Languages aren't purely functional. (Gah! That term's overloaded in this discussion. (Gah! So's that one!)) What I mean is that languages might have historical, cultural, anthropological or even sentimental value beyond their "use value" in the modern world. Many cultures preserve their history orally, and if young people stop learning the "useless" local language then thousands of years of history and mythology will be lost. Maybe that doesn't matter, but it seems to me that increased global communication, for all its good points, is bringing us closer to a monoculture of worldviews.
what would the explanation for the big cycle be in social terms?
I was wondering the same thing. The only explanation I could come up with was alphabetical ordering.
IIRC there was once a study of graduates from a police academy where rooms were allocated alphabetically. Years after they graduated, there was a strong correlation between two graduates' alphabetical proximity and the likelihood that they had remained friends. Conclusion: you can learn to like almost anyone if you're forced to spend time together. Getting back to the point, if groups were selected alphabetically for some activities, or if classes were seated alphabetically, you'd expect to see a long chain in the social network. I don't know why the ends of the chain should be joined, though.:-)
I subsequently skimmed the paper corresponding to the diagram. (Unfortunately all figures and tables are missing from the PDF - I can't find a complete version on the web.) They note the unusual structure of the graph - it's almost a perfect spanning tree, and is not a "small world" by any means - and explain it by a prohibition on "seconds relationships" - that is, a prohibition on dating the former partner of your former partner's new partner. They say this rule produces a graph with no 4-edge cycles. Cycles with an odd number of edges are rare because the reported relationships are overwhelmingly heterosexual, so this amounts to a prohibition on cycles with less than 6 edges. Applying this prohibition and controlling for degree distribution, randomly generated graphs are similar in structure to the one in the diagram.
However, it's possible to create a 4-edge cycle in a heterosexual graph without a "seconds relationship" - A dates B and then C, each of whom subsequently dates D. It's been a while since I had to deal with high school dating etiquette, but IIRC this situation is only taboo if B and C are friends - something which isn't recorded in the data. Nevertheless, while the social taboos might be more complex than those desribed in the paper, I'm convinced by the basic argument that they produce the unusual structure of the graph.
To confuse the issue further, pictures of not people having sex are sometimes porn - it's definitely porn if the not people are not naked. If the not people are naked but not having sex, it's not porn.
Expressed as a binary number where the first digit represents people, the second represents clothes and the third represents sex:
000 - not porn
001 - sometimes porn
010 - not porn
011 - porn
100 - porn (softcore)
101 - porn (hardcore)
110 - not porn
111 - porn (fetish and/or in a hurry)
No wonder the Supreme Court had such a hard time with it.
For those who are interested, a PDF of the paper mentioned in the article is here. Running time is O(n^2) versus O(n^3) for previous algorithms, so don't go applying it to the Google cache just yet.
Swing doesn't use pixel co-ordinates for layout like Windows (and IIRC Qt). Instead you place widgets into nested horizontal and vertical boxes like GTK, and the pixel positions are determined by a LayoutManager. It's a better approach when you don't know how large the fonts will be, etc, at compile time.
So it's not necessarily a bug if two different Swing implementations position things differently, as long as the overall layout matches what the programmer specified.
Why not just end the "fuss" right now and let Windows win the OS wars?
Because Microsoft's competitors will never choose Windows. Everyone (even Microsoft) is free to choose Linux, because it isn't controlled by a single vendor (not even Linus controls it - other individuals/companies can always fork development if they don't agree with his decisions).
Of course it's unrealistic to expect Microsoft to choose Linux, but they could do so if they wished. If someone had told me 6 years ago that one day MacOS would be based on BSD, I would have thought they were mad. In 6 years' time, will Windows be a proprietary compatibility layer running on a Linux kernel? Probably not, but it's possible.
And yes, the government (remember: not *a* government, or *our* government) is simply an agreement between people. Our agreement is called the U.S. Constitution. Other countries have their own agreements.
That's strange, I don't remember agreeing to anything. In fact I don't even remember being asked... of course, I was only a baby at the time.
A constitution isn't an agreement. It's a non-binding description of how you can expect to be treated (for example, whether you can expect to be consulted before the terms of the non-binding description are unilaterally changed).
If you think government is an agreement, try disagreeing and see what happens to you.;-)
Though expected to be around 100,000 genes, the human genome turns out to be 30-40,000 genes instead -- right around the level of bacteria, for one comparison.
I don't understand why this figure generated so much fuss. We're looking at a combinatorial system - you don't need many inputs to get an enormous number of outputs. It's like being amazed that telephone numbers in a large city "only" have eight digits.
Picture the human genome as a binary string 30,000 bits long. Each bit represents a gene: 1 means active, 0 means inactive (genes with more than 2 possible states can be represented by multiple bits). This gives us an upper limit of 2^30,000 possible phenotypes without even considering developmental influences. That number dwarfs the number of atoms in the universe, let alone the number of people that have ever lived. Even if only one in a trillion of those phenotypes is viable, we still have 2^29,960 to choose from. For me, the question is not "how can this be complex enough to create a human being?" but "how can we find the tiny subset of this information that actually corresponds to human beings?".
That figure of 0.69 looks pretty arbitrary... still I guess it proves that we're the only ones intelligent enough to rig the statistics in our favour.;-)
Law enforcement and justice are not political in nature.
We obviously have radically different conceptions of what "political" means.
Firstly, slavery is a separate issue from economic status. Just because you have no money, does not mean you are a slave.
That's true in some political systems, because you have rights that can't be taken away from you (or spent on necessities like food). As I demonstrated before, in a pure capitalist system the only right a destitute person has is the right to work, and if he can't earn more than the price of his food (and why should anyone offer him more?) then he will remain a de facto slave. Bear in mind that if someone offers him the price of his food in return for his labour, his choice is to take it or starve. He is not coerced by the other party but by the biological necessity of eating.
In a democracy you can spend your last dollar but you can't spend your right to vote.
In fact you can. You spend it on an election, and then you're stuck with your (or rather, everyone else's) choice for an entire term with no "out".
That's spending your vote, not your right to vote. If you spend your vote you still have the right to vote at the next election.
A free market has a constant feedback, and response is near-immediate (compared to politics anyway).
I'm not disputing that the market is a very efficient economic mechanism. I'm just saying that a market system is not suitable for choosing governments, because governments can change the law and thus the rules of the market. This allows the current market leaders to distort the market in their own favour, ensuring their continued dominance. Although we disagree on many points I think we actually agree on this: market leaders should not be allowed to distort the market to their own advantage. But if you allow the rich to have disproportionate political influence, that is what will happen.
If it's truly unregulated free market capitalism, then there are no rules to be changed, and no political favour to win; the rules are set in stone: property rights and contract law.
To have property rights and contract law you need a government, and that means political influence. If money can buy political influence then the rich can change the law to suit themselves (which will probably mean a swift death for free markets, for reasons you yourself pointed out).
Furthermore, the existence of strong property rights completely negates your fear of concentrated power. It doesn't matter how much money a person/corporation has: if you don't want them to utilize your resources, they can do nothing about it.
They can change the law and weaken your property rights (introducing new taxes for example). I'm not opposed to property rights or contracts, but they will mean nothing in the long term if the market is allowed to interfere in politics, because they will be distorted to suit the interests of the rich.
I find it amusing that "universal suffrage" and "equality before law" are both integral components of free market capitalism
Nonsense. In a pure capitalist system, a man with no money has no power: how is that universal suffrage? Perhaps he can sell his labour, but he can't negotiate the price. If he is offered food and lodging in place of wages, he will take it to avoid starvation. With no money coming in he can never improve his situation, so he will remain a slave for the rest of his life. There is nothing here that violates the principles of free market capitalism: a man voluntarily exchanges his labour for food and lodging, that is all.
Now imagine the same situation with universal suffrage (which means the right to vote, not the right to spend your last dollar on food). Even wage-slaves can use their votes to influence the law, perhaps to provide them with food and lodging so that they can demand money for their labour. (Yes, providing food and lodging will mean using the evil coercive power of government to steal somebody else's property, but that's the social cost of avoiding slavery. Perhaps you think it's nobler for each man to stand on his own two feet, in which case I say why not abolish property rights and contract law as well, and fight it out with our teeth and fingernails?)
In fact, free market economics are really a superset of democratic voting, since you can cast more than one "vote" on an issue which is really important to you, and withhold "votes" in areas which you don't care about
Only if you have "votes" to cast. In a democracy you can spend your last dollar but you can't spend your right to vote.
You'd have to use addresses derived from public cryptographic keys. If two people claimed the same address, whichever one held the private key corresponding to the address would be able to prove it with a digital signature, so you could avoid address spoofing. The address space would have to be large enough to avoid accidental collisions - 128 bits (same as IPv6) would be ample.
Unfortunately this would mean no address aggregation so you wouldn't be able to use prefix-based routing. It's unreasonable to expect every router to maintain an up-to-date routing table for the entire network without aggregation, so routes would have to be discovered on-demand, allowing routers to ignore hosts that weren't receiving any traffic.
It's depressing how often I come across this "vote with your dollars" argument. Corporations are only concerned with their image in the eyes of potential customers. Governments have to worry about the opinions of all potential voters, not just those with money.
"The system works", as you put it, only in the short term. The problem is that power brings more power: over time, unregulated capitalism tends to concentrate more and more power in the hands of fewer and fewer people, because those with money are able to buy political influence and change the rules in their favour, thus attracting more money and more influence. Unless you are the single richest individual, you will sooner or later be in a position where the people above you are rigging the rules against you and forcing you down. This is the simple fact that free market libertarians fail to grasp: unregulated capitalism is not in anybody's long-term self-interest, except the single richest individual in the world. Everyone else eventually loses what they've won so far. In order to prevent a spiral towards tyranny, money and political power must be separated. That means not relying on the market as a mechanism to distribute social justice.
The only system that benefits more than one human being in the long run is a system based on universal suffrage and equality before the law. The market is not such a system. People who oppose unregulated capitalism are not necessarily whining parasites or tree-hugging utopian idiots. They just realise that a game of five billion players in which the winner gets to change the rules is not a game you want to play. The free market is a useful mechanism, but to treat it as a substitute for democratic government is a recipe for disaster.
When you want to send a packet across international borders you'll need a form stamped by the information ministries of both countries and signed by both the sender and the receiver (in triplicate of course). And there will be a customs charge for every packet set.
Result: the MTU will increase from 1460 bytes to 1.46 GB, and the quickest way of reading a foreign website will be air travel.
You support government control because it's preferable to corporate control, but perhaps there's a third option: no control.
I'm talking about a completely decentralized network with no central body allocating addresses, with strong encryption at the link level and end-to-end, guaranteeing privacy and freedom of speech to anyone who can connect to it.
Freenet and the Freehaven project's second-generation onion router have laid a lot of the groundwork, but they're designed to be internet overlays. What we need is a truly decentralized packet-switching network, independent of the internet, capable of operating over an ad hoc collection of wireless, leased line, modem and (for the moment) internet connections. The internet can function as scaffolding but nothing in the new network's design should be internet-specific.
It's already possible to build small networks of this kind - see Mute, for example. Each machine's address is derived from its public key, and you find routes by broadcasting. But broadcasting every query isn't scalable, so in my PhD research I'm looking for scalable ways to route packets across a large, untrusted network with no address aggregation. If you have any ideas, please reply and I'll send you my email address.:-)
Point well taken - I had my teeth kicked in last night in an unprovoked attack, in a CCTV-monitored street in London. Cameras did not prevent the attack, they're also unlikely to help the police find the attacker because the image quality is so poor.
It's ironic that this story appeared today, because last night I was attacked by a stranger on a London street. He kicked out one of my teeth and I needed stitches in my lips. I'm pretty sure the whole incident was caught on CCTV - central London is bristling with cameras - but I haven't bothered to contact the police because I'm also pretty sure the video images won't be clear enough to identify him. He probably realised that too - he wasn't deterred by the presence of cameras.
Because of the poor image quality, video cameras are useful for substantiating eyewitness accounts, but they're not very useful for deterring or identifying criminals. The police still have to catch someone red-handed for the video evidence to be useful.
Compiling the kernel isn't the problem, although there should really just be one command that does bzImage, modules, and modules_install all together.
The command you're looking for is make bzImage modules modules_install.;-p
What I'd really like to see is a program that uses some of the hardware detection code from live Linux distros to generate a kernel configuration file that matches the current hardware. Of course you could go in and tweak it afterwards, but all your drivers, number and type of CPUs etc would be pre-selected.
If you only want a CLI, add the line BootGUI=0 to C:\MSDOS.SYS (you might need to uncheck the hidden, system and read-only attributes first). Thanks for the link to 98lite though.
Old kernels aren't necessarily the best choice for old hardware. My laptop is a P120 with 32MB RAM - i guess that fits most people's definition of "crappy hardware from years ago". I run the latest 2.4 kernel with the 2.6 scheduler and kernel pre-emption backported - more responsive than stock 2.4 but happier in low memory situations than stock 2.6. It probably uses less memory than a 2.0-series kernel because almost everything is modularized, and i'm willing to bet it has much lower scheduling latency.
Slashdot Mirror
Languages aren't purely functional. (Gah! That term's overloaded in this discussion. (Gah! So's that one!)) What I mean is that languages might have historical, cultural, anthropological or even sentimental value beyond their "use value" in the modern world. Many cultures preserve their history orally, and if young people stop learning the "useless" local language then thousands of years of history and mythology will be lost. Maybe that doesn't matter, but it seems to me that increased global communication, for all its good points, is bringing us closer to a monoculture of worldviews.
I was wondering the same thing. The only explanation I could come up with was alphabetical ordering.
IIRC there was once a study of graduates from a police academy where rooms were allocated alphabetically. Years after they graduated, there was a strong correlation between two graduates' alphabetical proximity and the likelihood that they had remained friends. Conclusion: you can learn to like almost anyone if you're forced to spend time together. Getting back to the point, if groups were selected alphabetically for some activities, or if classes were seated alphabetically, you'd expect to see a long chain in the social network. I don't know why the ends of the chain should be joined, though. :-)
I subsequently skimmed the paper corresponding to the diagram. (Unfortunately all figures and tables are missing from the PDF - I can't find a complete version on the web.) They note the unusual structure of the graph - it's almost a perfect spanning tree, and is not a "small world" by any means - and explain it by a prohibition on "seconds relationships" - that is, a prohibition on dating the former partner of your former partner's new partner. They say this rule produces a graph with no 4-edge cycles. Cycles with an odd number of edges are rare because the reported relationships are overwhelmingly heterosexual, so this amounts to a prohibition on cycles with less than 6 edges. Applying this prohibition and controlling for degree distribution, randomly generated graphs are similar in structure to the one in the diagram.
However, it's possible to create a 4-edge cycle in a heterosexual graph without a "seconds relationship" - A dates B and then C, each of whom subsequently dates D. It's been a while since I had to deal with high school dating etiquette, but IIRC this situation is only taboo if B and C are friends - something which isn't recorded in the data. Nevertheless, while the social taboos might be more complex than those desribed in the paper, I'm convinced by the basic argument that they produce the unusual structure of the graph.
Expressed as a binary number where the first digit represents people, the second represents clothes and the third represents sex:
000 - not porn
001 - sometimes porn
010 - not porn
011 - porn
100 - porn (softcore)
101 - porn (hardcore)
110 - not porn
111 - porn (fetish and/or in a hurry)
No wonder the Supreme Court had such a hard time with it.
Looks interesting... how do you determine the context of an article?
For those who are interested, a PDF of the paper mentioned in the article is here. Running time is O(n^2) versus O(n^3) for previous algorithms, so don't go applying it to the Google cache just yet.
*grumble grumble* Damn scientists spoiling my fun with your empirical facts. ;-)
So it's not necessarily a bug if two different Swing implementations position things differently, as long as the overall layout matches what the programmer specified.
Because Microsoft's competitors will never choose Windows. Everyone (even Microsoft) is free to choose Linux, because it isn't controlled by a single vendor (not even Linus controls it - other individuals/companies can always fork development if they don't agree with his decisions).
Of course it's unrealistic to expect Microsoft to choose Linux, but they could do so if they wished. If someone had told me 6 years ago that one day MacOS would be based on BSD, I would have thought they were mad. In 6 years' time, will Windows be a proprietary compatibility layer running on a Linux kernel? Probably not, but it's possible.
That's strange, I don't remember agreeing to anything. In fact I don't even remember being asked... of course, I was only a baby at the time.
A constitution isn't an agreement. It's a non-binding description of how you can expect to be treated (for example, whether you can expect to be consulted before the terms of the non-binding description are unilaterally changed).
If you think government is an agreement, try disagreeing and see what happens to you. ;-)
I don't understand why this figure generated so much fuss. We're looking at a combinatorial system - you don't need many inputs to get an enormous number of outputs. It's like being amazed that telephone numbers in a large city "only" have eight digits.
Picture the human genome as a binary string 30,000 bits long. Each bit represents a gene: 1 means active, 0 means inactive (genes with more than 2 possible states can be represented by multiple bits). This gives us an upper limit of 2^30,000 possible phenotypes without even considering developmental influences. That number dwarfs the number of atoms in the universe, let alone the number of people that have ever lived. Even if only one in a trillion of those phenotypes is viable, we still have 2^29,960 to choose from. For me, the question is not "how can this be complex enough to create a human being?" but "how can we find the tiny subset of this information that actually corresponds to human beings?".
That figure of 0.69 looks pretty arbitrary... still I guess it proves that we're the only ones intelligent enough to rig the statistics in our favour. ;-)
We obviously have radically different conceptions of what "political" means.
Firstly, slavery is a separate issue from economic status. Just because you have no money, does not mean you are a slave.
That's true in some political systems, because you have rights that can't be taken away from you (or spent on necessities like food). As I demonstrated before, in a pure capitalist system the only right a destitute person has is the right to work, and if he can't earn more than the price of his food (and why should anyone offer him more?) then he will remain a de facto slave. Bear in mind that if someone offers him the price of his food in return for his labour, his choice is to take it or starve. He is not coerced by the other party but by the biological necessity of eating.
In fact you can. You spend it on an election, and then you're stuck with your (or rather, everyone else's) choice for an entire term with no "out".
That's spending your vote, not your right to vote. If you spend your vote you still have the right to vote at the next election.
A free market has a constant feedback, and response is near-immediate (compared to politics anyway).
I'm not disputing that the market is a very efficient economic mechanism. I'm just saying that a market system is not suitable for choosing governments, because governments can change the law and thus the rules of the market. This allows the current market leaders to distort the market in their own favour, ensuring their continued dominance. Although we disagree on many points I think we actually agree on this: market leaders should not be allowed to distort the market to their own advantage. But if you allow the rich to have disproportionate political influence, that is what will happen.
To have property rights and contract law you need a government, and that means political influence. If money can buy political influence then the rich can change the law to suit themselves (which will probably mean a swift death for free markets, for reasons you yourself pointed out).
Furthermore, the existence of strong property rights completely negates your fear of concentrated power. It doesn't matter how much money a person/corporation has: if you don't want them to utilize your resources, they can do nothing about it.
They can change the law and weaken your property rights (introducing new taxes for example). I'm not opposed to property rights or contracts, but they will mean nothing in the long term if the market is allowed to interfere in politics, because they will be distorted to suit the interests of the rich.
I find it amusing that "universal suffrage" and "equality before law" are both integral components of free market capitalism
Nonsense. In a pure capitalist system, a man with no money has no power: how is that universal suffrage? Perhaps he can sell his labour, but he can't negotiate the price. If he is offered food and lodging in place of wages, he will take it to avoid starvation. With no money coming in he can never improve his situation, so he will remain a slave for the rest of his life. There is nothing here that violates the principles of free market capitalism: a man voluntarily exchanges his labour for food and lodging, that is all.
Now imagine the same situation with universal suffrage (which means the right to vote, not the right to spend your last dollar on food). Even wage-slaves can use their votes to influence the law, perhaps to provide them with food and lodging so that they can demand money for their labour. (Yes, providing food and lodging will mean using the evil coercive power of government to steal somebody else's property, but that's the social cost of avoiding slavery. Perhaps you think it's nobler for each man to stand on his own two feet, in which case I say why not abolish property rights and contract law as well, and fight it out with our teeth and fingernails?)
In fact, free market economics are really a superset of democratic voting, since you can cast more than one "vote" on an issue which is really important to you, and withhold "votes" in areas which you don't care about
Only if you have "votes" to cast. In a democracy you can spend your last dollar but you can't spend your right to vote.
Unfortunately this would mean no address aggregation so you wouldn't be able to use prefix-based routing. It's unreasonable to expect every router to maintain an up-to-date routing table for the entire network without aggregation, so routes would have to be discovered on-demand, allowing routers to ignore hosts that weren't receiving any traffic.
Thanks very much for the links, I'll give it a look.
"The system works", as you put it, only in the short term. The problem is that power brings more power: over time, unregulated capitalism tends to concentrate more and more power in the hands of fewer and fewer people, because those with money are able to buy political influence and change the rules in their favour, thus attracting more money and more influence. Unless you are the single richest individual, you will sooner or later be in a position where the people above you are rigging the rules against you and forcing you down. This is the simple fact that free market libertarians fail to grasp: unregulated capitalism is not in anybody's long-term self-interest, except the single richest individual in the world. Everyone else eventually loses what they've won so far. In order to prevent a spiral towards tyranny, money and political power must be separated. That means not relying on the market as a mechanism to distribute social justice.
The only system that benefits more than one human being in the long run is a system based on universal suffrage and equality before the law. The market is not such a system. People who oppose unregulated capitalism are not necessarily whining parasites or tree-hugging utopian idiots. They just realise that a game of five billion players in which the winner gets to change the rules is not a game you want to play. The free market is a useful mechanism, but to treat it as a substitute for democratic government is a recipe for disaster.
Result: the MTU will increase from 1460 bytes to 1.46 GB, and the quickest way of reading a foreign website will be air travel.
Yeah, because corporations with bad business models have never done any damage to the internet.
I'm talking about a completely decentralized network with no central body allocating addresses, with strong encryption at the link level and end-to-end, guaranteeing privacy and freedom of speech to anyone who can connect to it.
Freenet and the Freehaven project's second-generation onion router have laid a lot of the groundwork, but they're designed to be internet overlays. What we need is a truly decentralized packet-switching network, independent of the internet, capable of operating over an ad hoc collection of wireless, leased line, modem and (for the moment) internet connections. The internet can function as scaffolding but nothing in the new network's design should be internet-specific.
It's already possible to build small networks of this kind - see Mute, for example. Each machine's address is derived from its public key, and you find routes by broadcasting. But broadcasting every query isn't scalable, so in my PhD research I'm looking for scalable ways to route packets across a large, untrusted network with no address aggregation. If you have any ideas, please reply and I'll send you my email address. :-)
Animals don't consent to being eaten, but that's legal in every country.
Point well taken - I had my teeth kicked in last night in an unprovoked attack, in a CCTV-monitored street in London. Cameras did not prevent the attack, they're also unlikely to help the police find the attacker because the image quality is so poor.
Because of the poor image quality, video cameras are useful for substantiating eyewitness accounts, but they're not very useful for deterring or identifying criminals. The police still have to catch someone red-handed for the video evidence to be useful.
The command you're looking for is make bzImage modules modules_install. ;-p
What I'd really like to see is a program that uses some of the hardware detection code from live Linux distros to generate a kernel configuration file that matches the current hardware. Of course you could go in and tweak it afterwards, but all your drivers, number and type of CPUs etc would be pre-selected.
If you only want a CLI, add the line BootGUI=0 to C:\MSDOS.SYS (you might need to uncheck the hidden, system and read-only attributes first). Thanks for the link to 98lite though.
Old kernels aren't necessarily the best choice for old hardware. My laptop is a P120 with 32MB RAM - i guess that fits most people's definition of "crappy hardware from years ago". I run the latest 2.4 kernel with the 2.6 scheduler and kernel pre-emption backported - more responsive than stock 2.4 but happier in low memory situations than stock 2.6. It probably uses less memory than a 2.0-series kernel because almost everything is modularized, and i'm willing to bet it has much lower scheduling latency.