People are not "innocent until proven guilty". They are presumed innocent until proven guilty. Big difference.
If I kill someone, I'm guilty from the moment I do it, whether or not it's ever proven in court. Perfectly put.
And I'd add that the presumption of innocence only applies to the government, not to the whole world. We should not rush to judge, but neither should we pretend that somebody is innocent just because they were not proved guilty.
OJ may not have been convicted, but he is widely and correctly shunned as a murderer.
You can expect some very tense moments with officers when you deny them permission to search your vehicle or your property or if you remain silent. I think it helps if you can say this with full respect for the cops, and sympathy for their position. The one time I got asked this was in college after being pulled over for a bad headlight. I was polite but firm: no.
When they asked me why I wouldn't let them search, I told them (truthfully) that my US Government teacher told me to always answer questions like that with "no" unless I had consulted with a lawyer first. The cops had a hard time telling me not to listen to authority, and so gave up and let me go.
I don't know quite how it works, but I have some friends with chips on their shoulder about cops, and it seems like cops immediately sense the hostility and feel obliged to be hostile back. Being chill and friendly has definitely gotten me out of a couple of tickets that I probably deserved.
. I said that it doesn't make sense to start with tests when you understand the problem, but don't understand what the solution looks like yet. Again, I disagree. I do it all the time, and prefer it except for throwaway code. It may not make sense to you, but that's different than it not making sense.
when you have a straightforward problem with thorough design completed before you begin implementation. Sure, you can use it there, but that is not its strength. Go read a book, any book on TDD. Any of them will tell you that it is not just a testing technique, it's a design technique.
What do you use to measure your coverage? 100% of execution paths taken or 100% of lines executed is not sufficient to prove correctness. In this case, I was talking about LOC coverage, which is the common metric used. I agree that's not sufficient to prove correctness, but in my experience it's plenty good for easy refactoring and low bug rates. For recent projects, it has been somewhere under one bug in production per developer-month, and often much lower than that.
That's plenty good for consumer and most business apps. If I were doing life-critical development, I'd add more pratices, though.
As with any metric, though, you shouldn't use it as the sole driver of decisions. One should first strive for good tests, and only use a coverage metric to look for missed spots. If a manager uses a coverage number as part of an incentive plan or employee evaluation, it's an easy metric to hack by writing garbage tests that don't test anything.
It seems to me that if the same person is writing the code and the unit test, there is a high chance that errors will still slip through the system. If I write a function and forget to check for null, what are the chances that I'll remember to write a unit test that checks for failure with null arguments? There are a couple of common fixes for this problems in shops that like unit testing.
One is doing test-driven development. If you write the test before you write the function, you are much more likely to write good tests.
Another is pair programming. If you are there with another developer, they will hopefully remind you. And by reminding you, you'll learn, and remind other pairing partners in turn. Often we'll play a game where one person writes a bit of test and another makes it pass. Then the second person writes a bit more test and hands it back to the first. This forces you to think about the code from both sides, which is both fun and effective.
If you hate pairing, then code reviews are another option. Because the feedback comes later, people are less likely to apply it, but it's still much better than solo coding.
What I'm criticising is the idea that unit tests are some sort of 100% reliable safety net, such that you can refactor away without risk as long as you have them; this is a common claim by TDD advocates [...] If some TDD advocates say that, they are wrong. For near-riskless refactoring, you also need good acceptance tests.
Unit tests prove that the program does what the programmers think it should. Acceptance tests prove that a program does what the people using and paying for it think it should. Some people, like me, think automating the acceptance tests is well worth it. Others don't, and are happy with the level of safety full unit testing gets them.
Either way, the claim that strong tests mean riskless refactoring is almost tautological. If you refactor and break your system in some way that the tests don't catch, then that means you don't have everything tested. That probably means you added functionality without adding a test first, which means you're not fully doing TDD.
In practice, nobody tests quite everything, and nobody starts knowing how to test everything anyhow, so part of learning good TDD is learning how to make the risk trade-offs. As long as they're doing some short-cycle iterative process, teams learn this pretty gracefully; when you get hit with a production bug, you say, "Oh, we missed a test. Can we avoid that in the future?" and go from there.
And that school of thought is fundamentally unsound, because it relies on the automated unit tests to find any change in behaviour, but you'll never get near 100% coverage for most practical projects. You might not. I do, though.
If you start a code base with TDD and stick with it, though, it's pretty easy to get strong effective coverage. And both my gut feel and the stats from my teams suggest that unit testing is effectively free. Time spent on writing and maintaining tests is made up for in time saved through reduced debugging and reduced cost of change.
Getting 100% coverage is something I never care about, as there's some stuff that is rarely worth it, like simulating disk failures. But 95% coverage is certainly doable, and I've done it on projects up to the 5 developer-year size, and a team of mine is doing it now on a web project serving 3 million users a month that has released on average twice a week over the last year.
I also have this doubt about unit test: If the coder is already aware of the possible failures, then he should be fixing the code instead of writing a test. That's a false dichotomy. You should do both.
Every good programmer manually tries out bits of code they have just written so that they can be sure it works. Unit tests just capture that manual testing and automate it, so that you're sure the code works not just today but tomorrow, next month, and next year.
I think there are other problems with TDD. Namely that you can only use it when you don't need to learn how to solve a problem as you go... I mostly disagree. With TDD, you just have to know at least one thing about the output of whatever you're creating. You assert that, then you make it pass. Maybe you refactor a bit now that, through implementation, you understand the problem better. Then you say, "What's the next thing this should do?" and start the cycle over.
So I use it all the time when I'm figuring out what I'm doing, and in fact lean on it a lot to get me through tricky problems. It's especially helpful there because I can radically change key pieces and my tests tell me where I've gone wrong.
There are two times when I don't use TDD. One, where it's some quick script. The other is when I'm trying something radical enough that I know I'll be throwing the code away.
But for anything I intend to maintain and that I want to actually work, I love TDD + Refactoring. On teams where we've started that way from the beginning, our bug rates have been below one per developer-month, and sometimes well under that. TDD's not all you need for that, but it's a key piece.
I still think there is an entrenched class of people that own and control most of the world's resources, and there is far less social mobility in either direction between this class and the rest of us than many would like to admit. Worldwide, definitely; I was thinking mainly about the US. But even there, people certainly overestimate social mobility, and the big rise in income inequity lately is a big cause for worry. The Economist has had some great articles on that over the last few years, so if you're looking for good numbers back up your argument, it's worth a visit to the library.
I would have no problem with capitalism if there were some kind of base level guarantee of resource availability to everyone. As it is, I feel that the owning class uses economic coercion to keep resources out of the hands of the working class. From the studies I've seen, there are two main components to the kind of equality that breeds social mobility.
One is education. Making sure that every child is well educated regardless of background is a giant factor. The US is so-so at this; some places do well, and some poorly. I would love, love, love for this to be better.
The other is the ease of starting and running a business. Here, the US excels. The paperwork burden is pretty low, both for starting and running. Liquid capital and credit markets make it relatively easy to get money. Our legal and governmental systems are comparatively fair and well run, so a competitor can't just bribe somebody to have your business shut down or steal your assets. Labor and tax laws are pretty fair to small businesses, and culturally we're very accepting of pretty much anybody that can pay their bills and turn a profit. So my female latina pal that I mentioned had little trouble starting a business, focusing much of her energy on the stuff that matters, rather than bullshit that in other countries would stop her cold.
In short, I don't think problems of inequality and injustice will be solved by 'more free markets.' Less regulation of markets will only make it easier for the more ruthless to dominate. I'm sorry to say that you may have fallen for a bit of conservative propaganda here.
Efficient, effective markets are not necessarily unregulated ones. The US stock exchanges, for example, are among the most effective in the world, but they are heavily regulated. The trick is that they do it in such a way to maximize transparency, liquidity, participation, and velocity, while reducing transaction costs and non-core risks.
An effective market minimizes friction and maximizes competition. It's a fine line; you can fail by overregulating, which raises barriers to entry and reduces competition. Or you can fail by underregulating, creating markets where monopoly or oligopoly have equally bad effects.
So sometimes people who push for deregulation are up to something good. Airline deregulation, for example, has dramatically reduced ticket costs and increased options available, with safety improving, thanks to a good mix of safety regulation and enlightened self-interest.
But AT&T and Comcast pushing for non-neutral networks is a great example of bad deregulation. They used previous monopoly power to fight off fair competition, creating an oligopoly in ISPs. And they now want to use the oligopoly power to create barriers to entry on the Internet.
The way I look at it, although some rich people are perfectly civilized and use their fortunes to do good (like the billionaires pushing to restore the estate tax), many will always be venal bastards. The judo-like trick is to harness their venality for the public good, and well-run markets are stunningly effective at that.
A person is not truly free unless they feel secure in their ability to provide for themselves, and our economic system seems designed to make the majority of people feel so insecure in this ability that they simply give in and do whatever the owning class tells them to do. I agree about 90% with this. I think the owning class is not so big a problem these days, honestly: most capital is mediated through so many layers that it has very little direct influence on day-to-day business, and the post-industrial shift to knowledge and service work has pushed the balance of power away from pure capital in the direction of skilled labor.
But I agree the modern economic system is still a big problem. I think that's because of corporatism, though, and not capitalism as such. Modern corporations are internally feudal in nature, and they train people in unthinking obedience.
I think this is primarily a problem with large corporations, where most people, and especially people with power, are largely isolated from real-world input and the corrective effects of free markets. American government was built on the notion of a bunch of small independent land-owners and merchants, with widely distributed power and the inclination to tell tyrants and busybodies to go to hell, and corporatism runs directly counter to that.
However, I see some positive trends that are reversing that. The command-and-control approaches of Planet 1950s are becoming less and less effective. Building Oldsmobiles may have benefited from a lot of central control, but when your money comes from innovation, your best strategy comes from empowering small teams and turning them loose. Google, for example, is socially pretty flat and cellular.
Also, the Internet, in reducing the cost of communication and transaction, has further reduced the value (and therefore influence) of capital. I know somebody who is starting a product business, and she has outsourced almost all of it: manufacturing, wholesaling, distribution, retailing, shipping, accounting, and advertising. Twenty years ago she would have needed 20x the capital she's spending just to hire staff, get an office, build inventory, etc.
The same thing applies in Silicon Valley startups. The necessary amount of capital is probably 5-10x lower than it was in the last bubble. That means many more people are trying and selling a lot less of their equity. Finding good engineers, designers, and business people is much harder than finding the necessary capital right now.
In the third world, you're seeing a parallel success with microlending. Millions and millions of people are becoming much more independent thanks to small injections of capital allowing them to make much better use of their own skilled labor.
So in essence, I think the solution to the problem of corporatism is more energetic use of the free market to empower individuals to make their own way. Large corporations have a hard time fighting this in the way that you'd expect; because US capital is so mobile and so returns-focused, they will do anything that makes a buck, and that includes empowering a lot of entrepreneurs.
Gah. I wrote you a longer reply, but now the back button on Slashdot destroys your comments. Isn't AJAX wonderful? So this will be a bit rushed. Sorry.
The top ten American billionaires include four members of the Walton clan of Wal-Mart fame. Rather undercuts some of your arguments. Of the four, one is a chairman of Wal-Mart's board, so he's not deeply involved in making things happen. Another runs a modestly successful bank. The other two seem to do nothing of note. I think they're in the middle of the shirtsleeves-to-shirtsleeves chain.
The actual force here is not the Waltons but Wal-Mart, which is indeed something I worry about. But they're no more sinister than AT&T, and probably less so. Dynasty has little to do with it.
And the majority of American billionaires did not come from poor working class backgrounds. Many of them came from already wealthy families. I'm not denying that having some money isn't helpful in getting ahead in business. A stable home, a good education, and seeing people handle money well; those things will give you a leg up no matter what you choose to do. But you get that in the top 30%, not just the top 1%.
Take Bill Gates and Warren Buffett, the top two, for example. Both definitely got a leg up from their families, no denying. But both their families worked; Gates's dad was a corporate lawyer, and Buffett's dad was a stockbroker.
My point is that dynastic wealth is not a major force in American business. I can't think of any big name in American business who has a railroad tycoon or an oil baron as a great-grandfather.
That could change, of course, and we should be on guard against it. But so far, any positive feedback loops where wealth brings more wealth don't last across the generations in the US.
Even within a given generation, those positive feedback loops seem pretty limited in their power. Warren Buffett may have a lot of money, but it's 0.1% of what American citizens control. That leaves plenty of room for the titans to clash against one another; Gates and Ballmer are pretty wealthy, and so are Page and Brin. But they fight vigorously to limit each other's further gains in wealth. And of course the capital markets fund all sorts of lesser coalitions against the titans.
If you're worried about society-distorting dynastic wealth, turn your eyes to India, China, Russia, and Saudi Arabia. Their versions of capitalism and government are much less robust, permitting a lot more dubious behavior and dubious outcomes.
Nicely put. But there is a positive feedback loop in our economic system that is not damped by an effective check or balance. Money gives one more power to influence the economic system itself, to influence valuation for various goods and services. And that lets one make more money than others without benefiting society. This leads to a runaway concentration of wealth in the hands of those least likely to care about the community, and most able to treat ruthlessly with their fellow man. That's a totally legitimate worry, and it's something we have to keep our eyes on. However, I can think of four effective checks on this.
The most obvious is satiability. A lot of people get enough of whatever they are seeking, and don't work as hard. Most people are like this. Some are insatiably greedy or power hungry, of course. A lot of the ones who aren't are in it because they just like working and making things or doing deals. That's the classic Silicon Valley serial entrepreneur.
The second check is old age and death. As people get older, most don't perform as well, and many lose their ability to amass wealth. And corpses are notoriously bad deal-makers.
The third is the the effect behind the saying, "from shirtsleeves to shirtsleeves in three generations". A common pattern in the US is that some unprivileged person makes a bundle, and his children grow up in wealth. But because they've had a soft life, they don't have the emotional balance or the skills necessary to even maintain their wealth, let alone increase it. So the third generation (or fourth) is back working for a living.
The fourth is that a lot of people who have built up fortunes give them away. Warren Buffett and Bill Gates are two great examples, but there are plenty of smaller ones, too. The sociopath entrepreneur exists, but is relatively rare. Making money involves extensive and skilled collaboration, and since you can't take it with you, it's pretty natural that a lot of the great fortunes end up using it to improve the world they've studied so intensely.
Yeah, there are some neofeudalists who would love to change this. But historically, they haven't been a huge threat; I can't think of a major business player of today who has a name like Carnegie or Stanford.
As far as threats to democracy go, I worry more about large corporations and political dynasties. The large corporations aren't bound by any of those checks, and are frequently controlled by a series of dubious characters. Political dynasties seem to have power that's much more readily transferable. George W Bush strikes me as a bigger threat to democracy than the top ten American billionaires combined.
Yet our economic system is based on the premise that individual profit is most rewarding to individuals. You misunderstand. That's like saying our justice system is focused on criminals. While technically true, the point for both is still net shared benefit.
There are two reasons the economic system appears to focus on individual profit. First, that takes the minority of truly greedy and/or power-hungry people and channels their activities into something that's net positive. Second, if you aren't very careful to make it work for the individual, you open it wide up for all sorts of theft, gaming, and other anti-social behavior.
It's counterintuitive, I realize, but designing for greed permits niceness, while presuming niceness gives greed free reign.
can genes code for behaviors that are detrimental to the individual but good for the gene pool overall? Genes generally code for behavior that is good for genes. A gene for you to treat family well doesn't give a shit about you personally; your family members are likely to have the same gene, so it's just being good to other copies of itself.
Those seriously wondering about this topic should read The Selfish Gene (Richard Dawkins's first book, wherein he coined the term "meme"). Then follow that up with Good Natured: The Origins of Right and Wrong in Humans and Other Animals. They're two very readable books by two real scientists, and they have rocked the worlds of everybody I have lent them to.
the media has absolutely no ethical boundaries. The media is not a single thing. And the media is only selling what you're buying.
Count up where you spend your media money, and see how much of it is on high-quality news organizations that sell direct to the public, like the WSJ, the NYT, The Economist, or the Christian Science Monitor. If you're already paying good sources more than bad ones, only then do you have the right to complain like this. And don't forget to count the money you spend on cable and the time you devote to reading and watching ad-supported media, which is money to them.
A system of collective intelligence will emerge that will tap directly into the sources of news. The media as we know it will die. We care about the truth. Sure, that's brilliant. I mean look at the quality of independent journalism that has emerged on YouTube. The peak of which seems to be Leave Britney Alone.
The people in power, both in corporations and in governments, are already very good at manipulating TV journalists, and pretty good at manipulating print journalists. They will have no trouble manipulating solo, amateur bloggers for some time to come.
Although I certainly hope we as citizens manage collectively to out-compete traditional journalism, it's ridiculous to say we don't need the mainstream media until citizen journalism is regularly way better on crucial stories. Otherwise you're just arguing for being even less informed.
But you, yourself, are a member of the media? Reporting on a topic that paints a picture of the picture-painters to the American public? Like the blogosphere, the the mainstream media is a self-examining device. For both, the answer to Quis custodiet ipsos custodes? is supposed to be everybody.
The next story, if the media is up to it's usual games, would be to present a count of how many times Mr Barstow's own organization has used these same experts to sell it's own rags to the masses. Christ, did you even read TFA? That question is asked and answered in a linked article.
Where your pseudo-outrage is coming from, I have no idea. Is this some snide hipster pose that makes you feel part of the ironic elite? Or are you really opposed to the media trying to understand the largest media fuck-up of the decade?
Personally, I'd love to see more of this, so that next time we commit to spending tens of thousands of lives and trillions of dollars, we have some vague idea of what's really going on.
Quite the contrary, Dawkins was offering that life on Earth could have been designed... but that the life that designed it must have ultimately evolved from a spontaneously evolving form. You misunderstand his point.
Even if we can show intelligent interference in the course of evolution, that doesn't prove the creationists right. Extraterrestrials are just as good an explanation as Jehovah (or Allah or Brahma or Zeus). And in fact they're better, because even if we could prove that somebody messed with human genes, evolution is still the most parsimonious explanation of the source of the entity or entities designing things. It's a simple application of Occam's Razor.
The only way a Darwinian/Neo-Darwinian assertion could be demonstrated is to observe it through experiment. This has failed to happen. That's incorrect. For example:
Darwin himself never heard of DNA, but applying Darwinian theory would suggest that there would be a variety of interesting relationships between DNA patterns and our morphological and paleontological classification of animals. And that relationship has been proved out extraordinarily well, and ever more so with the rise of cheap sequencing.
Of course, it could have been very different. If we had found no evolutionary relationship in the DNA, it would have demolished Darwinian theory. If we'd found a copy of the Pentateuch, that would have been fabulous support for ID. Equally so if we'd found a map to Betelgeuse with a note suggesting we drop by for a visit. As it is, the neo-Darwinian synthesis is way ahead on points.
Agreed...but I don't recall censorship and refusal to allow a topic to be addressed as one of those. Censorship? You'll have to point me to where these science johnnies are burning books, confiscating the presses used by the Discovery Institute, and arresting people who suggest modifications to the modern evolutionary consensus.
Because what it looks like to me is that they're saying that in science journals, you should only publish solid science. And that when universities hire science professors, they should get ones doing good scientific work. And of course, that it's not worth wasting time on pseudoscience.
In my mind, that's not censorship; it's professional responsibility. If they spent all their time jawing with people who had wacky ideas and little to no proof, they'd be up to their ears no just in creationists, but psychics, crystal healers, UFO believers, ghost hunters, Loch Ness Monster proponents, astrologists, iridiologists, and people with perpetual motion machines and 250 MPG carburetors.
As a taxpayer, I'm paying scientists to get serious science done. That includes shutting the office door on kooks.
One could certainly be a competent physician, for example, and not believe in Darwinism (or neo-Darwinism). It seems to me that one could even be a quite competent practitioner of any of the biological sciences (other than the various sorts of paleontology) without necessarily agreeing with Darwinism. That is flatly incorrect.
Even a doctor has to understand evolved antibiotic resistance, pathogen evolution, genetic diseases, genetically engineered treatments, and animal testing of human treatments, none of which make much sense without evolutionary theory.
And that's the same theory that's woven into the core of pretty much every bit of modern biology. There aren't even any vaguely competing theories on offer any more. The last attempt I can recall is Lysenkoism, which was an obvious bust. So-called "intelligent design" is not even wrong.
But I certainly don't think that the loss of Darwinism would destroy American education or anything along those lines. So... people... GET A GRIP. The reason that people care so much is that because evolution is one of the most heavily established theories, you can't get it out of science education without getting rid of science itself. That's particularly true if your goal is to introduce creationism, which is terrible science (and pretty weak theology).
Our primary and secondary science education is already pretty weak, and scientific knowledge is becoming ever more crucial in the modern world. That's true not just for jobs, but for us to fulfill our duty as citizens. So yeah, in my view, this is a big deal.
Its the same thing. Anyone who takes issue with Steins message is being pretty petty and short sighted. That is not all Stein is doing.
If you catch any decent scientist casually and raise these points, yes, they will be happy to talk about them. Good scientists love other explanations. Even kooky ones can be fun to play with. And there are plenty of Christian scientists who do every bit of their work with the idea that God did it all. That view is entirely compatible with the theory of evolution.
But what scientists take issue with is creationism and/or "intelligent design", which is dubious theology and terrible science.
What the "intelligent design" crowd is up to is to get their theology taught as fact, when it is nothing close. By manufacturing the apperance of uncertainty they are trying to distort the political process.
If evolution or non-creationism is correct, and by having a dialogue people are convinced of this fact...then what is the problem? The problem is that's not what happens.
First, scientists generally should shun bad science. It's great to have wild ideas, but if people can't back them up, then they are just slowing down the people who actually have the chance of getting somewhere. Scientists should have to answer reasonable questions, but they are not obliged to argue with idiots all day.
Second, even letting the nuts in the room gives them unearned respect. An African witch doctor may think that Harvard is discriminating against him by not having a Department of Witch Doctoring, but that's tough. Funding and publishing that stuff gives it an appearance of legitimacy that is has not earned.
Third, kooks with a political agenda (<cough>Discovery Institute</cough>) use even a few accepted papers to create the appearance of a scientific controversy where there really isn't one. Then they use that to manipulate the political process. E.g., the "teach the controversy" bullshit. See The Manufacture of Uncertainty for more. It's the same trick tobacco companies used for years.
So it would be great if we could sit down and have a dialog about this. However, as far as science was concerned, the honest dialog around these points was over half a century ago. It's like some nut coming in and saying Einstein was wrong, and Newtonian mechanics works just fine. As one scientist said about something else, intelligent design is not even wrong.
test? test evolution how? evolution isn't really very good science. it is speculation based on observation. maybe that is why its defenders are so rabid. That's ridiculous. There are a zillion examples, but let me take an obvious one.
Darwin existed long before we knew much about the inside of a cell, or anything about genetics. It was decades later that Mendel's work was rediscovered, and many decades more before we found out that DNA was the core of genetics.
Once we learned about it, we could easily use evolution to predict that there would be certain relationships between the DNA of organisms that reflected their evolutionary history. For example, that humans would share a lot of DNA with each other, but some less with chimps. And we'd share less still with monkeys, then dogs, then snakes, then trees, then bacteria.
Those relationships have been found to exist over and over and over again. There is an entire new medical science, genomics built on teasing out those relationships and relating them back to physical correspondences, and then applying this knowledge to saving and improving human lives. And the theory of evolution is a keystone in that field, as well as many others.
So yes, evolution has been tested and retested. I started as "speculation based on observation", which is what science calls a hypothesis. But that's fine, because that's the first step in the scientific method. You should read about it!
MMCC as a theory will gain much more respect when it embraces challenges, instead of treating them in the same way ID treats challenges -- by throwing the scientific method under a bus. I agree totally. And some of the MMCC people are indeed zealots. But there are reasonable and science-oriented people on that side who still bristle at the challenges, because they see them as mainly political challenges, not scientific ones. The game is to create the appearance of scientific controversy, which is then used to undermine political action. It is a game the tobacco companies, for example, played for years.
I am all for being responsible with the environment but saying that "It can't hurt" is dead wrong. The less glib way to make his point is that given the cost and risk tradeoffs, it is worth some effort to cut carbon emissions until our science catches up and we know for sure. Think of it as buying insurance.
It can hugely hurt the economy[...] This is probably incorrect. You could substitute carbon taxes for other taxes with minimal economic impact, and a great deal of progress in reduced emissions. The trick is to start immediately, so that we don't have to take particularly drastic action. With research and technology adoption, slow and steady wins the race. Replacing all our power plants in one year is hideously expensive, but replacing them as they wear out is in effect free.
Slashdot Mirror
If I kill someone, I'm guilty from the moment I do it, whether or not it's ever proven in court. Perfectly put.
And I'd add that the presumption of innocence only applies to the government, not to the whole world. We should not rush to judge, but neither should we pretend that somebody is innocent just because they were not proved guilty.
OJ may not have been convicted, but he is widely and correctly shunned as a murderer.
When they asked me why I wouldn't let them search, I told them (truthfully) that my US Government teacher told me to always answer questions like that with "no" unless I had consulted with a lawyer first. The cops had a hard time telling me not to listen to authority, and so gave up and let me go.
I don't know quite how it works, but I have some friends with chips on their shoulder about cops, and it seems like cops immediately sense the hostility and feel obliged to be hostile back. Being chill and friendly has definitely gotten me out of a couple of tickets that I probably deserved.
That's plenty good for consumer and most business apps. If I were doing life-critical development, I'd add more pratices, though.
As with any metric, though, you shouldn't use it as the sole driver of decisions. One should first strive for good tests, and only use a coverage metric to look for missed spots. If a manager uses a coverage number as part of an incentive plan or employee evaluation, it's an easy metric to hack by writing garbage tests that don't test anything.
One is doing test-driven development. If you write the test before you write the function, you are much more likely to write good tests.
Another is pair programming. If you are there with another developer, they will hopefully remind you. And by reminding you, you'll learn, and remind other pairing partners in turn. Often we'll play a game where one person writes a bit of test and another makes it pass. Then the second person writes a bit more test and hands it back to the first. This forces you to think about the code from both sides, which is both fun and effective.
If you hate pairing, then code reviews are another option. Because the feedback comes later, people are less likely to apply it, but it's still much better than solo coding.
Unit tests prove that the program does what the programmers think it should. Acceptance tests prove that a program does what the people using and paying for it think it should. Some people, like me, think automating the acceptance tests is well worth it. Others don't, and are happy with the level of safety full unit testing gets them.
Either way, the claim that strong tests mean riskless refactoring is almost tautological. If you refactor and break your system in some way that the tests don't catch, then that means you don't have everything tested. That probably means you added functionality without adding a test first, which means you're not fully doing TDD.
In practice, nobody tests quite everything, and nobody starts knowing how to test everything anyhow, so part of learning good TDD is learning how to make the risk trade-offs. As long as they're doing some short-cycle iterative process, teams learn this pretty gracefully; when you get hit with a production bug, you say, "Oh, we missed a test. Can we avoid that in the future?" and go from there.
If you start a code base with TDD and stick with it, though, it's pretty easy to get strong effective coverage. And both my gut feel and the stats from my teams suggest that unit testing is effectively free. Time spent on writing and maintaining tests is made up for in time saved through reduced debugging and reduced cost of change.
Getting 100% coverage is something I never care about, as there's some stuff that is rarely worth it, like simulating disk failures. But 95% coverage is certainly doable, and I've done it on projects up to the 5 developer-year size, and a team of mine is doing it now on a web project serving 3 million users a month that has released on average twice a week over the last year.
Every good programmer manually tries out bits of code they have just written so that they can be sure it works. Unit tests just capture that manual testing and automate it, so that you're sure the code works not just today but tomorrow, next month, and next year.
So I use it all the time when I'm figuring out what I'm doing, and in fact lean on it a lot to get me through tricky problems. It's especially helpful there because I can radically change key pieces and my tests tell me where I've gone wrong.
There are two times when I don't use TDD. One, where it's some quick script. The other is when I'm trying something radical enough that I know I'll be throwing the code away.
But for anything I intend to maintain and that I want to actually work, I love TDD + Refactoring. On teams where we've started that way from the beginning, our bug rates have been below one per developer-month, and sometimes well under that. TDD's not all you need for that, but it's a key piece.
One is education. Making sure that every child is well educated regardless of background is a giant factor. The US is so-so at this; some places do well, and some poorly. I would love, love, love for this to be better.
The other is the ease of starting and running a business. Here, the US excels. The paperwork burden is pretty low, both for starting and running. Liquid capital and credit markets make it relatively easy to get money. Our legal and governmental systems are comparatively fair and well run, so a competitor can't just bribe somebody to have your business shut down or steal your assets. Labor and tax laws are pretty fair to small businesses, and culturally we're very accepting of pretty much anybody that can pay their bills and turn a profit. So my female latina pal that I mentioned had little trouble starting a business, focusing much of her energy on the stuff that matters, rather than bullshit that in other countries would stop her cold. In short, I don't think problems of inequality and injustice will be solved by 'more free markets.' Less regulation of markets will only make it easier for the more ruthless to dominate. I'm sorry to say that you may have fallen for a bit of conservative propaganda here.
Efficient, effective markets are not necessarily unregulated ones. The US stock exchanges, for example, are among the most effective in the world, but they are heavily regulated. The trick is that they do it in such a way to maximize transparency, liquidity, participation, and velocity, while reducing transaction costs and non-core risks.
An effective market minimizes friction and maximizes competition. It's a fine line; you can fail by overregulating, which raises barriers to entry and reduces competition. Or you can fail by underregulating, creating markets where monopoly or oligopoly have equally bad effects.
So sometimes people who push for deregulation are up to something good. Airline deregulation, for example, has dramatically reduced ticket costs and increased options available, with safety improving, thanks to a good mix of safety regulation and enlightened self-interest.
But AT&T and Comcast pushing for non-neutral networks is a great example of bad deregulation. They used previous monopoly power to fight off fair competition, creating an oligopoly in ISPs. And they now want to use the oligopoly power to create barriers to entry on the Internet.
The way I look at it, although some rich people are perfectly civilized and use their fortunes to do good (like the billionaires pushing to restore the estate tax), many will always be venal bastards. The judo-like trick is to harness their venality for the public good, and well-run markets are stunningly effective at that.
But I agree the modern economic system is still a big problem. I think that's because of corporatism, though, and not capitalism as such. Modern corporations are internally feudal in nature, and they train people in unthinking obedience.
I think this is primarily a problem with large corporations, where most people, and especially people with power, are largely isolated from real-world input and the corrective effects of free markets. American government was built on the notion of a bunch of small independent land-owners and merchants, with widely distributed power and the inclination to tell tyrants and busybodies to go to hell, and corporatism runs directly counter to that.
However, I see some positive trends that are reversing that. The command-and-control approaches of Planet 1950s are becoming less and less effective. Building Oldsmobiles may have benefited from a lot of central control, but when your money comes from innovation, your best strategy comes from empowering small teams and turning them loose. Google, for example, is socially pretty flat and cellular.
Also, the Internet, in reducing the cost of communication and transaction, has further reduced the value (and therefore influence) of capital. I know somebody who is starting a product business, and she has outsourced almost all of it: manufacturing, wholesaling, distribution, retailing, shipping, accounting, and advertising. Twenty years ago she would have needed 20x the capital she's spending just to hire staff, get an office, build inventory, etc.
The same thing applies in Silicon Valley startups. The necessary amount of capital is probably 5-10x lower than it was in the last bubble. That means many more people are trying and selling a lot less of their equity. Finding good engineers, designers, and business people is much harder than finding the necessary capital right now.
In the third world, you're seeing a parallel success with microlending. Millions and millions of people are becoming much more independent thanks to small injections of capital allowing them to make much better use of their own skilled labor.
So in essence, I think the solution to the problem of corporatism is more energetic use of the free market to empower individuals to make their own way. Large corporations have a hard time fighting this in the way that you'd expect; because US capital is so mobile and so returns-focused, they will do anything that makes a buck, and that includes empowering a lot of entrepreneurs.
The actual force here is not the Waltons but Wal-Mart, which is indeed something I worry about. But they're no more sinister than AT&T, and probably less so. Dynasty has little to do with it. And the majority of American billionaires did not come from poor working class backgrounds. Many of them came from already wealthy families. I'm not denying that having some money isn't helpful in getting ahead in business. A stable home, a good education, and seeing people handle money well; those things will give you a leg up no matter what you choose to do. But you get that in the top 30%, not just the top 1%.
Take Bill Gates and Warren Buffett, the top two, for example. Both definitely got a leg up from their families, no denying. But both their families worked; Gates's dad was a corporate lawyer, and Buffett's dad was a stockbroker.
My point is that dynastic wealth is not a major force in American business. I can't think of any big name in American business who has a railroad tycoon or an oil baron as a great-grandfather.
That could change, of course, and we should be on guard against it. But so far, any positive feedback loops where wealth brings more wealth don't last across the generations in the US.
Even within a given generation, those positive feedback loops seem pretty limited in their power. Warren Buffett may have a lot of money, but it's 0.1% of what American citizens control. That leaves plenty of room for the titans to clash against one another; Gates and Ballmer are pretty wealthy, and so are Page and Brin. But they fight vigorously to limit each other's further gains in wealth. And of course the capital markets fund all sorts of lesser coalitions against the titans.
If you're worried about society-distorting dynastic wealth, turn your eyes to India, China, Russia, and Saudi Arabia. Their versions of capitalism and government are much less robust, permitting a lot more dubious behavior and dubious outcomes.
The most obvious is satiability. A lot of people get enough of whatever they are seeking, and don't work as hard. Most people are like this. Some are insatiably greedy or power hungry, of course. A lot of the ones who aren't are in it because they just like working and making things or doing deals. That's the classic Silicon Valley serial entrepreneur.
The second check is old age and death. As people get older, most don't perform as well, and many lose their ability to amass wealth. And corpses are notoriously bad deal-makers.
The third is the the effect behind the saying, "from shirtsleeves to shirtsleeves in three generations". A common pattern in the US is that some unprivileged person makes a bundle, and his children grow up in wealth. But because they've had a soft life, they don't have the emotional balance or the skills necessary to even maintain their wealth, let alone increase it. So the third generation (or fourth) is back working for a living.
The fourth is that a lot of people who have built up fortunes give them away. Warren Buffett and Bill Gates are two great examples, but there are plenty of smaller ones, too. The sociopath entrepreneur exists, but is relatively rare. Making money involves extensive and skilled collaboration, and since you can't take it with you, it's pretty natural that a lot of the great fortunes end up using it to improve the world they've studied so intensely.
Yeah, there are some neofeudalists who would love to change this. But historically, they haven't been a huge threat; I can't think of a major business player of today who has a name like Carnegie or Stanford.
As far as threats to democracy go, I worry more about large corporations and political dynasties. The large corporations aren't bound by any of those checks, and are frequently controlled by a series of dubious characters. Political dynasties seem to have power that's much more readily transferable. George W Bush strikes me as a bigger threat to democracy than the top ten American billionaires combined.
There are two reasons the economic system appears to focus on individual profit. First, that takes the minority of truly greedy and/or power-hungry people and channels their activities into something that's net positive. Second, if you aren't very careful to make it work for the individual, you open it wide up for all sorts of theft, gaming, and other anti-social behavior.
It's counterintuitive, I realize, but designing for greed permits niceness, while presuming niceness gives greed free reign.
Those seriously wondering about this topic should read The Selfish Gene (Richard Dawkins's first book, wherein he coined the term "meme"). Then follow that up with Good Natured: The Origins of Right and Wrong in Humans and Other Animals. They're two very readable books by two real scientists, and they have rocked the worlds of everybody I have lent them to.
Count up where you spend your media money, and see how much of it is on high-quality news organizations that sell direct to the public, like the WSJ, the NYT, The Economist, or the Christian Science Monitor. If you're already paying good sources more than bad ones, only then do you have the right to complain like this. And don't forget to count the money you spend on cable and the time you devote to reading and watching ad-supported media, which is money to them. A system of collective intelligence will emerge that will tap directly into the sources of news. The media as we know it will die. We care about the truth. Sure, that's brilliant. I mean look at the quality of independent journalism that has emerged on YouTube. The peak of which seems to be Leave Britney Alone.
The people in power, both in corporations and in governments, are already very good at manipulating TV journalists, and pretty good at manipulating print journalists. They will have no trouble manipulating solo, amateur bloggers for some time to come.
Although I certainly hope we as citizens manage collectively to out-compete traditional journalism, it's ridiculous to say we don't need the mainstream media until citizen journalism is regularly way better on crucial stories. Otherwise you're just arguing for being even less informed.
Where your pseudo-outrage is coming from, I have no idea. Is this some snide hipster pose that makes you feel part of the ironic elite? Or are you really opposed to the media trying to understand the largest media fuck-up of the decade?
Personally, I'd love to see more of this, so that next time we commit to spending tens of thousands of lives and trillions of dollars, we have some vague idea of what's really going on.
Even if we can show intelligent interference in the course of evolution, that doesn't prove the creationists right. Extraterrestrials are just as good an explanation as Jehovah (or Allah or Brahma or Zeus). And in fact they're better, because even if we could prove that somebody messed with human genes, evolution is still the most parsimonious explanation of the source of the entity or entities designing things. It's a simple application of Occam's Razor. The only way a Darwinian/Neo-Darwinian assertion could be demonstrated is to observe it through experiment. This has failed to happen. That's incorrect. For example:
Darwin himself never heard of DNA, but applying Darwinian theory would suggest that there would be a variety of interesting relationships between DNA patterns and our morphological and paleontological classification of animals. And that relationship has been proved out extraordinarily well, and ever more so with the rise of cheap sequencing.
Of course, it could have been very different. If we had found no evolutionary relationship in the DNA, it would have demolished Darwinian theory. If we'd found a copy of the Pentateuch, that would have been fabulous support for ID. Equally so if we'd found a map to Betelgeuse with a note suggesting we drop by for a visit. As it is, the neo-Darwinian synthesis is way ahead on points.
Because what it looks like to me is that they're saying that in science journals, you should only publish solid science. And that when universities hire science professors, they should get ones doing good scientific work. And of course, that it's not worth wasting time on pseudoscience.
In my mind, that's not censorship; it's professional responsibility. If they spent all their time jawing with people who had wacky ideas and little to no proof, they'd be up to their ears no just in creationists, but psychics, crystal healers, UFO believers, ghost hunters, Loch Ness Monster proponents, astrologists, iridiologists, and people with perpetual motion machines and 250 MPG carburetors.
As a taxpayer, I'm paying scientists to get serious science done. That includes shutting the office door on kooks.
Even a doctor has to understand evolved antibiotic resistance, pathogen evolution, genetic diseases, genetically engineered treatments, and animal testing of human treatments, none of which make much sense without evolutionary theory.
And that's the same theory that's woven into the core of pretty much every bit of modern biology. There aren't even any vaguely competing theories on offer any more. The last attempt I can recall is Lysenkoism, which was an obvious bust. So-called "intelligent design" is not even wrong. But I certainly don't think that the loss of Darwinism would destroy American education or anything along those lines. So
Our primary and secondary science education is already pretty weak, and scientific knowledge is becoming ever more crucial in the modern world. That's true not just for jobs, but for us to fulfill our duty as citizens. So yeah, in my view, this is a big deal.
If you catch any decent scientist casually and raise these points, yes, they will be happy to talk about them. Good scientists love other explanations. Even kooky ones can be fun to play with. And there are plenty of Christian scientists who do every bit of their work with the idea that God did it all. That view is entirely compatible with the theory of evolution.
But what scientists take issue with is creationism and/or "intelligent design", which is dubious theology and terrible science.
What the "intelligent design" crowd is up to is to get their theology taught as fact, when it is nothing close. By manufacturing the apperance of uncertainty they are trying to distort the political process.
First, scientists generally should shun bad science. It's great to have wild ideas, but if people can't back them up, then they are just slowing down the people who actually have the chance of getting somewhere. Scientists should have to answer reasonable questions, but they are not obliged to argue with idiots all day.
Second, even letting the nuts in the room gives them unearned respect. An African witch doctor may think that Harvard is discriminating against him by not having a Department of Witch Doctoring, but that's tough. Funding and publishing that stuff gives it an appearance of legitimacy that is has not earned.
Third, kooks with a political agenda (<cough>Discovery Institute</cough>) use even a few accepted papers to create the appearance of a scientific controversy where there really isn't one. Then they use that to manipulate the political process. E.g., the "teach the controversy" bullshit. See The Manufacture of Uncertainty for more. It's the same trick tobacco companies used for years.
So it would be great if we could sit down and have a dialog about this. However, as far as science was concerned, the honest dialog around these points was over half a century ago. It's like some nut coming in and saying Einstein was wrong, and Newtonian mechanics works just fine. As one scientist said about something else, intelligent design is not even wrong.
Darwin existed long before we knew much about the inside of a cell, or anything about genetics. It was decades later that Mendel's work was rediscovered, and many decades more before we found out that DNA was the core of genetics.
Once we learned about it, we could easily use evolution to predict that there would be certain relationships between the DNA of organisms that reflected their evolutionary history. For example, that humans would share a lot of DNA with each other, but some less with chimps. And we'd share less still with monkeys, then dogs, then snakes, then trees, then bacteria.
Those relationships have been found to exist over and over and over again. There is an entire new medical science, genomics built on teasing out those relationships and relating them back to physical correspondences, and then applying this knowledge to saving and improving human lives. And the theory of evolution is a keystone in that field, as well as many others.
So yes, evolution has been tested and retested. I started as "speculation based on observation", which is what science calls a hypothesis. But that's fine, because that's the first step in the scientific method. You should read about it!
For more on this, see The Manufacture of Uncertainty.