Log in

No account? Create an account


DST is silly

« previous entry | next entry »
1st Apr 2008 | 10:45

Around the end of October I read David Prerau's book "Saving the Daylight". It changed my opinion of DST somewhat. I used to think that it was unmitigated stupidity, and that everyone should make seasonal adjustments to their timetables instead, or something. (For example, my secondary school had afternoon lessons just after lunch in the summer, but later after tea in the winter.) What I hadn't considered was the human-factors effects that this anarchy would have. In the USA for much of the 20th century DST was a popular idea but it was not consistently enforced by governments at any level. The result was prolonged chaos, documented at length in Prerau's book. DST only works well when applied consistently over large areas, and enforcing it by law is the best way of achieving it.

I'm not entirely convinced that the goals of DST are foolish. I'm by nature the kind of person that Benjamin Franklin and (with less humour) William Willett wanted to persuade to get up earlier. But I like long summer evenings, and it's hard to make the most of them without awkward winter daylight times if your clock is fixed relative to mean solar time.

However I think mean solar time is a fetish that should now be discarded, especially in the guise of UTC. Leap seconds are an abomination: civil time does not need to be synchronized that closely to the Earth's rotational angle. Ideally we'd just define civil time as a timezone offset from atomic time, without the intermediate layer that keeps the astronomers happy. (The UK abolished the RGO because responsibility for timekeeping had moved from the astronomers to the physicists. The rest of the world hasn't yet made the same break with history.)

Having abolished unpredictable leap second disruptions in our timekeeping, we would still be dependent for stability on politicians not arsing around with the timezones. They have proved themselves incapable of keeping their grubby hands off. Perhaps the solution is to make a further break from the noon-oriented mean solar time fetish and use a more practical benchmark as the basis of civil time: sunrise. If the sun always rises at 07:30, then we get nice long summer evenings, sensible daylight in winter, and a time system based on natural philosophy not political tinkering. Actually, rather than using local sunrise as your benchmark, you want northern, southern, and perhaps tropical sunrise time zones. For example, civil times in the northern hemisphere might be defined by one-hour offsets from local sunrise time at the intersection of the Tropic of Cancer and the Greenwich Meridian. Perfect!

So about five months ago I wrote a brief description of how this could work. But I quickly realised (with Prerau's book in mind) that the practicalities of persuading people to put this into practice made the idea ridiculous. So it became an April 1st RISKS Digest item, which I kept under my hat for all that time. I was greatly amused by simont's recent comment advocating something very similar, though I couldn't follow up until today :-)

| Leave a comment |

Comments {5}


from: cartesiandaemon
date: 1st Apr 2008 11:40 (UTC)

It's amusing how whimsical good ideas and pretend real ideas are so similar. This plan is awfully, awfully, seductive (apart from not working).

Reply | Thread

The Bellinghman

from: bellinghman
date: 1st Apr 2008 12:09 (UTC)

Not working shouldn't be taken as a reason for refusing it though. In my candid opinion, DST doesn't work either. (As may be evinced from the existence of double summer time and the like).

Reply | Parent | Thread

A zero-width positive lookahead assertion

from: pseudomonas
date: 1st Apr 2008 12:11 (UTC)

This reminds me somewhat of the (I think) Roman system, where you have 12 hours of daytime (by slicing dawn-dusk into 12 equal chunks) and twelve of night-time, likewise defined. This means that local noon is always the same time, 6 hours after dawn - just the hour lengths vary. I may have misremembered the particulars.

Reply | Thread

Tony Finch

from: fanf
date: 1st Apr 2008 12:13 (UTC)

My plan is to keep uniform-length SI seonds (and minutes and hours) but adjust the timezone offset each night, as we do for DST changes. For example, today's timezone is +00:52, as indicated in my RISKS item.

Reply | Parent | Thread


from: sion_a
date: 2nd Apr 2008 09:45 (UTC)

This would, on reflection, solve three problems we've been experiencing this week:
  1. I've gone back to my winter difficulty of getting up in the morning, because there's suddenly an hour's less daylight between sunrise and when I'm supposed to be getting up.
  2. 1ngi is having trouble getting to sleep, because there's suddenly an hour's less darkness between sunset and when we're trying to get to sleep; and most significantly
  3. Parsley has stopped coming in of his own accord before we go to bed because, being a cat, he's operating off solar time and doesn't know that "bedtime" is suddenly an hour earlier.

Reply | Thread