SF timekeeping systems

[Ultonius]

The problem of how to measure time on other planets and in space is something that most SF works not set on Earth have to deal with. I've been thinking about this myself, and I've come up with a couple of timekeeping systems, and tried to consider their ramifications.

The first system is intended for timekeeping on worlds with day and year lengths that are relatively similar to Earth's. It bears some similarities to the system used in Robinson's Mars Trilogy and Weber's Honorverse and Safehold series, where a planet's day is measured in hours, with the leftover time placed in an extra shorter hour known respectively as Timeslip, Compensate and Langhorne's Watch. One problem with this is that it prevents the day from being divided into two equal halves as Earth's can, since there is only one extra 'hour', and there may be an odd number of full hours.

In my system, a planet's day would be divided into an even number of normal hours. The leftover time would be divided into two halves. If each half was less than thirty minutes long, each would be added to one of the normal hours, creating two 'long-hours'. If the halves were between thirty minutes and one hour long, they would become 'short-hours'. Leftover seconds could be placed into 'long-minutes' and 'short-minutes' in a similar way, while leftover fractions of a second would be placed into 'long-seconds' and 'short-seconds'.

Taking Mars as an example, its day is 24 hours, 39 minutes and 35.244 seconds long. 24 is even, so dividing the 39 minutes and 35.244 seconds by 2 gives us 19 minutes and 47.622 seconds. Since 19 is less than 30, 2 of the normal hours will have 19 minutes and 47.622 seconds added onto them to make them long-hours. The 47.622 seconds would be placed into an extra short-minute, and the 0.622 seconds into an extra short-second. Mars would therefore have a day of 24 hours, 2 of them long-hours of 50 minutes length, each with a final short-minute of 48 seconds length, with the 48th second being the short-second.

One problem that this system would have is that it might be hard to represent the long/short-hours and long/short-minutes on a traditional clock face. However, there could be ways around this, such as having an additional partial circle of numbers beyond the starting point for long-hours/minutes, or marking the end of short-hours/minutes on the main circle. I'm also not sure how having two hours of a different length in the day would affect time zones or measuring longitude.

I'll talk about the second system in a later post, but in the meantime if anyone has any queries or has spotted any major drawbacks to the first system I'd appreciate the feedback.

[Purple]

Honestly, the whole long-X, short-X system seems very unpractical to me. The whole point of time keeping is to divide the day up into equal sized units for easy reference. You noticed some of the more obvious problems your self. But that is just the beginning. Remember, 24 is not only handy because it lets you divide the day by 2. It also lets you divide it by 3 into 3 equal sized work shifts. And your system with long and short hours would render any sort of non stop work arrangement a nightmare. And finally, the system would he hard to use period. After all, your colonists now have to keep track of long and short periods during normal routine day to day stuff. Next you have the issues of planets whose days are not neatly earth like. Worlds where a day is much longer or much shorter than our own.

I have to ask, what is your goal here? It seems that you want to standardize time across the whole nation/setting and use one earth day as a base or something similar. In that case, why not just have one official date and time that is set by the government and have each planet measure time in what ever way they want analogous to our time zones?

[Ultonius]

Honestly, the whole long-X, short-X system seems very unpractical to me. The whole point of time keeping is to divide the day up into equal sized units for easy reference. You noticed some of the more obvious problems your self. But that is just the beginning. Remember, 24 is not only handy because it lets you divide the day by 2. It also lets you divide it by 3 into 3 equal sized work shifts. And your system with long and short hours would render any sort of non stop work arrangement a nightmare. And finally, the system would he hard to use period. After all, your colonists now have to keep track of long and short periods during normal routine day to day stuff. Next you have the issues of planets whose days are not neatly earth like. Worlds where a day is much longer or much shorter than our own.

I have to ask, what is your goal here? It seems that you want to standardize time across the whole nation/setting and use one earth day as a base or something similar. In that case, why not just have one official date and time that is set by the government and have each planet measure time in what ever way they want analogous to our time zones?

I realize that the system isn't perfect, but it still seems to be better than the alternatives. If you want to keep normal hours, minutes and seconds you'll still have to deal with leftover time one way or the other, unless you want the clock time to drift out of sync with the time of day. If you want to keep a day of 24 hours of equal length on a world with a different day length, you're going to have to alter the length of the second, minute or hour, which would create a lot of confusion when dealing with off-worlders who use different measurements.

My main goal was to create a system that would keep the customary units as much as possible, while being adaptable. I don't think that the difficulties you talk about would be insurmountable. At most, one or two shifts would have to work half an hour or an hour longer. Keeping track of the long/short hours wouldn't necessarily be particularly difficult either. My personal preference would be to put them immediately after noon and midnight, so that all people would have to remember is that one o'clock comes sooner or later than the other hours.

As for worlds with very long/short days and standardized interstellar time, the second system, which I've yet to talk about, will be more useful for those.

[Baffalo]

I must agree with Purple on this. Numbers aren't just there for counting, they're there for several reasons. Why do we have 60 minutes in an hour? Because you divide it by 2 to get 30, 3 to get 20, 4 to get 15, 5 to get 12, 6 to get 10, and vice versa. It's a handy way of dividing up time, especially when the concept of exact seconds in a day mattered about as much as a gnat sneezing a thousand miles away. Numbers are important, but they're also arbitrary in many, many ways. Why do we use base 10? Why not base 16? Or base 5? Hell, base 12 would work nicely too, and it's just completely arbitrary.

Standardized time is something we take for granted, but it's a rather new concept. We didn't have timezones until the mid-19th century, we didn't have decent chronometers until a genius in England came up with it. Before that, water clocks were the only way to accurately measure time, and that was a very convoluted system involving tanks of water and little lines marked on the side. Yes, it's nice to know what time it is in London when you live several hours away, but it wasn't until the invention of the telegraph that this information became vital.

Now, yes, we're talking about sending ships halfway across the galaxy and we need to deal with planets that spin faster/slower, orbit faster/slower, etc. By far, what would be easier is to simply use Purple's strategy: set a standard time on Earth (say, GMT) and this is the standard for everywhere, whether on Earth (as it is now) or on a planet on the other side of the galaxy. Sure, there's a few issues such as time-dilation, but that can be solved by knowing the approximate distance from Earth (say, 10.2 lightyears) and simply setting certain planets at various points to have fixed chronometers set to Earth Standard Time, to serve as standard across the cosmos.

Now, as far as I'm concerned, planetary time should be broken down as such:

• For short-day planets (< 18 hours): Set a 12 hour day, with 30 minutes to an hour, adjusting seconds per minute accordingly.
  For Earth-day planets (18 to 30 hours): Set a 24 hour day, 60 minutes to an hour, adjusting seconds per minute accordingly.
  For long-day planets (> 30 hours): Divide the total number of hours by 12, rounding to the nearest number, and then multiplying that number by 12, with 60 minutes to the hour and adjusting seconds per minute accordingly.

What this does is make sure that each planet has a unit of time measurement approximately equal to an hour, while still being able to break the time of day down into handy blocks of time. Say a planet's day lasts 43 hours. Make it a 42 hour day, so the hours are slightly longer. However, if you want to break the day down into halves, you have 21 hours, or by 3 to make it 14 hours for three shifts. Of course, since the hours are longer, you could break it down by 6 to make it 7 hours. Since the hours are slightly longer, that gives you a shift just shorter than 8 hours on Earth.

It's not a perfect solution but it's a start.

[Darmalus]

If we are traveling to other planets, why are we shackles to light/dark cycles? Why not just make seconds/minutes/hours a constant (good enough for human scale activities) and let the computers keep track of weird 17.2 hour day planets doing business with 903.93 hour day planets. If my job is indoors, it really doesn't matter if it is light or dark outside as long as I get enough sleep beforehand. Heck, you'll probably have space stations where the length of the "day" is completely arbitrary and up to the whim of the local administers.

Edit: and if we don't have super fast FTL or instant communications, then it REALLY doesn't matter because each system becomes an island that shouldn't give a damn about the time systems of people with a multi-year communications lag.

[Ultonius]

Now, as far as I'm concerned, planetary time should be broken down as such:

• For short-day planets (< 18 hours): Set a 12 hour day, with 30 minutes to an hour, adjusting seconds per minute accordingly.
  For Earth-day planets (18 to 30 hours): Set a 24 hour day, 60 minutes to an hour, adjusting seconds per minute accordingly.
  For long-day planets (> 30 hours): Divide the total number of hours by 12, rounding to the nearest number, and then multiplying that number by 12, with 60 minutes to the hour and adjusting seconds per minute accordingly.

What this does is make sure that each planet has a unit of time measurement approximately equal to an hour, while still being able to break the time of day down into handy blocks of time. Say a planet's day lasts 43 hours. Make it a 42 hour day, so the hours are slightly longer. However, if you want to break the day down into halves, you have 21 hours, or by 3 to make it 14 hours for three shifts. Of course, since the hours are longer, you could break it down by 6 to make it 7 hours. Since the hours are slightly longer, that gives you a shift just shorter than 8 hours on Earth.

It's not a perfect solution but it's a start.

That sounds rather complicated, but I suppose it could work. I was thinking of days around the 18-30 hour range when I came up with the system.

Edit: and if we don't have super fast FTL or instant communications, then it REALLY doesn't matter because each system becomes an island that shouldn't give a damn about the time systems of people with a multi-year communications lag.

I was thinking in terms of relatively slow FTL, without FTL communications.

Now that we've got on to space stations and universal time, I'll describe the second system, since it is intended for space habitats and planetary colonies where the rotation and/or orbit are not convenient for earth-style days and years (as well as an official auxiliary interstellar dating system). It's a variant of the second-based 'metric time' used by Stross and Vinge in some of their works.

The main building blocks of the system are the kilosecond/kay (roughly 16.6 minutes), megasecond/meg (roughly 11.6 days) and the gigasecond/gig (roughly 31.7 years). A 'metric day' is 100 kiloseconds (roughly 27.8 hours). It is divided into 10 decakiloseconds/decakays/decs (roughly 2.78 hours), each of which is divided into 100 hectoseconds/hecs of 100 seconds each. Kays tend to be used to describe lengths of time between a hec and a dec. Clock faces are labelled 1 to 10, with 10 hec graduations between the numbers.

The meg functions as a 10-day week, and 5 megs as a month. A 'metric year' is 50 megs (roughly 1.6 Earth years). The age of majority is usually 600 megs (roughly 19.01 years). Finally, the gig is used instead of the century, though it is roughly 3 times shorter.

Dates would be counted from the beginning of the Common Era, and given in megs to one decimal place, while times would be in decs to two decimal places. For example, the metric date of noon tomorrow would be 63,480.8, while the time would be 8.80. A less precise method of dating, the equivalent of stating the year and season, would be to give the number of gigs to two decimal places, i.e. 63.48.

[Dr Roberts]

In a "realistic setting" a "universal" timekeeping system is irrelevant. However assuming FTL thus enabling trade (the only viable reason for a "universal" time) you could just have 'local time' which is based on their planets rotation and orbit etc then a "Trade Time" for companies to use for trading. The more complicated the system, the more pointless it is as there isn't exactly a large audience interested in knowing how time is kept in these settings.

[Simon_Jester]

Or people could just get used to knowing what time it is at the place you're calling/visiting, and adjusting plans accordingly. Do I call the corporate factor as soon as my ship arrives in orbit? Nah, it's the middle of the night at the longitude of his offices. Does he have a backup on the other side of the planet? Better find out...

Looking that up on a computer will be much easier than all the stuff you had to do just to get in touch with them in the first place; if you can't adapt your schedule to local time, maybe you should reconsider this whole 'Han Solo, interstellar trader extraordinaire' line of work.

[Batman]

Heck the same is true for this here lone planet.
The way I see it, keeping timestamps constantly consistent is actually only important if you actually have superfast FTL and/or communications. If all communications is by courier, adjusting to local time (like we do now) is no big deal. The head honchos at Wayland Yutani won't know you took a few zs before reporting back and it wouldn't have made a difference if you had done so immediately anyway what with the signal taking days (if not more) to arrive.
In fact, I think most people will ignore your standardized (and unworkable) time system in favour of one that actually works locally even then. Let the people who have to work with offworld contacts at any hour do so. You go with whatever is locally convenient (just like, you know, the real world?)
Where this gets interesting is planets that have years longer than their days or multiple star systems that don't really have night.

[Baffalo]

Where this gets interesting is planets that have years longer than their days or multiple star systems that don't really have night.

I am rather curious how you'd solve the problem of multi-star systems and localized times. Would you want to set a primary star as your main time reference? Or would you much rather try and work all the stars in?

[Batman]

What makes you think I have a solution? That's why I said this is where it gets interesting

[PeZook]

Seeing as clocks are dirt-cheap, and we have those wondrous things called "computers", it will likely be a complete non-issue. It might be a long shot, but an inerstellar trader will probably be able to afford a mobile app which will tell him EXACTLY what time it is at any of the colonies he wants to visit, translated into whatever time standard he uses. Maybe even display it as a cool graphical interface cross-referenced with his contact list on the planet/space station.

So the primary consideration when setting up time standards will be psychology (ie. keeping colonists from going insane) and economics (minimizing issues with planning work) rather than interoperability, because writing a conversion algorithm and distributing it to everyone who might want one is going to be trivial.

[Baffalo]

What makes you think I have a solution? That's why I said this is where it gets interesting

Well, normally planets in multi-system stars are portrayed as orbiting a single star predominantly over the other. However, depending on the mass of the stars themselves, the stars could be orbiting in any number of complex arrangements that lead to all sorts of headaches. Binary systems are the easiest to wrap our heads around, since there's a single common barycenter to the stars' orbits, but when you start to find triple and possibly quadruple star systems.. yeah, complex and interesting. Thankfully, we don't really need to worry about unstable systems, as most unstable systems typically will vie for position until they settle into a stable orbital system by ejecting stars that leave the system unstable.

In my head, I can picture one of three types of star system with multiple stars and planets:

• A primary mass star, with the planets and other stars orbiting it.
  An equilibrium star system where planets orbit all stars.
  A hierarchy system where planets shift orbits depending on their proximity to various stars in the system.

The first star system would have a fairly standard time-keeping arrangement, as would the second. The second would be more or less the same due to the planets orbiting both stars (say two stars of similar mass orbiting a close, common barycenter), so it would be much like Tattooine. The only one that gets complicated would be the planet constantly shifting orbit from star to star, and those types of planets aren't going to be stable enough to support life due to constantly moving in and out of habitable zones. Any instillation built on these types of planets would probably need to follow the standard for stations and other orbital facilities.

[Ultonius]

A common time standard could still emerge for political reasons. If colonies in the Asteroid Belt and Jupiter Trojans, for example, have been running on Earth time, but then become independent, there might be a move to develop their own calendar rather than using one based on the year of a planet that no longer rules them. They might adopt metric time both to distinguish themselves from Earth and to enhance the unity among themselves, since each asteroid and the habitats associated with it would have its own orbit and year (which might be too long to base a practical calendar around anyway). Revolutionary France experimented with a form of decimal time that divided the day into similar units to the ones I talked about earlier, but it didn't catch on, partially because it required a redefinition of the second to fit 100,000 of them into a day. Space habitats, on the other hand, could adjust the length of their day/night cycle by reprogramming the timers for their mirrors or artificial lights, allowing them to adapt to metric time relatively easily.

[PeZook]

'Relatively easily' if you exclude the parts about retraining everybody to use the new standard and adjusting the software on all electronic, mechanical and scientific equipment.

It's much easier to set up a new clock and calendar before your society is established. It's the same reason why the US isn't keen to adopt the metric system: what they have works, and changing everything would be just too much of a hassle and give no benefit whatsoever.

[Eternal_Freedom]

Frankly, I tihnk the system used in SW works well enough. As shown in X-Wing: The Krytos Trap you have clocks set to local time and Galactic Standard Time (Coruscant time, in other words). Some planets apparently set their clocks to GST and this is not unusual enough to arouse suspicion immediately in a detective-trained pilot.

I think that's all you need. Local time and a standard time for ships in space to set their watches by. It's what we have today if I'm not mistaken: ships at sea (especially subs) run on GMT rather than local times.

[Simon_Jester]

Recalibrating scientific equipment for a new 'second' is gonna be a bitch, as is rewriting all the reference tables to account for things like "this piston cannot safely descend at more than two meters per second" having changed...

[Jub]

Then have something like local time, galactic time, and scientific time. Two of them are for basically telling when you need to be places, and then the last is for calculations and the like.

[fgalkin]

Recalibrating scientific equipment for a new 'second' is gonna be a bitch, as is rewriting all the reference tables to account for things like "this piston cannot safely descend at more than two meters per second" having changed...

Why would you redefine a second- it's not based on Earth time any more? It's currently defined as "the duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom". The tables will remain constant throughout the galaxy, independent of local time.

Have a very nice day.
-fgalkin

[Simon_Jester]

Recalibrating scientific equipment for a new 'second' is gonna be a bitch, as is rewriting all the reference tables to account for things like "this piston cannot safely descend at more than two meters per second" having changed...

Why would you redefine a second- it's not based on Earth time any more? It's currently defined as "the duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom". The tables will remain constant throughout the galaxy, independent of local time.

That's pretty much my point- why would you?

You can look upthread for the context of what I said.

[Skgoa]

I would like to point something out: as Simon has touched apon, a shitload of stuff is based on the second being a second long. All computers internally operate not on "human" time but on seconds since X, with X most often being 1.1.1970. No matter what the GUI of your time keeping device will tell you, underneath it WILL count seconds. It might very well use complex conversion tables depending on your location and other factors, just like is the case today. (There will be apps for that.) A day isn't exactly 24 hours long and a year isn't exactly 365 days - but somehow we manage to get around that. So yeah, if it is practical (regarding rotation speed) to have a somewhat normal day-night cycle, we will see a local time. Spaceships and stations will most probably use "universal" time. And every important date will be given in both formats.

[Baffalo]

Computers run on internal clocks operating in the gigahertz range, with each pulse of the crystal causing the operations to cycle to the next operation called upon. Most computers do this independently of clocks that we care about, because the clock itself simply pulses. For the clocks we care about, most computers have a second clock that pulses and cycles registers, building up in a register until a pre-set number is reached and then it trips a register to cycle to the next second. The reason this is the case is that the second clock is actually a hardware accessory that gets called upon to deliver a number that is then added to the 1.1.1970 you mentioned Skgoa.

This isn't perfect, because these clocks sometimes skip. Most internet connected computers link up with a standardized clock, and if you pay enough attention, you'll notice your wrist-watch sometimes drifts compared to your computer. For most purposes, it's not worth noticing, but for precise scientific measurement, if you're off by enough, tiny deviations become large problems.

[Ahriman238]

Well, 40K uses the Terran calender no matter whether it fits or not. I do believe they make allowances for local days.

HH everyone uses the standard seconds, minutes and hours, but local days and years and a celnder usually based on the initial colonists' landing. For events in space or of galactic significance everyone uses the Earth Post-Diaspora calender.

There was a webcomic that recently ended, Escape from Terra, where Earth kept plodding along but all the space colonies (especially in the asteroid belt) use the Martian day and year, and replace hours/minutes with 'centimes' each 1% of a Martian day.

[andrewgpaul]

Actually, 40k has its own "space calendar". The official "1231239.M41"-type dates are for official record-keeping purposes, and it's stated that planets will use their own local timekeeping in day-to-day operation.

One thing the 40k date system does seem to do is take uncertainty into account. None of the proposed systems in this thread take account of the complications from FTL - or even relativistic STL - travel.

[Baffalo]

One thing the 40k date system does seem to do is take uncertainty into account. None of the proposed systems in this thread take account of the complications from FTL - or even relativistic STL - travel.

Good point! IIRC, the GPS satellites in orbit over Earth, just a short (relatively speaking) ways up, must account for a 6ms variation due to changes in space/time. They're actually experiencing time faster than we are on Earth.