I don't know in the UK, But most heating systems I saw in the US are terrible. Typically based on air conditionning with a single thermostat placed right over the main air arrival. Which means that if you set it to 70F, you get 70F at the center of the house and a gradient of temperature toward the walls and windows. And when it is 20F outside, you feel your pain. Also the thermostat is often ancient which means that the accuracy of the things has certainly decrease significantly over the years...

My mother in law still has not understood that and insist her house to be set at 65F. except that by the window, it is closer to 50F.

The main interest of Hadoop is that it makes it easy to do out of core computation if the computations are loosely coupled and are mostly IO-bound. For anything else, Hadoop is probably not the right tool and is overhyped and typically inefficient.

except the nearest bus station is not the world.

I am actually not sure how TFA comes to the conclusion that spanish would be a good second language. The question should be "assuming I already speak English, which second language should I speak." If 95% (pulled out of a hat) of spanish speakers also speak english, then learning spanish might not actually allow you to reach much more people.

Well, it is probably linked to the fact most of these companies are international companies with employees all over the world needing some form of interaction with the data.

If you really want to get an internal network that is disconnected from the internet, it means that you will need an army of monkey copying data using memory sticks to feed the data bank and bringing reports back to the employee that needs it. And that induces super high latency in the system.

The problem seems difficult to me. Completely isolated networks might have an unreasonnable operational cost. (Though a massive data breach might just be as bad.)

How does this change anything? the shareholders of Google are not in the UK and so would not be paying taxes there.

You are exposing yourself to Hollywood accounting by doing this.
Say you want to pay taxes in country FOO because the taxes are better there. You choose to subcontract your wallpainting to a company in FOO that charge you an insane amount. Your big company no longer makes any profit, and the wallpainting company is making tons of profit at FOO's rate.

Of course, this one is obvious and will certainly be considered fraud. But there are more subtle ways of doing the same thing that will look legitimate.

The summary mention an extension of the OPT visa which are essentially granted to foreign students after completion of a degree in the US. This extension of OPT will certainly benefit these students as the length of OPT typically leaves little margin of error to move to a different status.

I don't know whether it is a good thing or not for the tech industry, but there are lots of STEM student directly affected by that.

What makes it even worse is that Barbie is meant to be an image of a women that girl want to achieve.

It really is a matter of infrastructure. When I was living in france, I never drove a car. It was not useful. Driving was typically not much faster than taking the train. I could go to my university in 45 minutes while driving took about 35 minutes. But that gave me the opportunity to read in the train and to take a daily walk.
Later I was studying in Grenoble and my parents were living in Paris. To go and see my parents, public transportation (bus+train_tgv+train_city+bus) was taking about 4 hours and a half, 3 of them were in the "main train" which gave me time to do homework, read a book, whatever. The total cost was under 100 euros round trip. The same trip driving would have taken me 6 hours of actual driving (plus pauses) and cost at least 60 euros of gas.

Now, in the US, it is much more difficult becasue even if you had a good train, there would still be no public infrastructure one you arrive. But I guess you could rent a car.

exactly, if my student writes the paper, I might not read the latex file myself. What I typically do is that all things that are not meant to be part of the article is either a \note{} or a \todo{} which resolve to write in bold, red, and change background color to yellow (or green). That way, it is impossible for me to miss it before it is sent to the reviewers.

I am teaching one of these classes right now (but in a different institution). And I can tell you that there is a lot of variability in students submissions even for very simple algorithms. When I accuse a student of cheating (which happens unfortunately quite often), I usually have NO DOUBT that the code was copied from somewhere else.
You find groups of functions which match exactly code found online and which are written in a style completely different from the rest of the assignment.
You find functions with variable name different from an online template and that group of function is not indented at all, while the rest of the code is. Soon after you realize that copy pasting from that website screws up formating. No student would write it that way.
Sometimes you find comments in the code from a language the student does not speak.

In my opinion, we (whoever teaches computer science) are typically VERY conservative before saying that a code stems from plagiarism.

Since when does Intel ship devices that support nVidia CUDA?

AH! Maybe that's what I need! Bringing cookies to my class. That's smart!

Actually, this is misconception. The cloud can probably deliver 16Pflops. The problem with the cloud is not computation power. It is communication bandwidth and latency.

What makes a supercomputer is the balance between processing capability, communication capability and IO speed. For many applications, you need to be able to synchronize the processors with very little overhead. Many scientific application work under the following patterns: do a small computation, make a small communication with your neighbors, rince repeat for 10 hours. If you do not have balance capabilities, you are wasting lots of ressources. This is the type of computation the cloud can not really help you with.

Now if your application is: get 1MB of data, compute for 2 week, send 1MB of data. Then the cloud will be fine. Unfortunately, not many scientific applications follow that model.

I am not too familiar with US laws and regulations. But I assume that if the same company get caught twice or if DoL start catching one company like that every week, then the fines will become higher.