From a development perspective it certainly sounds easier to have one global timezone with DST than a bunch of smaller ones without it. Would that make sense in reality? Probably not but I definitely think timezones take more work to compensate for properly.
Lets just have 2 timezones, Chinese time and EST w/ permanent DST. The most populated timezones for Eurasia and the americas, and they’re both 12 hours apart, so nobody has to do timezone math, just swich AM and PM.
There was actually a really interesting idea I heard to have no time zones. And I actually think it could be a good idea. It’ll never happen because people would need to re-learn time but if it was always the same time everywhere it would make scheduling and business so much easier. No one would need to convert between different zones or be late because of an incorrect conversion. The downside is that times which are conventionally morning or evening etc, would no longer would be so people would have to get used to time just being a construct for scheduling and not a representation of the natural day/night cycle…but it actually doesn’t sound like a half bad idea.
Problem you run into is the areas where we need to tie things to solar days across an area.
You end up with places having to regulate that school starts at 22:00, and gets out 05:00 the next day.
Businesses close for the night at 06:00 and open bright and early later that day at 22:00.
You have places where one calendar day has two different business days in it, so the annoyances faced by people who work overnight shifts spreads to everyone, and worse gets spread to financial calendars, billing systems and the works.
Time is an air bubble trapped under a screen protector. It’s annoying, and you can push it around to try to keep it out of the way, but you can never really fix it.
There’s just too many inherently contradictory requirements for us to end up with a “good” system, and we just need to settle for good enough.
My dream is that we stop changing things. Whatever we have in time zone database today is what we stick with going forwards. No more dst shifts, no more tweaks to the zones, no more weird offsets and shifts, because we don’t get to stop dealing with the old layout when we change, we just add a new one that we think is better.
For the most part, dealing with this stuff is a solved, shitty problem. It’s when we change the rules that problems come up. Worse when we change them retroactively. (Territory disputes between nations have been resolved with the conclusion that land was actually in a different time zone in the past because it was actually in another country. Not a problem usually, unless there’s a major stock exchange in an island that was transferred between nations and retroactively changing what time it was affects what laws were valid at the time certain transactions took place.
Not really. Timezones, at their core (so without DST or any other special rules), are just a constant offset that you can very easily translate back and forth between, that’s trivial as long as you remember to do it. Having lots of them doesn’t really make anything harder, as long as you can look them up somewhere. DST, leap seconds, etc., make shit complicated, because they bend, break, or overlap a single timeline to the point where suddenly you have points in time that happen twice, or that never happen, or where time runs faster or slower for a bit. That is incredibly hard to deal with consistently, much more so that just switching a simple offset you’re operating within.
What matters is consistency and our time system has tons of crazy inconsistent shit in our. Everyone knows about leap years, but do you know about leap seconds? Imagine trying to write a function to convert unix time to a current date and suddenly all your times are a second off.
From a development perspective it certainly sounds easier to have one global timezone with DST than a bunch of smaller ones without it. Would that make sense in reality? Probably not but I definitely think timezones take more work to compensate for properly.
Lets just have 2 timezones, Chinese time and EST w/ permanent DST. The most populated timezones for Eurasia and the americas, and they’re both 12 hours apart, so nobody has to do timezone math, just swich AM and PM.
There was actually a really interesting idea I heard to have no time zones. And I actually think it could be a good idea. It’ll never happen because people would need to re-learn time but if it was always the same time everywhere it would make scheduling and business so much easier. No one would need to convert between different zones or be late because of an incorrect conversion. The downside is that times which are conventionally morning or evening etc, would no longer would be so people would have to get used to time just being a construct for scheduling and not a representation of the natural day/night cycle…but it actually doesn’t sound like a half bad idea.
Problem you run into is the areas where we need to tie things to solar days across an area.
You end up with places having to regulate that school starts at 22:00, and gets out 05:00 the next day.
Businesses close for the night at 06:00 and open bright and early later that day at 22:00.
You have places where one calendar day has two different business days in it, so the annoyances faced by people who work overnight shifts spreads to everyone, and worse gets spread to financial calendars, billing systems and the works.
It’s not better.
🤔 that’s a fair point…
Time is an air bubble trapped under a screen protector. It’s annoying, and you can push it around to try to keep it out of the way, but you can never really fix it.
There’s just too many inherently contradictory requirements for us to end up with a “good” system, and we just need to settle for good enough.
My dream is that we stop changing things. Whatever we have in time zone database today is what we stick with going forwards. No more dst shifts, no more tweaks to the zones, no more weird offsets and shifts, because we don’t get to stop dealing with the old layout when we change, we just add a new one that we think is better.
For the most part, dealing with this stuff is a solved, shitty problem. It’s when we change the rules that problems come up. Worse when we change them retroactively. (Territory disputes between nations have been resolved with the conclusion that land was actually in a different time zone in the past because it was actually in another country. Not a problem usually, unless there’s a major stock exchange in an island that was transferred between nations and retroactively changing what time it was affects what laws were valid at the time certain transactions took place.
Not really. Timezones, at their core (so without DST or any other special rules), are just a constant offset that you can very easily translate back and forth between, that’s trivial as long as you remember to do it. Having lots of them doesn’t really make anything harder, as long as you can look them up somewhere. DST, leap seconds, etc., make shit complicated, because they bend, break, or overlap a single timeline to the point where suddenly you have points in time that happen twice, or that never happen, or where time runs faster or slower for a bit. That is incredibly hard to deal with consistently, much more so that just switching a simple offset you’re operating within.
What matters is consistency and our time system has tons of crazy inconsistent shit in our. Everyone knows about leap years, but do you know about leap seconds? Imagine trying to write a function to convert unix time to a current date and suddenly all your times are a second off.
Just look at this insane bullshit nonsense. The added complexity of time zones and daylight saving time is nothing compared to simply supporting our time system.
We need to synchronize all computer times with that one clock that can stay accurate to within 1 second every 40 billion years.
We do? With NTP
I’m referring to this one, the most bleeding edge of accuracy. I don’t think NIST would have implemented this particular clock (yet).
Incredible list, the scale.
hmm
A little aspirational?