Neither this law, nor the original version of it, would have retroactive applicability; in other words, you can't make something illegal today, and then prosecute the guy that did it yesterday

It also wouldn't apply outside the US either, which is sort of a problem. What's to stop a Wikileaks clone starting in say, Iran, and doing the same amount of damage as Wikileaks?

It's weird in a way, I thought America came out of the Wikileaks cables pretty well. America acted, for the most part, in private exactly as they acted in public. It's everyone else that looked like a bunch of douche bags.

As a Brit, it's done more to repair the reputation of the US than, say, the election of President Obama. I've actually surpised myself with that sentence, but it's actually true!

Mass bends space-time, right? So why not define it as a certain amount of curvature - say the mass needed to bend a light beam in vacuo by some measurable amount, divided by a chosen constant to give 1kg according to the theory.

That would couple the kilogram to G, which we know less _much_ less precisely than the current kilogram.

A better, but equally flawed, solution would be redefine the kilogram in terms of the electron. After all, we can measure its rest mass to a great degree of accuracy. But then usuable amounts of mass would be inaccurate to the level of precision given to Avogadro's constant.

It's a harder problem than it looks.

What part of this does the US legal system not understand?

Excessive bail shall not be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted.

And I say this as a Brit!

A test suite that guarantees 100% coverage is called formal verification

While I agree with most of your comment, that part is not correct. Having 100% statement coverage is not enough to achieve a formal verification.

There could be:

1. Missing code - e.g. an exception that isn't caught that leads to the crash of the program in some circumstance outside your test cases (see Ariane 5).
2. A sequence of calls that leads to a problem. e.g. A procedure has three upstream callers, and you got 100% coverage using two of those upstream callers. The third one doesn't satisfy the pre-condition and "boom."
3. The code might actually be incorrect with respect to some edge case. For example, a 30 year old bug was recently discovered in a sort program because it used 32-bit ints. It failed to correctly sort a list larger than four billion items.

In short, you can never test you way to perfect software, unless you can prove your test cases constitute a formal verification. It shouldn't surprise you that this is as hard as simply verifying the program.

Those who ride longer distances/more strenuous routes seem to think it is worth wearing even when having to put up with ribbing/heckling from people like you, so apparently the benefits are noticeable enough to make it worth their while to pay the extra expense of cycling gear

I cycle 13 miles to and from work every day wearing Lycra. I know I look like an idiot but wearing that stuff gets me home about ten minutes faster than wearing t-shirts and shorts.

I'd rather have 50 extra minutes with my wife a week than Kowtow to the hecklers. Besides, most of them are throwing stones in glass houses anyway - they're hardly supermodels are they?

How do you compile a compiler written in the language it compiles...

Enjoy

Would this ever happen here in the US (you know, the home of the free)?

A month ago we were the laughing stock of the world. Every (learned) American knew the UK as the place where the nightmare of 1984 was coming to pass. The slow drip, drip, drip of authoritarianism would erode one of the world's oldest democracies.

Then came the election and the hung parliament. Suddenly the Conservatives and the Liberal Democrats joined together in to a machine that seems to be greater than then sum of its constituent parts.

Overnight, the country has gone from a 1984 wannabe dictatorship to a liberal paradise. They even have a plan to pass an act that prevents unnecessary criminal statues. WTF?

If this works like the Parliaments act in that the same legislation must be passed in two Parliaments for a criminal statue to become law, then this would be a giant constitutional innovation. Forcing there to be a general election between criminal bills would prevent some of legislative diarrhoea that we saw towards the end of the last Parliament.

I feel proud to be a Brit again!

Simon

Which is why I don't like push-button ignition. If my car ever goes into hyperdrive because of a stuck throttle, I take comfort in knowing I still have a kill switch, and I grew up driving tractors and cars without power steering or power-assist braking, so I can cope.

If you gas pedal ever got stuck down, the solution is not to turn of the engine. This can cause all sorts of additional problems.

You put the car in to neutral. This disconnects the engine from the transmission allowing you to stop safely. I believe even automatic cars have a neutral gear, so this technique will work for those cars too.

This technique should be taught as part of learning to drive, whereever you live. It takes ten seconds to explain and could save quite a few lives.

and thats just the problem, they rely on models, which are hopeless (i write software that produces resource industry models for a living, so i know some of the pit falls). let me put it to you this way - the weather man can't predict the weather for the comming week. but for some reason you think they can predict the weather 100 years into the future accurately?

Are you surprised casinos make money? After all, I can't predict every spin of the roulette wheel, every throw of the dice, every card that comes out. Even so, casinos do in fact make money! They do because over many iterations the games are balanced in favour of the casino.

The weather is much like the individual spin of the roulette wheel or the throw of the dice; it is inherently unpredictable. However, just like the casino games, the overall *statistics* of the weather is not unpredictable at all. We call these statistics "the climate."

So not being able to predict the weather has absolutely no bearing on whether we can predict the climate. The two problems are entirely separate.

However, it is obviously much harder to model the statistics of climate than it is to model the statistics of the roulette wheel; and the accuracy of such models is currently what is under scrutiny.

Simon

That's why so many sites are adopting OpenId.

I'm not sure why people are adopting OpenID. It requires all this extra overhead of going to and from an additional authentication server. It's a complicated protocol and complexity breads insecurity.

If I use OpenID I've gone from one point of failure (the compromise of my computer) to two points of failure (compromise of the OpenID provider and compromise of my computer). There's actually a third potential point of failure in that the OpenID protocol could be flawed in some way, which compromises all OpenID providers.

What's wrong with entering a entering a username, the site replying with a challenge token? I then sign the token with my PGP key and access is granted. You could make this extremely painless by making a browser plugin that handle most of the leg work.

Now I'm back to a single point of failure and the security of the login authentication has been substantially improved. With OpenID I've created a separate point of failure and I'm still stuck using crappy password authentication.

OpenID is a pretty crap solution to this problem.

Simon

The kid will get more fun doing simple web pages that he can see results in right away (important to get that positive feedback going, or he'll go back to his psp or wii or xbox or whatever). html+javascript gives him that, plus you don't need anything to start except a web browser and a text editor, - not even a web server.

Perhaps even more importantly, he'll be able to show his friends and his friends will understand what he's done. They might even think it's cool.

A tiny console app will not have the same effect.

What obnoxious pricks some people are who think that we are somehow more or less a part of nature than any other dumb animal. Our tool usage, inventiveness, resource gathering and exploitation, sexuality, cruelty, and civilizations are all perfectly natural behavior for the critter we are.

For most of civilized history, you are correct. Natural selection was still busy selecting for the local maxima of reproductive success - despite our outwardly civilized appearance.

However, we're very close to rooting the whole set-up. No other species has a welfare state! No other species remodels the entire Earth to suit them! No other species has become capable of directly modifying their own genetic structure! We're getting better all the time at subverting the natural order of things.

Yet, that's probably a good thing. Natural does not always mean good. In the wild, natural selection is a cruel executioner. Most infants die before adulthood. Many adult die of disease or malnutrition. Even fewer of these adult actually have offspring. And only a elite handfull become, in the fullness of time, the common ancestor of the whole species.

"Natural" is exactly the opposite experience we should wish for our children and for each other.

Simon

There are a couple of things that might prevent this:

1. The sneakernet
2. Wifi-net.

The Sneakernet

Disk capacity is now so large that now you can carry terrabytes of data around with you in your backpack. It will be quite feasible in a few years to transport every top twenty game, music album and movie for the last five decades in a single backpack. This amount of data is so large and so easily duplicated that it would be almost impossible to stop.

Unlike packets, that go through certain "choke-points" the sneakernet really is a peer to peer network where each node is created equal. Even if they won the battle to censor the Internet, it would be no use.

The Wifi-net

There is another peer to peer network being created right under our own noses!

How long it is before people realise that we can chain our wifi access points together to make a shadownet. An Internet that isn't run by big business but individual access points relaying traffic to each other? How on earth do you even start to control that kind of network.

I find that the attitude "This system is rubbish, we need a rewrite" is the surest sign of a junior developer. Everyone who has read anything about the history of building software systems know that this attitude will only lead to failure.

ALSA is a classic example of the second system effect. They took something that worked but had some limitations and replaced it with an absolute monster that doesn't really work, even today.

It is often the case that the most important work in making a great application is the most dull. Desktop Linux projects have a lot to learn in this respect.