radtea wrote:First, anyone proposing a "simple" solution to a problem that has remained irritating and unsolved for decades is missing something.
Often what they're missing is a historical accident that resulted in self-defeating criteria.
So for example, most people didn't care about synchronizing clocks until the railroads came in. People liked the idea of having 12:00 noon come when the sun was highest, wherever you were. But railroads wanted to run schedules that would work across the country, and they wanted to be able to tell people in each railroad town when the train would leave. They each used their own time system. In the USA somebody who was in a central position for communication among railroads suggested a system they could all follow, and enough of them did that everybody else fell into line. Having time zones 1 hour apart meant that nobody had to have 12:00 noon be more than half an hour off, in theory. It was an ugly compromise that worked adequately. Then state governments got involved, etc.
Any time system that tries to provide consistent times and yet also match the days and seasons, must have ugly compromises. And that's why the problem has been irritating and unsolved for decades.
There is a simple system available to provide consistent time all around the world. Set an arbitrary beginning point and then count seconds. The size of a second is also arbitrary but you can use a second that is in common use. Count up seconds until you get an overflow, then start over. A clock that can keep more bits will be different in the high bits that you don't care about. There's no particular reason to ever have a leap-second.
If you care whether it's night time where somebody else is, then figure out the difference in longitude. It has nothing to do with the clock.
If it's sunrise now and you want to know when sunrise will be tomorrow, it might be convenient if the number of seconds it takes for that is approximately some nice round number. But it isn't a great big deal.
Why not do something like that? Because what we already have is clumsy and inconvenient, but not bad enough that we feel much need to change it. China covers 5 times zones but they say they are all the same time. It only causes them inconvenience when bureaucrats insist on uniform behavior. There's no particular necessity for a department store in Urumqi to open at 8 AM Beijing time, unless somebody says it has to.
Second, only that very small fraction of computers connected to the 'Net are addressed by your scheme, leaving out all the vast multitude of embedded systems, including the ones in watches.
Those are sunk costs. Use your wristwatch until its battery goes bad. Then throw it away.
I'm generally in favour of 12 months of 30 days each with a 5 day Saturnalia at the New Year, but I recognize it's never going to happen. Calenders have almost nothing to do with time--which is handled quite nicely by Julian Day Number--and everything to do with human convenience, and "what we have now" is almost always more convenient than "what we would really like to have plus all the work necessary to get there."
Yes. But after some giant catastrophe, or after the revolution (practically the same thing) would be a good time to change. They did it after the French revolution. We might as well spread the ideas now, in case some survivor remembers.
