More DST Fun
“This is like Y2K except this one is really happening,” said [Purdue University] IT spokesman Steve Tally.
Currently, most Indiana computer users set their PCs to a special “Indiana East” setting — Eastern time that doesn’t spring forward every April. Starting this April, however, they’ll change their PCs to Eastern Daylight Time. The few who observe Central time set their computers to Central, and will also make the switch. Tally predicts the changeover will create havoc with the widely used Microsoft Outlook calendar application. When the time changes, he said, appointments will still be listed according to the old Indiana East time. The calendars of Central time Outlook users, in turn, will continue to list appointments according to Central time.
With a nationwide shift in daylight-saving scheduling slated for next year, Indiana’s experience offers a preview of potential glitches in store for the rest of the country. Starting in 2007, daylight-saving time will begin on the second Sunday of March rather than the first Sunday in April, as it does today. Daylight-saving time will end the first Sunday of November, a week later than it does now.
I heard on the radio yesterday that computer technology actually plays a much bigger roll in the growing gap between high-wage and low-wage employees than does immigration, such that those opposed to immigration should also be opposed to computers. I suppose one could look at the legislature mucking around with timekeeping as a way of creating demand for IT jobs, and thus slowing, ever so slightly, the rate at which IT efficiency disempowers low-wage workers.
My favorite DST bug was when Windows first started doing DST compensation automatically. The first time Windows computers were trusted to “fall back” an hour there was a bug such that several computers set their clocks back, and then set their clocks back, and then set their clocks back again . . . some computers ended up three, four, five, six hours behind . . .
Unix computers, as a rule, keep heir own time in UTC, and then calculate time for the user’s locale based on their current offset from UTC. As far as the computer is concerned, the past always occurs before the present, which always occurs before the future. Time does not change, except for small corrections to compensate for skew. The computer can represent time in whatever bizarre-o way mandated by the user, assuming the files defining “human time” are up to date with the latest legislation.
The fun part comes when you have a computer like my workstation, that can boot into Linux or Windows. Windows stores time as “local” time, and has no proper concept of Universal time. It just resets its clock based on local rules. For this reason, I tell Windows to use UTC as its time zone, so it never adjusts the clock, and the clock is always right, even if it is not my local time zone.
Anyway, the whole DST issue wasn’t such an issue back in the “analog” days, because back then, time was a pretty fuzzy concept anyway. But ask yourself: what time will it be precisely one month from now? If you look at your watch and your watch says 3pm, and you answer that a month from now it will be 3pm, then you may be wrong, because there’s a good chance that one month from now 3pm will actually be 4pm. When the time changes in Indiana, and you ask yourself what time was it exactly one year ago, your question will be different from most of the Eastern time zone because last year, Eastern time changed but not parts of Indiana. Further, Congress has recently changed when DST comes and goes, such that we have to recalculate what time it would be this time next year, the year after, and so on.
The best way to solve a problem is to not have it in the first place. Before the advent of railroads, time zones didn’t exist–people looked at the sky and figured that noon was when the sun was overhead. Once we were able to move across the planet quickly we found it difficult to coordinate schedules because time was different as you moved East-to-West, so we adopted standard time zones, one hour apart. Now, noon is within an hour of when the sun is in the middle of the sky, unless its DST, when noon comes an hour before the sun gets to the middle of the sky, give or take an hour, or more, if your time zone has been shifted for political and economic reasons.
At my company, our employees typically start work anywhere between 7am and 10am, and go home between 3pm and 6pm. And they are spread across three time zones. And when I fly to DC to meet my boss, the five hour flight actually takes all day. I like to think that now that we can easily compute offsets from a single standard time source, that we will eventually start telling time this way. I like to think that when my grandchildren are born, the time will be recorded in UTC. And they will be perfectly comfortable with a notion that the sun might rise at 1800. They’ll find it perplexing when we talk about the dark ages of the twentieth century when people had to keep doing math to figure out what time it was in different parts of the world, even though the time bore only a fuzzy resemblance to Solar Time anyway. They’ll feel some gratitude that their watches tell them the correct solar “human” time for whatever longitude they may be at, and also remind them when they have a meeting scheduled in the standard “machine” time.
Alas, my crackpot philosophy of time is ahead of its own time. Patience . . .