It is 5:40:79 Simple Universal Time - change your time thinking now | Hacker News Reader

It is 5:40:79 Simple Universal Time - change your time thinking now

telmich.github.io

16 pointstelmich59 comments

Open articleView on HN

Swatch Internet Time (or Beat) all over again: http://en.wikipedia.org/wiki/Swatch_Internet_Time

This idea was simple to understand, but didn't catch on!

If it used Greenwich as 0 and there had been more users back then, it would have caught on.

You are right. I actually like that approach as well, but that system had at least two major drawbacks:

- being pushed like a marketing thingy for a specific company - not resembling hours/minutes

I guess the latter point was crucial for it not being adapted. Or it was also too early (early internet time)

Not resembling traditional time notation is good thing IMHO, as one can't even accidentally confuse local time with non-local one.

i recall that even as of ~2000 php3 had @beat support via its date() function. I just checked to see if it had been removed, but it's still (surprisingly) there.

At least you can't say it failed for lack of ubiquity :)

do you know how to get SUT support in PHP? ;-)

Adopting metric time, while a neat idea, would probably have a lot of consequences for all the SI units derived[1] from the current standard 'second', no?

On an unrelated note, in standard gravity (1𝘨) a pendulum with length 1m will complete half an oscillation in fractionally more than 1s (~1.006s), an annoyingly close relationship for one that's not supposed to correlate exactly.

[1]: https://en.wikipedia.org/wiki/SI_derived_unit

You are right, you'd have to use a new second definition. That's why I am referring to "simple seconds" (you can debate the name...) in the article.

I guess it's easy to find another molecule/atom to relate the time to - there are many around.

This solves nothing - person X still has to take into consideration the daily routine of person Y on the other side of the planet. Announcing that you're disregarding time-zones does not have the effect of making time-zones irrelevant.

Something like SUT is only really suitable for things which are completely independent of human behaviour (i.e scheduling automated tasks for regular intervals).

Whenever one uses universal time or local time they still have to calculate (or look up some table) before they can consider whenever it's day or night in some unfamiliar remote place.

Universal time system solves the problem of discussing time. It works when you're talking to a person, negotiating time of some event - you can both be sure that you use same notion of time.

It's especially good if notation radically differs from usual HH:MM:SS (like swatch @BBBB), as one can accidentally misassociate time systems in their head.

But the world still spins regardless of how you end up talking about time - you will still need to figure out the human behaviour at the specific location on earth at that specific time. It is useless to schedule a conference call when the participants are all in bed. SUT does not solve this problem - people will still need to take into account the fact that night and day still exist and most people work during daylight hours.

You still have to negotiate when one's available. I.e. ask and tell time values - there's no other way of figuring out others' schedules.

And this is easier when you don't have to do any calculations on those values. This is the part which SUT (and alike) solves.

It solves a problem which does not exist - people organise things based on an agreed timezone already. The pain of transitioning to SUT is not worth the (debatable) gains.

Actually I don't agree with you - I think if humans got used to SUT, they can (in their own time!) easily recognise the shifts around the globe. I.e. 2:00 SUT is the start of the day in Europe, 5:00 SUT is start of the day in US and 7:00 SUT in Asia.

I believe that due to talking about the same time, people can more easy remember the other times.

Compared to time zones, which are based on learning them by heart without a relation to your current time zone (only relates to UTC and your time zone to UTC again -> 2 calculations).

Yes, but you're missing the point - people talk about the time of day, almost universally in relation to the standard 9-5 working pattern that most of the industrialised world follows.

What does 'this morning' mean in SUT? 'This evening'? 'I'll speak to you tomorrow'?

So use UTC instead. It is already there, your computer, phone and alarm clock can already handle it, and the issues of leap seconds, leap years, a changing length of day and accurate time keeping have already been addressed.

It doesn't solve the stated aim of having to work out a good time to call someone in a different part of the world. They have the same time (no timezones), but what time of day is OK to call them? They don't work 9-5 they work 3-11.

But if you think this idea is good, let's agree on a worldwide standard for electrical outlets while we are at it? You've got about as much chance of that proposal succeeding.

electrical outlets, language, voltages, train systems, side on which you drive are actually things on my todo list.

Point taken, I'll continue my work on making the world a bit better.

Right. The problem is that there are two fundamental reasons for dealing with time across the globe, and their solutions are generally mutually exclusive.

Reason 1 is synchronization. Knowing if an event happens in place X at time T when that happens in place Y. This is what universal time simplifies.

But universal time messes up the other reason for dealing with time across the globe, which is knowing what part of the day it is in other places.

While timezone conversion pretty readily allows synchronization across the globe, universal time doesn't have a readily-available solution for knowing what's going on in New Zealand right now. (It's 2:30 AM there, so probably not a whole lot, but you knew that as soon as I said 2:30 AM, which is the whole point.) You are still going to need to have some way to say what the current time means in other places. This is automatic right now. Two in the morning means the same thing everywhere. A universal time doesn't. So you'd have to say something like, "It's currently time T, which in New Zealand is several hours before sunrise." So you're basically coming up with ad-hoc local times anyway.

Won't using UTC globally be enough?

Converting to decimal time will render most current time-aware devices obsolete, from wrist watches to microwaves to airport terminal software.

Also, scientific calculations rely on metric time[1], and need not to deal with minutes/hours.

[1]: http://en.wikipedia.org/wiki/Metric_time

The problem is that you cannot modify people's behaviour by pointing them to something existing:

harry: "hey, utc would be great, let's drop CEST." joana: "UTC was designed for something else ... <more offtopic arguments"

whereas with SUT, you have the chance to change behaviour:

harry: "There's a new time format, SUT. If we adapt to it, the world will become much simpler." joana: "Wow, so cool. Let's do that!"

And then you keep a list of work hours in each friend's contact card. So Sally is in New York, her work hours are 14:00–22:00. Frank is in Amsterdam, so his work hours are 09:00–17:00.

It is exactly the same as UTC, just using decimal. So just use UTC, or the time zone support in your calendaring software.

I'm not sure why the proposal uses "simple seconds", "simple hours" etc. We didn't call them "metric inches" or "metric yards" but rather centimeters and meters. A new naming system will be important otherwise some people are going to be really peeved to learn their "8 hour work day" is a lot longer now.

"Simple" is just the reference to the name - in my opinion the name of the new hours/minutes/seconds does not really matter - it just needs to have hours/minutes/seconds in their name and needs to be distinguishable from the current scheme.

Obligatory xkcd: http://xkcd.com/927/

I think that it is a bad idea because these measurements are not duodecimal by accident.

They are quantities and notions that we use every day, extremely frequently, and the mathematical properties of duodecimal bases versus decimal are useful in this context. I think this is why these things kinda evolve towards different bases.

I can't get the logic here. Why this switch will be helpful? Does this mean : If I live in Europe (e.g.) and the sunrise is at 3 SUT, I'll have to go to work at around 4 SUT, but in the states the sunrise is at 7 SUT and people will go to work at 8 SUT?

Yes, and it gets really bad for people who live in places where which day and date it is changes in... the middle of the day, for a lack of a better word.

> Why this switch will be helpful?

I was raised in Britain towards the end of the transition from imperial to metric units. I would NOT go back.

So it worked for yardsticks. Why not clocks?

> If I live in Europe (e.g.) and the sunrise is at 3 SUT ...

You're catching on. It's simpler than Daylight Saving Time, don't you think?

From "Motivation" [1]

> To place a call, you have to find out

> in which time zone the target person is > how much offset to UTC this particular time zone has > remember the delta from my time zone to UTC > create the delta from both time zones > find a good time for a call > This is quite cumbersome and wastes a lot of energy world wide, every day.

So now I will not have to find the timezone, but I'll have to take a look at something like "SUT Sunrise World Map" and figure out which part of the day-night cycle is the destination.

I agree, this will help me when I travel from Europe to the States - since I will not need to adjust my mind and freak out when I see my watch, but when I go there I'll have to adapt to something else, different than timezone.

P.S. I like the idea of 1 min = 100 seconds, My programmer's background and SUT will give me something like 10 megaseconds = 1 hour, which is nice.

[1] http://www.nico.schottelius.org/docs/sut/

That isn't the key property here - the same sort of effect would be seen if we all just agreed to settle on UTC, for instance, instead of having local timezones.

The proposal is also decimalising time: 1 day is 10 hours of 100 minutes of 100 seconds instead of 24/60/60.

You are absolutely right!

Here's my (hopefully unique) take on this. A part of what makes analog watches beautiful is the number of hours (12) - hexagons (6 segments) and hexagon-like structures are aesthetically pleasing.

With SUT and its 10 hour system, clocks and watches would start looking / be reminiscent of car wheels (most car wheels are 5 symmetrical segments [1]).

My knee-jerk reaction is that I'm not willing to give up hundreds of years of watch face refinement for easy math conversions.

[1] https://www.google.ca/search?q=car+wheel&tbm=isch

You can always use a circle (as with current watches) and just add the number of hours per day (does not matter if it is 12 or 10).

You could probably quite easy relabel analog clocks to use SUT, btw...

Not quite what I meant. Lots of watch faces play on the fact that 12 is divisible by 2, 3, 4, and 6. For example, take this [1] Ulysse Nardin watch. It has six nice "hexa"-markings (a common theme among watches). Some other watches go for the cross-hair (division by 4). Neither one of the styles would be possible with the 10 hour system.

Note: those watches do not reflect my personal taste, they were just the easiest to find.

[1] http://www.ulysse-nardin.ch/ulysse-nardin-webseite/uploadfil...

[2] http://www.tissot.ch/imageslib/cache/2011/png/image225png_T0...

There are so many proposals for a different, better time system. This one is actually identical to my proposal for a change to O'Harean time (it's based on 20 hours, but for no especially good reason; I suggested changing it to 10). And I too invented my own kind of date/timestamp, which is somewhere midway between readable for humans and computers: https://lucb1e.com/rp/othertimes/lucb1e-time.php

Thanks for the pointer! It's always great to read about other ideas. Maybe you can explain how your proposal would make the world more simple?

Well unix timestamps are not quite readable for humans, this is a bit better. It lacks a system of weeks though, so I guess it's only a midway solution, not great for human usability either. And it's not so much of a system as it's an idea.

Can someone explain the importance of local times needing to be the same? Who cares if I go to work at 3, while you start at 5. Do we think it will force a creep of the work day's starting and stopping time? Some of us already stay up (or get up) to deal with other parts of the world. Is it that we need to compare our bedtimes with others?

I actually wouldn't mind this. None of our relatives live in the same area as us and anything that requires coordination (e.g., planning to video-chat), requires specifying a time zone.

Major problem would be counting days. You know that 3 nights equal three days. But if day changes during your afternoon nap, that can lead to some unexpected complications. Eg the natural day would be out of sync with official day.

That's actually only correct if you think locally - your night does not actually match with your neighbour's timezone night.

but yes, it means some people will sleep during "day" while some will sleep during "night". It may be appropriate to think about their definitions when using SUT as time base.

That's a very good point - getting up is always relative to the daylight AND culture. That's also why messing around with summer time just makes things worse (and many countries stopped utilising it).

It's also way easier with one time format to communicate:

john: "let's meet tomorrow virtually, I'll be reachable from 4-7 SUT" marry: "great, I stand up at 5:90 SUT, let's meet at 6:15 SUT"

It also visualises the differences between regions better and does not force people to think in local time:

Instead of getting up at 8 local time, you stand up whatever time makes sense, not locking down to the "usual 8 o'clock" rythm.

Shouldn't they use a single unit and metric prefixes, instead of hours, minutes seconds?

For instance, meters are always meters. Whether millimeters, centimeters, decimeters, meters, kilometers, et al.

SUT proposes:

- A day has 10 hours

- An hour has 100 minutes

- A minutes has 100 seconds

Instead, using using hour as the single unit.

1 day = 10 hours

1 hour = .1 day = 100 minutes = 100 centihours

1 minute = .01 hour = 1 centihour

1 second = .01 minute = .01 centihour = .0001 hour = 100 microhours

I don't think SUT fits its own motivation: instead of comparing time zones (for which numerous apps already exist), I have to compare local working times (for which no apps exist ;-).

Further, having seconds, minutes as multiples of 10 only makes very few things easier. Or when was the last time you had to find out how many hours are in 1234567 seconds? Did you have to do that more often than finding out how many months are in 1234567 hours?

Last, it totally ignores the concept of leap seconds.

This is a great idea. The French tried this after the Revolution. They also came up with a "metric" calendar: 12 months of 3 weeks of 10 days, with the extra 5 days per year (6 on leap years) being a holiday. The holiday was taken at the end of the year. Sort of like between Christmas and New Years now.

http://en.wikipedia.org/wiki/French_Republican_Calendar

It seems the French have more than just good wine to offer!

The French tried something similar after the revolution (Decimal Time). It didn't catch on either.

It's essentially pointless. Yes, it's a bit easier to represent times and handle them differently (5:50 = 5.5h = 550m) but it's not that hard with standard time either, as evidenced by the fact that I and everyone else in the world managed to learn to tell the time about the same time we figured out how to tie our shoelaces.

Actually it would be interesting to know whether they aborted it. I guess it was "too much use of 24/60/60 format" problem.

I do though think that after we get used to SUT, our lifes will be much easier.

Its unlikely we will ever changing our system of measuring time until we find ourselves living on other planets, at which point the standard clock will feel inadequate and require a more universal solution. In terms of earth though there simply isn't enough of a catalyst to make the change.

Actually telling the time to somebody else can be even more consuming. Some languages say half past 4, some half to 5, some even say "quarter to 5" when it's 16:15 and "three quarters to 5" when it's 16:45. In some languages you can say stuff like "3 minutes before 3 quarters to 5" for 16:42. People just like to make things interesting, I guess.

</OT>

I guess these cultural differences will continue to exist - you even have 3 different versions for 1645 in German language.

I don't think this is a problem, as cultural differences are important to exist, as they resemble local differences.

The problem is that 100-minute hours isn't better than 60-minute hours by one particular metric: divisibility. 60 is highly composite (divisible by 2, 3, 4, 5, and 6), and one of the most common operations performed on an hour is division. Perhaps dividing an hour by 5 isn't particularly common, so 24 would have worked -- as such it's not surprising that there are 24 hours in a day.

24/60/60 weren't chosen willy-nilly, and replacing them with 10/100/100 isn't an improvement.

The nomenclature invites unnecessary confusion, as the terms are all already used ("How many simple seconds are in an actual second?").

Why not use metric prefixes on a day to come up with the names? This requires no new definitions, just a new application of existing terminology.

simple hour = deciday

simple minute = milliday

simple second = 10 microdays

If I travel at a constant speed of 100 km/h (~62mph), it takes me one hour to travel 100km (~62 miles). This is not even math.

Do you want me to

a) convert between km/h and km/"simple h"

b) replace all tachometers, speed-limit signs and laws?

c) define "simple km" to fit?

Oh, it's 5:98:67 now - have to leave work before 6 SUT today, though usually leaving at 6:30.

When do you usually leave work in SUT?

Just getting rid of DST would solve a lot of what this is trying to accomplish.

I agree - that was one of the initial things we were discussing when developing SUT.

This is much cooler than Joseph Franklin's Metric Leisure Week proposal.

My computer just went from 3 GHz to ~2.6 Ghz... This may not fly... ;)

Reply on news.ycombinator.com