But does solar time matters nowadays? In modern day-to-day life noon is not 12:00. People wake up at 5-6-7. And go to bed at 10-11-12. Even for early risers, middle of the day is well past 12:00. Having DST time makes solar mid-day closer to factual mid-day. I'd even argue to for +2 hours DST :D

Yes, solar time matters as your biological rhythm depends on it. The most important part of it being able to have daylight when you either rise in the morning or at least have the sun get up quickly after that. Another important point is, that there is plenty of experimental evidence that shows that starting schools before the pupils are really awake is not effective.

It is a misconception, that "noon" is the middle point of your awake time. That was never the case. Our day usage most of the time is asymmetric: start the day and especially the work with the sunlight, spend the evening time off after work after sun set.

Using the high point of the sun as the reference for the daytime is a very practical choice, as this is the one easily observable time point which doesn't move much over the course of the year (a few minutes, but that is symmetric).

Do my biological rhythms know when is the middle of solar day? Or maaaaybe the stable day regime?

In my whereabouts sun raises at 3am around midsummer (it would be 2am without DST) and at 9am around christmas. Nowdays I wake up at 7am around the year and somehow my biological rythm is fine. It used to be just fine when I waking up at 10am too. The core thing seems to be stable day rythm, not adhering to solar time.

What you're saying is sunrise matters, not solar time. Although even for sunrise, I'd say students' daily routine matters much more. Back in high school, when I spending my late evenings and nights on computer, I didn't care if it is bright or dark in the morning, i was sleepy anyway...

I'd also argue that in today's sedentary world luring people outside is very important. And bright spring/summar/fall evenings is very important at that. There's plenty of evidence for the benefits of exercise, socialising and so on. And with forever-non-DST time, we would loose that for couple weeks improvement in the morning.

IT matters when you have an economic zone in which some have 9am being 'basically mid-day', while the others have '9am being 'oh, the sun just went up'.

This traditionally has been solved in two ways:

1. do the split and accept that you will have costs associated with different time zones and sleep/wake cycles in your population (e.g.: USA / Soviet Russia)

2. don't do the split, and have the cost by putting a time/light adjustment burden on the population (e.g.: PRC, European Union)

My preferred solution would be to abandon all time zones and use something like UTC, or yet better: .beat / Biel Mean Time [1], which does not come with a 'oh, 12 means midday' penalty.

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

> My preferred solution would be to abandon all time zones and use something like UTC

Which means that

a) textual time references in stories would stop working unless also specifying a location, which would cause a lot of weirdness, especially in fiction

b) perhaps more importantly for day-to-day life, it causes a lot of weirdness around the change of the calendar day and the weekday. It's a rather useful property of the current system that the change of calendar day happens at a time when most people are asleep or otherwise not too fussed about the exact calendar date.

If you completely abolish time zones, that means that for everybody not lucky enough to live near wherever the meridian ends up (which would always be a significant fraction of the world's population, no matter where you'd place the meridian), the change of calendar date would happen during the middle of the waking hours, which causes a lot of weirdness (do public holidays now actually start and stop in the middle of the working day?) and makes date references vastly more cumbersome because a single calendar date can now refer to both solar n in the afternoon/evening and solar date n+1 in the morning.

So you'd have to start using rather awkward double dates for referring to a certain solar day ("the event will take place from the 22/23rd to the 24/25th", "see you on Thursday/Friday" [which isn't the same thing as "see you on Thursday or Friday" today, because that'd become "see you on Thursday/Friday or Friday/Saturday" under the new system properly]), or else develop some convention based on the local solar noon/midnight, which is basically reintroducing time zones through the back door.

That only is a problem if you don't separate the 'time number' from the observed local time.

You could easily have a 'multiple calendar days during one solarday' situation. In fact, many countries already do this for their TV programming, in which a show might screen on "Thursday at 23:55 until 03:15".

For common things, you could just refer to as 'tomorrow morning'. If you need more precision, you can fall back to .beats. "School starts @740" Incidentally, that's exactly why I liked .beats: no false associations.

But for most people those associations are very useful and actually a desirable property.

Besides, even if you switched to some new, completely-blank-slate system of tracking time, those associations would quickly redevelop since presumably most people will predominantly deal with local events within their own time zone. Only now with added confusion when you do encounter a foreign time reference after all…

The thing is, those "global time" proposals don't change anything about the fact that you have to look up local time and/or local customs when scheduling events in foreign timezone – you have to do it today, and you still have to do it even with a single global wall clock time.

The only thing it does – resolve some potential (potential, not definite!) ambiguity around what (wall clock) time exactly cross-time zone events are taking place, but at the price of completely breaking any association between clock time and the solar day. Personally, I don't think that's worth it.

> "School starts @740"

Which precisely illustrates the problem – because you didn't mention a location, that sentence is now completely meaningless. If I encountered it in a random story – does it mean that it is in the morning (but early morning or late morning or what?), is it in the afternoon…?

Like I said, I posit that most people regard the association between clock time and solar time as an advantage, not as a drawback.

> You could easily have a 'multiple calendar days during one solarday' situation.

I didn't say that this doesn't already happen today, but today it affects far fewer people in mostly less important ways.

You're wrong. EU already has the split.

China is an interesting example as well. With unofficial local times in far-away provinces.

If solar time does not matter, why do you care about at when number you get up, go to work, come back and get back to sleep? This would be even more ridiculous. People care about the time, because it's a proxy for sun.

People care about time because it's a proxy for pretty much any common activity.

Most people start and endworking at common time. If you want to spend time with other people for hobbies or socializing, you need to find a matching spot when people are done working.

Sure we can stick to solar time and adjust office hours instead. But then we're back at square 1...

But then you don't care about the position the number of a specific time occurs in. 16:00 could be at any time. The reason to lock it at at a specific position of the sun is because of the position of the sun. That's what I meant with "why do you care...".

Technically yes. We could switch to perma-non-DST and just move all office hours one hour earlier. And I'd be fine.

But I'm pretty sure perma-non-DST people would not be happy with such solution.