I wish I'd had the equivalent when starting at my first University back in 1975.

Typo: Back in 1965.

By the way: Charter Oak's mandatory "cornerstone" class was a wonderful experience, and just what I needed. Think of it as a boot camp on how to research and write academic papers (and read them). VERY enabling.

I wish I'd had the equivalent when starting at my first University back in 1975.

I'm using Charter Oak State College. This is because they are an accredited and respected school, specializing in in distance learning, which accepts a high enough score on a recent Computer Science GRE exam for credit-by-exam on many of the course requirements, and (if certain requirements are met, such as being active in an industry using them) accepts older credits (i.e. math, distribution requirements) that many other institutions would consider "stale" and timed-out. They do not currently offer a BSCS, however, so I'm going for a BSGS with a CS concentration. I've taken three classes from them so far, and am currently taking a math class (needed but not offered there) from the University of North Dakota for transfer credit. Several of my few remaining requirements I expect to complete by exam (and a couple - public speaking, English - by waiver due to documented work experience).

Another excellent school (with a more technical orientation) that also specializes in distance learning is Thomas Edison. They do offer a BSCS, but stopped accepting the CS subject GRE for credit (because too few students used it for them to continue the effort of keeping it qualified against their own requirements). In my case this made a big difference in how much work would be required to reach the diploma. For some others, especially those who need many of the classes (or test-out equivalents), are tech focused, and would find the more directly applicable degree a benefit, Thomas Edison would be a good choice.

These schools are oriented toward people who wish to complete their degrees but are employed or located where going to a classic college is impractical. Examples: Deployed military personnel, low-level medical employees seeking higher certifications for career advancement, workers located far from a good subject-appropriate school or working schedules that interfere with school scheduling.

Two credit-by-examination programs are also related to this. DANTES is one - driven by the military's need to provide education for their soldiers without interfering with their duties (but open to all). Another is CLEP (College Level Examination Program), a product of the College Board which provides 33 subject tests which are accepted as proof of accomplishment by most universities. Each program lets you receive college credits without actually taking a class, by testing whether you've successfully taught yourself the subject.

Whenever someone trots out that tired old tripe... I'm reminded of the evolution of submarine ballast and trim systems. The earliest were supremely simple (read obvious), but had deadly (and non obvious) failure modes... (Long story short, they were unstable longitudinally and could easily pitch the boat onto it's beam ends.) John Holland's great breakthrough was to make the system more complex - and his concepts from the late 1800's can be found replicated almost exactly (though more complex in detail) in submarines being built today.
 
And you see the same thing replicated in engineering again and again, the trend is towards more complexity (and higher performance) - not towards simplification.
 
The moral of the story? Don't take engineering advice from non engineers.

I've done that. Ancient General Automation GA-16/440 minicomputer using their native Fortran 66.

I'd set a constant = 5, then inadvertently used the constant as the result variable in an expression, and, sure enough, it was 5 that I'd redefined. Incredible whacked-out results from then on.

Took one hell of a long time to debug.

I think that bug, and all the accumulated strangeness of writing for that platform, is why I feel relieved that I don't write software any more.

No, I didn't. ar is the real part of a, ai is the imaginary part. Same for br, bi and cr, ci.

But sometimes spreading an operation over several lines makes it harder to read. For example, given the lines

ar = br*cr - bi*ci;
ai = br*ci + bi*cr;

I can spot immediately that this is a complex multiplication. Spread the same code over several lines each, doing only one operation per line, and It will take me some time to figure out.

As long as the data doesn't have to be turing-complete I'm satisfied.

I see, you wouldn't want to write a compiler. ;-)

The is a good reason I need a separate class defined to output each character in the string "Hello World!"

I see, you have to work on your abstraction. You certainly have to separate your character classes from your output classes. Don't forget to make proper base classes for both. For the abstraction, you know? Who cares that you'll only ever have one output class?

Learn some quantum mechanics. And no, I don't mean popular science versions of quantum mechanics. I mean the real theory.

And yes, I am a physicist. I have learned that stuff. And no, the uncertainty relation is not just a limit on measurement. The quantum state itself does not and cannot contain information about exact position and exact momentum at the same time. Indeed, even though you can formally write down states where at least one of them is exactly defined, those states are not physical. Physical states have neither a well-defined position nor a well-defined momentum. Independent of any measurement.

Of course you can beat it. It just doesn't make much sense.

Ah,. now I get how you got relativity in. Well, if we measure the same particle, we necessarily do it at the same location, and the relativity of "same time" only happens for measurement on different places; more exactly, for spacelike intervals. Operators belonging to spacelike separated points commute, and therefore there's no uncertainty relation between them (well, formally there's the uncertainty relation with >=0, but since on the left there are only nonnegative quantities anyway, that inequality doesn't restrict anything).

Because the only thing that relativity changes in respect to the uncertainty relation is is that the velocity is no longer proportional to the momentum, so you cannot say "velocity" instead of "momentum" here (in particular, the uncertainty relation also holds for the photon, but that doesn't change the fact that the photon goes exactly with c, because for photons a momentum uncertainty does not translate to a velocity uncertainty). But other than that, relativity doesn't add anything relevant to the uncertainty relation.

Times a million. Chas just won the entire Internet with that post.

PC manufacturers like HP used to void warranties when clients installed GNU/LInux, not anymore.

Just a year or so ago I bought my wife an HP laptop specifically for a sysadmin class where she'd be installing Linux on it. Got it home, had a question for HP about it, and discovered in the process (from the phone support) that installing Linux would void their warranty. Checked the paperwork: Yep! So we returned it to Staples for a full refund and went with something else.

When did HP change this policy?