Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
DEAL: For $25 - Add A Second Phone Number To Your Smartphone for life! Use promo code SLASHDOT25. Also, Slashdot's Facebook page has a chat bot now. Message it for stories and more. Check out the new SourceForge HTML5 internet speed test! ×

Comment Re:Subtraction... (Score 1, Funny) 124

Yeah, you're butt-hurt because you didn't do the word problem right originally, but that's no reason to keep spouting nonsense.

Keeping the same number of significant digits in the lower and higher efficiencies, a 3 percent increase from 25.6 would round to 26.4, and a 2 percent increase from 25.6 would round to 26.1.

This means that to describe a percentage increase that properly rounds to 26.3, you need one more digit, and once you decide to add the extra digit, you want to make it as accurate as possible. The 2.7 percent given in the article rounds to 26.3, as accurately as possible.

Comment Re:Are our lawyers really this clueless? (Score 1) 58

WTF are you talking about, Willis? Open source licenses are only effective because of copyright (and patent, in some cases).

If you tell someone that there's no warranty, that's a disclaimer. No license or contract required, and the disclaimer may or may not protect you, but that's another issue.

If you tell someone "you can only do 'x' with this" then you have to have a right to create that restriction. If we're discussing tangible property, the right is usually pretty obvious, e.g. you own the property. If we're discussing intangible property, then there are very few valid rights that allow imposition of restrictions, such as copyright or trade secret. In the case of true open source, trade secrets are pretty much non-starters, so that basically leaves copyright.

Without a valid right to impose a restriction, the restriction is void and unenforceable.

Comment Re:Are our lawyers really this clueless? (Score 1) 58

It says, if you're in a country where copyright doesn't apply then there is a contract enforcing the same conditions as the license would.

Except that doesn't apply in the U.S. Copyright 101 (and a zillion court rulings) say that you can't use a contract to do copyright-type things -- unless you actually have a copyright.

Comment Re: I blame Trump. (Score 1) 1149

Ahh, the old "you're too stupid to understand," coming from someone who apparently doesn't realize that when you slot in any reasonable approximation of the current US population, the resulting numbers (2.4e-6 and 5.1e-6 for black-on-white and white-on-black, similar order of magnitude for the ones you claim are necessary for proper reasoning) are so small that any effect from the conditional probabilities will be well down in the noise.

Comment Re:I blame Trump. (Score 1) 1149

Another way is that black people do a lot more murder on a per capita basis.

Absolutely. And they kill much more inside their race than outside it, by a factor of 5:1. Proximity appears to be a major factor in the statistics.

the chances of getting killed by a white guy are less for a black person than the reverse.

No, that's not what your numbers show. It's a subtle distinction, but while your numbers do, in fact show that on a per capita basis, blacks commit more murders of whites than whites do of blacks, they also show that per capita, more blacks are killed by whites than whites are killed by blacks.

There are fewer whites murdering blacks than blacks murdering whites. But there are so many fewer blacks than whites that each black has a higher chance of being killed by one of the few murderous whites than a white has of being killed by one of the few murderous blacks.

Comment Re:I blame Trump. (Score 1) 1149

No, but your statement means that you don't understand conditional probabilities, otherwise you'd have realized that his numbers are not enough to draw that conclusion.

Conditional probabilities? If a white is murdered by a black, what are the chances that he is also murdered by a white? Is it large enough to really matter?

Otherwise, if these conditionals did not matter, a random person, irrespective of skin color, would be about 5 times more likely to be murdered by a white person than by a black one from population size alone.

You're the one extrapolating to a perfectly smooth, round, random person, not me. My statement said nothing about such a critter, and follows from the data given. Assuming people are murdered but a single time by a single perpetrator (which is by far the usual case), then you can indeed make such statements, just as you can make statements about the black death rate from diabetes and the white death rate from heart attacks without knowing about the black death rate from heart attacks or the white death rate from diabetes.

Comment Re:I blame Trump. (Score 5, Interesting) 1149

It means you are innumerate.

Although the math doesn't require an absolute population value, it's a bit easier to reason about it if one is used, so I'll round the population of the country to 300 million.

If whites are 70% of this population and 500 whites per year are murdered by blacks, then the white population is 210 million, and the rate of whites murdered by blacks is 500 per 210 million -- about 2.4 whites per million are murdered each year by blacks.

If blacks are 13% of this population, and 200 blacks per year are murdered by whites, then the black population is 39 million, and the rate of blacks murdered by whites is 200 per 39 million -- about 5.1 blacks per million are murdered each year by whites.

Ergo, as I said, these numbers, if believed, show that a black person is twice as likely to be killed by a white person as a white person is to be killed by a black person.

Well, only approximately, since 5.1 is not exactly twice 2.4, but that wasn't your quibble now, was it?

As other commenters have pointed out, that statement is fairly meaningless without further context, but I'm not the one who posted the context-free numbers. All I did is point out one conclusion from them. Another conclusion, of course, is the one probably intended by the original poster, e.g. that on average, black people commit more interracial murders than white people, but that conclusion likewise requires additional context before you could consider it to be actionable data.

Here's one piece of context for you -- "on average" has its own problems. As Dylan Roof illustrates, a single white guy easily supplied 4.5% of the carnage in that white-on-black number.

Should that make blacks feel safer or less secure?

Slashdot Top Deals

"Life sucks, but death doesn't put out at all...." -- Thomas J. Kopp

Working...