Engineering, at its root, is the practice of taking abstract reasoning into physical form. Nobody might have ever seen a certain kind of widget before, but if you know the right equations and do the math right you can make that widget and know what it will do. This leads to a tendency to take beliefs seriously and to apply them consistently that can be dangerous when mixed with the wrong kinds of beliefs.

People are good at wearing beliefs like clothing to impress others and not really acting on them. Christians might believe "Its good to give all your money to the poor" without actually believing that they should give all their money to the poor. We're taught one thing explicitly, but by watching how other people act we learn to do something else implicitly. Its non-trivial to learn to be an engineer and take explicit ideas seriously in your professional life while not doing so in your religious life, but we as a culture have generally learned how this is possible and Christian engineering students grow up with lots of good role models showing them how to compartmentalize their beliefs. Sometimes it doesn't work, though, and the student becomes Bible literalists.

Muslims studying engineering in other countries, however, don't have the advantage of role models in how to continue believing-but-not-believing and so its far easier than it would be in the West for someone to come along and persuade them that they have to take their religion seriously.

Freedom of speech means sometimes being offended. The fact that some people are offended by four letter words doesn't mean you can't use them, but by the same rules the fact that you are offended when others write censoring software doesn't mean that they can't do that either.

It might not protect that much against high explosives, but the main danger facing military personel sitting in rooms on bases is mortar or artillery fire. Those mostly kill through shrapnel which I presume this wallpaper would stop rather effectivly.

More that I think that the one good thing about capitalism is that it works so well. If it gets beaten in some special circumstances (like the production of non-rivalrous goods) then I'm not going to be shedding any tears despite the fact that I might call myself a Libertarian. Free markets, on the other hand, _are_ something I'd attach moral significance to.

When socialism can defeat capitalism in the free market it deserves to win.

I've never seen a political group where all or even a decisive majority of its members were reasonable people.

It seems to me that rather than the identity and timeframe for the different technology nodes (which anyone who knows Moore's law could have given in advance) the interesting thing from that slide is what it says about delay scaling and energy scaling. Whenever you shrink your process you have a certain amount of gain that can go into either making the chip faster or making the chip more power efficient. For a long time back in the day people wanted to stay at 5 volts to preserve compatibility, so everyone just kept putting it into going faster. Nowadays chipmakers try to go for a more balanced strategy.

But here, on this chart, Intel is saying that they're going to a delay scaling of "~1", staying at pretty much the same speed. And they're looking to increase their energy scaling from "~.5" to ">.5". So it looks like we really have topped out in terms of GHz.

To quote TFA

"I thought it wouldn't make much sense to go through all features of the phone just yet - I'm not lazy, I just have to leave some bread for Nokia's announcement guys!"

So I suspect you didn't read TFA very well.

Presumably the oxygen is going to released again when you recharge the battery. That's what recharging is, reversing the chemical reactions that took place during the discharge.

Actually, when the FDA delays the introduction of a new drug there generally isn't much outrage, at least among the general public in the way that there would be if they let through a drug that causes negative side effects. There are some cases, drugs relating to AIDS or abortion for instance, where there are organized political groups trying to get the FDA to approve new drugs, but otherwise denying approval costs the FDA nothing and its a testament to the fact that government bureaucrats aren't entirely selfish that new drugs get approved at all.

Still, the research I've seen says that FDA approval process for new drugs (we're ignoring food and quality control here) saves about 2,000 lives in the US every year, but at the same time causes between more 5,000 and 20,000 people to die for lack of FDA controlled drugs. Thats only one study and if other people can come up with others I'd be happy to see them.

I suppose it was appropriate that they didn't include Emacs in the list of cults, since it actually has its own church.

While I agree with a good bit of what you said there, you have one thing reversed. While no knock warrants and such mainly entered the public conciousness with the Patriot act, they had been on the books and been used for quite a while before that in drug policing. I suppose democrats, who do most of the worrying about civil liberties, were less inclined to worry when there was a democrat in the White House.

And Diebold ATMs cost ten times as much as Diebold voting machines.

Since the layer is implemented in hardware its performance impact isn't much, just a pipeline that's a bit longer resulting in a higher penalty for a missed branch prediction. The fact that they've decoupled the instructions they use inside their chips from the instructions people put their programs in is actually a big advantage in one sense: they get to mess with their internal instructions whenever they can gain some advantage by doing so and not worry about the compatibility implications.

I'm sure the instruction translation does have some impact on power, but I'd be flabbergasted if it was as expensive as, say, branch prediction or scoreboarding. So for the Atom (which I'm pretty sure has a scoreboard, but no register renaming or out of order execution) x86 is still competitive, but x86 would probably have big problems moving to a lower performance/power architecture.

Maybe that's how they run things at coushy institutions like Harvard, but at MIT we have higher standards than that. If someone can't hack MIT its better that they either fail quickly before they've wasted too much money or learn how to study effectively early so they end up learning stuff all the time they're there.