On 2026-11-18 08:15 AM, James Bellaire wrote:
On 11/18/2026 7:06 AM, Brooks Harris via tz wrote:
I can't help but think "PDT", "MDT" and "CDT" might turn out to be the better choices when Canadians find themselves waking in the pitch dark and reverse their "permanent daylight saving" decision, as happened in the USA in 1974.

"USA in 1974" was a mess. At the time, DST was lastSun April to lastSun Oct. The bill for permanent DST was introduced November 7th (after the end of DST) and became law December 15th, 1973. Effective date in the law was written as the fourth Sunday after enactment, which was January 6th and was intended to be "permanent" through April of 1975 (when normal DST would begin then end in October of 1975).

That created a situation where the country fell back as usual in October then sprang forward on January 6th. Early January happens to be the time of the year with the latest sunrise (not December 21st, the shortest day). So in effect, they unintentionally picked the absolute worst week if not the worst day of the year to spring forward with the biggest shock value. Sensationalized news stories about early morning darkness (thanks to a spring forward in January) ignored any benefit of afternoon sun. The overall accident statistics for the winter of 1973 and the winter of 1974 were about the same. Safer travel in the afternoon balanced the additional morning accidents that made the news.

Political pressure led Congress to suspend "Permanent DST" from October 27th 1974 to February 23rd 1975 (a law passed October 5th, 1974). The US did not get a gradual shift in the fall of 1974 nut at least lastSun February was not on the worst day of the year.


The transitions in Canada this year will be smoother. There was no "fall back" where people gained an hour of morning sun November 1st and lost an hour of evening daylight. Sunrise time is gradually moving toward the latest sunrise date in early January and then will gradually move back to earlier sunrises. No "jump scare" on one of the the worst days of the year. In Vancouver the sun will rise at 9:07am on December 27th through January 6th. Not 8:07am on January 5th followed by 9:07am on January 6th. A gradual change. (If my calculator is correct, today's sunrise is at 8:26am - so Vancouver residents only have 40 minutes to lose over the next 40 days before sunrises start getting earlier again.)

The shortest day in Vancouver is about 8 hours 10 minutes (regardless of what time the sun rises). This year they will miss the 4:13pm sunsets in December. Hopefully people will see the benefit of having an hour in the evening to get things done (traveling home, shopping, etc.) in daylight.

Further north will have the same length of days as last year. When the shortest day is around 7.5 hours (Edmonton) or 5 hours (Yellowknife) does it really matter what time the sun rises?

I expect Canada remaining on DST this year will have far less impact than the fiasco in the US in 1974 and 1975. Yukon has worked fine with permanent UTC-07 since 2020. Northeast BC has had permanent UTC-07 since 2015 (Fort Nelson) and 1972 (Dawson Creek). Saskatchewan's "CST" is permanent UTC-06 since 1960 (1972 in Swift Current). DECADES of experience leaving the clocks on what would be DST in a strict solar time world. I hope people will give it more than a four month trial.


A few words to help those who only know 1974 from a history book, or worse, a face book.

I've investigated the 1973/1974 fiasco to some degree. I agree that it was especially bad timing. There were many things different at that time, especially no computer timekeeping, as today.

I'll accept your opinions about Canadian's adjustments to these changes. I don't live there.

We are still left with the dilemmas of what the "time" is called and the POSIX abbreviated names, at TzDb, downstream, and applications that depend on them. Let us know what display names appear on the various systems.

If the Sunshine Protection Act were to pass it looks like the confusions are going to be much worse and more widely distributed.

Reply via email to