Comment Re:don't drop the soap! (Score 2) 126
If he was semi-smart he would have walked down to the harbour, paid cash for a berth on a cargo ship to Panama and then disappeared.
If he was semi-smart he would have walked down to the harbour, paid cash for a berth on a cargo ship to Panama and then disappeared.
Yeah, I don't get the problem. It sounds like Redis wrote a lot of code that people find useful and released it as open source. They've decided now that they no longer wish to contribute to their open source project.
Is the author pissed off that they quit? Does he get pissed off at every open source developer who stops making contributions?
The paper itself is much more interesting than this "summary."
Most music systems make extensive use of simple harmonic ratios, or close approximations. The results of their simple pleasantness vs. harmony seem to pretty strongly support the idea that simple ratios are "more pleasant."
The study itself is interested in competing psychoacoustical theories, one of which predicts that relationship is unaffected by timbre, and the other that predicts timbre has an effect. They found that sounds with more high harmonics could differ more from the simple harmonic ratios and still be pleasant, but roll off of those harmonics did not matter much, showing that timbre matters, but not in exactly the same way either model predicts.
Also, yeah, westerners are and were totally unfamiliar with things like bells. It's not like there's a massive western cultural institution that's wielded insane power for a couple thousand years that likes to install them in all their buildings or anything.
What do you mean? They are "moving fast and breaking things", they should be the model for the future. Unfortunately in this case, it was their own cash flow, but hey, that's the price of being "disruptive". If their CEO and CFO wore a black turtleneck shirt, they would be the perfect company.
Accounting is for boomers!
I didn't say it was.
That's called solar time, or sidereal time, depending on whether you define "a fixed alignment as the start of the day" based on the stars or the sun. Both of those are kept, and are available. You can set your watch to either, if you choose. You can also set your watch to Greenwich Mean Time, which is the yearly average.
It's often impractical to synchronize clocks so frequently, so we invented UTC, which is the atomic time synched to a modern equivalent of GMT whenever they drift about a second apart.
Close.
TAI is kept by atomic clocks, one second per second. Solar time is kept by observation of the sun. The two aren't synchronized, they drift apart. There's also sideral time, which is kept by observation of the stars, which also drifts relative to the other two.
Then there are a bunch of things in between.
UT is based on a particular measurement of sidereal time.
UTC is based on TAI but fudged periodically to approximate UT.
GPS time is used by US GPS satellites, equal to TAI plus the UTC offset at the time GPS came online.
UTC is a decent thing to set your watch to. It's easy to calculate from GPS or TAI, the calculation only needs to be updated occasionally, and it approximates sidereal time well enough to navigate or point a small telescope. UTC is an approximation that's reasonable for human scales. Timing critical applications should use TAI, GPS or similar, not UTC. Anybody bitching about leap seconds shouldn't be working on such things.
A second is defined as 9 192 631 770 "ticks" of a cesium atom. You don't want to base it on something that only ticks once a second because it's hard to measure small inaccuracies with a standard that coarse. Even mechanical watches typically tick twenty or thirty thousand times per second, and that cheap Timex quartz is 32 768 / s.
The second is defined by taking the fixed numerical value of the caesium frequency , the unperturbed ground-state hyperfine transition frequency of the caesium 133 atom, to be 9 192 631 770 when expressed in the unit Hz, which is equal to s1.
Not that melting ice. The melting ice at the end of the last ice age:
the trend toward needing a negative leap second is clear, but he thinks it's more to do with the Earth becoming more round from geologic shifts from the end of the last ice age.
If your tech is using UTC for anything other than displaying the current time to the user, it's doing it wrong. There are several time standards already that don't have leap anythings.
Sadly those who own the robots are the ones who benefit the most. I seem to recall something about owning the means of production, but my American public education didn't seem to cover that subject.
With 70+ years of examples of how collective economies work, are you seriously arguing for Marxism?
First, water vapour is opaque at visible frequencies, but the sun emits over a very wide band, most of which will totally ignore clouds. For something like this, you'd use radio astronomy. The sun is a superb radio source, easily strong enough to compete with terrestrial sources, especially if you pick the right frequency.
Second, no it wouldn't. It wouldn't be triggered by the presence of detecting light on the optical spectrum, but the presence of a specific alignment. So you only ever synchronise when you hit exactly the correct point.
You then want an interferometer to observe the sun, where the centre of your virtual dish is precisely the meridian. When the centre of the sun is precisely midway across your telescope's field of view, you transmit a synchronisation pulse.
Thirdly, a leap second is introduced every 21 months or so. If we assume a month of 30 days, that means you have clock drift of 1/630th of a second per day. You could miss a lot of days before the maximum possible error exceeded the minimum error in the clocks used. The risk of time travel would be zero.
The chief benefit of a quantum gas clock is that it is able to measure time accurately enough to measure even small relativistic effects. (The change in the speed of time from a vertical displacement of ten feet is observable.)
Redefining a second according to such clocks won't cause any serious headaches - most users won't notice the difference, and those who would likely already use quantum gas clocks.
But it would make a difference in what they publish, as the intervals of time would have genuine meaning. (It's hard to talk about attosecond intervals when the error bars of what a second actually means will swamp the interval you're measuring.)
The chief benefit for synchronising to a quantum gas clock is that the error you inevitably introduce into the system is much smaller than the error of the atomic clock.
I would call your criticisms fair. And, yes, 192k would be sensible for the reasons you mention.
Instead of faffing around, do this:
1. Install a quantum gas clock as the planetary timekeeper. It's much more accurate than an atomic clock. Place this at the Greenwich Observatory.
2. Agree on a fixed alignment as the start of a day.
3. Place a sensor at Greenwich which detects when that alignment occurs. This resets the counter for the quantum gas clock, so a second always starts when the alignment occurs.
4. Synchronise all atomic clocks to the quantum gas clock.
There is now a daily reset which is at a much higher resolution than most computer clocks (but within the nanosecond resolution Linux supports). The system itself adds and removes fractions of a second as needed with no need of political decisions. This correction will always be smaller than natural clock drift, so is inside the error bar software is already designed to handle.
fortune: No such file or directory