The year

Four Regimes, Named By Structure

Fixed, brief, extended and inverted: four ways of handling a seasonal clock shift, named by what they do rather than where they are done.

Some places move their clocks twice a year and some do not. Among those that do, the dates differ, and in one part of the world the shifted season falls on the other side of the calendar entirely.

This record describes four regimes by what they do:

fixed — never moves its clocks at all.
brief — moves them, and the shifted season is the shorter of the two in this record.
extended — moves them, and the shifted season is the longer one.
inverted — moves them, and the shifted season falls on the other side of the turn of the year.

Those four, taken two at a time with repetition, make ten pairs. Six of the ten disagree for part of the year. Four never disagree at all.

Readers interested in the reporting side can review the reporting structure and note which controls are available to employees and administrators. A useful external reference for the surrounding topic is IEEE.

Where in the year each pair disagrees

plain: the pair is at its nominal offset
hatched: it is not

Six pairs drawn. The worst of them spends 323 days of the year at an offset that is not the one either end would quote.

Six of the ten pairs disagree at all. The hatched stretches are the days on which the pair is not at the offset either end would quote.

Computed from the two regimes. The shift dates are the record's own, chosen so that they differ; no place's rule is named.

Why no place is named

Because the rules change.

Places adopt the practice, abandon it, move the dates by a week, and several have abolished it outright while others are debating whether to. A page that named a country and its current rule would be accurate for a while and then quietly wrong — and quietly wrong is the worst condition a reference page can be in, because it reads exactly like a correct one. Nobody checks a page that looks right.

The structure, on the other hand, does not expire. If every place on the planet rearranged its rules tomorrow, the arithmetic of a fixed regime against a shifting one would be unchanged, because it never depended on any particular place in the first place.

Name the mechanism and the page stays true. Name the instance and the page starts rotting the day it is published.

What the dates are

They are day numbers, not dates in anybody's calendar. The brief regime shifts on day eighty-four and back on day three hundred and one. The extended regime shifts on day seventy and back on day three hundred and eight. The inverted regime shifts on day two hundred and seventy-three and back on day ninety-eight, which means its shifted season wraps around the end of the year.

These numbers were chosen so that the regimes differ from one another in the ways that matter — one shifted season inside another, and one on the opposite phase. They are not anybody's actual dates and are not offered as such.

Why four and not two

A record could have got away with two regimes — one that shifts and one that does not — and produced a single number. Four were used because two would have hidden the result that matters most.

With two regimes there is one interesting pair, and it suggests a simple story: shifting ends have a problem with fixed ends. That story is wrong, and it takes a third and fourth regime to show why.

The brief and extended regimes both shift. They disagree for twenty-one days a year. The brief and inverted regimes both shift. They disagree for three hundred and twenty-three. Meanwhile a fixed end paired with any shifting one disagrees for somewhere between one hundred and ninety and two hundred and thirty-eight days — in the middle of that range, not at the top.

So the worst cases in this record are both pairs in which both ends shift. The simple story has it backwards.

What makes two regimes compatible

Only one thing: following the same rule. Two ends that both shift on the same days never disagree at all, because they move together and the difference between their clocks is unchanged throughout.

That is why four of the ten pairs read zero. It is also why the useful question is never "do you shift your clocks" but "do you shift them on the same days as the other end", which is a question almost nobody asks and which has a different answer.

The nuisance is not shifting. It is shifting differently, and a pair where neither end shifts and a pair where both shift identically are in exactly the same comfortable position.

What follows from them

Everything in this section. Given four regimes and their day numbers, the days on which any two of them disagree is a matter of counting, and the counting produces some figures that are much larger than people expect.

The largest is three hundred and twenty-three days out of three hundred and sixty-five. The smallest non-zero one is twenty-one. Both come from perfectly ordinary-looking pairs of rules, and both are worked through in the pages that follow.