It isn't so much that underclocking will give you the benefits, but undervolting.

Power, if I could remember the damn equation through this cold, is depending on voltage^3.

Lowering the frequency will allow you to use less volts to keep the transistor stable. Underclock, and lower the voltage, test the stability. Find a undervolted value that you like, cook it with a stress test for 24 hours or so, and then watch the difference.

The rest of your reply after 0.25% doesn't really apply because this is a phone, using an ARM chip that doesn't provide nearly comparable power to a desktop.

As an previous end user of a Symbian phone, the phone was slow. It started up programs slowly, it handled task switching slowly. That's the part that matters to the end user, not stats and arguments made with the modern desktop hardware in mind.

And this is part of the reason why I stopped using my Nokia N82. The other part was that it crashed so damn often when I'm using gmail and opera. To add insult to injury, microkernels are suppose to allow more graceful crashes of kernel level components.When gmail / opera crash, it's white screen, and the only input that it would take would be to power off, and then to power on. So where's the microkernel advantage here? Clearly something was wrong with the network code, otherwise it wouldn't have crashed like that. But I don't exactly call a pure white screen with no options but to power-cycle "graceful handling of crashes".

One of the most overlooked computer problems are faulty power supplies that cannot give power near what the specs says.

CPU may be the brains of the computer, but the power supply is the heart, supplying vital electricity to all the component. Too often, I've worked on machines where as soon as I plug in a cheapo tester, nothing lights up as the proper voltage, yet the machine still manages to "run".

That could possibly be monitored with software, if the BIOS supports voltage monitoring.

They have been endangering my bank account and my credit rating for years. Half the time they release books that suck, at ridiculous prices, and since the bookstore doesn't have a used copy, I end up paying for the full price.
Textbooks that are worth their sticker price are rare. The majority of the text aren't worth half of that sticker price.

It is a rather poor explanation...
But their compressor is catered to executables, so that it transmit a sort of primitive assembly language, so that the diffs are smaller, transmit that, and have the client end to apply and reassemble.

So their compression algorithm has little applications outside of executables.

Are you trying to get experience so that you are more marketable in the job market? If so, and you're in a post-secondary, try internship / co-operative education.
If you just want to program for the sake of programming, try contributing to an existing project.
If you're trying to learn how to run one, start a project.

This is by no means the definitive way to do things, nor correct, but I think you get the idea.

Seems like NSA is constructing a much bigger site according to the Salt Lake Tribune.

Try RTFA. The attack is against AES-256, which although the attack is still theoretical, 2^119 is more than collapsing the keylength to less than half its original size.

That actually doesn't look too good, because as pointed out by Scheier, attacks are only going to get better.

NSA can't guarantee that they're collecting American citizens' email, even if they try.

Anyone want to propose a solution?

If your book's already being distributed illegally, that means that a) you've got good info that people want, or at least a professor use your book
and b) people didn't care to fork out money for the paper copy, or they don't want a paper copy.

You can create your book, without the publisher and distribute on line without the cost of going to the printing press, for much less. I'd easily fork over $10 for a good digital copy of a book.