Science Changes to the world's time scale debated.

[T.Neo]

I would like to know why a method of keeping time that is totally arbitrary (due to its source of time zero of the epoch)

To be fair, the start date of a calender is arbitrary as well.

This is Special Relativity 101.

But... wait. Isn't time consistent for the observer in their own reference frame?

I.e. to me a second is always a second, whether I'm on Earth or Jupiter or on a starship sailing to Delta Pavonis at 92% c, even if it isn't a exactly a second to someone else.

 

[Cras]

To be fair, the start date of a calender is arbitrary as well.

That is true, so why ditch one arbitrary model for another that has no connection to human history?

But... wait. Isn't time consistent for the observer in their own reference frame?

I.e. to me a second is always a second, whether I'm on Earth or Jupiter or on a starship sailing to Delta Pavonis at 92% c, even if it isn't a exactly a second to someone else.

Yes, time is always time for the observer. Special relativity only applies when dealing with two observers, and two observers who are far apart, and/or moving at vastly different speeds will never agree on what "now" entails. It is the "time slice" that constitutes what makes up the presence universe.

So to go back to the point that the current calender is worthless when applied to a far off space colony because of all the earth cycle references, I take it a step further and say that the point to moot, since no calender, one with terrestrial references or not, will never be able to apply to both places, since the further they are apart, the most disagreement will arise on what entitles today. Granted, for these time slice to grow significantly out of whack, the distances must be very very large or their speeds must differ tremendously, but nevetheless, the point is the same. The concept of the present, such as a hope to have the same day, time, second for us on Earth and our far away space colony is not at all a worthy cause since neither of us will agree what occures at each interval.

We will always agree on the order of things, the cause will always come before the effect, this is preserved by the fixed speed limit of C, but when and which event occur at the same time will not. In summary, what the far away space colony will consider as the presence could entail events that have occured in our past, or even in our future, and makes the concept of the "Stardate" an impossible dream.

 

[T.Neo]

That is true, so why ditch one arbitrary model for another that has no connection to human history?

But that's not necessarily the case, the examples given in this thread do have a connection to human history (even if it isn't linked to a major or globally important event).

Also, why not just have clocks elsewhere in space that are fast or slow to compensate for relativity differences?

I've always imagined an interstellar starship having an onboard clock (say, with mission elapsed time) and then an Earth/destination clock, sped up to compensate for the difference in speed due to relativity. :lol:

 

[Arrowstar]

Airliners tend to get their clocks set by the GPS.

GPS time, if I recall, does not incorporate "leap" anything. A conversion is necessary to go between local civil time and GPS time. (Someone correct me if I'm wrong.)

 

[RisingFury]

Can you prove this? TBH, it sounds more like something you've made up rather than something intrinsic to the human body.

Oh, I forgot. You don't need to sleep...

 

[Cras]

Also, why not just have clocks elsewhere in space that are fast or slow to compensate for relativity differences?

I've always imagined an interstellar starship having an onboard clock (say, with mission elapsed time) and then an Earth/destination clock, sped up to compensate for the difference in speed due to relativity. :lol:

I guess it all depends on how accurate one wants to be. To envision a single mission to a far away place that alters its clock to compensate for relativity so it knows what time it is on Earth is all well and good.

But the the issue is what is happening now. Now we can alter the thing and say that things happen on this day, and that is the way it is. And for a set of circumstances this can work and would be well enough. But is would not tell the true and whole story. When distances grow, the very act of living, walking around doing chores in the space colony, the fact that the distance between the Earth and the colony will be in constant flux from orbits, what you will find is that the other guy's present will fluctuate as well up and down your "time scale".

So if the purpose of the time keeping system is to determine what happens and a certain time, and that would entail that if two people sit down on a chair at exactly star date 5432.432, then both people would have sat down on their chair at the exact same moment in time. But relativity says no. If the guy on Earth had super vision and was able to see the guy at the colony, depending on if their distances were growing or shinking in relation to each other, he could see either the guy has yet to sit down, or has already sat down, because his present universe, again depending on the change if he is moving towards or away from the other observer, incorporates the other observer's past or future.

So the point I guess I am trying to make is that the present is something that is not only relative, but becomes fluid, it fluctuates between two obeservers as their distance and velocities move, so as the Earth moves away from the colony on its orbit around the Sun, and then begins to move back again, the vector that determines the present time and the present universe in relation to the colonial observer's time line, will sweep from past to future, and make any hope of making an absolute notion of what happens on what day like we do on earth, be totally impossible.

And this is not a by-product of information only able to travel at C, this is assuming we can take snap shots of the entire universe that incorporates what we see as the present, or as I referred to them earlier as a "Time Slice". It is just the manifestation of time being a relative idea, not absolute. If Newton's laws were the universe we lived in, then time is time or each and everyone of us, and there is a cosmic clock out there that we can all set our watches to and agree on what the present is for now, then, and always, but alas, this is not our universe. We are stuck with this crazy relativistic one that simply refuses to be easy to understand.

 

[Linguofreak]

The problem is that the "epoch plus seconds" timekeeping system is absolutely and utterly meaningless in the real world, no matter how "science-y" it sounds.

That's why, in my system, we'd only use it for things requiring precision and predictability (and thus no leap seconds). Everyday timekeeping would still be done with traditional date formats.

Seriously. I started typing this post at 1 320 506 964. What?

That would be (assuming you added the missing leap seconds to the Unix time) the time scientists would use in my system for an event in an experiment requiring precise timekeeping that occurred at that time. In everyday use it would still be "Saturday, November 5 2011, 15:29:00 NPUTC".

Huh?

That 'time' makes absolutely no sense to me. It is just a random number. It might as well be 74 088. Or 761 390. Or 555 614 042.

And that's why everything but precision timekeeping would still use traditional dates.

It is without reference to anything. Humans are wired to think in terms of days and months and years.

Perfectly fine for computational or astronomical applications, but not for human ones.

Well, in the unlikely even that we ever do colonize space, traditional date formats will go completely out the window. Planetbound people will still measure time in days, months, and years, but their lengths will be different on each planet, necessitating a seconds-only interplanetary/interstellar time standard (of course, then you'll start needing to apply relativistic corrections, etc, and it gets really hairy).

People inhabiting self-contained space habitats might very well take a seconds only time format as an everyday time format (perhaps with breaks in places to make it resemble traditional formats/be more readable), so you might have something like 132/05 0:69:64 (Y/D H:M:S), with a "minute" being 100 seconds, an "hour" being 100 "minutes", a "day" being 10 "hours", and a "year" being 100 "days" (or a "month" being 100 "days" and a "year" being 10 "months", with a format of Y/M/D H:M:S with the above time being given as 13/2/05 0:69:64).

Preferably, don't use the UNIX epoch. I dislike UNIX.

As Jarvitä said, the epoch is negotiable. You can use the Gregorian epoch, the Anno Mundi epoch, the Ab Urbe Condita epoch, the epoch of the Heisei period, the J2000 epoch, the date on which the new system is planned to be introduced, or whatever else you please. Just as long as an epoch is chosen and time is counted in seconds from that epoch for precision timekeeping.

I chose the Unix epoch (plus leap seconds) for my example because I'm a Linux nut and Unix time is an established time format widely used in computing with all of the required properties (except for the fact that it's 24 seconds off from the actual time elapsed from its epoch because it ignores leap seconds).

Seconds are derived from the day. If I were choosing a time unit based on a universal physical phenomena, I would not choose something as arbitrary as the second.

It's just that the SI defines the second based on a specific physical phenomenon, it isn't derived from it.

Seconds have the advantage of being an entrenched standard, and the fact that they are now defined in terms of a universal physical phenomenon (and thus won't change) makes them acceptable, even if something actually derived from a physical phenomenon would be optimal.

My own preferred alternative to the second, though it's as bad as the second for arbitrariness (and has zero chance of ever being implemented) is the gigafoot / c (or light-gigafoot, the time it takes light to travel 1 billion US feet). It works out to slightly more than a second.

 

[Urwumpe]

GPS time, if I recall, does not incorporate "leap" anything. A conversion is necessary to go between local civil time and GPS time. (Someone correct me if I'm wrong.)

Wrong. GPS does actually transmit the number of leap seconds that apply since 1980 (It does so every 12.5 minutes, together with ionospheric information). But the GPS time signal for synchronization is without such leap seconds, it is pure atomic time.

 

[Linguofreak]

And the year or the reference we use is nonsense also (I'm sceptical if that certain person called Jesus even did exist).

Even secular historians generally agree that he existed, they just don't agree that the Bible's claims about him are true.

In any case even *Christian* historians generally agree that the AD epoch is several years off from his actual date of birth, so an argument for abandoning the current year numbering could be made in a way not offensive to (most) Christians. Which isn't to say that a lot of people (both Christian and non-Christian) wouldn't resist a year numbering change that was carried out just for the sake of getting rid of the current year numbering, since that would add to the math they'd have to do in discussing dates on opposite sides of the change for very little benefit.

 

[T.Neo]

People inhabiting self-contained space habitats might very well take a seconds only time format as an everyday time format (perhaps with breaks in places to make it resemble traditional formats/be more readable), so you might have something like 132/05 0:69:64 (Y/D H:M:S), with a "minute" being 100 seconds, an "hour" being 100 "minutes", a "day" being 10 "hours", and a "year" being 100 "days" (or a "month" being 100 "days" and a "year" being 10 "months", with a format of Y/M/D H:M:S with the above time being given as 13/2/05 0:69:64).

Why not just define a "day" as ~86 kiloseconds? Why the obsession with round numbers?

I don't state my height as only a meter, or the distance from my residence to Cape Town as only a 1000 kilometers, so that such measurements can be in tens only... :shifty:

As Jarvitä said, the epoch is negotiable.

I was more making a jab at Jarvita's dislike for 'mythology' being the basis of calendar dates.

My own preferred alternative to the second, though it's as bad as the second for arbitrariness (and has zero chance of ever being implemented) is the gigafoot / c (or light-gigafoot, the time it takes light to travel 1 billion US feet). It works out to slightly more than a second.

:lol:

 

[Urwumpe]

Can't we just skip the mess and count years since Ford?

 

[Cras]

yes

 

[Jarvitä]

Why not just define a "day" as ~86 kiloseconds? Why the obsession with round numbers?

Because every other SI quantity has "an obsession with round numbers". And the second already defines the SI units Ampere and Metre, so the cleanest solution would be to adopt the second + the usual SI prefixes as the sole unit of time. As for the epoch, maybe zero could be the second the change takes place - that way negative time is "old system era", and positive time is "metric time era".

It's marginally more sane in the same way that the SI is marginally more sane than the US customary system.

And it's not like this would make you abandon your circadian rhythm. You'd learn to think in terms of kiloseconds instead of hours and megaseconds instead of dates eventually, and the generation that grows up with the system wouldn't even miss the old one.

 

[Urwumpe]

Can't we define time as local sun azimuth and measure it in degrees, like our ancestors did?

 

[Jarvitä]

Can't we define time as local sun azimuth and measure it in degrees, like our ancestors did?

This system does solve the problem of telling time on planetary bodies in orbit around a star (as long as they aren't tidally locked), but it creates additional trouble in communicating the time to people who don't occupy the same space you do. Discrete time zones are useless enough as it is. Do you really want to make them continuous?

 

[T.Neo]

Because every other SI quantity has "an obsession with round numbers". And the second already defines the SI units Ampere and Metre, so the cleanest solution would be to adopt the second + the usual SI prefixes as the sole unit of time. As for the epoch, maybe zero could be the second the change takes place - that way negative time is "old system era", and positive time is "metric time era".

It's marginally more sane in the same way that the SI is marginally more sane than the US customary system.

Sorry, but I do not think you understand.

The issue is not the system of measurement, the issue is what is being measured.

The day is a unit of time, yes. But it is also a cycle that is pretty intrinsic to human life. Why obliterate it to fit in with your entirely arbitrary measurement system?

I do not restate my height as a meter just so that it can be a round number.

Round numbers in actual reality are pretty much nonsense. So why move away from a relevant measurement just to fit in with a round number or a prefixed version of an SI measurement?

Because it's "marginally more sane"? I am sorry. You may like the idea, but it seems like anything but sane to me.

And traditional dates and time are not necessarily as bad as the imperial system, the problem with which is the totally inconsistant measurements and silly benchmarks (like Farenheit. I hate Farenheit).

The reason that days and years are shaped the way they are, is because they correlate to relevant events. A date counting up from an arbitrary epoch in arbitrary units does not.

Despite some internal inconsistancy, the fact is that traditional time works. It works as a system of timekeeping and a means to set dates. Converting in some instances is annoying, but there are alternatives and for high-precision scientific uses for example, a more exact system can be used. But there is no need to force said system on greater society just because you think the idea is cool.

And seriously, if I were to define a universal base unit of time, I would base it on the day, because the day is a pretty relevant cycle.

Except... whoops! The second is based on the day. So wait: you're proposing we abandon the day, for a decimal derivative of an arbitrary time unit based on the day.

Marginally sane? :beathead:

You'd learn to think in terms of kiloseconds instead of hours and megaseconds instead of dates eventually, and the generation that grows up with the system wouldn't even miss the old one.

I am positive I would miss it, when my clock tells me to go to sleep because it's the "sane" kilosecond to do so, but it is midday outside.

Also, it is generally more important to me whether something is during a particular day or not, not whether it is during the 45th or 120th hour of the week.

Kiloseconds are too short to describe days, megaseconds far too long.

And also something you should be pretty scared of: the day, the "minute" and the "hour" are accepted for use with the SI.

(as long as they aren't tidally locked)

Do you see any tidally locked planets around here?

I don't. I see even fewer people living on them.

Discrete time zones are useless enough as it is.

Why then, does everyone use them? :dry:

 

[jangofett287]

Remember that all time measurements MUST have an epoch.
Seconds: since the start of the current minute
Minutes: since the start of the current hour
Hours: since the start of the current day.
Date: number of months and days since the start of the year.
year: number of full years since an arbitrary point in time 2011 years ago.

 

[Linguofreak]

Why not just define a "day" as ~86 kiloseconds? Why the obsession with round numbers?

Because it would allow keeping of regular schedules without having to do a lot of math. 100 kiloseconds is about 27.7 hours, which seems to be within the range that humans can adjust to sleep schedule wise (there are experiments on record involving 28-hour sleep schedules with a 24-hour light/dark cycle, 28-hour sleep schedules with 28-hour light/dark cycles should be easier to adjust to).

So, sleeping on a 100000 kilosecond schedule, you could just take the last 5 digits of your second count from <insert epoch here> and (for instance) sleep from 0:00:00 to 3:35:00, work from 3:75:00 to 7:10:00 (with lunch at 5:00:00), eat supper at 7:50:00, and have from 7:90:00 or so till 0:00:00 free.

With an 86.4 kilosecond schedule, the numbers would be different each day (though I suppose you could use traditional hours:minutes:seconds below the day, but then that messes up the neatness of your timekeeping on scales longer than a day).

Another alternative would be to use a 24 hour day and divide it up into 100000 "seconds" (each being exactly .864 current seconds long).

I don't state my height as only a meter, or the distance from my residence to Cape Town as only a 1000 kilometers, so that such measurements can be in tens only... :shifty:

It's more equivalent to deciding to grow to the next round meter in height, or sell your house and buy another that's exactly 1000 km from Cape Town.

Or to making the second a bit longer so that it can be "the duration of 10,000,000,000 periods (rather than 9,192,631,770) of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom."

 

[T.Neo]

Because it would allow keeping of regular schedules without having to do a lot of math. 100 kiloseconds is about 27.7 hours, which seems to be within the range that humans can adjust to sleep schedule wise (there are experiments on record involving 28-hour sleep schedules with a 24-hour light/dark cycle, 28-hour sleep schedules with 28-hour light/dark cycles should be easier to adjust to).

Yes, but why change your sleep schedules just so that your day-cycle can be defined by a round number of seconds? It makes no sense at all.

Especially if you are sleeping on cycles of 100 kiloseconds, and the Earth is turning every 86.4 kiloseconds or thereabouts...

With an 86.4 kilosecond schedule, the numbers would be different each day (though I suppose you could use traditional hours:minutes:seconds below the day, but then that messes up the neatness of your timekeeping on scales longer than a day).

Another alternative would be to use a 24 hour day and divide it up into 100000 "seconds" (each being exactly .864 current seconds long).

Or you could just describe your day in kiloseconds, tenths of kiloseconds, and tens of kiloseconds.

Kilosecond: ~16.6 minutes.
Tenth of a kilosecond: ~1.6 minutes.
Ten kiloseconds: ~2.7 hours.

That would make 02:21:00 as 0:8:4:60. 13:45:00 would be 4:9:5:00. 23:59:59 would be 8:6:3:99. I think, my math is bad.

Or you could just measure time of day as the number of kiloseconds that have elapsed. Seconds and kiloseconds are nice measurements on their own. I guess you could also use... hectoseconds? Does anyone ever use anything with a "hecto" on the front?

It's more equivalent to deciding to grow to the next round meter in height, or sell your house and buy another that's exactly 1000 km from Cape Town.

Exactly. But now think about it: why should I do something as serious as sell my house and move hundreds of kilometers away, just so that my distance from Cape Town can be a round number?

Isn't that absurdly silly?

And if you think about it: why change a long-standing, ingrained and relevant cyclical time period to match a a round decimal number of an arbitrary fraction of that same time period?

I can't help but find it absurdly silly.

 

[Linguofreak]

Sorry, but I do not think you understand.

The issue is not the system of measurement, the issue is what is being measured.

The day is a unit of time, yes. But it is also a cycle that is pretty intrinsic to human life. Why obliterate it to fit in with your entirely arbitrary measurement system?

Outside of planetary environments, a day can be (within certain limits) what you make it. I'm certainly not advocating that a day on Earth be defined as anything other than 86400 seconds for the forseeable future (eventually, due to tidal deceleration, it will be 86,401 seconds, but we'll cross that bridge when we come to it in ~50,000 years).

But 100 kiloseconds is close enough to a day that, if you're living on a self-contained space colony (assuming that such things ever will exist) far from any inhabited planet (planetary colonies will doubtless use their planets rotation to define their day), and you already have a decimalized second-count time standard for scientific use (and for coordination between planets with different day lengths), you may as well use the 5 low order digits of that second count for your "time" and the high order however-many digits for your "date". I'm not advocating 100 kilosecond days for any other situation than such space colonies (and maybe as a standard for ships in transit as well).

I do not restate my height as a meter just so that it can be a round number.

Nor can you arbitrarily grow or shrink on a whim. If you could (within certain limits) you might round yourself off to one meter (or two) if it did away with the need to make certain calculations in every day life.

Round numbers in actual reality are pretty much nonsense. So why move away from a relevant measurement just to fit in with a round number or a prefixed version of an SI measurement?

I don't advocate doing so when there actually is such a relevant measurement, such as if you are living on Earth, which has a day that is 86,400 seconds in length.

When there is no such relevant measurement, such as when you're living in a space colony far from any inhabited planet, and the 10 nearest planets have 10 different day lengths based on their own day/night cycles, but you get to choose when to turn your lights on and off, why not slip in to a nice round 100 kilosecond schedule?

I am positive I would miss it, when my clock tells me to go to sleep because it's the "sane" kilosecond to do so, but it is midday outside.

So would I.

But the story is slightly different when you're living in an environment where the time for midday is arbitrarily chosen by a human being (or a human government) rather than defined by the sun being overhead.

Also, it is generally more important to me whether something is during a particular day or not, not whether it is during the 45th or 120th hour of the week.

Kiloseconds are too short to describe days, megaseconds far too long.

Which is why a convenient value for the "day" in situations where a planetary environment doesn't define it for you is 100 kiloseconds.

Of course, the really messy situation would be when you *do* have a planetary environment, and it throws a week-long day at you.