You advocate a ________ approach to calendar reform

You advocate a

( ) solar ( ) lunar ( ) lunisolar

approach to calendar reform. Your idea will not work. Here is why:

( ) solar years are real and the calendar year needs to sync with them
( ) solar days are real and the calendar day needs to sync with them
( ) the solar year cannot be evenly divided into solar days
( ) the solar day cannot be evenly divided into SI seconds
( ) the length of the solar day is not constant
( ) the length of the solar day is more than 86,400 seconds on average
( ) the length of the solar day is ultimately unbounded
( ) the length of the solar day frequently decreases
( ) there can be negative leap seconds

( ) the lunar month cannot be evenly divided into solar days
( ) the solar year cannot be evenly divided into lunar months
( ) having months of different lengths is irritating
( ) having months which vary in length from year to year is maddening
( ) having one or two days per year which are part of no month is stupid
( ) your name for the thirteenth month is questionable
( ) that would destroy numerous birthdays and retcon the rest

( ) the lunar month cannot be evenly divided into seven-day weeks
( ) the solar year cannot be evenly divided into seven-day weeks
( ) every civilisation in the world is settled on a seven-day week
( ) having one or two days per year with no day of the week is asinine

( ) requiring people to manually adjust their clocks is idiotic
( ) local time should not be discontinuous
( ) local time should not go backwards
( ) local time should not repeat itself
( ) local time should not be ambiguous
( ) "7am" is a social construct
( ) social constructs are actually very important
( ) no amount of clock-moving can increase the amount of solar energy received by Earth
( ) "daylight saving" doesn't

( ) UTC already solves that problem
( ) zoneinfo already solves that problem
( ) rearranging time zones yet again would make the zoneinfo database larger,
    not smaller
( ) the day of the week shouldn't change in the middle of the solar day
( ) this would further ambiguate the word "day"
( ) local "midnight" should be the middle of the local night
( ) no matter what, there will always be at least one time zone
( ) nobody would agree to pick your time zone over theirs

( ) there needs to be a year 0 and negative year numbers
( ) no, we don't know what year the Big Bang happened
( ) years which count down instead of up are not very funny
( ) by that logic, 2000 was the final year of the Nineties
( ) Discordianism is not very funny
( ) planetary-scale engineering is impractical
( ) not every part of the world has four recognisable seasons
( ) "sunrise" and "sunset" are meaningless terms at the poles
( ) Greenwich is not unambiguously inferior to any other possible prime meridian

( ) time zones exist specifically to resolve the problems with that scheme
( ) coordinating geographically dispersed events would become nearly impossible
( ) I shouldn't need to adjust my wristwatch every few miles
( ) many clocks do not have line-of-sight to GPS satellites
( ) you can't put a GPS chip in a mechanical clock
( ) clocks on planes would literally run backwards
( ) what happens during an inserted leap second if the offset isn't a round number of minutes?

( ) the Earth is not, in fact, a cube
( ) time travel is impossible
( ) high-tech applications need far more accuracy than your scheme allows
( ) Unix time doesn't respect leap seconds
( ) leap seconds have been a fact of life for more than forty years
( ) leap seconds are more frequent than leap years
( ) 13-digit numbers are difficult for humans to compare, even qualitatively
( ) TAI already solves that problem
( ) most of history can't be renumbered with atomic accuracy
( ) clocks based directly on astronomical observations are inaccurate
( ) everybody in the world is already used to sexagesimal time divisions
( ) date formats need to be unambiguous
( ) abbreviated date formats should be possible and still unambiguous
( ) a leading zero on the year number only delays the inevitable
( ) date arithmetic needs to be as easy as possible

Specifically, your plan fails to account for:

( ) humans
( ) clocks
( ) computers
( ) space
( ) the inconsistent rotational and orbital characteristics of Earth
( ) rational hatred for arbitrary change
( ) unpopularity of weird new month and day names
( ) total incompatibility with the SI second
( ) general relativity

and the following philosophical objections may also apply:

( ) BC and AD aren't
( ) technically, that's how it already works
( ) they tried that in France once and it didn't take
( ) nobody is about to renumber every event in history
( ) good luck trying to move the Fourth of July
( ) nobody cares what year you were born
( ) time cannot be stopped, nor can clocks be abolished entirely
(x) the history of calendar reform is insanely complicated and no amount of
    further calendar reform can make it simpler

Furthermore, this is what I think about you:

( ) sorry, but I don't think it would work
( ) this is a stupid idea, and you're a stupid person for suggesting it
( ) please just shut up and fix your broken date/time code

Back to Time
Back to Things Of Interest

Discussion (24)

2013-12-09 11:47:52 by qntm:

The past will always exist, and they'll never adopt your new calendar.

2013-12-09 12:42:19 by jymbob:

Dang. It appears my work at (in the late 90s, not really touched since) was all for naught

2013-12-09 12:44:34 by jymbob:

Oh, and by the way: Your twitter icon is 18 months out of date:

2013-12-09 13:09:20 by Zache:

No reference to voted 1

2013-12-09 13:22:06 by Rafe:

This is surprisingly relevant to me. I made a calendar reform once, although it was part of a bigger flight of fancy and not meant seriously. I was thinking about all the systems that would be different if we were on base-6 or base-8 arithmetic and I started to invent the fictional countries of "Senaria" and "Octalia" where these systems would somehow have a home. I worked out a highly regular solar calendar that would work for both radices with one small adjustment, although it doesn't even try to have lunar units like weeks or months.

I could go on but I'll just say that at first glance these are the complaints that seem to apply to my fantasy calendar:

2013-12-09 13:39:25 by dsr:

Does it count as calendar reform if your suggestion solely affects time zones?

My proposal which is mine is that we stop using time zones. Simply[0] pick an existing time zone, and have everyone use that around the globe.

The negative is that people will have to stop associating arbitrary time numbers with local sun visibility (600 no longer necessarily means early morning, noon is no longer a near-synonym for 1200PM) -- and, of course, the flag day will be hard on 23/24 of the world or so.

The positive is that long-distance coordination problems (such as conference calls, or specifying when your flight arrives, or arranging for everyone in Asia to jump up and down at once) no longer need to spend time an effort on understanding where everyone involved is and what sort of weird Daylight Saving time scheme they have going on.

2013-12-09 14:10:57 by jymbob:


I'd forgotten about Swatch time. My issue with it is: why on earth would you pick UTC+1 as your base?
Other than that, my take on it corresponded to Beat*10 (divide the day up into 10 units rather than 1)

The lack of smaller subdivisions irked me as well. What's the point in designing a standard less accurate than the current system?

2013-12-09 14:14:24 by qntm:


(x) the date shouldn't change in the middle of the day
(x) zoneinfo already solves that problem

2013-12-09 14:43:01 by Bauglir:

Aw, all I want to do is redistribute the number of days per month so that they alternate 30/31 properly, with January getting the leap day (so it breaks the alternation in all non-leap years, but it's inevitable that it breaks somewhere).

2013-12-09 14:45:36 by Bauglir:

Wait, I wrote that wrong. February needs to get the extra day, not January. You'll never get people to give up December 31st, and so leap months have to be even, and since it can't be the last available choice, it needs to be the first. And it reduces the amount of actual change here, anyway.

2013-12-09 15:51:43 by Michael:

What year zero?
Also, fuck the past. I say we round up all those people and send 'em back. Then we won't have to worry about them and their pesky notions of "but base 60 makes sense, err, because, err, ah! it divides evenly by heaps of numbers, unlike, say, err, 100" and "but I don't understand decimals, that's why I find Celsius too hard".

2013-12-09 16:58:27 by ducken:

I'm with bauglir on this one, and always have been. however,

( ) the history of x reform is horrifically complicated and no amount of
    further x reform can make it simple

is always the most compelling reasoning against any reform. information inertia is a very real and very terrifying thing. I work in a property tax office, and the switch from one property designation paradigm to another 20 years ago still screws up my work daily. if historical record matters, once you get above an arbitrary amount of collected info, any hope for fixing a broken system is slaughtered more viciously than minorities in an FPS.

2013-12-09 17:49:54 by IdolRevolver:

Lots of these objections apply to the current calendar too.
Personally, I advocate an ordinal system, where you start counting at 1 and stop at 365 (366 in a leap year). No need for months at all.

2013-12-09 21:51:50 by Moti:

Needs a "the following installments of XKCD may also apply" section.
Here it's of course 927 :)

2013-12-10 03:35:01 by Sean:

I second Michael's question. 1 B.C.E was followed by 1 C.E., so we don't have a year 0 in the current calendar... (We're lucky that it's far enough in the past that not many people have to deal with the resulting off-by-one errors.)

2013-12-10 05:23:46 by peitro:

ISO 8601 week date: date +%G-W%V-%u. With this format ...

* Every year contains an integer number of weeks.
* Determining the day of week from a date string is trivial.
* The calendar looks the same every year, save for leap weeks.
* Adding or subtracting a number of days and weeks is trivial.

2013-12-10 05:54:00 by Artanis:

Every time I see one of these, I always wonder if I can write something that would cause every single option to be checked.

2013-12-10 09:09:41 by jonas:

See also Claus T√łndering's Calendar FAQ on .

2013-12-13 04:45:29 by EternalDensity:

I'm going to stick with NewPix

But yeah, any calendar system will be broken in some way so we might as well stick with the one we (think we) know.

2014-01-01 21:07:36 by DragonDave:

As far as I know, there is no good standardised way of expressing, to a computer: "Our meetings are at 10am everyday, London local time".

We can't express this with numerical or abbreviated timezones (since it oscillates between GMT and BST on a yearly basis). We can't guarantee that the government won't decide to change when daylight saving happens, or even that it happens at all, so it's always going to need to refer to a lookup list.

There's the tzinfo library, which provides that information. ISO8601 tells you how to write datetimes in an interchangable way, including repetition and partial dates and periods. We need an extension that lets us write something like "2013-01-01T10:00:00[GBR-London]".

Evolution, not revolution.

The devil, as always, is in the details.

2014-01-15 03:31:05 by Randall:

My personal favorite calendar reform deals with expressing dates in the far future and past unambiguously and without inordinately long expressions: Y10K. It's even an RFC standard:

2014-02-11 23:01:25 by ANONYMOUS:

It is obviously time to destroy the Moon.

2014-02-20 05:42:04 by BlitzGirl:

Someday, one of us will destroy that stupid skycircle.

2015-05-11 21:36:28 by Argent:

The Discordian calendar resolves all these problems by declaring them features. Hail Eris!

This discussion is closed.