The 23:59:38.7 value is the true solar day for September 16th, as compared to the mean solar day which is defined to be 24 hours exactly when averaged over the course of a year. The article for Equation of Time describes it in more detail, with rather wordy explanations, I'll try to simplify. http://en.wikipedia.org/wiki/Equation_of_time
It's a sine wave with two components. The larger comes from the Earth's elliptical orbit around the Sun. When the Earth is near perihelion in January, it moves faster and farther each day along its orbit, so needs an extra few seconds for more rotation to bring your spot on the surface back under the noon Sun. And the converse in July, where the slower orbital speed at aphelion means each solar noon occurs a few seconds earlier.
The smaller component comes from the axial tilt and runs at twice the frequency (two cycles per year). Near the solstices, the Sun's declination (angle above the horizon at noon) remains nearly constant day to day, so all of its daily motion is in the east-west direction. Near the equinoxes, the Sun's apparent path is more oblique, so some of its apparent velocity occurs in the north-south direction leaving its apparent east-west motion slower. That shortens the solar day since the Earth has to rotate less far to the next solar noon.
The slope of the equation reaches its negative maximum on Sept 16th, so that is the shortest solar day. On this day, the Earth is roughly halfway from aphelion to perihelion, so experiences its greatest orbital acceleration. And it is near the autumnal equinox, where the Sun's apparent east-west motion slows, additively with the same sign as the orbital component, thus creating the global maximum.
It's worth noting all of this refers to "day" as noon to noon or midnight to midnight, affecting daytime and nighttime equally. The phenomenon doesn't affect the daylight/nighttime ratio or insolation or temperatures like the seasonal cycle does.
I studied this stuff a few years ago when I made a sundial. I understood this better - and now understand why that figure 8 symbol is printed in globes - from learning about an analemma:
Given the superlative, I assume that's the earliest at the earliest location in the entire world (wherever that is, but it clears up the ambiguity of multiple locations).
So. How do I compensate for this if I wish to convert a milliseconds since epoch value to an accurate datetime type? Are these the concerns that smart people who write DateTime types in frameworks deal with?
You don't have to; solar days are not the same as "normal" days (defined by TAI and used by UTC). This is why we have leap years, and leap seconds in UTC.
The effect discussed here averages itself out over the course of one year.
Leap years and leap seconds are required due to other effects that don't (leap years because of the non-integer ratio between year length and day length and leap seconds because the length of day is increasing slowly)
Well you do sorta have to deal with it; leap seconds aren't fun, and are actually quite stupid. Time's already complicated enough without adhering to a few folks that care about "solar" time. Either kill them off, or move to leap hours (which puts the next jump so far out, no one will care).
Leap seconds are not accounted for in epoch-time. Basically, 23:59:59 and 23:59:60 have the same value (this also makes it easier by not having to handle the 60 :) )
We disabled leap-seconds at work in our NTP servers, and just use the normal mechanisms to compensate slowly.
Ah ok, I think I get it. It is talking about a complete revolution of the earth. It is not talking about a day as in day/night. And it is not really a revolution, it is about the time it takes for the sun to be in the same position from one day to another.
2.- When we talk about the longest day, we are talking about daylight. That changes according to time of the year because of the inclination of the earth. This is what I originally thought they were talking about.
3.- The usual way to define a day is exactly 24 hours. But this is a little bit different. It is not even exactly a turn of the earth of 360 degrees.
Edit: Which 1 minute after posting I realize is the same thing, so forget that half-baked response and instead look at these different ideas of the measure of a day:
Your first answer wasn't wrong, it just defines a "siderial day", using other stars as a reference point instead of the sun. After rotating 360 deg with respect to the background stars (which we assumed fixed for our purposes), the earth then has to rotate a few minutes more to get the sun back in the same position.
And thus "right ascension" is measured in minutes (and hours). Astronomers don't use degrees to measure "longitude" (right ascension) on the celestial sphere.
Face North (at night) and look right. Everything on that side ascends. In astronomy, East-West distances are measured by how long it takes things to rotate past a fixed observational direction.
Every night, a slightly different chunk of sky will also be directly overhead at midnight, due to the difference observed earlier between solar (sun-relative) day length and sidereal (star-relative) day length.
Please no! Even the default 24 hours is too little for spending time with my girlfriend, attending class, doing work, thinking about the side project, learning Haskell, planning my fiction book...
And for one day my sleep deprivation is also increased by 21.3 seconds.
It's a way to help you prevent yourself from spending too much time on News.YC. If you turn it on you'll only be allowed to visit the site for maxvisit minutes at a time, with gaps of minaway minutes in between. The defaults are 20 and 180, which would let you view the site for 20 minutes at a time, and then not allow you back in for 3 hours. You can override noprocrast if you want, in which case your visit clock starts over at zero.
The problem I've found with things like noprocrast and assorted browser extensions is finding the exact right balance between making it enough of a hassle to get me to go away and do something more productive most of the time, but not intrusive enough that I'll end up disabling it permanently when it blocks me when I really want to have some downtime and "deserve it".
Noprocrast simply doesn't work for me for that reason.
What's worked best for me is the "Delayed Gratification" extension for Chrome, that delays certain websites for a specific amount of time. I found about 30 seconds is sufficient for me to go be able to convince myself to go back to work most of the time when I find myself just randomly going to a site like Reddit or HN, yet short enough that I can wait out the delay without being too tempted to disable it when I really want to.
My comment was just a little snarky sharing a feeling that I could use more time to produce more value. I'm not sure where you get the idea that I'm spending too much time on Hacker News; it's certainly not the case.
Obviously someone got really offended by my amusement.
This is the sort of thing which encourages people who exult science to act as if those who don't are stupid.
Where is the word "sidereal" in the headline? Leaving it out is part and parcel of scientism's self assertion as the one true belief.
That's not what people ordinarily mean by "day" - the boxes on a calendar are all the same size. It's not even what people ordinarily mean by "shortest day" - i.e. the day with the least amount of daylight.
Not only is it a fact of little or no practical utility, it is nearly impossible to verify by reasonable means and therefore for the vast majority of those who assert it to be true it is a revelatory truth accepted on faith. Only those with full notebooks, big telescopes and accurate chronometers have a chance of verifying the hypothesis. Otherwise, it is just part of a canon.
> scientism's self assertion as the one true belief
There is no such thing as 'scientism' and science is a method, not a belief system.
There are people who go through life without dogma. Every idea they have can be challenged. Therefore, if you define 'belief' as 'dogmatic belief', as you seem to, they don't have any beliefs at all.
It's a sine wave with two components. The larger comes from the Earth's elliptical orbit around the Sun. When the Earth is near perihelion in January, it moves faster and farther each day along its orbit, so needs an extra few seconds for more rotation to bring your spot on the surface back under the noon Sun. And the converse in July, where the slower orbital speed at aphelion means each solar noon occurs a few seconds earlier.
The smaller component comes from the axial tilt and runs at twice the frequency (two cycles per year). Near the solstices, the Sun's declination (angle above the horizon at noon) remains nearly constant day to day, so all of its daily motion is in the east-west direction. Near the equinoxes, the Sun's apparent path is more oblique, so some of its apparent velocity occurs in the north-south direction leaving its apparent east-west motion slower. That shortens the solar day since the Earth has to rotate less far to the next solar noon.
The slope of the equation reaches its negative maximum on Sept 16th, so that is the shortest solar day. On this day, the Earth is roughly halfway from aphelion to perihelion, so experiences its greatest orbital acceleration. And it is near the autumnal equinox, where the Sun's apparent east-west motion slows, additively with the same sign as the orbital component, thus creating the global maximum.
It's worth noting all of this refers to "day" as noon to noon or midnight to midnight, affecting daytime and nighttime equally. The phenomenon doesn't affect the daylight/nighttime ratio or insolation or temperatures like the seasonal cycle does.