I just checked the onavo site on the wayback machine from march 27th in 2018 and it doesn't say anything about Facebook. So why do you think "those people" knew they were buying VPN service from Zuckerberg?

With your attitude towards them I cannot blame anybody for ignoring your input. Let me guess, you cannot be bothered to report the version or the steps necessary to reproduce. Do your bug reports contain more then something along the lines of "You morons, a three year old wouldn't make this bug. Fix it."?

You are misunderstanding how these models work. They exploit deep reinforcement learning. This algorithm basically turns the image into small blocks, trains partial nets on the blocks and reintegrates those nets into one and copies it to all the parts. Perhaps you remember that before deep learning was invented the image recognition programs could recognise for example a cat only when it was in the same spot of an image as it learnt from its training data. By copying the trained net to all positions this was solved and learning generally became much faster. But you now have the problem, that data distributed through out an image can influence locals parts of the net (by the integration process).

So you cannot use these images to train a network using deep learning to ignore the data, because the data is created only by the learning process. I am no expert in this field, but I understand that in image recognition only a limited number of block sizes is used and a lot of tools therefore use the same block size and are so susceptible to the same manipulations.

The last nuclear power plant will go offline next year. There is no way to stop that now, nuclear is slow with everything, so the preparations for shutdown are mostly done. You can't just simply restart it.

I am German, so I will see first hand what happens, when 10% of baseload capacity go offline. But I have the feeling, that the fans of nuclear energy are most afraid of the event, because if nothing big happens then nuclear energy looses another argument.

So in other words you can't explain it, but for some reason the correlation which seems to exist can in no case be any hint, that there might be a cause. Correlation is a first step in trying to find a cause, isn't it? Or would you go looking for a cause when it contradicts the correlation. Taking your examples, both Singapore and Hong Kong are basically cities, so not really comparable to most other countries. So those examples are straw men. Israel is a contrary example, but if it is so easy in that climate, why are all the neighbouring states so much worse. Couldn't that indicate, that Israel is also an exception? The other contrary indication you give is AC. AC comes with some cost. With rising temperature and rising fuel costs, what will happen in those areas, which can only be productive because of AC? So there is at least a possibility, that the long term productivity is linked to the climate.

So you intend to sign up to a study where you have a 50% chance to receive radiation which has a high likelihood of causing thyroid cancer? No? Why not, as you believe there is no proof it should be harmless, no?

As you only accept one standard of proof, care to show me a setup to proof the existence of gravitation. I mean you just have to do the same experiment, one time with and the other without it, as obviously only a double blind test can show that there really is an effect.

in C you know, at least, that j is being multiplied by five and the results stored in i.

Oh, really?

unsigned short i;
double j;

j = -1.0;

So where does this code store the value of j multiplied by five? How do I not have to know, what types I use. Also what about stuff like: #define j call_expensive_function()

Sorry, but the idea, that you can tell from a line of C what is happening is wrong. It is not as complex as C and often the likely result is correct, but there is nothing in the language which precludes very stupid code.

Perhaps because there are no "Miranda Rights" in Britain?

Yeah, because the previous state was much better. That area was a military training ground of the red army. It couldn't be used because of remains (chemicals, hand grenades, amunition and other nice stuff), cleaning it was deemed too expensive. Now the area has been cleaned by the company which built that solar park. They have leased the area and intend to use it for 20 to 30 years for solar power. Then the panels will be removed and recycled.

Oh, and it is indeed very friendly to the environment, as now the environment has time to regenerate after all the nice remains have been removed. The area in question is supposed to be heathland, so no high plants growing.

That is the gist of the text, which goes with your photo.

The monstrous earthquake/tsunami combo the Fukushima reactor was hit by was "obvious in hindsight?"

When there are signs, that in historic times an equally strong tsunami (and therefor probably similar quake) took place and a recommendation was made to have higher walls against floods, then yes, it was "obvious in hindsight". The obvious flaw was, that the diesel generators were located in a place which is about the worst possible place for a plant near the sea.

No longer, the pipeline through the Baltic Sea is finished and the first deliveries of gas have started. So the next price dispute between Russia and Ukraine will have much less impact

Sorry, but I very much doubt that Heise would sell that information. First they would probably get into trouble with German privacy laws and their users would be furious if that would become known, I certainly would be. Isn't it possible, that someone just does the right thing once and doesn't see any reason, why some other party (it isn't only facebook, also google+ and twitter are handled the same way) should receive nearly complete information what its users are doing on site?

While it is true that the Phong solution is still likely "wrong" due to being not perfectly accurate, it's still a lot less wrong than thermal effects uncorrected

The main problem with Phong is, that it can create energy depending on the parameters. Meaning the emitted light can be stronger than the incident light and so in that calculation create thrust out of nowhere.

Additionally the Pioneer probes are made out of metal, Phong is derived from a model of plastic. The properties of those two materials are quite different, one being a conductor the other an insulator, so the Fresnel equation gives quite different values for the reflective properties, additionally metals are often anisotropic in their reflection capabilities. This has influence on the direction and form of the lobe for the first order effects. I also don't understand, why they didn't use one of the established BRDFs which are at least physically correct.

I looked through the paper and I see no prove, that the parameters they assume for diffuse and specular reflection don't violate the laws of physics.

There is perhaps a simple explanation for it, as the script only checks the return code and not the contents, an error page created by a corporate firewall (or something similar which blocks facebook with some explaining page) could also return 200 instead of an error and it would show you as logged in.

I am not so sure, don't you think it would feel creepy, when you just looked for a book, lets say about a travel to the moon on one site and then suddenly in all the tabs open you start to see offers to name a star, buy some piece of some space mission, a DVD with footage from the Apollo-mission and so on? Everything targeted not quite right, but obviously close to what you just did. I think it will be very hard for targeted advertising to avoid, that people feel like some creepy guy is following them.

And honestly, the advertising business never had any kind of restraint. Advertising on the web got very fast very obnoxious to the point, where I won't browse without an adblocker. So I doubt, that the advertisers will be careful this time.