I'm getting the impression from the article that they are proposing to use this technique to build semiconductor *components* such as standalone transistors, diodes, etc., etc.

That seems much more feasible than what is implied by the title of this post.

I *still* haven't seen any explanation of how this will avoid violating causality, in the way that *all* methods of getting from one reference frame to another faster than light does.

Of course, maybe that just means the universe is acausal. Weird, and a bit troublesome for our puny simian brains to wrap themselves around, but I suppose the universe doesn't care.

Why not just say they have to show their work by submitting a file saved periodically during the process?

Huh... irony. Moore's Law is about gates, and Gates' Law is about moors (in software performance).

While it's absurd to think about solar powering an airliner's *flight*, it's not *entirely* crazy to think that solar power could take up some of the slack for non-critical on-board power needs, such as lights, entertainment electronics, etc., at least during the day, and reduce the power load on the APU, and thus cut fuel consumption a tiny bit.

Current solar panels wouldn't work due to weight, but you could imagine the roll-to-roll printed photovoltaics that have been talked about being doable.

Eric Nylund's Signal to Noise http://www.amazon.com/Signal-Noise-Eric-S-Nylund/dp/0380792923 explores a lot of the same kinds of mathematical concepts as Cryptonomicon, but in a Space Opera Thriller. The sequel is quite good as well.

Unfortunately, it's impossible to say *anything* about this story without spoiling it, so I'll just leave it as a bare recommendation.

This is a really bad misunderstanding of how fractional reserve banking works. Banks *don't* loan money that they don't have, they can only loan out of their reserves.

The reason money gets multiplied is that money moves so fast in the system that the million dollars coming in today can also go out today. The money a bank loans out today is deposited tomorrow (somewhere), and that bank now can reloan out 90% of the $1 million hypothetical loan, and all of this happens so fast that it *looks* from the outside like they just create money

Fractional reserve banking is entirely feasible conducted purely with real gold, though would multiple money slightly more slowly due to the physical transfer times of the gold. Which is why gold-based currencies never actually use the physical gold as a counter any more (technologically advanced counterfeiting making currency validation almost impossible at any reasonable speed, aside).

Indeed, we know this, because FRB was *originally* invented by goldsmiths accepting deposits of physical gold as currency and loaning it out (though, for the above reasons, usually as receipts for delivery of gold rather than physical gold).

It's not really the math itself that you use in programming. It's the mental processes that you learn learning math that are *also* useful in learning (and doing) programming.

I won't go so far as to say that anyone who hasn't done advanced math *can't* be a good programmer, but I will say unequivocally that it's highly unlikely.

But more importantly, being a really good programmer absolutely requires that you be *curious* about algorithms. It's mind-boggling to me that anyone sufficiently curious about algorithms to be a good programmer would even *ask* the question "do I need all this math?". Who cares? Why don't you *want* to learn it?!?!? Hint: "Just for the joy of learning it" is the right answer.

The rubber hose cryptanalysis won't be able to get you to *tell* them the password, but I see no reason why they couldn't get you to *reveal* it by playing the game.

Since it's subconscious, you won't even know that you're revealing it.

The basic problem with this is that cops' job is to collect evidence of crimes and arrest those they think are responsible. It's not the public's job to do this.

When they release their evidence, they taint it for any future court proceeding. Even if it were allowed, it probably shouldn't be in general because the public (including the jury) can be presumed to have seen it.

Of course, if they aren't planning to use it for evidence, it's technically ok for them to release it (though there are the same concerns about everyone's privacy as when the public releases it), but it doesn't seem like a good idea...

Because it sends the message that there is no chance that they will be prosecuted for any misconduct regardless (otherwise this would be tainted evidence for *their* trials).

I really wish that people who used this analogy would go look up the case where it was created.

You know, the one in which publishing a communist newsletter was deemed equivalent to shouting fire in a crowded theater.

It's very dangerous to create slippery slopes, and pointless when there are adequate ways to deal with the problem already.

So, in this day of actuators and microcontrollers, how would you actually do this without getting rid of *all* patents?

It's become trivially easy to implement *any* part of any mechanism in software.

It's easy to make platitudes about this stuff... actually coming up with a rigorous legal definition that would rule out the things you like while still allowing the "good" patents is really hard, if not impossible.

Of course, perhaps all patents have outlived their usefulness... but that's a different discussion.

Readily apparent to one's intuition.

The intuition of a person that has taken (and paid attention to) college-level courses is *of course* more efficient at comprehending things that are the topics of the courses.

Perhaps you were thinking of common sense?

The vast majority of people have *crappy* intuitions.

Proof that reddit is faster than Slashdot, I suppose...

I still say it should be "Santorum squeezed out after he repeatedly comes in number two".

Yes, that's cute, but the differences are *far* more than just the hybrid system. The $16k Civic has a manual transmission, a different body style, no Bluetooth, no auto-up windows, no power locks, no cruise control, a much less sophisticated information cluster, and it doesn't even come with a *radio* or *air conditioning* standard.

In order to get anywhere close to the same car minus the hybrid system, you have to go up to the EX level, which starts at $20,655.

The hybrid system itself, therefore, is about a $3500 premium.

Now, of course, you may not *want* all those extra features (though most people want most of them)... and if so, then the Civic Hybrid model isn't for you... but it's not because a hybrid costs 50% more for 13% extra mileage. It's because of marketing.