It's certainly better than nothing and will likely deter all but the targeted industrial (or governmental) espionage attacks. Then again, i'd rather use the best solution, rather than a "good enough" one.
If you are really concerned about industrial/governmental attacks, you should definitely read the "Maintenance" section in the TPM standard. It specifies an optional backdoor which allows the manufacturer to extract your keys from the TPM. Of course it's not called a "backdoor", it's just a convenient way to retrieve the so-called "non-migratable" keys when your motherboard dies, but you can imagine other uses.
I'd rather use a "good enough" solution with no known weaknesses against my threat model, than one which gives a false sense of absolute security. The key should not be in the laptop, even in a TPM, period.
I could easily install a keylogger or data dumper in that initrd
You'd have to borrow my laptop, get access to the disk (i.e. reboot from a CD or disconnect the disk), install the keylogger, and return the laptop without me noticing. If you can do that, you could as well disable the secure booting code in the BIOS or hack the TPM itself.
Or you could exploit a security hole in my OS to install the keylogger with e.g. a virus. But if your virus has enough privileges to alter the boot partition, why bother installing a keylogger ? Just steal all the files you need. And this would work against a TCPA system too if its OS is unsafe.
The OS has to be able to boot without the user entering anything.
So you would like your computer to boot without requiring any authentication while the thief is sitting in front of it ?? No, I suppose you actually expect the user to enter a password on some kind of login screen. How is this different from entering it earlier in the boot process ?
encryption keys are locked in a TPM chip of some sort, and you can't retrieve them with software
And how do you unlock that TPM when you need to access your files ? I suppose you enter a passphrase or plug a USB dongle at some point during the boot process.
Encrypted filesystems require your boot partition have the encryption keys unencrypted
No. Just set up an unencrypted kernel+initrd which will prompt you for a passphrase or read a key from a USB dongle.
You don't need TCPA to protect your data against physical theft. Actually, a TPM is not even required to resist physical attacks (just read the specs). TCPA mostly aims to protect against software attacks (i.e. malware and unsophisticated users).
My definition: a system in which one agents gain is only possible through an equal expense on a seperate agent.
Fine with me, but the problem is how we measure perceived gains and losses. If we deal with sums of money, you will certainly agree that we must correct for inflation. But aren't there other things that we must factor in, like currency exchange rates ? Why not also correct for the average global growth that everybody is taking for granted ?
if everyone has a 2nd gen drug that allows you to live 20 years, and then I personally invent a 3rd gen drug that allows 30 years, but keep it to myself... are all the other people somehow poorer?
If everybody takes it for granted that medicine is supposed to progress at a certain rate, then yes, they are now lagging behind the normal pace of innovation, and you are ahead, and that gives you power over them.
I said there was another option: "you have $100 taken from you when someone else earns $100."
OK, that's strict zero-sum, but my constant-sum examples are pretty much the same, and that's why you chose $100/$100 (i.e. 50%/50% out of a fixed 100% of purchasing power) rather than $150/$200 (i.e. 43%/57%).
You only showed how political power is zero-sum... but that is not economic
In economics everything has a value: money, goods, services, currencies, social status, political power. The perceived value is not the same for everyone, though, and it doesn't scale linearly with the amount of money/goods/whatever you have; that's why game theory uses utility functions rather than sums of money. Before you claim to have proven something, would you please define your utility function ?
If everyone is growing then the global economy can not be a zero-sum game by definition
Could be that the utility function evaluates to 0 when you grow at the same rate as everyone else.
But 3rd generation isn't 'all that matters' if the choice is between 2nd gen and death.
You are implying that some people should consider themselves happy to be the losers in your game. But they won't, because their choice is between 3rd gen, 2nd gen and death, just like yours.
Of course I'd rather have $100 when everyone has $100
Personally I'd rather adopt the $150/$200 system, because it means there is a game, and I can learn to play better and aim for $250 on the next round.
You can whine about the global economy not distributing wealth equally...
Are you labeling me a crazy communist ?:-) Have I been whining ? Let me summarize my position, and I bet you'll agree:
The global economy is a zero-sum game, if you choose any realistic utility function.
Those who play "against" each other in the stock market create wealth for themselves collectively (because competition stimulates progress, and so on).
The loosers are those who choose not to play (or who can't afford to make risky investments). They may benefit from overall progress and productivity gains, but who cares that the Playstation 2 dropped from $300 to $130 in 6 years, when the only thing that counts is the new $500 Playstation 3 ?
My point since the beginning is that it doesn't make sense to call the stock market a game (be it zero-sum or not), because it is not even a closed system. The wealth it creates is only relative to the rest of the economy. That's why I insist on considering things in the context of the global economy.
Now assume everyone on earth adopts capitalism, so that we have a closed system, and everyone invests in the same stocks and indices in order to benefit from the 10% long-term average ROI. Since everyone enjoys the same growth, there's no relative benefit in playing. In this context the best ways to beat the system are: Being a venture capitalist; or becoming the CEO of a Fortune 500 company and paying yourself a huge salary; or acquiring wealth by military force. Oh wait, doesn't all this sound familiar ?
Assume that the earth was split into two planets with *NO* communication whatsoever and no trade whatsoever between the two halves.
Yeah right... In case you haven't noticed, there's something called "globalization" at work. The West forced Japan to open to free trade in 1853. Even North Korea doesn't live in complete isolation today.
if we were talking about politics this entire time, your comment would be relevant.
Well "power" is a generic concept. It works on a local scale too.
By the way, why you didn't answer the question ? (Would you rather take the $100 like everybody else, or $150 when your neighbour makes $200 ?) You said that there are other choices. Were you thinking about playing harder and getting $250 when the neighbour still makes $200 ? Then wouldn't that mean you are afraid of losing something (what?) even though you claim this is not a zero-sum game ?
Anyway this has little to do with the original debate. Obviously it's a good idea to invest in the stock market, for the same reason that it's a good idea to invest in government bonds. Whether the trading is done by humans or by programs is a minor detail.
Can you at least admit the global economy isn't a zero-sum game?
It depends on your utility function.
If you only care about raw numbers in US$, then it's not a zero-sum game. But then 20% inflation should make you very happy, and I bet it doesn't.
If your utility function is purchasing power then it's still not a zero-sum game, thanks to technical progress which benefits everybody in the long term.
But if your utility function is the power that economic wealth brings in world affairs, then I say the global economy is a zero-sum game.
would you rather eat less as long as everyone else eats less too, or be well nourished while some people eat gourmet every night?
I wasn't really implying a conscious choice, but your point is interesting. According to experiments on the Ultimatum game in the USA, plenty of people would rather starve than allow their neighbour to receive $70 while they only get $30. What do you think their utility function is ?
If it is a zero sum game, then the global average growth would be zero. What the parent was pointing out is that this is not the case.
Value is relative. If you correct your and your neighbours' growth for inflation in your local market, you will begin to see a zero-sum game. If you are interested in your standing in the global economy then you need to correct for global growth, and you get a zero-sum game.
Of course the standard of living improves for everybody. But it improves faster for the winners.
So if a $10^12 country grows only by 3% a year, it must be a loser compared to all the $10^9 countries that manage a 15% growth ?
Yes, because its place in the global economy decreases. If there are 100 of these 10^9 countries, the share of the 10^12 country drops by 0.95 percentage points the first year. And it gets worse: 1.04 points the second year, 1.13 points the third year, etc (until the smaller countries become too large to sustain 15% growth).
if the stock-market was a zero-sum game where the only way to win was to have someone else lose the same amount, then there'd be no point in playing it. Your average return would be zero.
Should be: the average return of all the players would be zero. But the good players would still consistently trounce the newbies. Think about poker, which is a zero-sum game.
But the sum total of the profits (or losses) of all companies is hugely positive.
Sounds like global economic growth. It may be huge, but it's only about 4 %. Explain to me how any country, company or individual that grows slower than the global average is not a loser in this game.
Bad idea. If you hold ISPs responsible, they will have no choice but to interfere with what you do with your computer. They won't let you connect to the net unless you are running Windows build #XXXX with firewall Y and antivirus Z. This can easily be done with Trusted Computing and Trusted Network Connect.
Both of these technologies are great for corporate networks, but I hope you can see where this leads if they become mandatory on your personal machine.
According to Cornell's lawyers, the DMCA was not a concern
Are these the same lawyers who sued the Norvegian guy who published the DVD obfuscation data ? And who patent the DNA sequences of medicinal plants from abroad ?
Forget about copyright - what about industrial trade secrets ? Would they mind if someone published details of the GPS military-grade signal ?
I'm PAYING twice for content? Once to download, and a second time to upload
Depends entirely on your subscription. Cable might bill you for each byte sent or received, but DSL is typically "1 Mbit/s downstream, 256 kbit/s upstream, unlimited volume". In the later case, bittorrent uploads neither increase your bill nor consume scarce resources.
How many amateurs and computer engineers are there around the world picking their computers apart?
Yeah, how many pick the ethernet chip apart, transistor by transistor ?
Plus, note that a backdoor does not necessarily result in observable information flow. If I wanted to bug a PC, I'd patch its hardware random number generator.
The HDCP handshake needn't be strong. We all know that cryptography is of little use to DRM.
The point is, it's painted as a technological measure for protecting copyrighted content. That's enough to get you arrested under the DMCA if you willfully break it.
Can you think of a way to win against rootkits without TCPA?
Yes, and I'll write it down before the Trusted Computing Group and their MSR minions come and brainwash me: Just don't make it possible for malware to drop a VMM underneath the OS in the first place. At least not after the boot process has reached some point of no return.
Or, require physical confirmation from the user. Do you remember when motherboards had jumper-controlled reflashing ? We didn't have BIOS-based viruses back then.
Why do most built-in WiFi cards have a hardware override switch ? Because users don't trust software anyway, even drivers signed by Microsoft, especially when they are running in a TCPA black box that nobody can decompile.
Oh, and TCPA itself requires a means for establishing the physical presence of the owner (aka a switch or jumper). There's no way around it.
Slashdot Mirror
It's certainly better than nothing and will likely deter all but the targeted industrial (or governmental) espionage attacks. Then again, i'd rather use the best solution, rather than a "good enough" one.
If you are really concerned about industrial/governmental attacks, you should definitely read the "Maintenance" section in the TPM standard. It specifies an optional backdoor which allows the manufacturer to extract your keys from the TPM. Of course it's not called a "backdoor", it's just a convenient way to retrieve the so-called "non-migratable" keys when your motherboard dies, but you can imagine other uses.
I'd rather use a "good enough" solution with no known weaknesses against my threat model, than one which gives a false sense of absolute security. The key should not be in the laptop, even in a TPM, period.
AC
You're giving them the algorithm used
Obviously you need to learn about Kerckhoffs' principle.
I could easily install a keylogger or data dumper in that initrd
You'd have to borrow my laptop, get access to the disk (i.e. reboot from a CD or disconnect the disk), install the keylogger, and return the laptop without me noticing. If you can do that, you could as well disable the secure booting code in the BIOS or hack the TPM itself.
Or you could exploit a security hole in my OS to install the keylogger with e.g. a virus. But if your virus has enough privileges to alter the boot partition, why bother installing a keylogger ? Just steal all the files you need. And this would work against a TCPA system too if its OS is unsafe.
The OS has to be able to boot without the user entering anything.
So you would like your computer to boot without requiring any authentication while the thief is sitting in front of it ?? No, I suppose you actually expect the user to enter a password on some kind of login screen. How is this different from entering it earlier in the boot process ?
AC
encryption keys are locked in a TPM chip of some sort, and you can't retrieve them with software
And how do you unlock that TPM when you need to access your files ? I suppose you enter a passphrase or plug a USB dongle at some point during the boot process.
Encrypted filesystems require your boot partition have the encryption keys unencrypted
No. Just set up an unencrypted kernel+initrd which will prompt you for a passphrase or read a key from a USB dongle.
You don't need TCPA to protect your data against physical theft. Actually, a TPM is not even required to resist physical attacks (just read the specs). TCPA mostly aims to protect against software attacks (i.e. malware and unsophisticated users).
AC
the depth of 0 km makes it pretty clear this wasn't a natural quake.
Before someone decides to "liberate" North Korea based on this intel, please read until the bottom of that page where it says "location uncertainty: depth fixed by location program", and read what this means here.
My definition: a system in which one agents gain is only possible through an equal expense on a seperate agent.
Fine with me, but the problem is how we measure perceived gains and losses. If we deal with sums of money, you will certainly agree that we must correct for inflation. But aren't there other things that we must factor in, like currency exchange rates ? Why not also correct for the average global growth that everybody is taking for granted ?
if everyone has a 2nd gen drug that allows you to live 20 years, and then I personally invent a 3rd gen drug that allows 30 years, but keep it to myself... are all the other people somehow poorer?
If everybody takes it for granted that medicine is supposed to progress at a certain rate, then yes, they are now lagging behind the normal pace of innovation, and you are ahead, and that gives you power over them.
I said there was another option: "you have $100 taken from you when someone else earns $100."
OK, that's strict zero-sum, but my constant-sum examples are pretty much the same, and that's why you chose $100/$100 (i.e. 50%/50% out of a fixed 100% of purchasing power) rather than $150/$200 (i.e. 43%/57%).
Wikipedia has an example of a realistic non-zero-sum transaction: a country trading its excess of bananas for another's excess of apples. That's a benefit of free trade. Does playing in the stock market produce that kind of mutual benefit ?
AC
Then, after all this, you say... I'd rather have the 3rd option (which wan't mentione before)
Hey, you're the one who said those aren't the only two choices. (But then you declined to clarify what you had on your mind)
You only showed how political power is zero-sum... but that is not economic
In economics everything has a value: money, goods, services, currencies, social status, political power. The perceived value is not the same for everyone, though, and it doesn't scale linearly with the amount of money/goods/whatever you have; that's why game theory uses utility functions rather than sums of money. Before you claim to have proven something, would you please define your utility function ?
If everyone is growing then the global economy can not be a zero-sum game by definition
Could be that the utility function evaluates to 0 when you grow at the same rate as everyone else.
But 3rd generation isn't 'all that matters' if the choice is between 2nd gen and death.
You are implying that some people should consider themselves happy to be the losers in your game. But they won't, because their choice is between 3rd gen, 2nd gen and death, just like yours.
AC
Of course I'd rather have $100 when everyone has $100
Personally I'd rather adopt the $150/$200 system, because it means there is a game, and I can learn to play better and aim for $250 on the next round.
You can whine about the global economy not distributing wealth equally...
Are you labeling me a crazy communist ? :-) Have I been whining ? Let me summarize my position, and I bet you'll agree:
My point since the beginning is that it doesn't make sense to call the stock market a game (be it zero-sum or not), because it is not even a closed system. The wealth it creates is only relative to the rest of the economy. That's why I insist on considering things in the context of the global economy.
Now assume everyone on earth adopts capitalism, so that we have a closed system, and everyone invests in the same stocks and indices in order to benefit from the 10% long-term average ROI. Since everyone enjoys the same growth, there's no relative benefit in playing. In this context the best ways to beat the system are: Being a venture capitalist; or becoming the CEO of a Fortune 500 company and paying yourself a huge salary; or acquiring wealth by military force. Oh wait, doesn't all this sound familiar ?
AC
Assume that the earth was split into two planets with *NO* communication whatsoever and no trade whatsoever between the two halves.
Yeah right... In case you haven't noticed, there's something called "globalization" at work. The West forced Japan to open to free trade in 1853. Even North Korea doesn't live in complete isolation today.
AC
if we were talking about politics this entire time, your comment would be relevant.
Well "power" is a generic concept. It works on a local scale too.
By the way, why you didn't answer the question ? (Would you rather take the $100 like everybody else, or $150 when your neighbour makes $200 ?) You said that there are other choices. Were you thinking about playing harder and getting $250 when the neighbour still makes $200 ? Then wouldn't that mean you are afraid of losing something (what?) even though you claim this is not a zero-sum game ?
Anyway this has little to do with the original debate. Obviously it's a good idea to invest in the stock market, for the same reason that it's a good idea to invest in government bonds. Whether the trading is done by humans or by programs is a minor detail.
AC
Can you at least admit the global economy isn't a zero-sum game?
It depends on your utility function.
would you rather eat less as long as everyone else eats less too, or be well nourished while some people eat gourmet every night?
I wasn't really implying a conscious choice, but your point is interesting. According to experiments on the Ultimatum game in the USA, plenty of people would rather starve than allow their neighbour to receive $70 while they only get $30. What do you think their utility function is ?
AC
What is the value of money ?
Would you rather earn $100 in the US where it buys you 39 Big Macs, or in China where you get 83 for the same amount of money ?
Would you rather earn $100 when everybody is making $100, or $150 when everybody else is making $200 ?
AC
Value is relative. If you correct your and your neighbours' growth for inflation in your local market, you will begin to see a zero-sum game. If you are interested in your standing in the global economy then you need to correct for global growth, and you get a zero-sum game.
Of course the standard of living improves for everybody. But it improves faster for the winners.
AC
Yes, because its place in the global economy decreases. If there are 100 of these 10^9 countries, the share of the 10^12 country drops by 0.95 percentage points the first year. And it gets worse: 1.04 points the second year, 1.13 points the third year, etc (until the smaller countries become too large to sustain 15% growth).
AC
Should be: the average return of all the players would be zero. But the good players would still consistently trounce the newbies. Think about poker, which is a zero-sum game.
But the sum total of the profits (or losses) of all companies is hugely positive.
Sounds like global economic growth. It may be huge, but it's only about 4 %. Explain to me how any country, company or individual that grows slower than the global average is not a loser in this game.
AC
Both of these technologies are great for corporate networks, but I hope you can see where this leads if they become mandatory on your personal machine.
AC
So lets hope the private key wasn't on the hacked server, or that the intruder didn't bother backdooring GPG for passphrases.
AC
Are these the same lawyers who sued the Norvegian guy who published the DVD obfuscation data ? And who patent the DNA sequences of medicinal plants from abroad ?
Forget about copyright - what about industrial trade secrets ? Would they mind if someone published details of the GPS military-grade signal ?
AC
Depends entirely on your subscription. Cable might bill you for each byte sent or received, but DSL is typically "1 Mbit/s downstream, 256 kbit/s upstream, unlimited volume". In the later case, bittorrent uploads neither increase your bill nor consume scarce resources.
AC
So you think you control your PC or tape deck ? Think again.
It's called "spam".
Yeah, how many pick the ethernet chip apart, transistor by transistor ?
Plus, note that a backdoor does not necessarily result in observable information flow. If I wanted to bug a PC, I'd patch its hardware random number generator.
AC
The HDCP handshake needn't be strong. We all know that cryptography is of little use to DRM.
The point is, it's painted as a technological measure for protecting copyrighted content. That's enough to get you arrested under the DMCA if you willfully break it.
AC
Granted, but check this helicopter robot armed with a shotgun and tell me you are not scared. AC
Yes, and I'll write it down before the Trusted Computing Group and their MSR minions come and brainwash me: Just don't make it possible for malware to drop a VMM underneath the OS in the first place. At least not after the boot process has reached some point of no return.
Or, require physical confirmation from the user. Do you remember when motherboards had jumper-controlled reflashing ? We didn't have BIOS-based viruses back then.
Why do most built-in WiFi cards have a hardware override switch ? Because users don't trust software anyway, even drivers signed by Microsoft, especially when they are running in a TCPA black box that nobody can decompile.
Oh, and TCPA itself requires a means for establishing the physical presence of the owner (aka a switch or jumper). There's no way around it.
AC
Exactly. Yet another example of the kind of "innovation" that gave us NTP vs RIM.
Email, IM, PC, phone, TV, tivo, mp3, web, P2P - Pick two buzzwords, write a press release.
The real question is: why do journalists (and bloggers) propagate this clueless marketing, instead of debunking it ?
AC