Except that in many cases arming our allies to be able to fight off their immediate threats is much better for us than putting ourselves in a great position to be able to fight in their stead. I mean, which would you rather have: (a) a South Korean air force powerful enough to gain air control against North Korea, or (b) the US Air Force and Navy be forced to scramble a lot of their assets to Japan and fly long-range sorties from there while the North Koreans push the Southern army down to Busan?

Of course, this is not hypothetical; one of the reasons South Korea did so badly against the North in the first few days of the 1950 war was that the North had an air force and the South didn't. If the US Navy hadn't intervened promptly to provide carrier-based air patrol over Korea, the North Koreans would have won quickly.

First of all, Unix did not invent file permissions. Not by a long shot.

Other than that, well, this is a permission scheme, but finer grained and smarter than Unix permissions. Unix permissions are user based; it's primarily "who owns this file, and which users' processes are allowed to do what with it," with a few extensions like suid that are about escalating an app's privileges. So for example, vanilla Unix permissions will allow any process to access any file owned by the process' user.

The OS X application sandboxes being discussed, in contrast, are about what files or resources an application is allowed to touch, even if the application's process is owned by the same user as the file. So the sandboxes can forbid an application from surreptitiously opening files even when they're owned by the same user running the application. I.e., if one of the applications you're running gets compromised by a buffer overflow that allows arbitrary code execution, the injected code is not allowed to read arbitrary files in your account.

This of course exists in OS X together with Unix-style user-based file permissions. So in short, OS X can forbid a process from opening a file on two kinds of grounds: either (a) the process' user doesn't have permission to access that file, or (b) the process' application doesn't have permission to access that file. And in the latter case, there's also a trick that the OS and libraries work together to identify which files the user has explicitly opened through the user interface, and dynamically grants permission for the process to open those files.

Read the damn press on the case. Assange has been charged with rape. The claim that he was only charged with some absurd and untranslatable Sweden-only thing called something like "sex by surprise" is bullshit that his defense lawyers made up.

Oh, and a note: while you're reading the press on the case, be very careful to note which claims are coming from which party. A lot of the "facts" that people recite about this case (especially early on) come from Assange's legal team. Who, again, made a lot of shit up early on—at the same time when the Sweden prosecutors were being really quiet about it.

This is a misleading statement. No court ever "threw out" Assange's case. What you mean is that one prosecutor didn't think it merited further investigation, and another one believed it did.

What matters is the concrete allegations as to who did what, the credibility of the allegations, and the law. If Assange used force to push a woman down into bed because she insisted that he use a condom and he wanted to go bareback against her consent, that's rape. If Assange started having sex with a woman while she was asleep because that way she couldn't stop him from going bareback, that's rape too. It doesn't matter what label these women apply to the incident; that's up for the law to decide.

If we applied your standard, we'd do terrible things, like let physically abusive husbands off the hook because their battered wives defend them and blame themselves for the violence. (Yes, battered wives do this. Stereotypically.)

Stop making ignorant requests and read the actual allegations against Assange. For example, one allegation is that he used physical force (pushed the woman down into the bed) when one of the women insisted that he use a condom. Another allegation is that he initiated sex without a condom on one of the accusers while she was sleeping—when she had successfully insisted the previous night that he use a condom.

Your experience is insufficiently broad, and suffers from a terrible selection bias. How much experience do you have, for example, with women who are raped by their own husbands and don't report it? How many women do you estimate will admit to you that they had unwanted sex with a man who used alcohol and a little bit of physical force to overcome her resistance?

You're citing an article that starts off by citing Assange's original lawyers' bullshit about "sex by surprise." This is something that Assange's defense team made up. Assange was never accused of such a thing in Sweden. Basically, early on in this case Assange's lawyers said a lot of falsehoods about the case, and people have swallowed it.

Second, the sad fact is that a large proportion rape victims don't refer to it as "rape."

Neither. That's "why the hell don't we have socialized healthcare."

You're misinterpreting this a bit. You don't have an obligation to return an item that you find; you can always leave it where you found it. You are obligated to return the lost item only if you take it.

Informally put, the law codifies the idea that returning lost property to its owner is a favor that you can do them. But as such, the only things you can do with lost property are those that you can justify as helping the owner reclaim their property. So, for example, if you take a guy's lost phone to return it to them, you can inspect and operate the phone to identify the owner—and even call the owner using their own phone to tell them that you found it—but you can't use it to go on a 900 number phone sex orgy at the owner's expense.

Another example: you're generally forbidden to sell lost property, but there's an exception that if the property is perishable, you may sell it and later give the proceeds to the owner. Again, the reasoning is that selling the items before they spoil is in the interest of the owner, so you're doing them a favor by selling it on their behalf.

As I recall it, in California law indeed nobody is obligated to retain lost property or do anything to restore it to the owner—but the law does spell out obligations that you assume if you take the lost property. In every case you can always turn the item over to the police to rid yourself of any obligation. Also, the only way you're allowed to claim somebody else's lost property is to go through a process that involves turning it in to the police for them to hold for a specified time period before surrendering it to the finder.

There's even more detail than this in California law, for things such as lost perishable goods (you can sell them and give the proceeds to the owner) and recovery of the finder's reasonable costs (e.g., if you lose your horse and I find it, choose to house it and feed it for a week until you're reunited).

You're looking at it through the lens of survivorship bias. How do you think you'd have done if you'd invested in 1913 on the stock markets of Spain, Germany, Japan or Argentina?

No, not necessarily. You're extrapolating from the US stock market's performance in the 20th century. This is not a representative sample. Stocks may very well go down long term. Of course, if that happens, you will probably have bigger problems to worry about than "my stock index funds have been tanking for 20 years," so I suggest don't let this stop you from long term stock investing—which is meant to protect you from the risk of long-term inflation, not from long-term societal collapse.

I think I should expand on this. Back in 1902, a stock broker in the NYSE could calculate the DJIA by pencil in the back of an envelope: add the current prices of those 12 stocks up in the board and divide by 12. When it expanded to 30 stocks in the 20s it became that much harder, but you still could pay a guy to do it by hand and put up the updated values periodically up in the board. The jump to capitalization-weighted indexes with 500+ stocks was only possible with computers.

IIRC the DJIA only applies weighting to stocks that split after inclusion in the index, in order to keep the index price reflecting the actual changes in the companies.

I feel inclined to call the DJIA "retarded," but that's unfair to the people who invented it in 1896 and expanded it in the 1920s (from 12 to 30 stocks). Back in those days, it wasn't that terrible of an idea. But really, since the introduction of the S&P 500 in 1957, and total market indexes in the 1970s, there's really no point to it, and the fact that people keep reporting it is just retarded. The Wilshire is a more reasonable index—it just includes basically every US domestic stock.