Heart attacks do indeed come out of nowhere. You can have a good idea of the chances, but not whether or not one will happen. Mine sure surprised me. If it had been a more serious one, and I hadn't gotten prompt treatment, it could have killed me. People do die of single heart attacks.

No large species we know of, past or present, is as adaptable as we are, or able to survive in so many different places. I'd bet that the human species would survive any of the common 60-megayear ELEs. I don't know how many would survive, so the 6 billion might be accurate.

On the other hand, all that energy for cooling has to come from somewhere. If it's fossil fuels, we're better off reducing it. I think insulation is a good idea, and it often makes economic sense.

For one thing, the Antarctic land ice has been melting, lowering the salinity near Antarctica, and making it easier for water to freeze.

You know, reading through that stuff, I see that models are doing pretty well. There is a discussion of the slowdown of observed warming, including speculation on where the extra energy could be. Overall, it looks like a good scientific discussion, with confidence levels and admissions of anomalies, and it comes out concluding that models have improved since 1990.

Hey, you hire some of the unemployed people to go around throwing rocks at windows and the rest to fix the broken windows. Everybody's got a job, they can afford their drugs and alcohol, and the prostitutes can charge more. Everybody wins! Or maybe we shouldn't set people up permanently with makework, when what they're doing isn't useful any more.

Anybody who would draw up a detailed plan for everything you say would be an idiot, since most of that is better done by the market. Government can help out the structurally unemployed with their problems, and everything else works reasonably efficiently.

Raw pointers work nicely as non-owning pointers. If the raw pointer goes out of scope, normally or through an exception, nothing happens to the data pointed to. In many situations, this is what you want.

In my code, there are raw pointers, but no explicit use of "delete" outside of destructors.

I'm not understanding this. It was my impression that Java rigidly defined elementary data formats. In Java, an "int" is a 32-bit twos-complement number. In C or C++, an "int" is some sort of signed number at least 16 bits long.

C and C++ are more transparent languages, in that you can see more easily the general sort of machine code they're going to produce. Java is very dependent on the quality of the JVM in use, more than C and C++ depend on things.

Thing is, it's easier to for a stupid developer to get something apparently working in Java than in C++ or some other languages. It attracts stupid developers more than most other languages.

Operator overloading is very useful, when done properly. It can make things much easier to read and write. It's really easy to abuse, and then it can make things unreadable. In general, it takes a lot more knowledge and experience to use C++ safely than to use Java safely, but it's possible to express things better. (C++ has become much easier to use properly with C++11 and C++14 features, also.) One of Stroustrup's design principles was to keep features on the basis of how they could be used, not how they could be abused.

I really, really can't get excited about automatic getters and setters. Getters and setters are bad for encapsulation and promote unhealthy dependence on internals. I have no problems about writing them when I need them. (I do like the C# feature that allows me to have a virtual variable that has programming logic behind it.)

There isn't much that Java is "very good at", but overall it is a good language for some purposes. It is a decent procedural language slanted towards O-O programming (nominally it is an O-O language, but it's easy to use as a procedural language). It's something of a B&D language, so mediocre programmers are unlikely to mess things up too much. It has too much C syntax (including the stupid switch statement), so it looked familiar. It had a lot of advantages in the 90s, being early to implement garbage collection and to provide a very large standard library. It got a leg up with Java applets on the web (although that use is pretty well dead, it helped Java get popular).

It's not a great tool for expert programmers, but it's adequate and fairly safe for mediocre programmers. In general, it's a good business-oriented language.

The US Federal government is set up to require a broad agreement to make laws, either a majority of both houses and presidential approval, or two-thirds of both houses. The filibuster is a mechanism for a single Senator to impede the progress of a bill, provided he or she is willing to keep standing and talking for a very long time. This gives Senators a chance to express extreme opposition to a bill, as long as they're willing to do the work.

The compensation program has a couple of purposes. First, pharma companies make very little on vaccines, so any legal trouble would likely drive them out of the business, and we'd have no vaccines. Second, in the rare case that somebody has a problem with a vaccine, it is possible to compensate them without going through the ordeal and delay of a lawsuit. Since any medical procedure has dangers, the government should be ready to help out people who were hurt by a mandatory procedure.

Sure, the people distrust the media. That doesn't really help.

Given one unreliable source for facts, and nothing conflicting, people drift into believing the unreliable source even while they note its unreliability. Read some ancient history if you like; the field is rife with accounts of events that have no conflicting accounts extant. Check on how many people repeat the story about the 1.7 million strong Persian army invading Greece because, although Herodotus was obviously wrong, there is no other figure from ancient sources.