Additional leap second to be added to UTC

[Halcyon]

 

I got the following e-mail from the Australian National Measurement Institute today.


Dear NTP user

As you may be aware, a leap second will be inserted into Coordinated Universal Time, and thus Australian time, at the end of December.
This will occur at 11 am AEDT on 1st January, 2017.

NMI NTP servers will begin announcing the leap second 24 hours before it occurs so you should make sure that your NTP servers are polling sufficiently frequently to catch the announcement.

Best regards

Dr Michael Wouters
Leader, Standards for Time and Frequency
National Measurement Institute
Department of Industry, Innovation and Science

[TheSteve]

This thread probably deserves more attention then it received. It isn't too often you get to see the seconds on a clock go to "60":

Clock: inserting leap second 23:59:60 UTC

[awallin]

our Meinbergs only showed 23:59:59 twice - what a dissapointment 


On servers with mostly a variety of Ubuntu installs with ntpd 4.2.6 and newer we had no issues. On some machines I specified a "leapfile" in /etc/ntp.conf and in others not - this didn't seem to make much of a difference (these are all close to the Meinberg stratum-1:s).

[jc101]

I managed a screen shot from my phone running an NTP app...

[helius]

What a terrible and pointless process
UTC is useless. It is neither accurate temporally, nor astronomically. Just abolish it completely.

[Ampera]

Time is useless. Why the hell do we have it anyways? I say we take all our stupid clocks and throw them in a pit together with some gasoline and a few matches.

[joseph nicholas]

I threw away my watch after seeing Easy Rider.

[slicendice]

HAHAH, time is not useless. Every(at least most) server infrastructure on the planet would crash if we had no time. They heavily depend on time to stay in sync.

[suicidaleggroll]

What a terrible and pointless process
UTC is useless. It is neither accurate temporally, nor astronomically. Just abolish it completely.

How do YOU suggest we resolve the discrepancy between the irregular period of the Earth's rotation and absolute time?  Either we use the Earth's rotation as the absolute time standard and screw everything up, or we only use a real absolute time standard and eventually the sun is rising at midnight and setting at noon, or we keep track of both, and slowly add to the offset between them so we can keep track of both absolute time and local time on Earth with minimal effort.

[Ampera]

HAHAH, time is not useless. Every(at least most) server infrastructure on the planet would crash if we had no time. They heavily depend on time to stay in sync.

You can add servers and infrastructure to that pile.

[slicendice]

So...do you think this domain would work and be accessible if we had no time? Are you saying this site is a piece of garbage? If so, what on earth are you doing here then?

Without time this excellent Blog/Forum would not be available, nor would YouTube or Google!

[helius]

How do YOU suggest we resolve the discrepancy between the irregular period of the Earth's rotation and absolute time?  Either we use the Earth's rotation as the absolute time standard and screw everything up, or we only use a real absolute time standard and eventually the sun is rising at midnight and setting at noon, or we keep track of both, and slowly add to the offset between them so we can keep track of both absolute time and local time on Earth with minimal effort.

The "discrepancy" did not ever need to be resolved, and it simply perpetuates confusion between two entirely separate concepts (time, and astronomical position in space). These are properly speaking, incommensurable, and so there was not a "discrepancy" to begin with.

This thinking that the neolithic "time" concepts of "noon", "midnight" and so on, are in any way important or worthy of being maintained is a folly and shows how far we have yet to go culturally.

[awallin]

This thinking that the neolithic "time" concepts of "noon", "midnight" and so on, are in any way important or worthy of being maintained is a folly and shows how far we have yet to go culturally.

FWIW, if you prefer TAI, a continuous time-scale without leap-seconds, just use PTP (aka IEEE1588) instead 

[Ampera]

So...do you think this domain would work and be accessible if we had no time? Are you saying this site is a piece of garbage? If so, what on earth are you doing here then?

Without time this excellent Blog/Forum would not be available, nor would YouTube or Google!

we would still have rocks we could bang together.

[Len]

This thread probably deserves more attention then it received. It isn't too often you get to see the seconds on a clock go to "60":

Clock: inserting leap second 23:59:60 UTC

Yes. In particular, CloudFlare should have paid more attention to it.

[helius]

FWIW, if you prefer TAI, a continuous time-scale without leap-seconds, just use PTP (aka IEEE1588) instead 

Yes, of course. TAI is time, within the frame of the Earth's gravity well. (Please, let's not spread ignorance even more by making stupid statements about "absolute" and so on.) However, it's not feasible to only interact with TAI-respecting systems, as mongrel UTC is the basis for legal "time" and its own patchwork of inconsistent differences. But TAI should unreservedly be the only way that time is measured; when anyone forgets that, it inevitably comes back to bite them. In POSIX, the correct way to measure time is using clock_gettime() with CLOCK_MONOTONIC.
"Monotonic time" is pleonasm: time is by definition monotonic, unless you're H.G. Wells.

[slicendice]

we would still have rocks we could bang together.

Yes and with time added we could have timing which in turn could give us rhythmic music produced by multiple rocks timely banged together. FANTASTIC! 

[suicidaleggroll]

How do YOU suggest we resolve the discrepancy between the irregular period of the Earth's rotation and absolute time?  Either we use the Earth's rotation as the absolute time standard and screw everything up, or we only use a real absolute time standard and eventually the sun is rising at midnight and setting at noon, or we keep track of both, and slowly add to the offset between them so we can keep track of both absolute time and local time on Earth with minimal effort.

The "discrepancy" did not ever need to be resolved, and it simply perpetuates confusion between two entirely separate concepts (time, and astronomical position in space). These are properly speaking, incommensurable, and so there was not a "discrepancy" to begin with.

This thinking that the neolithic "time" concepts of "noon", "midnight" and so on, are in any way important or worthy of being maintained is a folly and shows how far we have yet to go culturally.

Again, what do you propose as a solution?  Believe it or not, but a LOT of fields depend on these "neolithic concepts" of solar local time, such as tidal behavior, weather forecasting, etc.  I can't think of any replacement for UTC that would address these issues without vastly compounding the problem.  I mean seriously, we're talking about adding a second to the offset between UTC and GPS once every couple of years, is it really such a big deal that fixing it requires abolishing the entire concept and replacing it with something significantly more complicated?

[timb]

Swatch Internet Time should be the standard. Discuss.

[Ampera]

we would still have rocks we could bang together.

Yes and with time added we could have timing which in turn could give us rhythmic music produced by multiple rocks timely banged together. FANTASTIC! 

your no fun.

[Jeroen3]

Beats are still attached to solar days, which isn't stable and will require leap beats.

Leap seconds and leap years can be omitted, but your 8 AM start of your workday would slowly shift.
It's a bit backwards though, we change the time instead of our schedule when moving to another "time zone" or performing these small corrections. It does not make much sense, why not start work an hour later? Are these "special" (00:00, 6:00, 12:00: 18:00) moments of time of such psychological importance?

I guess we'd have to wait for interplanetary standard time or universe standard time.
Which is not required until you can take the next taxi to Mars.

[sleemanj]

why not start work an hour later? Are these "special" (00:00, 6:00, 12:00: 18:00) moments of time of such psychological importance?

It is convenient in some regards to have fixed time points around common human activities.

For example, if you book an airline ticket with an arrival time of 13:40, you know with a reasonable degree of certainty that this is in the day, that businesses will probably be operating, that you might have time to do some shopping etc, because 13:40 is "early in the afternoon" and you know what "early in the afternoon" is like.  When you are at this place, it is also easy for you to schedule your day, because you don't have to do any mental gymnastics such as "what time is midday here?"

If we all just observed UTC (which in a multiplanetary situation really would be "UNIVERSE"al even if it was adjusted for Earth days) your arrival time of 13:40 could have you arriving in the middle of the night which would be a bit inconvenient.

So, as annoying as it is to have timezones (and it's very annoying for a programmer!), it does make life easier for people to be able to specify time-of-day around the world (as opposed to moment-in-time which is why we have UTC).

[slicendice]

why not start work an hour later? Are these "special" (00:00, 6:00, 12:00: 18:00) moments of time of such psychological importance?

It is convenient in some regards to have fixed time points around common human activities.

For example, if you book an airline ticket with an arrival time of 13:40, you know with a reasonable degree of certainty that this is in the day, that businesses will probably be operating, that you might have time to do some shopping etc, because 13:40 is "early in the afternoon" and you know what "early in the afternoon" is like.  When you are at this place, it is also easy for you to schedule your day, because you don't have to do any mental gymnastics such as "what time is midday here?"

If we all just observed UTC (which in a multiplanetary situation really would be "UNIVERSE"al even if it was adjusted for Earth days) your arrival time of 13:40 could have you arriving in the middle of the night which would be a bit inconvenient.

So, as annoying as it is to have timezones (and it's very annoying for a programmer!), it does make life easier for people to be able to specify time-of-day around the world (as opposed to moment-in-time which is why we have UTC).