Archive for April, 2008


Tibet has been in the news recently, as people are using the fact that the Olympics are taking place in China to protest against Chinese control. I’ve never been to Tibet, and I don’t know very much about it. Tibetans are an ethnic group, they have a long history, they have their own religion, they speak their own language. From my Western perspective, it seems natural that they should have some ability to determine their own fate. Right now they don’t seem to have that. That’s too bad. It’s also not that different from a lot of other ethnic groups around the world–the Kurds come to mind.

Why does Tibet have such strong support in the west? Why do we see so many bumper stickers saying “Free Tibet” and so few saying “Free East Turkistan?” My wholly superficial knowledge suggests that the two causes are very similar: ethnic groups which want more freedom from the central Chinese government. I see two distinct features of Tibet that make it more popular in the West: Tibetans have a charismatic leader, the Dalai Lama, and Tibetans aren’t Muslims. There is no denying that a charismatic leader is very important for these sorts of struggles. Unfortunately it seems that such people are born, not made. Tibet was very fortunate in the man who became the Dalai Lama.

The Chinese government naturally resists any arguments in favor of Tibetan independence. From a Western perspective they do so in a very ham-handed way: restricting journalists, making statements that seem obviously false, accusing the Dalai Lama of being a terrorist. More interesting is that most Chinese people outside of Tibet appear to support the government. It’s hard to imagine that Chinese people trust their government. But it seems that at least on the subject of Tibet, they do. Is Chinese propaganda more successful within China than it is outside of China? Is there a peculiar Chinese perspective that makes them see the Tibetan issue differently from most people in the West? Is it simply that China’s strict control over information prevents people from seeing an opposing point of view?

From a geopolitical perspective, Tibet is unimportant. I hope that Tibetan people get more freedom and the right to elect their own government. For that matter, I hope that everybody in China gets that right. But the brutal truth is that is that the Tibetans are one abused ethnic group among many, and there are several who have it a lot worse.

I find the apparent success of Chinese propaganda within China to be much more troubling. The increased economic freedom in China was expected to bring increased political freedom. I see no signs of that. China is going to increasingly dominate the world economy–or Chinese society is going to fall apart, one or the other. If China doesn’t fall apart, and if Chinese propaganda continues to be as successful as it has been so far, our ideals of free speech and free elections will matter less and less. This is very speculative, and not at all a doomsday scenario. But I think that if we want to retain what we believe to be the strengths of our society, we need a better understanding of how China is keeping itself immune to them–something the Soviet Union was increasingly unable to do.

Comments (18)

Digital Copyright

Control over copying of analog data is, in practice, made possible by cost: it’s a pain to transcribe a book, it’s hard to make a perfect copy of a record. The cost of copying digital data, however, is near zero. This means that control over copying of digital data is, in practice, impossible.

It’s not hard to control whether anybody can make money using digital data. If somebody is making money, customers are finding them, and that means that you can find them too. If they are making money, you can sue them in court and take away that money. However, there is no practical way to prevent somebody from making a copy without making any money.

There are many industries which exist on the basis of charging per copy–most obviously, the industries which sells CDs, DVDs, and books. When a digital copy is just as good as any other, those industries are dead. They will take a while to die, but they will still die in the reasonably near future.

Historically speaking, none of these industries are very old. So the fact that they are going to die need not be a major problem for most people. I’ve seen occasional arguments that it’s important to find a way to compensate people who create data, as otherwise they won’t do it. That is nonsense. People create things all the time. Art existed long before copyright was invented. The people who will be hurt are the pop stars and the pop star wannabees. It will no longer be possible to make a large fortune in the data creation business.

Interestingly, it will still be possible to make a large fortune in some closely allied business, such as brand creation or jingle writing or video advertising. These are the areas where art finds a direct commercial application. Digital or digitizable art which is sold for its own sake is going to become the province of patronage, performance, and personal creation. There will still be stars, but they won’t be created by a studio system, and they won’t necessarily be rich.

None of this need be bad. But it does mean that our cultural landscape will be quite different from what we are used to.

Comments (1)

Financial Complexity 2

Almost six months later, I’ll look back at my earlier post about financial complexity. Considering the continuing troubles of the financial markets, it seems clear that I underestimated the degree of the problem. But I think I was generally correct in pointing out that the complexity of modern markets helped hide the nature of the bets that institutions were making. Many institutions thought that they had only a controlled amount of exposure to the mortgage market, only to find out that they were wrong. They had lent money to other institutions, which had made derivative bets, which were founded on the mortgage market, and problems rebounded back to institutions which thought they were acting soundly.

I believe the complexity has exceeded the ability of people to fully understand it at this point. That may be a general trend for financial markets–they naturally push the limits of complexity, limits which do increase over time as people learn more and develop better tools.

It remains to be seen how much this will affect the real economy. While it seems quite likely that the U.S. is now in a recession, it could still be a mild one. I suspect that some of the problems are facing now are really due to stagnant middle class income and increased health care costs–the ongoing issue of increasing inequality. At this stage of the business cycle, more people are vulnerable than usual, because fewer people did well when the economy was relatively strong.

But that need not be a bad thing. If few people do well when the economy is strong, it is possible that few people will do poorly when the economy is weak. Not very likely, perhaps, but possible. We’ll have to see.



One of the drawbacks of our secular age is that it is less clear what we should strive for in our lives. In medieval Europe it seems as though life goals were pretty clear for most people: keep your nose clean, and you get to go to heaven after you die. Heaven is a good place, and the details are left undefined because they are sort of hard to imagine.

Most people want to lead a good life, but my guess is that these days most people don’t really believe in a literal heaven. Without a set of rules handed down from the priests, how do we know what a good life consists of? Based on what I see in bookstores, a lot of people—at least enough for a niche publishing market—are aiming for happiness. A good life is a happy life.

This does not imply an outbreak of hedonism and selfishness, as many people are made happy by altruism and quiet reflection. However, to me, it doesn’t seem like the right sort of goal. I don’t think I have the goals of a good life fully pinned down, and it’s not like I think one should strive to be unhappy, but happiness as a goal doesn’t seem right to me. It seems to me that happiness should be a result of a good life. We should aim to do the right thing; if we succeed, we can expect to be happy. Aiming for happiness seems to put it the wrong way around.

Happiness by itself just seems too disconnected, too prone to short-circuiting a deeper examination of the situation.

Comments (5)


CERT recently issued an unfortunate advisory about gcc.

The advisory is about an optimization which gcc introduced in gcc 4.2: given a pointer p, then a comparison of the form p + C1 < p + C2 can be reduced to C1 < C2. This is always valid in standard C and C++, because those languages say that if you have a pointer to an object, and you add an integer to that pointer, then the result must be a pointer to that same object, or must point just past the end of the object. Any other result is undefined. This means in particular that p + C can not wrap around the end of memory, so the compiler can ignore wrapping issues.

Now, I don’t know who would have thought otherwise. Checking for overflow is always a good idea when validating user input. But checking for pointer overflow using wrapping arithmetic does not make sense. What matters is whether the offset is within the bounds of the object the pointer points at. A test which relies on wrapping overflow for pointer bounds seems rather strange.

However, if you do write such a test, you are indeed in trouble. It is perfectly reasonable for CERT to warn against doing so. gcc implements this optimization because it is very useful for determining how often a loop will iterate when the loop is written using a pointer. So people should be aware of it.

Where CERT went wrong is in only mentioning gcc in their advisory. In fact all popular optimizing compilers implement the same optimization. The CERT advisory suggests that you can avoid this problem by avoiding new versions of gcc. That is just plain incorrect. If you switch to a different compiler, you will have the same problem.

As the CERT advisory is written today, they warn that gcc 4.2 and later are a potential hazard, without warning about any other compiler. That seems very unfair and wrong. I certainly hope that CERT will correct or withdraw the advisory.

Also, CERT describes the problem in a rather vague manner, which has led several people to misunderstand it and think that the potential problem is even worse than it seems–i.e., that it might apply to code that people might actually write.

All that said, I am working on adding a warning to gcc for this case. I doubt it will ever be useful to any real programmer, but I hope that it will pacify CERT.

Comments (13)

« Previous Page« Previous entries « Previous Page · Next Page » Next entries »Next Page »