Yeah, that table is based on LNT, a "theory" with less supporting evidence than Santa Claus. Actually, that's not fair to Santa, since the evidence directly contradicts LNT. But LNT is mandated by law in many cases, which you should keep in mind the next time someone tells you that the left is pro-science.

LNT is "Linear, no threshold". According to that nonsense, a radiation dose expected to cause cancer in a person, but distributed over 7 billion people still causes 1 "extra" cancer in the world. This dose may not even be detectable, by the way, and would be far smaller than the ordinary background radiation levels.

In reality, people with occupational radiation doses have lower cancer rates than the general population.

I still have, but no longer need, a relay box wired to a parallel port pin on my router.

The router sniffs conntrack looking for recently used peers, and pings them. If they ping well, they get stuffed into a list of good pingers. It then hits a good pingers every few seconds. If it can't hit anything for a while, it restarts DHCP. If that doesn't work, it flips the relay for a few seconds.

Worked like a charm on my cable modem. Now that I have fiber, the ONT is rock solid. If I'm down, there is a fiber cut somewhere, or a router leaking magic smoke in their head end.

P.S. The other pins are wired to an old school LED bar display, cycling lame patterns. One is "cylon".

What is the importance of being 5 years late?

Costs Of Nuclear Power Plants - What Went Wrong?

You've done the experiment? This experiment? The one that can't work?

Or perhaps, after reading that section again, he meant an unserialized gun, rather than a self-made gun.

There are a bunch of other mistakes in the article too. The biggest one that I noticed was that he thinks that it is illegal to sell a self-made gun. It isn't.*

US gun laws are complex. I doubt that many lawyers could make it through an article of that length without a bunch of mistakes.

* Don't try it. The law allows you to make guns for your own use. You can dispose of them as you please later, so long as you intended them to be for your own use at the time you made them. If you do try to sell one, the prosecutor will argue that you always intended to do so, and the jury will believe him instead of you, and you will end up in prison.

Just to clarify, making a gun is legal in the US, for anyone legally able to possess one.

Provided that the gun in question is not automatic. Automatic means either that specially trained ATF technicians are able to coax it into firing more than once with a single pull of the trigger, or that it uses an open breech design.

Open breech means that the gun rests with the bolt back. Pulling the trigger releases the bolt to move forward into battery, where the gun fires. Special parts are needed to cause the gun to stop after a single shot, and the easy removal of those parts makes the gun automatic, even if those parts are present.

By contrast, in a closed breech gun, the trigger releases a firing pin or striker, starting the cycle. Special parts are needed to prevent the gun from firing again until the bolt returns to battery. Removal of those parts turns it into a nightmare machine, unable to reliably contain the pressure of the burning propellant.

But a closed breech allows a disconnect in the action, requiring that the trigger release for each cycle. Without those parts, or with worn parts, the gun is an automatic. With those parts, it is semi-automatic, or self-loading. Or, a lever allows the user to select between the two, making a select-fire gun.

To summarize:
Open breech = automatic (by decree)
Closed breech, disconnector = semi-automatic
Closed breech, no disconnector, or selectable disconnect = automatic.

Private ownership of automatics requires special licensing of the owner, and a special tax stamp paid on the gun. Those stamps have not been issued since 1986, but owner licenses are available.

An ordinary person can also get licensed to manufacture automatics, but because the ATF won't issue a stamp for their product, they can't make an automatic for personal use. They can only use that license only to make guns for entities that do not require NFA stamped guns, which basically means military and law enforcement.

Or, a person can get licensed to possess an automatic, and purchase a pre-1986 stamped gun. (Note that conversion devices like the Lightning like and the Drop-In Auto Sear [DIAS] count as guns here, as far as the law is concerned, even though they aren't guns.) Expect to spend about $10k getting started in this hobby.

Luty's SMG is an open breech design. Don't even think about building one. But the book is a good read. It will help you understand how the Taliban held off two global super-powers mostly using guns they made themselves. In caves. With hand tools.

Note 1: Conversion of an AR-15-clone is simple. Drill one hole in the right place, drop in one part, one spring and one roll pin, potentially swap out a few other parts, depending on the exact design of your clone, and you are done. But drill that hole without proper authorization and you are looking a 10 year felony sentence.

Note 2: Since we are living in a post-Constitutional, post-Rule of Law era, any owner of a semi-automatic gun can be arrested and charged for NFA violations at any time. The ATF technicians have years of experience getting guns to double fire, and access to soft primers that will fire nearly unprovoked. They also have all the time in the world to tinker with your gun, and they get paid a salary to do it. They will get your gun to double fire at least once, and away you go.

Sounds a lot more like "remove" to me. Deprecate is when you warn that something is going away in some future release, including possibly the next release. Remove is when you take it out of the current release.

Shit. Did I miss the memo that today was opposite day?

For the last few years, Slashdot has been running slightly to the left of Mao. Around here, conservatives are all but extinct, and libertarians are endangered. Republicans are routinely mocked and vilified. Scroll through the comments on this story and you'll see plenty of clever new ways to say "This Republican is vile, but..."

Rule of Law is pretty much dead, and not everyone is happy about it. It is a broad enough issue that it can transcend the left-right spectrum.

Ask yourself what amount of your money is it safe for you with withdraw from the bank. The law says $10,000 will get you attention for sure. How about $5000 twice? How about $2000 every day for a week? $1000 each day for two weeks? $500 daily for a month? Where did you draw the line, and why?

$10,000 in a single transaction is Rule of Law. You may disagree with it, as I do, but what you can and cannot do is written in law, and not a surprise to anyone. Structuring is Rule by Masters, with a federal investigator or prosecutor deciding if you've broken the law or not.

Spotting a car is easy. Spotting people, or other random obstacles, not so much.

Cars tend to be large and made of hard reflective surfaces. 2 or 3 ultrasonic sensors at fixed locations in the bumper is enough to notice a car and avoid hitting it. Those sensors are cheap, and you can probably run them with an 8-bit PIC.

A system to detect random objects is much more involved. More and better sensors, vastly more complicated program and a real CPU to run it. In this case, radar and a camera, both of which require lots of processing to use. All quite expensive.

Even better, the car wasn't parking itself. From the two articles, it sounds like the driver hit the gas with the expectation that the car was going to prevent him from running into people. It wasn't capable of that, and wouldn't have overridden his explicit action even if it were.

Science is a tool that we use. The concept of "wrong" doesn't apply to it.

To use science, we compare an idea to reality by making a prediction. If the prediction doesn't come out, the idea is wrong.

You may notice that the credentials of the people that support the idea are not a factor. The Royal Society took "Nullius in verba" for their motto. It means "on the word of no one", an explicit rejection of authority, and a reminder to verify all statements by an appeal to facts determined by experiment.

My field is operational. It is tested against reality daily. When I get things wrong, which happens, everyone knows. There are no boogeymen to blame it on, and no one would doubt the plain evidence of their own eyes if I started saying that it was just laymen and conmen telling them that things were broken while I, the expert, say that things are fine.

No one should be coerced to plead guilty against the threat of huge sanctions.

Prosecutions are stacked against the defendant, particularly federal prosecutions. They are alone with their own resources against buildings full of government lawyers drawing a salary, with no incentive to seek justice, just convictions to pad their stats.

By forcing him to plead guilty to a lesser charge (to avoid something silly like 18 consecutive death sentences, or whatever they came up with), the rest of us are duped into believing that he actually did something wrong. Pleas should only be allowed on all charges, or none. Anything in between is institutional coercion, a corruption of justice.

Further, there should be a very, very high bar against charging someone for going about their ordinary business. If his business isn't illegal in general, it shouldn't be illegal when government agents lie to him.

If you pre-pay at a gas station and tell the cashier that you are filling up because you like your getaway car to be in top condition before you rob a bank, is that guy now a felon for not refusing your business? By the logic of this case, if you are an undercover cop he is.

We should be pissed about this. But we aren't. Why not?

You can't tell if a prediction came true or not? Not even when that prediction was blasted from every rooftop and television on the planet for 25 years?

I'm pro-science, but only when science is the method for deciding if an idea is good or not. I have no care for science-that-falls-out-of-the-mouths-of-authority.

The predictions failed. The models are wrong. Not inaccurate, wrong. Inaccurate would be if they were off by a few percent. Wrong is when they are off by x2 or x3 or x10. Science has spoken clearly here.

And my views are anti-intellectual? :(

You could be the honest man, if you wanted to be. Pick a model, and compare the prediction for 2014 with the "data" from 2014.

Note that this has already been done for various ensembles of models. You can see the chart for yourself, on page 3 of the paper that gives you badfeels.

What's really funny about this is that the climate debate really picked up when people with actual statistics skills showed up and started looking around.

Essentially, you have it backwards. The professionals showed up and started talking about the Excel macro-level work they were finding.

Not that they needed to. When the model is wrong, it doesn't matter how skillfully it was written. It only takes an honest man to say what everyone can see.