XKCD imitates MeFi (Again) September 24, 2018 8:12 PM   Subscribe

I think the cartoons are coming from...inside the site.
posted by kirkaracha at 10:06 PM on September 24, 2018 [6 favorites]

A webcomic powered by its own sense of self-satisfaction
posted by thelonius at 7:15 AM on September 25, 2018 [7 favorites]

I enjoyed that thread the other day! Time is weird. A couple years ago, an AskMe sent me down a rabbit hole of reading about early clocks in Europe and how they divided time before that, and it's such a cool strange subject - to think how differently we could do this kind of bedrock seemingly-unchangeable thing. Fun to see this xkcd -- thanks for pointing it out.
posted by LobsterMitten (staff) at 7:21 AM on September 25, 2018 [1 favorite]

about early clocks in Europe and how they divided time before that, and it's such a cool strange subject

I recently watched Tudor Monastery Farm, and there was a part about keeping time and clocks. The entire show is worth watching, but I believe it is episode 3 where time and clocks are discussed.
posted by terrapin at 7:56 AM on September 25, 2018 [5 favorites]

My partner and I sometimes watch college football and whenever they bring out the chains to measure for a first down, she groans and complains. It’s so arbitrary! The ref has been putting the ball down wherever all game, and NOW they are going get all precise?

True yeah, I get you, but isn’t that like everything else and also there’s no time like the present?
posted by notyou at 8:00 AM on September 25, 2018 [3 favorites]

Randall's not quite right about everyone's clock running at the same rate on the equinox. Daytime, if defined as the length between sunrise and sunset, is longer than 12 hours (using conventional timing) on the equinoxes, and the effect is more pronounced at higher latitudes than at the equator.

Two reasons for the day being longer than 12 hours on the equinox: First, sunrise is defined as the time the first limb of the sun (not the center) appears above the notional horizon, and sunset as the time the last limb disappears. Second, atmospheric refraction causes the sun to appear slightly higher in the sky than it would on an airless earth when near the horizon, including appearing above the horizon when it would otherwise be slightly below it.

Per Wikipedia, the combined effect of these two is that a bit of the sun appears above the horizon when its center (on a theoretical airless earth) would be 50 arcminutes (0.83°) below the horizon. This makes the length of the day longer than 12 hours on an equinox, but also mean the length varies by latitude. At the equator, the sun rises and sets perpendicularly to the horizon, but at other latitudes it rises and sets at shallower angles, the angle becoming less at higher latitudes, meaning it spends a greater time within those critical 50 arcminutes. Playing around with the US Naval Observatory's Complete Sun and Moon Data for One Day page, I get the length of the day on the equinox to be 12:07 at the equator, 12:09 at 40° latitude (north or south), and about 12:14 at 60°.

So if you set sunrise to be 6:00 a.m., and sunset to 6:00 p.m., the equinox is only the day all clocks on earth run at most nearly the same rate, not exactly the same rate.
posted by DevilsAdvocate at 9:50 AM on September 25, 2018 [7 favorites]

I swear I saw an XKCD closely related to this xkcd-ish joke in a similar timeframe. Darned if I can find the XKCD now though.
posted by quacks like a duck at 10:49 AM on September 25, 2018 [1 favorite]

quacks like a duck: here. It travels over water too.
posted by matthewr at 3:10 PM on September 25, 2018 [4 favorites]

Is this 6a/6p thing compatible with the International Fixed Calendar? We could be on the verge of enjoying dependable and predictable structure in our days!
posted by rhizome at 3:18 PM on September 25, 2018 [1 favorite]

My mum used to be an admin for a big paper company in the PNW. I'd occasionally go in on the weekends and help her catch up on filing, especially at the end of the month when she had compile monthly reports. But due to who knows what reason, each reporting "month" had exactly 30 days in it, so the first day of "February" was January 31. I never did figure out what happened to the extra 5 days at the end of December - maybe they just didn't count them since the plant was closed over the holidays.
posted by muddgirl at 3:47 PM on September 25, 2018 [2 favorites]

In retrospect I am probably getting details about this calendar wrong. Maybe it was each month had 4 work weeks in it? Anyway it was strange and there were definitely missing days but I was 15 and deep into teen ennui so I never asked.
posted by muddgirl at 3:49 PM on September 25, 2018 [3 favorites]

The whole problem with the 6 to 6 calendar is that you end up working 12 hours in the summer and 4 hours in the winter. That doesn't sound ideal. In fact, people above the Arctic circle would have to work ceaselessly in the summer months.
posted by Literaryhero at 3:59 PM on September 25, 2018 [1 favorite]

Kodak had a 13 equal month calendar (this was a recent pub trivia question).
posted by rodlymight at 4:22 PM on September 25, 2018

Kodak had a 13 equal month calendar (this was a recent pub trivia question).

That's the International Fixed Calendar above. Fun fact: Kodak used it internally until 1989!
posted by rhizome at 5:37 PM on September 25, 2018 [2 favorites]

Time is such a weird thing. I've done work off and on for years on navigation systems, and they all rely on time to give you an accurate position fix. I don't pretend to understand most of it, but I see the extraordinary effort people smarter than me put in to eliminating nanosecond clock drifts or asynchronous periods measured in days and it astounds me every day that I can open an app on my phone which will tell me where I am to within a couple of inches.
posted by backseatpilot at 5:57 PM on September 25, 2018 [2 favorites]

> there's no reason not to move to Real Solar Time™, where the sun rises at 6 AM and sets at 6 PM all year round

Isn't this how sundials work?
posted by ardgedee at 5:59 PM on September 25, 2018 [2 favorites]

*bursts in from the early 2000s*

posted by cortex (staff) at 6:19 PM on September 25, 2018 [15 favorites]

Not to steal your thunder at all, but I wanted to know more about the International Fixed Calendar and found a good article about it, so I made this FPP to continue discussion specifically about that.
posted by limeonaire at 6:27 PM on September 25, 2018 [4 favorites]

I'm still trying to make Swatch Time happen.
posted by bongo_x at 1:42 AM on September 26, 2018 [2 favorites]

Speaking of Swatch, did everyone know that the Smart in Smart Car is short for Swatch Mercedes ART?
posted by zamboni at 3:25 AM on September 26, 2018 [3 favorites]

> *bursts in from the early 2000s*


Swatch Internet Time is going to be 20 years old next month. Somebody should do an anniversary post about it on October 23.
posted by ardgedee at 4:24 AM on September 26, 2018 [2 favorites]

Isn't this how sundials work?

No; since the sun rises and sets at different spots on the horizon at different times of the year, a sundial will not show sunrise, nor sunset, as being at the same time year round.
posted by DevilsAdvocate at 6:49 AM on September 26, 2018 [2 favorites]

I have one of those Swatch watches launched with internet time. It also has a little animation of a dog peeing on a lamppost.
posted by chavenet at 4:32 PM on September 26, 2018 [2 favorites]

I have a similarly practical solution to the peskiness of time zones - eliminate them by fixing timekeeping geographically to every point on earth, linked to solar noon where you are. A gradient at a resolution of, say, 1 square(ish) foot (come on, this thing is going to be imperial measurements. USA!). Instead of passing from one time zone to another and having an abrupt shift from 3:15 to 2:15 (or or or) your clock would track where you are, lat and long, at any given time, and tell you the exact time for that location. So you could theoretically keep moving due westward at a constant speed (between 675 and 1035 mph, depending on latitude) and stay at precisely the same "time."

You'd want to notate time as not just the time of day, but the location, down toa pretty high degree of precision. More work and money for clock, map, surveillance business.

I'll let some eggheads sort out the practical details.
posted by dirtdirt at 12:46 PM on September 28, 2018 [1 favorite]

Interestingly, the Japanese (pre-Meiji) used a similar temporal hour system, in which time was divided into six daytime periods, and six nighttime periods, which varied in length according to local day/night cycles. This meant that the length of the "hour" was seasonally and geographically variable.

Even more interestingly, they didn't use the numbers 1, 2 or 3 for religious reasons so a clock would have time slices from nine to four, counting backwards (since their earliest timekeeping was by monitoring incense as it burned down).

...then it gets even more complex for reasons I don't recall (or understand anyway), so ultimately you ended up with a system in which the variable-duration units of time were:

6: sunrise
5: morning
4: morning
9: noon
8: afternoon
7: afternoon
6: sunset
5: evening
4: evening
9: midnight
8: before dawn
7: before dawn
posted by aramaic at 6:14 PM on September 30, 2018

fixing timekeeping geographically to every point on earth, linked to solar noon where you are

That's where I started as well, but the trouble is that human circadian rhythms are synced by the first appearance of morning light, not by the most-nearly-overheadness of the sun. This is why flabdablet radically local time takes sunrise at 06:00:00 rather than solar noon at 12:00:00 as the fixed reference point for the violence that needs to be performed on the length of a second to make every nominal day occupy exactly 24 hours of wall clock time, and specifically does not require sunset to be reckoned as 18:00:00.

This requires only imperceptibly more violence than it would to keep the days lined up with solar noon, and it also means we get all the benefits of daylight saving time with very few of the downsides.

The other key idea there is using different units for personal/commercial/legal timekeeping (seconds) from those used for scientific/technical timekeeping (secs). The latter can and should be derived from the most accurate atomic clocks available and used as the underlying basis for the location-dependent derivation of the former.
posted by flabdablet at 6:45 AM on October 3, 2018

« Older Metafilter Spotify account?   |   Ettiquette on reposting an entire, ongoing website Newer »

You are not logged in, either login or create an account to post comments