back in the 20th Century, when these events happened almost every year. But after 16 years with only 3 of these added seconds, too many programmers have become unaware of them, and have built their systems around the wrong clock (UTC), even though more appropriate clocks (TAI and GPS) have long been available.
More complex than precession, but not as esoteric as frame shifting. Tides naturally show down earth's rotation, so a natural second now is longer than it was eons ago. The scientific definition of a second was fixed to an atomic standard a half century ago, but the earth continues to slow down. On top of that, there is a lot of other causes of variation in the earth's spin rate -- atmosphere, winds, currents in the molten rock core, shifts in the crust (earthquakes), even human-made water reservoirs that shift enough mass to alter the planet's rotational moment of inertia. While the changes are far far below natural human perception, differences become blatantly obvious when scientists start using atomic clocks for astronomy, surveying, GPS, and other things.
if we are calibrating leap seconds what's on first? solar? atomic? GPS? I hope there are ntp mirrored backups somewhere...
The underlying atomic time standards don't have leap seconds. Only the solar-oriented derivatives have leap seconds. For more detail, read ETC(SS)'s very first linked article, Leap second. NTP is part of the problem. PTP (Precision Time Protocol) is more oriented towards using the atomic standards, but I don't believe is used in actually keeping the many atomic clock standards linked.
True, but it is a little easier to handle than leap months (Mercedonius). The real use of the leap months was to keep you friends in office longer or you enemies reign shorter. I think that is the real plot behind leap seconds.
To me, abolishing the leap second seems equivalent to abolishing the Gregorian and Julian calendars because they are too hard to deal with, and returning to the Roman calendar because it is simpler.
