3 ms·

Do tell, then, how a remote attestation system differs from my computer at present - other than the fact that under it, I would have less control over exactly w

Do tell, then, how a remote attestation system differs from my computer at present - other than the fact that under it, I would have less control over exactly what bits exit my network in response to some query. If this were about genuine security, there would be no talk of hardware. The hardware component is there to "secure" the system against me, the owner. It may have other advertised purposes, but that is its only essential one. The only one, in fact, which could not be achieved simply by writing properly secure software (start with eschewing Microsoft and anything it has ever touched.) Any technology which seeks to enter my home encased in tamper-resistant epoxy does not have my best interests in mind. No exceptions.

tptacek 17y ago
I think you think that you don't haven any control over the contents of a TPM, because you've never read anything technical about them. In your universe, the TPM is a scheme to allow Intel to control your computer, and could never be used to allow a Linux kernel you compiled to sign your trusted known running image of a security VM. I also think you're a troll, but I'm waiting for IE7 to install on a VM, so I don't mind batting you around a bit.
- asciilifeform 17y ago
  > I think you think that you don't haven any control over the contents of a TPM You think wrongly. The user is permitted to control some, but not all of a TPM's contents. I have read the publicly available specs of a number of TPM modules. Each contains a portion designed to resist readback and/or modification by the owner. Almost every spec touts the chip's cutting edge tamper-resistance properties. I challenge you to show me a counter-example.
  - tptacek 17y ago
    They're tamper proof to the extent that they can verify their own integrity. They aren't tamper proof to the extent that they can secretly relay your Gmail password back to Portland. The core of the TPM provides a very simple method of verifying memory contents, and enough cryptography to build a remote attestation scheme --- which you could just as easily use to manage 10 of your own machines as Intel could use to secretly control the world's PCs. The comical part of this argument is that in the TPM-free world, we have basically no privacy, because our computers are all controlled by people with enough time and energy to find secret flaws in the software we run. But hey, everyone's gotta have something to get all crazy about. God bless you. With me, it's how crappy xkcd and BSG are.
    - asciilifeform 17y ago
      > They're tamper proof to the extent that they can verify their own integrity. They're tamper proof in that they are designed to physically resist decapping and electron microscopy. Don't dance around this issue, it is a key one. The phrase "tamper-resistant" has a rather specific meaning among chip vendors. I have read a substantial body of Treacherous Computing-related literature (I don't keep up with the field very diligently, I admit - my stomach is only so strong.) There is unending talk of encrypted electrical buses and display links. Would you like to tell me with a straight face that these ingenious measures are meant to protect against virii or intrusion by script kiddies? > in the TPM-free world, we have basically no privacy Speak for yourself. I suffer no intrusions into my privacy which a TPM would prevent. > our computers are all controlled by people with enough time and energy to find secret flaws in the software we run Even if this were true, it is much preferable to a world where computers are controlled by those with access (legitimate or otherwise) to a TPM master key. (Rest assured there shall always be a master key, regardless of what the spec sheets say.) I'll take traditional malware chaos over media cartel totalitarianism any day of the week.

tptacek 17y ago

I think you think that you don't haven any control over the contents of a TPM, because you've never read anything technical about them. In your universe, the TPM is a scheme to allow Intel to control your computer, and could never be used to allow a Linux kernel you compiled to sign your trusted known running image of a security VM. I also think you're a troll, but I'm waiting for IE7 to install on a VM, so I don't mind batting you around a bit.

asciilifeform 17y ago
> I think you think that you don't haven any control over the contents of a TPM You think wrongly. The user is permitted to control some, but not all of a TPM's contents. I have read the publicly available specs of a number of TPM modules. Each contains a portion designed to resist readback and/or modification by the owner. Almost every spec touts the chip's cutting edge tamper-resistance properties. I challenge you to show me a counter-example.
- tptacek 17y ago
  They're tamper proof to the extent that they can verify their own integrity. They aren't tamper proof to the extent that they can secretly relay your Gmail password back to Portland. The core of the TPM provides a very simple method of verifying memory contents, and enough cryptography to build a remote attestation scheme --- which you could just as easily use to manage 10 of your own machines as Intel could use to secretly control the world's PCs. The comical part of this argument is that in the TPM-free world, we have basically no privacy, because our computers are all controlled by people with enough time and energy to find secret flaws in the software we run. But hey, everyone's gotta have something to get all crazy about. God bless you. With me, it's how crappy xkcd and BSG are.
  - asciilifeform 17y ago
    > They're tamper proof to the extent that they can verify their own integrity. They're tamper proof in that they are designed to physically resist decapping and electron microscopy. Don't dance around this issue, it is a key one. The phrase "tamper-resistant" has a rather specific meaning among chip vendors. I have read a substantial body of Treacherous Computing-related literature (I don't keep up with the field very diligently, I admit - my stomach is only so strong.) There is unending talk of encrypted electrical buses and display links. Would you like to tell me with a straight face that these ingenious measures are meant to protect against virii or intrusion by script kiddies? > in the TPM-free world, we have basically no privacy Speak for yourself. I suffer no intrusions into my privacy which a TPM would prevent. > our computers are all controlled by people with enough time and energy to find secret flaws in the software we run Even if this were true, it is much preferable to a world where computers are controlled by those with access (legitimate or otherwise) to a TPM master key. (Rest assured there shall always be a master key, regardless of what the spec sheets say.) I'll take traditional malware chaos over media cartel totalitarianism any day of the week.