My five-year experiment with UTC
timestripe.com134 points by adamci 10 months ago
134 points by adamci 10 months ago
IMO the best argument for time zones is it keeps calendar days roughly[1] aligned with “natural” days, such that day changes happen when it’s dark outside and most people[2] are asleep and most people won’t be entering a new day during a work meeting.
Why? Because it is confusing if most people’s natural days are divided into different calendar days. If I wake up and the calendar says Tuesday, then I can be sure that I don’t need to worry about the dental appointment on Wednesday afternoon. But if Wednesday starts midday, then I have to spend extra brain cycles to work out which parts of my natural day belong to Tuesday and which parts are Wednesday. It’s a lot of hassle avoided by having days change over when I’m asleep.
If you want real life example of such confusions, look up “early morning flight confusion” on your favourite search engine. It turns out a lot of people are confused by flight times like “Wednesday 1:20am”, because if you are catching an 1:20am flight, you are heading to the airport on the previous day before midnight, and mentally the flight feels like a part of the previous “natural” day. What sometimes happens is people will book a flight on Wednesday 1:20am, head to the airport on Wednesday evening because they think “well the flight is on Wednesday right”, and find out that it is actually Thursday after midnight.
By the way, one successful attempt at addressing this problem is Japanese late night anime schedules. If an anime is aired on Wednesday 1:20am, the TV station will instead write “Tuesday 25:20” on the schedule. It makes no sense from a technical point of view, but feels right for the human because what is the early morning of Wednesday if not just Tuesday night prolonged?
[1] Even extreme cases like Spain are only a couple of hours out of sync, it’s not like the sun shines on Madrid during “midnight”.
[2] the article mentions remote workers and frequent travelers, which I am willing to wager are a tiny (but over represented in nerd spaces) fraction of the society.
Incidentally, this works for someone close to the UTC time (3h off for Moscow) for the reasons you mention.
It likely does not generalize at all.
They also haven't faced an issue where an event is booked by someone else at a DST-affecting time: as that time changes between two days, you now need an extra "if" in your brain (is it DST there? Europe and US are on different cycles there too).
I've tried booking all the cross-timezone meetings in UTC, but didn't get far with it: it was more work for everyone compared to somewhat reduced confusion and suddenly overlapping meetings.
> It turns out a lot of people are confused by flight times like “Wednesday 1:20am”, because if you are catching an 1:20am flight, you are heading to the airport on the previous day before midnight […]
If you use 24-hour ("military") time—00:00 to 23:59—this becomes less confusing. Perhaps an attempt to culturally shift from using 12-hour to using 24-hour time would be useful. (In the US it'd probably less of a challenge than metrification.)
[1] https://en.wikipedia.org/wiki/Date_and_time_representation_b...
[2] https://en.wikipedia.org/wiki/12-hour_clock#Use_by_country
Maybe it’s less confusing in practice but it’s not clear to me how in theory. It seems like the core issue is still present: an event so early in Wednesday’s morning feels like an extension of Tuesday night so one has to remember “Tuesday before I sleep” when the event details say “sometime Wednesday”.
> one successful attempt at addressing this problem is Japanese late night anime schedules
Depending on the channel, the "new date cut-off" (more accurately: 日付変更) sometimes varies. Some channels start the day at 5am, others at 4am, and so on. Because of this, one gets to see wild things, such as a time-table scheduling a show at "29:30" whereas another channel would write "05:30".
In any case, your claim is correct. The people who stay up to watch late-night shows do not consider their day finished around those hours, and the times are intended to reflect that. However, there is still inconsistency around the hour of the new day.
A similar issue exists in video games. I know of one game that "starts the new day", so to speak, at 5am, and another game that does so at exactly 12am. To add further confusion, many Japanese websites and games are designed with the assumption that the user is always in Japanese Standard Time. (Most Japanese do not travel abroad, and Japan itself has exactly one time zone, thus Japanese programmers rarely ever think about time zones). Thus, when overseas, you may see the new day in a video game start at 1pm local time!
> because what is the early morning of Wednesday if not just Tuesday night prolonged?
This is what really grinds my gears with the official definition of "night" and "morning".
Tuesday 21:00-23:59 is considered "night", and then Wednesday 00:00 is considered "morning", which is highway insanity to me. No one would say they were going out 'till "Wednesday morning" if they went to the pub until 02:30 Wednesday.
IMO it only starts being "morning" around 05:00, maybe even 06:00.
Due to the tilt of the earth the length of a solar day is not constant (unless you're very near the equator along a specific, probably wobbly, line).
Even now, obtaining solar noon is comparatively easy. Then add half of an earth's rotation* (slightly affected by the earth's orbital period too), and that's midnight.
In the days of candle light, midnight wasn't a crazy time to draw that diving line. However these days even people with crazy early AM weekdays to avoid commutes sometimes cross over midnights on the weekend.
+
You are correct that the modern world really should just start counting the day at like 3 or 4 am, and continue counting past 2359 to almost 2700 or 2800.
Haha, don't get me wrong, I am not asking for official days to be anything but 24 hours.
I guess I am arguing that "morning" "midday" "evening" and "night" should hold no official definition, since they are much more about the vibe rather than an absolute position on the clock. However, if they do remain as official definitions, they should be heavily revised.
It's usually morning after you sleep.
"I woke up very early in the morning at 4am to start work at 5."
I like to say things like "13 in the morning" to say "1pm but I just woke up".
Its possible your example is wrong and the confusion arises in the "twelves" (i.e. 1230 AM being 0:30 which makes no sense since anyone that didn't learn the quirk would expect the AM/PM bit to roll over at the same time as the hours, not one hour off the mark).
I believe all these problems are caused by time zones, not solved by them. If the current date/day of week wasn't vaguely connected to sleep/wake cycle, we wouldn't be conditioned to assume it was.
For instance, we handle 24 hour clock (with 60 minute hours?! WTF?) fine, even if it's insane and we use the decimal system elsewhere. Because we got used to it when we grew up. Many even have the clock reset in the middle of the day, ending up with things like two hour activities spanning from eleven to one! (Don't get me started on shit like "half four" - it means different things depending on where you are, but it never means two!)
Heck, there's a whole, huge country that gets by on inches and ounces and pounds and miles and whatnot - a mind boggling amount of extra mental energy required, and with the exception of some recent lapse when selecting their boss dude they seem to be doing mostly as well as (and over the previous century arguably somewhat better than) the rest of us in the thinking department.
If we grew up with date/day changing in the afternoon, every kid in preschool would have it figured out already. "Early morning flight confusion" would seize to exist (except, perhaps, for the poor bastards who'd happen to live close to wherever we decided to place the prime meridian), because everybody would know what "1:20" meant, and not randomly guess at something based on when they happen to sleep.
Yes and no, locally it would work because that's what you grew up with, but traveling, and especially communicating would be equally confusing. Currently we have that problem on globe-spanning video calls where "let's schedule another meeting tomorrow" could mean very different things. In the global utc scenario, tomorrow would still mean different things to someone growing up with "midnight" during the day vs. When you're asleep.
Again, I believe the problem you're talking about is caused by time zones.
If you're used to date changes happening "whenever", you'd stop assuming they happen at a fixed point during the day.
Any argument that starts with "Today, we're already confused about ... because we're used to date changes occurring while we sleep" is meaningless, since we wouldn't be used to date changes occurring while we sleep.
The whole terminology thing would be solved by having two words; one meaning "after next sleep cycle" (which would probably be more common when speaking to people in your own geographic region), and one meaning "after next date change" (which would probably be more common when speaking to people in other geographic regions). The appearance of this second word would likely happen naturally within weeks, if not days after abolishing time zones. As usually happens with language.
The argument against abolishing time zones always seems to assume that humanity would be incapable of adapting to a time zone free world. Like this would mysteriously be the first and only change in the history of mankind society couldn't adapt to.
> If you're used to date changes happening "whenever", you'd stop assuming they happen at a fixed point during the day.
But that's not the case. We're not all cosmopolitans traveling the world all the time. For anyone living in a timezone where 00:00 happens when most people are asleep, there'd only be one word for tomorrow, as it'd be synonymous for "after date change" and "after the next sleep cycle". There would be no other word, because there is no need for that. Except in circles where coordinating meetings with people across the globe happens on a regular basis. But we already got that figured out by explicitly naming the time zone or using UTC.
The vast majority of people don't have to deal with this, so having the date change at night is just more convenient, as it completely avoids this ambiguity in day to day life.
But obviously as a nerd you get hung up on that one aspect and go "but imagine we had only one timezone, we could just say 13:00 and there'd be no ambiguity, no chance for a misunderstanding, it would be so much more efficient and coherent!"
> The argument against abolishing time zones always seems to assume that humanity would be incapable of adapting to a time zone free world.
Again, you need to take a step back and try to look at this from the average person's pov, not as a scientist trying to find the optimal solution for a problem 99% of people don't have. And "people would be able to adapt to this" Is a pretty weak argument. You could probably also change the clock itself so that there's 63 seconds in a minute, 59 minutes in an hour, and 195 hours in a day. I'm sure people could adapt to this. Doesn't make it a good argument.
On the other hand, just imagine the sheer chaos this switchover would cause worldwide, for what, marginal gains in convenience for a miniscule amount of people.
It reminds me of "that guy" who constantly wants to refactor half the codebase because something would look a little cleaner then, and if you let them, they'll end up refactoring the same part of the code over and over again every six months because they had another idea how to make it even more streamlined and elegant. And then someone else comes in wanting to implement a simple feature, but is greeted with an incomprehensible construct, while the refactoring-guy is repeatedly and excitedly reassuring them that it's "really simple and intuitive once you get the hang of it".
> But that's not the case. We're not all cosmopolitans traveling the world all the time. For anyone living in a timezone where 00:00 happens when most people are asleep, there'd only be one word for tomorrow, as it'd be synonymous for "after date change" and "after the next sleep cycle". There would be no other word, because there is no need for that.
Meanwhile, I know what the word desert means, even if I have never been in a country that has one. I know what a cyclone is, even if I've never seen one. We're affected by, and have knowledge of more than our local environment. We read books. We watch TV and movies. Heck, we're having this discussion in a language that that isn't spoken in the country I'm living in. Most people do, in fact, interact in some way with people in other time zones right now. Perhaps not on a daily basis, but at least a bit.
It just occurred to me; if we placed the meridian in the Pacific, very few people would have the date change during their sleep cycle. Sucks to be Niue, I suppose ;)
By the way - I'm not even for abolishing time zones. I'm just spectacularly tired of the straw man arguments against it. It's like saying "we couldn't build a city that's on both sides of the river, because people would drown when they go to work", as if building a bridge would be an impossibility.
There is a perfectly good argument against it btw, which is the one you propose at the end. The change itself wouldn't really be worth it. But that's simply because people are confused by changes. Not because the end result would be worse.
In that particular case, "another meeting tomorrow", I think my expectation is pretty much not affected by time zones. I expect it to be 24 hours from when the current meeting started, plus or minus 2-4 hours.
i mentally translate tomorrow in a business meeting usually means sometime in the next 8-40 hours...
You argue that any system would be easier if we had all learned it in preschool, then turn around and claim some massive, mind boggling amount of effort that must be expended by Americans to use ounces, miles, and all sort of other yucky units you might need a lot of energy to use, but Americans have been learning, alongside SI, since preschool. Just like how you don’t likely struggle with 7 day weeks of 24 hour days with 60 minute hours.
So which is it?
I don't get the flight example? Wednesday 1:20 AM = Thursday 1:20 (24 hr notation)?
Anyway it seems to me this whole problem arises with 12 hour notation.
I switched permanently to 24hr notation after my buddy asked me to pick him up from the airport at "one-o'clock on Thursday". No problem, I didn't have any classes Thursday afternoon, so "see you then". You see where this is going? At 0-dark-thirty I get a call: "dude, where are you?" Yeah. I still give him shit about it, thirty years later.
24hr time, folks. 24hr time.
Wednesday 1:20 AM (12 hr notation) = Wednesday 1:20 (24 hr notation), or is there another convention I’ve never heard of? Either way, this seems to further validate the original confusion.
> book a flight on Wednesday 1:20am, head to the airport on Wednesday evening because they think “well the flight is on Wednesday right”, and find out that it is actually Thursday after midnight.
This bit here, I don't understand what Thursday is supposed to do here. If my flight leaves Wednesday at 1:20 AM, I'm leaving for the airport Tuesday evening.
But if I'm interpreting OP correctly, people are booking flights for Wednesday 1:20 AM but leaving for the airport Wednesday evening? Why would you be confused about that?
You don't see why people would confuse "Wednesday in the middle of the night" with "Wednesday night"?
But the ticket doesn't say "in the middle of the night" or "night", it just has the date and time.
People read 1am and then they think and remember "middle of the night"
So if you happen to think of it as "middle of the night, specifically 1am", and you fall into mental shortcuts a little bit too hard, oops you might end up there on the wrong day.
The correct 24h notation would be: Wednesday 01:20.
By the way, if you’re used to 12h time, you can get a similar experience to that described in the article simply by setting your watch/phone to 24h time. After a while your brain just starts to recognize 15:00 and 3pm as the same thing, and there’s no explicit conversion required for you.
Wednesday 01:20 doesn't really help. They weren't accidentally showing up in the afternoon. The passenger knows it's the middle of the night.
Tuesday 25:20 has a much better chance of helping.
If we’re not accustomed to this “overflow” notation, it only looks like a typo and an error. In my world, the day-of-week begins/cycles at 00:00, so don’t tell me that is happening on a Tuesday!
It also does not help for entry into a programmatic form or database field. That sort of notation is only good for output to a human who knows this custom.
That's why I said chance of helping.
But the important part is that 01:20 does not solve the problem.
Well, I'm interested to know how you define "the problem" then, and what "solving" it would look like, and why that "solution" can be approached by unfamiliar notation.
The problem is that 01:20 on Wednesday is on Tuesday night, which trips people up.
There is no perfect solution, so I can't tell you what solving it would look like. But I'd say that putting "(Tuesday night)" next to where it says "01:20 Wednesday" would be helpful to many of the people making this mistake.
How would you define the problem? It's not people mixing up 1:20am and 1:20pm.