Actually all but one of the reactors were capable of normal operation. I know it's a bit of a risk picking the one that you'd leave activated, but still. Don't shut down your only working power generator if you absolutely need it.
And what do you think about the small solid state nuclear reactor for emergency power idea ?
So which is it : 1) we know all the little feedback mechanisms in our athmosphere (and yes - there are millions, if not billions of those) and so we can predict the consequences of geoengineering and we have reasonable evidence for global warming 2) we do not know all the little feedback mechanisms, meaning that we cannot predict the consequences of geoengineering AND we do not know anything about global warming, not even if it happens at all
If you "believe" in global warming, you're pretty much forced to accept geoengineering will work. And of course, it won't.
Just shoot up a lot of water 1km into the air. The amount of water that has to make it to 20km is not that great. Some of the drops will get lucky and get to 20km.
Anyone with a plane and basic chemical knowledge who feels particulary "green" at some point in time.
Or a foreign power, or even terrorists that feel like causing a cooling disaster would benefit them, or alternatively would benefit <insert imaginary child-raping friend/prophet/mass_of_earth_with_a_name/... of choice here>
Undoing geo-engineering is basically impossible, so as soon as someone makes the decision for you, you're pretty much fucked. We're hanging by a rope from a cliff face and geo-engineering is researching scissors. I'm not saying it is necessarily going to blow up, just that it's pretty fucking likely.
Just wait until politicians realize that geo-engineering solutions 1) don't require global coordination (in fact they'd probably be detracted by it) 2) don't cost all that much, negating the need for huge budgets and the power concentration they cause
Politicians won't like this. Forced CO2 reductions, especially in the form of increased taxation, on the other hand, embody all bad things that government is : it's hugely invasive, expensive, requires massive bureaucracies, and requires regulatory bodies with worldwide power. Caesar should have considered "veni, co2'ed, vici"... much more effective.
In addition to that there's of course the huge legal question. Right now climate change is the result of natural processes. Suppose the US starts changing climate, obviously against the wishes of at least a few parties.
Suppose the US were to fix global warming by actively regulating climate. Obviously this is equivalent to "stealing" all agricultural production from Greenland, parts of Canada, Siberia, even Tasmania will sufffer on the other end of the globe. Less obivous, but much more massive losses will occur in Africa. Furthermore, there will be massive increases in required heating in large parts of the world, basically returning to the situation of the 1970s.
Since the US directly and deliberately caused these damages, is it now not responsible for paying damages for the losses suffered ?
To illustrate : in 1940s it was perfectly normal for a winter to kill > 1000 people a year in Moskou. Today, the only dead are from alcohol poisoning and traffic accidents. Whoever fixes global warming through geoengineering is obviously directly responsible for their deaths. What with the half trillions dollars of additional heating required for people to survive ?
I'm not complaining, I like the thumbnails. But they do constitute exactly the situation you asked me to demonstrate : 1) they're showing the site's content, often with the sites' advertisements downplayed 2) google's ads, by contrast, are superimposed on that
You know what else would have prevented a meltdown ? NOT shutting down the reactor. If the tsunami had killed the operators of the plant, the plant would have survived intact. The plants were built with the assumption that it would not be possible during a disaster to cut off the plant's access to the grid. Then the grid connection was made along a single long line built over the sea.
I mean, come on.
I'm not suggesting the following is a good idea, but a nuclear reactor is a hell of a lot more stable than a diesel generator. So in a nuclear power plant, why not have a reactor, running on spent fuel for example, which can run entirely cold (cold meaning ~200 degrees, the point being that the reactor is perfectly safe without cooling), and have that pump provide power to the cooling system for the real reactors ? It wouldn't need to provide more than a few kilowatts.
That way even if the plant and it's operators are completely cut off, the cooling system will remain in operation for years after the shutdown order is given - and it only needs to run for 24 hours.
Everyone else has exactly the number of guns that Google does,
Riiiiight... you sound like the Iraqi information minister here.
1) Does google have the same technical options as everybody else ? No, they have more (e.g. they use undocumented, or badly documented features of their own software, but it doesn't quite end there) 2) Does google have the same clout as everybody else ? Definitely not.
Of course it makes sense. The TPM itself has they keys, and you can use them to encrypt and decrypt things, you just cannot read the keys. Here's how it works.
The TPM makes a signature of the "trusted computing base". This is the code of the bootloader + operating system that's executing + the current program that's executing. Basically a checksum of all code that can influence the execution of the program. This checksum itself forms the basis of an encryption key.
Then the program executes, and produces some output. The program sends this output to the TPM, which signs it using the checksum key, then signs the whole thing using it's private key.
The manufacturer of TPM's makes a public key available (the manufacturer's private key is in a TPM, and is destroyed after a production run), and the TPM itself has it's public key signed with the manufacturer's key. This is not meant to be a 100% correct representation of the process, because there's a bunch of caveats, mainly to do with protecting the privacy of the TPM owner, but a general operating principle.
Ok, so you make a deal. The remote party agrees to execute your program, and it makes a tcp connection to your server (or establishes another means of communication). Ok now you receive output from the program. You know the checksum from the execution environment of your program on the remote computer, and you can verify it's genuine by examining the certificate chain. You look up this checksum and see if you can agree with it (basically, if there's no VM or debugger or strange dll overrides or some such in the remote execution environment, you accept*). Otherwise, you refuse to communicate and ask for your money back.
* you don't know the exact layout of the remote environment, only it's checksum. So you can verify that against the "expected" execution environment that the guy promised at contract time. E.g. VMWare on bare metal.
One of the things this gives you is a secure communications channel with your program. You are 100% sure it's your program on the other end, and due to the encryption nobody can insert or delete stuff from the stream (unless you haven't done your homework). Now you use the program, secure in the knowledge that the parties that can fuck you over are very minimal : 1) intel (or AMD, or Snapdragon, or...) 2) the memory manufacturer 3) the hardware producer of the TPM 4) VMware (in this example) Specifically, you have the same security environment you would have if the server was your own server. (think about it, just typing the response to this post into your computer requires a bigger set of trusts : your OS, browser and slashdot would be included, each of those could change your post in transit).
Since the site clearly embodies the fact that they find the apple situation acceptable, clearly they have no problem with draconian, completely one-sided DRM. They have no problem with their computer obeying someone else's wishes over their own.
Yet this is supposedly what they're attacking.
This is like Chavez' famous speech, you know, the one where he attacks the rich for wasting money for 30 minutes, and the first question from one of the journalists is "Isn't that an armani suit with italian shoes you're wearing ?". Or the picture of Al Gore's vacation getup : not just one huge trailer, but actually 2 wagons. Behind the end of wagon 2 there is a car with a boat mounted on it, and the trailer actually has a built-in garage, presumably containing a car. The same can be said about his garage : a big collection of gas guzzlers. It makes the statement "Al Gore cares about wasting CO2" completely untenable (well, he's a politician, of course).
The same is true here. This person clearly doesn't mind proclaiming in public that he's in favor of DRM, or at least not willing to make sacrifices to prevent DRM. And yet he's attacking a very mild form of DRM that is reciprocal.
And the second thing is that this video was clearly made with a budget. Which begs the question : who paid for this ? Hell, maybe it's Apple.
Okay, search for a term which has ads. Don't make it complicated, make it "car". There you go : content from external sites, google's ads. But this is quite tame, right ?
Now hover your mouse over one of the results. Boom. All content of the external site, rendered. Google's ads still visible (and more prominent than those on the external site).
Alternatively, click on the ">>" icon to the right of a result entry.
It's probably the amount of RAM required to have every line of source code + one device image file + gcc + java + linker +... all in ram at the same time.
It'll probably compile on 640k of ram. It just won't be finished by Christmas.
The issue is not to make systems perfectly secure, but simply hard enough to crack to stop 99.99% of attackers.
Bank cards have been cracked much more than TPM's, and they are much easier to fool too. Last crack I heard of what of VISA chip cards. Simply tell the bank the transaction is verified with the card, and tell the card the bank verified the transaction. *boom* Transaction passed, and you can pull this all but the very latest models of cards today. Morons. Yet they're still in use, and nobody's seriously discussing replacing them any time soon.
Also, it's not like a cracked TPM would let you do anything at all to a bank. You would need to crack *someone else's* TPM, without him noticing, and before he has a chance to enroll another TPM with the bank, or warn them. Then you'd have to execute code with the cracked TPM still running, so you'd be able to get at the actual signature, to get the data to encrypt.
While it's true that equipment capable of doing this exists, I'm pretty sure I will notice anyone bringing in chip slicing and electron microscope technology in my living room. Why, the boatloads of liquid nitrogen would probably not fit in Obama's largest living room. Also, you'll sure as hell not be able to give me back my laptop undamaged after that, so chances are you'll be noticed even if you have these resources.
Security has long since given up the idea of making cracks impossible. We're making them unfeasible.
What he is referring to is "trusted execution" technology, which is a different beast entirely. It basically means that the OS will refuse to execute any non-signed binary. Currently the only use of this I know of is on linux systems. Redhat, for example, supports this.
Sadly trusted execution security is as vulnerable to government morons, and others, as everything else : sign anything that loads code, from ld.so for the only slightly moronic types, to a perl or python binary for the truly gifted retard and it's all over.
But let's get real here : microsoft will (eventually) implement trusted execution on corporate domains whether or not TPM gets implemented. If you buy a computer that boots in trusted execution mode, it's all over. This has nothing to do with TPM, except that TPM allows remote systems to verify if you run a trusted execution environment or not.
I think the issue is much more about that this movie is made in a DRM'ed format that implements much more invasive stuff (in software) against you.
Sadly, the movie is propaganda. A TPM cannot take over a computer, it never will be able to do that (Ironically, quicktime has implemented mechanisms in the past that *do* take over computers at the lowest levels).
I don't get how someone who uses *anything* apple can seriously criticize trusted computing. Why criticize the danger of the needle in a mean and smiling kids' hand when you've already got harpooned in the ass ? Apple's DRM is the worst on the market, by far, and is 100% unilateral. Just try installing non-approved software on apple devices.
Trusted computing allows anyone to participate in both ends of the trust relationship, unlike apple.
Offline payments also seem largely unnecessary given how the internet is increasingly available anywhere.
I can attest to plenty of places where it isn't.
Also there are a lot of potential pitalls. If you transfer money to me offline, can the money disappear if the computers are never synchronized?
The money cannot disappear. There are 3 possible cases : Both people sync, there is no problem. Either one syncs, and the money's there. Both neglect to synchronize, and the money isn't transferred at all.
Loss of control. My stuff is mine, period, and I don't have to give lenghty explanations of that. But no, it's not piracy. On my hardware, which I paid for, I should have absolute access to every single bit of it.
You do, you just only have access to other people's data on your machine in encrypted form (this is not even necessarily so, but presumably this is how they'll want it).
You can touch every last bit on your machine including those on the TPM, you just cannot read them (only erase + rewrite). You have full access on your machine and your say-so determines 100% if something happens or not. TPM simply means it'll be impossible/hard for you to lie about having done something.
The basic application of the TPM is simply that. Send someone a program, receive back the signed output of that program. It is impossible for the owner of the computer to forge that signature, and that signature identifies exactly which series of machine instructions was used to arrive at that data. Everything else remains fully within his control.
I will certainly not buy anything that implements such a scheme.
1) you probably already have, do you have any American-made cpu produced in the last 2 years ? 2) even if you don't care for it, the ability to offer trusted relationships with other entities is a very useful one
The argument is that google is perfectly willing to add it's own adds to views of other people's webpages, yet refused the reverse (e.g. someone showing google with an add on top).
I guess the hypocrisy accusation comes from the fact that your argument applies equally well to just about any evil organisation. The problem is simple. If everybody is allowed to shoot and kill, those with guns have the obvious advantage. Since google >>>>>>>>>>> other websites, a similar principle applies here.
I wonder how many times this has to be said, but a TPM does not allow unilateral control. It allows bidirectional trust. I can send you a program, ask you to run it, get results back, and then verify that you indeed produced those results using my program. (e.g. think of it this way : I could send you a program that "is" a credit card. You could buy and sell stuff with it, transfer money between different instances of the program over the internet without contacting any bank, and then sync this program, say, weekly, back with the bank. Right now such a program is utterly impossible because banks do not trust you not to tamper with the data. TPM would allows banks (and that program) to detect (NOT prevent) you've tampered with the data. You can delete the data without anyone's permission. You can tamper with the data without anyone's permission, but you cannot lie about the fact afterwards.
If you could use trusted computing today, bitcoin would be no more complex than a program that can correctly execute transactions against an internal counter over tcp. Encryption would be implicit, and nobody would need to think about it.
One application I'd consider very nice would be this : you send a vmware image to a provider. It puts a site online at https:/// something. You can make the setup such that if they try to read the disk or memory of the vmware image, the site will go offline and they'll be completely unable to bring it back up, or read the disk, because they lack the encryption key. You would be able to let external parties process highly sensitive information without having to fear for the safety of your data.
I find the different applications possible with this technology to be both positive and negative from the government's point of view. It can partially prevent piracy (though no better than starcraft II's online only play can do so). It can just as well be used to create an online kiddy porn site that "customers" wouldn't be able to expose even if they wanted to. It could make a guaranteed 100% safe Tor network.
A TPM isn't good or evil, it's just a technology to prevent idiots from exposing their private keys. The thinking is simple : given that out of 5 major american banks 3 managed to expose their root keys for the payment system, 2 of which found fraudulent transactions (due to dishonest employees), what makes you think you can do better ?
But can we please stop the idiotic idea that some other party has access to TPM private keys ? This is not the case.
The whole point of the TPM is that it has a private key which NOBODY controls. It is in the device exclusively. There is circuitry inside the TPM to run key generation and to export the public key. There is not circuitry to export the private key at all.
Nobody has the key to this device. Not you. Not the manufacturer. Nobody.
That is why they're used at banks. Banks can have a root private key they control, and no matter how idiotic/malicious their employees, the key cannot be compromised (it can be erased by nitwits though, which sadly frequently happens, which usually causes an outage on inter-bank payment traffic until a new TPM can be shipped to them (bank TPMs are 6 kg black boxes that have additional security, mainframe compatibility and cobol and java interface libraries - see "Night of the living dead 184" for how well it works in practice). But nobody has the keys at all, which means they cannot be stolen without someone wondering why their inter-bank traffic just started getting all sorts of "permission denied" errors.
Those aren't gift economies, they're dictratorships. Obviously you don't pay for stuff in a dictatorship, but I seriously doubt that should be called gift economy. If anything, it's communism.
You are basing this on a wrong assumption. There is *nothing* in TPM that prevents you from creating or booting a custom OS. The ONLY thing TPM does is that it prevents you from lying about which OS you're running (you can refuse to say, you can give obviously wrong information, but you can't give convincing-but-wrong information).
Actually TPM allows protection in both directions. It works a bit like banks' systems. With a TPM you can secure a laptop, give it out to anyone, and you can set it up so they won't be able to break the encryption even if they know the passwords.
If you work for a company, you can give out VPN credentials to idiots that are uncopyable. If they get infected with a virus, the VPN won't come up.
I've consulted for a bank, and here's the dream : full offline money. If you have a TPM they will manage your account in your laptop (or phone, or...) and have full offline payments. Because the TPM will only give their program access to the data, they can still prevent you from simply adding money in your own account, while allowing fully disconnected payments to occur which the bank will only find out about weeks after the fact (and so can you on other's computers of course).
In general TPM's allow fully disconnected trust relationships.
Surely such features are worth something ? Several linux companies are already using them.
All it does is simply making sure that if you tell some company you're going to take good care of their data, you have to actually do it (or delete the data, you're perfectly at liberty to do that). I mean what do you have against this ? Other than "I want to pirate stuff" (which will still be perfectly possible, just slightly more involved).
Slashdot Mirror
Actually all but one of the reactors were capable of normal operation. I know it's a bit of a risk picking the one that you'd leave activated, but still. Don't shut down your only working power generator if you absolutely need it.
And what do you think about the small solid state nuclear reactor for emergency power idea ?
So which is it :
1) we know all the little feedback mechanisms in our athmosphere (and yes - there are millions, if not billions of those) and so we can predict the consequences of geoengineering and we have reasonable evidence for global warming
2) we do not know all the little feedback mechanisms, meaning that we cannot predict the consequences of geoengineering AND we do not know anything about global warming, not even if it happens at all
If you "believe" in global warming, you're pretty much forced to accept geoengineering will work. And of course, it won't.
Just shoot up a lot of water 1km into the air. The amount of water that has to make it to 20km is not that great. Some of the drops will get lucky and get to 20km.
done/done
Anyone with a plane and basic chemical knowledge who feels particulary "green" at some point in time.
Or a foreign power, or even terrorists that feel like causing a cooling disaster would benefit them, or alternatively would benefit <insert imaginary child-raping friend/prophet/mass_of_earth_with_a_name/... of choice here>
Undoing geo-engineering is basically impossible, so as soon as someone makes the decision for you, you're pretty much fucked. We're hanging by a rope from a cliff face and geo-engineering is researching scissors. I'm not saying it is necessarily going to blow up, just that it's pretty fucking likely.
Just wait until politicians realize that geo-engineering solutions
1) don't require global coordination (in fact they'd probably be detracted by it)
2) don't cost all that much, negating the need for huge budgets and the power concentration they cause
Politicians won't like this. Forced CO2 reductions, especially in the form of increased taxation, on the other hand, embody all bad things that government is : it's hugely invasive, expensive, requires massive bureaucracies, and requires regulatory bodies with worldwide power. Caesar should have considered "veni, co2'ed, vici" ... much more effective.
In addition to that there's of course the huge legal question. Right now climate change is the result of natural processes. Suppose the US starts changing climate, obviously against the wishes of at least a few parties.
Suppose the US were to fix global warming by actively regulating climate. Obviously this is equivalent to "stealing" all agricultural production from Greenland, parts of Canada, Siberia, even Tasmania will sufffer on the other end of the globe. Less obivous, but much more massive losses will occur in Africa. Furthermore, there will be massive increases in required heating in large parts of the world, basically returning to the situation of the 1970s.
Since the US directly and deliberately caused these damages, is it now not responsible for paying damages for the losses suffered ?
To illustrate : in 1940s it was perfectly normal for a winter to kill > 1000 people a year in Moskou. Today, the only dead are from alcohol poisoning and traffic accidents. Whoever fixes global warming through geoengineering is obviously directly responsible for their deaths. What with the half trillions dollars of additional heating required for people to survive ?
I'm not complaining, I like the thumbnails. But they do constitute exactly the situation you asked me to demonstrate :
1) they're showing the site's content, often with the sites' advertisements downplayed
2) google's ads, by contrast, are superimposed on that
Yeah but you have to admit it hurts their credibility pretty badly.
If code pink co-rallied with the NRA, you wouldn't call that "rather irrelevant", yet it's the same thing essentially.
You know what else would have prevented a meltdown ? NOT shutting down the reactor. If the tsunami had killed the operators of the plant, the plant would have survived intact. The plants were built with the assumption that it would not be possible during a disaster to cut off the plant's access to the grid. Then the grid connection was made along a single long line built over the sea.
I mean, come on.
I'm not suggesting the following is a good idea, but a nuclear reactor is a hell of a lot more stable than a diesel generator. So in a nuclear power plant, why not have a reactor, running on spent fuel for example, which can run entirely cold (cold meaning ~200 degrees, the point being that the reactor is perfectly safe without cooling), and have that pump provide power to the cooling system for the real reactors ? It wouldn't need to provide more than a few kilowatts.
Something like this would be more than adequate
That way even if the plant and it's operators are completely cut off, the cooling system will remain in operation for years after the shutdown order is given - and it only needs to run for 24 hours.
Everyone else has exactly the number of guns that Google does,
Riiiiight ... you sound like the Iraqi information minister here.
1) Does google have the same technical options as everybody else ? No, they have more (e.g. they use undocumented, or badly documented features of their own software, but it doesn't quite end there)
2) Does google have the same clout as everybody else ? Definitely not.
Of course it makes sense. The TPM itself has they keys, and you can use them to encrypt and decrypt things, you just cannot read the keys. Here's how it works.
The TPM makes a signature of the "trusted computing base". This is the code of the bootloader + operating system that's executing + the current program that's executing. Basically a checksum of all code that can influence the execution of the program. This checksum itself forms the basis of an encryption key.
Then the program executes, and produces some output. The program sends this output to the TPM, which signs it using the checksum key, then signs the whole thing using it's private key.
The manufacturer of TPM's makes a public key available (the manufacturer's private key is in a TPM, and is destroyed after a production run), and the TPM itself has it's public key signed with the manufacturer's key. This is not meant to be a 100% correct representation of the process, because there's a bunch of caveats, mainly to do with protecting the privacy of the TPM owner, but a general operating principle.
Ok, so you make a deal. The remote party agrees to execute your program, and it makes a tcp connection to your server (or establishes another means of communication). Ok now you receive output from the program. You know the checksum from the execution environment of your program on the remote computer, and you can verify it's genuine by examining the certificate chain. You look up this checksum and see if you can agree with it (basically, if there's no VM or debugger or strange dll overrides or some such in the remote execution environment, you accept*). Otherwise, you refuse to communicate and ask for your money back.
* you don't know the exact layout of the remote environment, only it's checksum. So you can verify that against the "expected" execution environment that the guy promised at contract time. E.g. VMWare on bare metal.
One of the things this gives you is a secure communications channel with your program. You are 100% sure it's your program on the other end, and due to the encryption nobody can insert or delete stuff from the stream (unless you haven't done your homework). Now you use the program, secure in the knowledge that the parties that can fuck you over are very minimal : ...)
1) intel (or AMD, or Snapdragon, or
2) the memory manufacturer
3) the hardware producer of the TPM
4) VMware (in this example)
Specifically, you have the same security environment you would have if the server was your own server. (think about it, just typing the response to this post into your computer requires a bigger set of trusts : your OS, browser and slashdot would be included, each of those could change your post in transit).
That's trusted computing.
Since the site clearly embodies the fact that they find the apple situation acceptable, clearly they have no problem with draconian, completely one-sided DRM. They have no problem with their computer obeying someone else's wishes over their own.
Yet this is supposedly what they're attacking.
This is like Chavez' famous speech, you know, the one where he attacks the rich for wasting money for 30 minutes, and the first question from one of the journalists is "Isn't that an armani suit with italian shoes you're wearing ?". Or the picture of Al Gore's vacation getup : not just one huge trailer, but actually 2 wagons. Behind the end of wagon 2 there is a car with a boat mounted on it, and the trailer actually has a built-in garage, presumably containing a car. The same can be said about his garage : a big collection of gas guzzlers. It makes the statement "Al Gore cares about wasting CO2" completely untenable (well, he's a politician, of course).
The same is true here. This person clearly doesn't mind proclaiming in public that he's in favor of DRM, or at least not willing to make sacrifices to prevent DRM. And yet he's attacking a very mild form of DRM that is reciprocal.
And the second thing is that this video was clearly made with a budget. Which begs the question : who paid for this ? Hell, maybe it's Apple.
Okay, search for a term which has ads. Don't make it complicated, make it "car". There you go : content from external sites, google's ads. But this is quite tame, right ?
Now hover your mouse over one of the results. Boom. All content of the external site, rendered. Google's ads still visible (and more prominent than those on the external site).
Alternatively, click on the ">>" icon to the right of a result entry.
It's probably the amount of RAM required to have every line of source code + one device image file + gcc + java + linker + ... all in ram at the same time.
It'll probably compile on 640k of ram. It just won't be finished by Christmas.
The issue is not to make systems perfectly secure, but simply hard enough to crack to stop 99.99% of attackers.
Bank cards have been cracked much more than TPM's, and they are much easier to fool too. Last crack I heard of what of VISA chip cards. Simply tell the bank the transaction is verified with the card, and tell the card the bank verified the transaction. *boom* Transaction passed, and you can pull this all but the very latest models of cards today. Morons. Yet they're still in use, and nobody's seriously discussing replacing them any time soon.
Also, it's not like a cracked TPM would let you do anything at all to a bank. You would need to crack *someone else's* TPM, without him noticing, and before he has a chance to enroll another TPM with the bank, or warn them. Then you'd have to execute code with the cracked TPM still running, so you'd be able to get at the actual signature, to get the data to encrypt.
While it's true that equipment capable of doing this exists, I'm pretty sure I will notice anyone bringing in chip slicing and electron microscope technology in my living room. Why, the boatloads of liquid nitrogen would probably not fit in Obama's largest living room. Also, you'll sure as hell not be able to give me back my laptop undamaged after that, so chances are you'll be noticed even if you have these resources.
Security has long since given up the idea of making cracks impossible. We're making them unfeasible.
Like many others he is an idiot.
What he is referring to is "trusted execution" technology, which is a different beast entirely. It basically means that the OS will refuse to execute any non-signed binary. Currently the only use of this I know of is on linux systems. Redhat, for example, supports this.
Sadly trusted execution security is as vulnerable to government morons, and others, as everything else : sign anything that loads code, from ld.so for the only slightly moronic types, to a perl or python binary for the truly gifted retard and it's all over.
But let's get real here : microsoft will (eventually) implement trusted execution on corporate domains whether or not TPM gets implemented. If you buy a computer that boots in trusted execution mode, it's all over. This has nothing to do with TPM, except that TPM allows remote systems to verify if you run a trusted execution environment or not.
I think the issue is much more about that this movie is made in a DRM'ed format that implements much more invasive stuff (in software) against you.
Sadly, the movie is propaganda. A TPM cannot take over a computer, it never will be able to do that (Ironically, quicktime has implemented mechanisms in the past that *do* take over computers at the lowest levels).
I don't get how someone who uses *anything* apple can seriously criticize trusted computing. Why criticize the danger of the needle in a mean and smiling kids' hand when you've already got harpooned in the ass ? Apple's DRM is the worst on the market, by far, and is 100% unilateral. Just try installing non-approved software on apple devices.
Trusted computing allows anyone to participate in both ends of the trust relationship, unlike apple.
Offline payments also seem largely unnecessary given how the internet is increasingly available anywhere.
I can attest to plenty of places where it isn't.
Also there are a lot of potential pitalls. If you transfer money to me offline, can the money disappear if the computers are never synchronized?
The money cannot disappear. There are 3 possible cases : Both people sync, there is no problem. Either one syncs, and the money's there. Both neglect to synchronize, and the money isn't transferred at all.
Loss of control. My stuff is mine, period, and I don't have to give lenghty explanations of that. But no, it's not piracy. On my hardware, which I paid for, I should have absolute access to every single bit of it.
You do, you just only have access to other people's data on your machine in encrypted form (this is not even necessarily so, but presumably this is how they'll want it).
You can touch every last bit on your machine including those on the TPM, you just cannot read them (only erase + rewrite). You have full access on your machine and your say-so determines 100% if something happens or not. TPM simply means it'll be impossible/hard for you to lie about having done something.
The basic application of the TPM is simply that. Send someone a program, receive back the signed output of that program. It is impossible for the owner of the computer to forge that signature, and that signature identifies exactly which series of machine instructions was used to arrive at that data. Everything else remains fully within his control.
I will certainly not buy anything that implements such a scheme.
1) you probably already have, do you have any American-made cpu produced in the last 2 years ?
2) even if you don't care for it, the ability to offer trusted relationships with other entities is a very useful one
http://www.youtube.com/watch?v=DTXO7KGHtjI
Virtual pop stars can't be that far off in America.
May I advise Belgium ? The food is good, the beer is great and so is the social security.
The argument is that google is perfectly willing to add it's own adds to views of other people's webpages, yet refused the reverse (e.g. someone showing google with an add on top).
I guess the hypocrisy accusation comes from the fact that your argument applies equally well to just about any evil organisation. The problem is simple. If everybody is allowed to shoot and kill, those with guns have the obvious advantage. Since google >>>>>>>>>>> other websites, a similar principle applies here.
I wonder how many times this has to be said, but a TPM does not allow unilateral control. It allows bidirectional trust. I can send you a program, ask you to run it, get results back, and then verify that you indeed produced those results using my program. (e.g. think of it this way : I could send you a program that "is" a credit card. You could buy and sell stuff with it, transfer money between different instances of the program over the internet without contacting any bank, and then sync this program, say, weekly, back with the bank. Right now such a program is utterly impossible because banks do not trust you not to tamper with the data. TPM would allows banks (and that program) to detect (NOT prevent) you've tampered with the data. You can delete the data without anyone's permission. You can tamper with the data without anyone's permission, but you cannot lie about the fact afterwards.
If you could use trusted computing today, bitcoin would be no more complex than a program that can correctly execute transactions against an internal counter over tcp. Encryption would be implicit, and nobody would need to think about it.
One application I'd consider very nice would be this : you send a vmware image to a provider. It puts a site online at https:/// something. You can make the setup such that if they try to read the disk or memory of the vmware image, the site will go offline and they'll be completely unable to bring it back up, or read the disk, because they lack the encryption key. You would be able to let external parties process highly sensitive information without having to fear for the safety of your data.
I find the different applications possible with this technology to be both positive and negative from the government's point of view. It can partially prevent piracy (though no better than starcraft II's online only play can do so). It can just as well be used to create an online kiddy porn site that "customers" wouldn't be able to expose even if they wanted to. It could make a guaranteed 100% safe Tor network.
A TPM isn't good or evil, it's just a technology to prevent idiots from exposing their private keys. The thinking is simple : given that out of 5 major american banks 3 managed to expose their root keys for the payment system, 2 of which found fraudulent transactions (due to dishonest employees), what makes you think you can do better ?
But can we please stop the idiotic idea that some other party has access to TPM private keys ? This is not the case.
The whole point of the TPM is that it has a private key which NOBODY controls. It is in the device exclusively. There is circuitry inside the TPM to run key generation and to export the public key. There is not circuitry to export the private key at all.
Nobody has the key to this device. Not you. Not the manufacturer. Nobody.
That is why they're used at banks. Banks can have a root private key they control, and no matter how idiotic/malicious their employees, the key cannot be compromised (it can be erased by nitwits though, which sadly frequently happens, which usually causes an outage on inter-bank payment traffic until a new TPM can be shipped to them (bank TPMs are 6 kg black boxes that have additional security, mainframe compatibility and cobol and java interface libraries - see "Night of the living dead 184" for how well it works in practice). But nobody has the keys at all, which means they cannot be stolen without someone wondering why their inter-bank traffic just started getting all sorts of "permission denied" errors.
Those aren't gift economies, they're dictratorships. Obviously you don't pay for stuff in a dictatorship, but I seriously doubt that should be called gift economy. If anything, it's communism.
You are basing this on a wrong assumption. There is *nothing* in TPM that prevents you from creating or booting a custom OS. The ONLY thing TPM does is that it prevents you from lying about which OS you're running (you can refuse to say, you can give obviously wrong information, but you can't give convincing-but-wrong information).
Actually TPM allows protection in both directions. It works a bit like banks' systems. With a TPM you can secure a laptop, give it out to anyone, and you can set it up so they won't be able to break the encryption even if they know the passwords.
If you work for a company, you can give out VPN credentials to idiots that are uncopyable. If they get infected with a virus, the VPN won't come up.
I've consulted for a bank, and here's the dream : full offline money. If you have a TPM they will manage your account in your laptop (or phone, or ...) and have full offline payments. Because the TPM will only give their program access to the data, they can still prevent you from simply adding money in your own account, while allowing fully disconnected payments to occur which the bank will only find out about weeks after the fact (and so can you on other's computers of course).
In general TPM's allow fully disconnected trust relationships.
Surely such features are worth something ? Several linux companies are already using them.
All it does is simply making sure that if you tell some company you're going to take good care of their data, you have to actually do it (or delete the data, you're perfectly at liberty to do that). I mean what do you have against this ? Other than "I want to pirate stuff" (which will still be perfectly possible, just slightly more involved).