I hate marketing stuff, I hate business stuff, and I really hate "networking" .. what I love is building software.

Me too. And that’s exactly what I do. Where I work, management handles the business stuff. You just have to keep looking for the right company where they realize that a first rate programmer is a valuable resource even if they aren’t into the business stuff.

The problem lies in the fact that doing well in those types of courses requires a certain type of analytic thinking that is simply not that intuitive for most people.

When I started college I majored in Physics. It was hard. I switched to the much easier (but still hard for some people) Computer Science major. After I graduated I got a job in software development. It was hard. These disciplines are inherently hard. Making the curricula easier may help graduation rates but it is not going to prepare students for careers in STEM. Most of the people who drop out of STEM and major in something else are going to lead happier and more productive lives than if we somehow kept them in a discipline that they are not suited for.

You obviously don't need computers to teach, but to claim that can't be helpful is just Luddism.

I agree. It's obviously a question that could be decided by a few carefully designed experiments. To make a blanket assertion without any evidence is not what you would expect from an educator. Given the importance of the question, I'm surprise that someone hasn't done the research.

When I am developing software, I have a lot of new ideas, many of them at least as good as some of the software patents that I have seen. My motivation for coming up with these ideas is to make my software more efficient and more reliable, not to patent them and keep the company lawyers employed. If I had to worry about whether someone had already patented one of my ideas, my productivity would come to a halt.

he earns like 3x normal pay if it's a country with an active war.

They would have to pay me a lot more than that.

They do not get twice the pay. No way in hell.

Right. I think the contractor's employee gets paid about one third of the billing rate. The rest of the money goes to paying for office space, taxes, support people (managers, accountants, secretaries), company profit, training, etc. Hiring contractors gives the government a lot of flexibility. If they need a system that uses technology X, they can hire a contractor with the relevant experience for the project, and when the project is over, the contractor is gone. Also, it is a lot easier to avoid a contractor who has done a poor job in the past than it is to fire a government employee.

Professors may need to ponder the "shared responsibility in the collective metropolitan realm", or at least pretend to, but engineers need to develop innovative products. Different jobs need different environments.

I can understand and empathize with that, because I was in that situation myself, at the time. About half the summer hires quit after a couple of weeks. Those of us that stuck it out were not the strongest or the toughest, but the ones that needed the money. I applied at the same place next summer, but the job I had been doing had been automated. Outsourcing is not the only way you lose manufacturing jobs.

We once were a manufacturng giant. People had decent middle class jobs.

When I was in college, I had a manufacturing job one summer. I stood on an assembly line doing an incredibly boring task until my hands were blistered. Being a manufacturing giant has its draw backs.

I live in the Washington D.C. area. I survived the deadly earth quake that ravaged the area last week. We are currently getting some rain and light wind, so I think I will survive the hurricane, also. What has me worried are the locusts and frogs that are scheduled for next week.

If I add a new public function to a class why should have to re-compile all the old code that doesn't use that function?

C++ requires that the size and layout of an object in memory (its vtable) be known at compile time for efficiency. Otherwise, there would be another level of indirection requiring a second memory read. From what i have read, Stroustrup was very aware of the concern for efficiency among C programmers and wanted C++ to be on a performance level with C.

you're being ridiculous. Anyone who types that is an idiot

There are are a lot of examples for type inference that do exactly that. I would still rather have a more verbose and informative typedef in code that I am reading than the word "auto".

OK, so you consider concurrency useless?

No

And lambda functions?

Yes

And type inference?

Especially type inference. Now we can write

var area = 0;

Instead of

double area = 0;

Did you catch the problem with the first statement? You can’t from context. It would pass the compiler and introduce subtle errors in you code as the “area” variable is always truncated to int. All so you can type var instead of double.

Have you SEEN the stuff they're adding in the C++0x standard?

I have the same impression. A bunch of useless, error prone, legibility degrading changes to save a little bit of typing. In my opinion this is the worst standards committee ever.

And how much does a ticket to Japan cost?

Yea, we need to get out of here before this place fills up with Brits.