vilhelm_s

Why would it be a GDPR violation? As I understood it, the point of the GDPR is to prevent people from being tracked without knowing about it and consenting to it. Here the consent button is as obvious as it could be, you really can't miss it.

Surely the air is still interacting with the airplane.

The shockwave generated by a body shaped like an acute cone is a (somewhat less acute) cone with the same apex. Inside the shock wave, the air gradually turns around to move parallel to the body, like so [0]. Each streamline that enters the shock structure keeps turning around, so each bit of air keeps feeling a force even after it passes through the original cone-shaped shock. In the case of a cone-shaped body, the interaction happens to create a single "straight" wave, but it's still generated by the entire surface of the cone, not just the sharp point of it.

As a rule of thumb, the magnitude of the sonic boom is proportional to the weight of the plane. That is because most of the interaction between the plane and the air is from the wing, it needs to continuously deflect down enough air to generate a lifting force. So the total boom has one contribution from the cross-section of the fuselage, and another from the wings, and the latter is determined by the weight.

[0] https://www.researchgate.net/profile/Richard-Benay/publicati...

I really doubt that. If we focus in particular on bike theft (one of the examples above), some webpage says that Norway had 60k bike thefts in 2004, 20k of which was reported to the police[1]. This is out of a population of 4.5 million. Some other page says that in New York City, there are 75k bikes stolen per year with 15k reported to the police[2]. This is out of a population of 8.4 million.

So all of Norway (rural and urban areas together) had a higher per-capita rate of bike theft than New York City (which is all urban).

[1] http://www.sykkeltyveri.no/bicycle_theft.html [2] https://www.bicyclehabitat.com/articles/security-pg80.htm

The Nature paper that this press release talks about is [0], and even better, the same issue of Nature also published an accompanying "News and Views" explanation of the paper for non-experts [1] which I thought was clarifying:

> A report by Wang et al.5 ( page 277 of this issue) now demonstrates a very large superluminal effect for pulses of visible light, in which a pulse propagates in a specially prepared medium [...] not only faster than a pulse travelling in a vacuum, but so fast that the peak of the pulse exits the medium before it enters it! [...]

> their light pulses do not violate causality. [...] They use smooth, well-defined light pulses, so that the peak of the pulse at the output results from the forward rising edge of the input pulse, which occurs far earlier in time, making it consistent with causality. An abrupt feature in the light pulse would not be able to travel faster than c.

[0] https://www.nature.com/articles/35018520?proof=t [1] https://www.nature.com/articles/35018657?draft=journal&proof...

The original post also says

> The Pendolino itself is a fine product, with the tilt removed. Alstom uses it as its standard 250 km/h train, at lower cost than 350 km/h trains. It runs in China as CRH5, and Poland bought a non-tilting Pendolino fleet for its high-speed rail service.

I guess this is a reference to "New Pendolino" (https://en.wikipedia.org/wiki/New_Pendolino), one of the most recent versions of the train which omits the tilting.

Yes, making a mistake when writing down the C semantics is definitely possible. The point of the work announced here is to get rid of that risk, so that the proof only relies on the semantics of the RiscV instructions (and the C semantics become an intermediate thing which are proven rather than trusted). It's still possible that they made a mistake when writing down the RiscV semantics, but machine code is much simpler than C so the risk is smaller.

I think these objects are maybe true but not relevant.

The leading example of the article was about Zooom videos, not single frames, and in a school setting you presumably know what culture the students are from, so I don't think you need it to work cross-culturally. And I don't doubt that people can deliberately hide their emotions, but many settings are not adversarial, so there would be no reason to do so. E.g., if you imagine that a teacher is teaching a lecture to a large audience over zoom and is using a "puzzlement meter" to see if they seem to be getting it, then any audience member who does feel puzzled will only gain from frowning and making the lecturer slow down.

In general, I think from my experience talking to people over video chat, some emotional information does get communicated over the video, so in principle computer programs should be able to pick up on it too.

In the next issue, the journal published a bunch of letters noting that she had rediscovered the trapezoid rule.

https://kconrad.math.uconn.edu/math1132s20/handouts/taicomme...

Mary Tai herself also replied and explained the background to the paper a bit more:

> While a doctoral candidate working on my dissertation at Columbia University in 1981, I needed to calculate total area under a curve. During a session with my statistical advisor, and after examining several alternative methods, I worked out the model in front of him. The concept behind it is obviously common sense, and one does not have to consult the trapezoid rule to figure it out. [...]

> Why I call it Tai's model. I never thought of publishing the model as a great discovery or accomplishment; it was not published until 14 years later, in 1994. Because of its accuracy and easy application, many collegues at the Obesity Research Center of St Luke's-Roosvelt Hospital Center and Columbia University began using it and addressed it as "Tai's formula" to distinguish it from others. Later, because the investigators were unable to cite an unpublished work, I submitted it for publication at their request. Therefore, my name was rubber-stamped on the model before its publication. [...]

Yes, that's right.

> The first stage of this project will explore the feasibility of exposing healthy volunteers to COVID-19. In this initial phase, called the Virus Characterisation Study, the aim will be to discover the smallest amount of virus it takes to cause a person to develop COVID-19. [...] Once the characterisation study is completed, this Human Challenge Study model can be utilised to conduct COVID-19 Human Challenge Studies from mid-2021 subject to regulatory approvals.

There are some more detailed writeups at https://www.gov.uk/government/news/worlds-first-coronavirus-... and https://hvivo.com/the-human-challenge-programme/

Similarly, newly built or renovated houses will have some "new house" smell, which is also VOCs. I have been told there is a cultural difference between China and the U.S., where Chinese people will let a house be empty for a few weeks to get rid of the smell before moving in (since the chemicals are believed to be bad for your health), while in the U.S. people don't seem to be concerned by them.

One thing I don't understand is that the dataset already comes with a "train dataset" (120 recordings) and "test dataset" ("Recordings of 22 repetitions of 7 narrative test set videos"). Why are they not just setting aside a randomly selected subset of the training data for testing?

I'm unclear on whether the testing data contains some of the same video clips that were shown in the training data, repeated literally? It seems it would be best to test the model on a previously unseen image, to ensure that it really learnt the mapping and is not just "memorizing" the responses.

Telescopes etc are already designed to give the same perspective and foreshortening effects as a "simple" lens or a pinhole camera. There can be some imperfections (e.g. "barrel distortions"), but these are pretty small. Compared to a single lens, a fancy telephoto lens captures more light (producing a brighter image), it may blur the out-of-focus part of the picture more, and it has less chromatic aberration, but the perspective projection is the same.

The perspective distortion comes for landscape photos are due to the field of view. So the absolute distance between lens and sensor doesn't matter, what matters is the ratio between that distance and the width of the sensor. As long as that ratio is the same as the ratio of the eye-to-picture distance and width of the picture, the perspectives will match.

The problem is that if you are looking at a landscape with your bare eyes, you will look at a quite large field of view. To reproduce that field of view when looking at the picture, you probably need to print it several meters wide, and look at it from a meter away or so. If you take a photo with the same field of view but print it at a handheld size, it will look "too small" and the perspective will seem exaggerated.

Oh I missed that!

I think you overestimate the energy in uranium. 8000GWh per tonne and year would be 28.8 TJ/kg, while the numbers I have seen is more like 0.7 TJ/kg (http://www.plux.co.uk/energy-density-of-uranium/).

Googling a bit: it takes about 13 MJ to desalinate 1 m^3 of water by reverse osmosis. Seawater contains 3mg uranium per m^3, which gives 2 MJ, So it's not energy positive.

In an ideal frictionless gas (zero viscosity), the flow after the object has passed through will be identical to the flow before the object, so in that sense the air moves "seamlessly" around the object. But during this process, the pressure of the gas still changes. You can think of this as just newton's second law: in order to change the velocity of an air packet, you must apply a force to it, and pressure is just force per area. So any time a gas flows in anything other than a straight line, you know that there was a pressure gradient involved and the pressure was not constant.

It's a common example in fluid mechanics textbooks to derive the flow of a zero-viscosity gas around a cylinder, you can get an exact solution e.g. here: http://brennen.caltech.edu/fluidbook/basicfluiddynamics/pote... . Note that in this example there is no drag, but there are two zones of high pressure in front of and behind the cylinder. Those zones would have higher-than-ambient temperature.

I thought this would be http://hackersdictionary.com/html/Some-AI-Koans.html .

From the NIF website,

> The facility is funded by the U.S. Department of Energy’s National Nuclear Security Administration (NNSA) and is a key element of NNSA’s science-based Stockpile Stewardship Program to maintain the reliability and safety of the U.S. nuclear deterrent without full-scale testing.

Most of the Department of Energy budget goes towards nuclear weapons, although they also do some work on civilian nuclear power and pure science. The name was always a bit of a euphemism.

Yeah, the Esri fillOpacity property just becomes an SVG property, and the SVG specification requires that it gets clamped to [0,1] before rendering.

In addition to computing the quartiles correctly, I also realized you should probably use (AREALAND+AREAWATER) rather than just AREALAND... in the above image the great lakes counties look suspiciously overemphasized. :)

I experimented a bit with the source code at https://github.com/tuttlepower/VotingMap , you can add an area correction by changing the line

    fillOpacity: (us_votes[i].total_votes / 59828),

to something like

    fillOpacity: (us_votes[i].total_votes / (59828 * 50 * (feature.properties.AREALAND / 5.195e10))),

But I think in practice, this effect is swamped by the "upper fence" effect described in the original post. In other words, the scale is very far from linear anyway, by changing the magic number 59828 above you can make it look dramatically different. I should really calculate what the (Q3 + 1.5 * IQR) value is, but above I just put in a "* 50" to make the overall impression of the map similar.

The resulting image: https://imgur.com/3K8Wwan

I was thinking the same thing. According to Wikipedia, the county areas vary by three orders of magnitude, from San Bernardino County, California (51,947 km^2) to Kalawao County, Hawaii (31 km^2), so it seems it can be very misleading.

For example, North Carolina and California have similiar population densities (80 people/km^2 versus 95 people/km^2), but on the "muddy" map North Carolina looks almost unpopulated while California is very emphasised---I guess this is because North Carolina is divided into smaller counties.

I guess in the case of San Bernardino it's a giant grey rectangle, so that particular one doesn't shift the red-blue impression very much, but still... :)

I think you are thinking of the accident in Goiânia (Brazil) in 1987, but this article is actually about the accident in Ciudad Juárez (Mexico) in 1983.

There have been quite a few accidents involving radiation sources from disused radiotherapy machines, see https://en.wikipedia.org/wiki/List_of_civilian_radiation_acc... . In particular, there was an incident in Taiwan in 1982 which seems extremely similar to the one described in this article, it also involved a Cobolt-60 source contaminating recycled steel.

The announcement is given at a press conference, so the reporters learn it at that time. The link above says that they try to call the winners a few minutes before it is announced at the press conference, but like your quote says, Doudna missed that call because she was sleeping.

I think at least the sales pitch is that the new magnets is all you need, and apart from that it uses the same well-understood technology that ITER is based around.

https://www.psfc.mit.edu/files/psfc/imce/research/topics/spa...

I mean, you began this thread by saying that they've added a new regulation on waste heat and nobody knows how to build a cooling system that satisfies it. Maybe we can begin by repealing that one and go from there...

And indeed, in Scheme (the language that originally made tail recursion famous) you can get stack traces from loops too! Well, at least in one Scheme implementation, MIT Scheme.

The MIT Scheme debugger has a notion of "reductions and subproblems" [1], which basically correspond to tail calls and non-tail calls. You can step "backwards" to the previous reduction, or "upwards" to the previous subproblem. In terms of other programming languages, that corresponds to going back to the previous loop iteration, or going up to the previous call frame.

Behind the scenes, this is implemented using a "ring-buffer or ring-buffers" [2], which stores the values of local variables for, e.g., the last 10 iterations through a loop. So if you step back further than that you will lose history and jump back to the start of the function call.

(The second link goes to the same blog as OP, I guess he has been writing a lot on this topic!)

[1] https://www.gnu.org/software/mit-scheme/documentation/stable... [2] http://funcall.blogspot.com/2009/05/you-knew-id-say-somethin...

Back then you probably couldn't electronically withdraw money from someone's account with just the account number---it has become more sensitive later.

Right. There was some previous discussion on Hacker News about this under the headline "Polio eradication program faces hard choices as endgame strategy falters" [https://news.ycombinator.com/item?id=21920406].

The article seems nonsensical in the same way, though.

> For plants, blue is achieved by mixing naturally occurring pigments, very much as an artist would mix colours. The most commonly used are the red pigments, called anthocyanins, and whose appearance can be changed by varying acidity.

It's just saying the same thing, "a red pigment appearing blue", without any explanation. (Also, as artists know, blue and red are primary colors, you can't get blue by mixing red paints.)

Maybe there's some kind of definition that make this make sense? On the face of it it seems very weird to claim that "this is actually a red pigment, it just looks blue". Why not say that it's a blue pigment which looks red in certain pH conditions?