TCPA Support in Linux

kempokaraterulz writes "Linux Journal is reporting that "The Trusted Computing Platform Alliance has published open specifications for a security chip and related software interfaces.". In the latest Gentoo Newsletter they talk about a possible 'Trusted Gentoo', and possible uses for hardware level security."

Comments

Finally ready for the main stream

[kaustik]

It really makes me happy to see that Linux distributers are finally seeing the light and providing the community with things we need in an Operating System. Hopefully this will lead to other advances in the wonderful world of DRM.
sigh

Re:Finally ready for the main stream

[yason]

It really makes me happy to see that Linux distributers are finally seeing the light and providing the community with things we need in an Operating System. Hopefully this will lead to other advances in the wonderful world of DRM.

It has been my understanding that trusted computing equals not DRM automatically. Trusted computing is initially neutral technology: the barriers are built up only after the chip gets to choose a side. You can let Microsoft turn your PC into a DRM environment using TCPA's technology but that's the Microsoftish / {MP,RI,??}AA'ish approach. You can also use TCPA to turn your Linux box into a hardware-reinforced installation of your choice. If TCPA was widespread, you could for example control how the bastard big co. digitally uses, views and copies personal information when you buy something on their website.

Re:Finally ready for the main stream

[Minna Kirai]

You can also use TCPA to turn your Linux box into a hardware-reinforced installation of your choice.

If you have the technical brainpower to use TCPA + Linux to build yourself a secure hardware platform, you could also more easily build an equally secure all-software Linux platform.

The only advantage the TCPA gets from using hardware is it's a big barrier-to-entry for reverse engineers with physical access to the machine: they can't just load it up into an emulator/debugger, they also have to dissect the CPU under an electron microscope.

TCPA, at its core, is a way for you to prove to remote companies that you haven't modified the behavior of your own computer. It accomplishes this with a combination of cryptography and tamper-resistant chips.

Tee hee... published before editing was finished

[PornMaster]

From a programmer's perspective, the IBM version of the TPM (or TCPA chip) looks like Figure 1. Garrick, please crop the caption out of the figure itself.

Garrick? Garrick? McFly? McFlyyyyyyyyyy?

Do we really need it ?

[CineK]

I mean - there are a lot of hardware security modules that can be used for building trusted systems right now.
Isn't the only purpose of pushing things like TCPA locking the platform down ?

Re:Do we really need it ?

[danheskett]

A locked down platform is very useful for some things.

One thing TCPA provides that many alternatives do not is a system of sealed storage. In this scheme, an application run under the TCPA feature set can access storage that is guaranteed by hardware to be only accessible by that one application, and no others. This storage is protected by hardware encryption, and cannot be accessed directly, even by the OS. If the application itself or any component is tampered with the sealed storage is inaccessible, since the Nexus, or hardware security manager, recognizes the binary itself as the key to the sealed storage. If that binary is modified, it can no longer access the sealed storage.

Sealed storage like this is useful in a lot of ways. Combined with a strongly encrypted internet communications a highly secure messaging system could be devised where the encryption was physically end-to-end. Since TCPA provides encryption from the keyboard, to the memory, to the Nexus to the CPU and every point in between, the plain text is only exposed when it is physically being typed - it never exisits in unecrypted digital form.

Re:Do we really need it ?

[niiler]

As I really have no problems with viruses, trojans and the like, the only purpose I see to this is restricting content. I, for one, feel that the current Disney/Bono copyright law is decimating the public domain and that fair use is in danger of going extinct. An open source version might allow one to simulate trusted computing when necessary and build in a back door so that fair use can still apply. Last I recall, the Constitution mandated fair use, not the content providers. Of course, this sort of programming would probably violate the DMCA (which as we know, supercedes the Constitution).

Re:Do we really need it ?

[ultranova]

No, of course we, the users, don't need DRM. It's entire purpose is to take control of a computer away from the user and into the hands of some other entity, who can allow or disallow any given function remotely. In such a system, DRM's role is to ensure that the user cannot regain control.

This, of course, is enormously usefull for entertainment and software industries. Especially the latter can force any license terms it pleases after DRM becomes required by law (it will - the two industries combined have deep enough pockets to force the issue). Up to this point, it has been "buy once, run as long as you please"; with DRM, it becomes "buy once per reboot" - especially Microsoft would benefit from this ;(.

So no, we don't need DRM, but they do, and they have all the money and power, so we are fucked.

Re:Do we really need it ?

[bechthros]

Seems to me it's a lose-lose situation. On the one hand, until it's hacked, you have users not being able to have their machine do what they want it to do. That's obviously bad thing number one.

But number two comes a couple years down the road from widespread adoption, when some critical flaw in TCPA is found by hackers, TCPA is hacked, and innocent businesses that have come to depend on it for security are disrupted and exploited. And then we're looking around all doe-eyed, like, "but they said it was unbreakable security, they said it was trusted computing!" TCPA is just antoher level of command heirarchy, and subject to hack.

"Trusted computing" has got to be one of the most insidious marketing doublespeaks I've ever heard in my life. All "Trusted Computing" consists of is computers who don't trust me.

Re:Do we really need it ?

[Grishnakh]

And how exactly is this useful to the user? Why would I want to run an application that has its own private storage which can't be accessed by other applications or the OS? The only argument I can see here is that you don't have to worry about some spyware sending your personal data to someone. But this only makes sense if you're running an insecure OS known for frequent vulnerabilities to spyware and viruses. For the rest of us, this is a non-issue. Why do we need special hardware to make up for the shortcomings of one OS?

In reality, this hardware just opens up a big can of worms in the form of applications that don't trust the user, and prevent the user from accessing data in any way except those approved by the application maker. Why would I want to run such an application on my computer?

Re:Do we really need it ?

[novus ordo]

If that binary is modified, it can no longer access the sealed storage.

This is good news for data corruption. All your data is fscked.

Re:Do we really need it ?

[SiliconEntity]

The only argument I can see here is that you don't have to worry about some spyware sending your personal data to someone. But this only makes sense if you're running an insecure OS known for frequent vulnerabilities to spyware and viruses.

Oh, sure. Linux is perfectly secure, right? Keep on dreaming, it must be nice.

Actually, Linux security bugs are found all the time. Dan Bernstein had a bunch of undergraduates look and found 44 Linux security bugs in just a few weeks. The only reason Linux users are safe is because there are too few of them to be an attractive target. If and when Linux becomes popular for applications like banking, there will be plenty of malware stealing people's account numbers and draining accounts.

Re:Do we really need it ?

[Anonymous Coward]

Well, this would be useful in an environment where you want to make sure that unauthorized apps or computers can't access data, such as financial records or credit card numbers. Yes, an insecure OS is one reason you might worry about this, but do you really know if your OS is secure or not? With sealed storage, as long as you can be reasonably sure that only one app can access the data, then the security of the OS becomes less of an issue. Also, if the data is encrypted end-to-end, then you can worry less about someone installing a hardware or software sniffer on your computer and scooping up data.

Don't get me wrong. I'm not saying this kind of setup is always desirable, since it can work both for and against a user's interests, but there are legitimate uses for it.

Re:Do we really need it ?

[Too Much Noise]

If that binary is modified, it can no longer access the sealed storage.

This is good news for data corruption. All your data is fscked.

If it were true, it's good news for a lot of corporations, too. Update WMP due to some security bug and you won't be able to access the authorisation data for playing the songs you purchased online. OOps! time to re-buy them! and even if you use a friendly store that will give you extra free downloads for purchased songs to cover that situation, you end up with: 1. parts of the disk space being lost (I assume that if you can't read the protected area you can't delete it either) and 2. reliance on the store not closing or losing their (presumably protected, too) customers' past transaction information due to simlar TCPA glitches.

Re:Do we really need it ?

[loqi]

But this only makes sense if you're running an insecure OS known for frequent vulnerabilities to spyware and viruses.

And therefore that "unnamed OS" is the only one with security issues? Face it, there are vulnerabilities in almost every desktop system in use today.

And how exactly is this useful to the user?

I'm sorry if the idea of a keylogger doesn't freak you out, but it should. Someone with your email password can do quite a lot (think about how many "I forgot my password" systems rely only on access to your email to effectively authenticate your identity). Businesses working with very sensitive data or even just security-conscious consumers that would see their credit card info encrypted the second it leaves their fingers may see the value in something like this.

Re:Do we really need it ?

[Minna Kirai]

Oh, sure. Linux is perfectly secure, right? Keep on dreaming, it must be nice.

Oh, sure. TCPA can protect against OS bugs? Keep on dreaming, must be nice.

TCPA means that signed software can run with full permission. It only stops intentional exploits (programs specifically designed to infringe copyright), not accidental ones (buffer overflows or cross-site scripting).

To block such things, there are many well known techniques that can be applied- priviledge separation, data-tainting, external-error trapping, etc. But all of those can be implemented in sofware alone, without help from TCPA or any other hardware. Conversely, TCPA without those signficiant software changes gives zero benefit.

The only people TCPA might protect is those who put themselves at risk by running slapdash amateur software like Linux and OpenBSD, instead of staying with known quality brands like Microsoft, where security is job N!

PS. Incidently, the flaws in your argument are directly analogous to those in George W. Bush's social security plan. In both cases, to prevent a vague danger, he suggests doing 2 different activities, when really only one of them goes towards solving the difficulty at all- the other just serves his ideological agenda (and is more elaborate and expensive, to boot).

Re:Do we really need it ?

[dgatwood]

The way I read it is that you would have a file or files encrypted with a key. That key would be stored in the chip. That chip would only provide the key to an application that matched the one that added the key to the chip.

The encrypted contents of the file would be visible to any other process. It just wouldn't be particularly useful without the key.

That said, one could easily write an OS that would prevent you from deleting a protected file. One could also write a different OS and hook the disk up to an unprotected computer running that OS and delete the file from there....

Re:Do we really need it ?

[Anonym0us Cow Herd]

And how exactly is this useful to the user? Why would I want to run an application that has its own private storage which can't be accessed by other applications or the OS?

I might want only a limited set of applications accessing a certian storage area.

• P2P application
• XMMS

Then in a different secured storage area, I only want a limited set of applications accessing....

• Usenet downloader
• Pr0n Viewer

Since I can trust the software within each group, I know that no evil RIAA people will be accessing my sacred secured storage. (Of course, torture may be allowed in the US -- after all -- think of all the poor record executives.)

Imagine a trusted P2P application that will only interconnect with the same trusted application? The trust works both ways. Just like the RIAA thinks they can "trust" their software running my computer to not be of my own creation, or a tampered version of their software, I can "trust" that MY software running on the RIAA's computer is similarly my original code, not tampered with or substituted.

Re:Do we really need it ?

[Anonym0us Cow Herd]

Yes we really need it.

The trust works both ways.

The RIAA / MPAA wants to be able to "trust" that I have not tampered with their application, or substituted my own application.

Similarly, I want to be sure that I can "trust" that the RIAA / MPAA has not tampered with my, nor substituted their own P2P application.

Imagine a GPL "trusted" P2P application. You can see the source. You can compile it on a UN-trusted GCC compiler -- but only with very specific compiler options such that it produces the exact trusted binary I have certified.

The P2P app will only connect with a trusted copy of the same application. I can "trust" with the same level of trust that other copies of my P2P app are genuine non-RIAA tampered copies.

(This leaves unaswered the issue of traffic analysis. The IP addresses of the nodes that you end up connecting to. But then, this could be addressed by a swarming-routing protocol like MUTE uses.)

Here's a different argument....

Just like the bad guys can trust my computer against me to ensure closedness (i.e. can't copy / print / save an e-mail), I can similarly trust the bad guys' computers to ensure openness. For instance, I can be assured that my open source can only be processed or manipulated using tools that I trust to keep it open.

My point is that the trust can be used to our benefit. We can "trust" the bad-guys' computers to do our will.

Re:Do we really need it ?

[Anonymous Coward]

Everything you list here could be achieved as easily with software.

So why does DRM do it in hardware? Simple. Because you have no control over the hardware; you cannot create your own hardware without millions or billions of dollars to invest; with DRM, someone else (someone richer, more powerful, more wise than you) has ultimate control of your computer.

Re:Do we really need it ?

[Anonymous Coward]

Of course, torture may be allowed in the US -- after all -- think of all the poor record executives.

Please, please, I'll talk, I'll talk! Just don't play any more Britney Spears!

Re:Do we really need it ?

[EsbenMoseHansen]

I'm sorry if the idea of a keylogger doesn't freak you out, but it should.

It should, indeed. And you precious "Trusted" (by who?) computing will not help you. Current keyloggers do not change one bit of software. It would be rather difficult to design a keyboard that would be proof against recording the keystrokes.

One obvious kind of tampering that this keyboard should be able to resist would be to install a small camera and a storage device, to be picked up later.

From a consumer angle, trusted computing is only good for two things as far as I can see:

1. providing a storage facility for keys and the like, which then have 1 (one) additional layer of security.
2. preventing viral infection of the OS.

2. is, as you might know, a non-issue for all operating systems currently in use in PC, bar Windows and DOS. 1. is just plain silly: It would be dead-easy to provide the same functionality with a small network-enabled device, with less complexity=fewer bugs.

Re:Do we really need it ?

[Minna Kirai]

I might want only a limited set of applications accessing a certian storage area.

You can accomplish all those things in a 100% software implementation of priviledge separation. No special TCPA hardware is needed.

However, if you did have the special hardware, you would still need modified TCPA-aware applications and OS to make it work.

So let's consider the two paths towards reaching your goal:
A) A modified OS that restricts which of your applications are allowed to access which parts of your file system.
B) A completely new PC and peripherals that have more expensive TCPA-compliant hardware, plus everything already listed in (A)

Hopefully, you can see that the cost (in both money and complexity, which translates to opportunity for errors) of A+B is higher than the cost for A alone.

The only thing TCPA's hardware modules do that couldn't be accomplished with pure software is make reverse engineering prohibitively expensive. It's designed solely to prevent you from knowing how to fully control your own PC.

Re:Do we really need it ?

[Minna Kirai]

I can "trust" that MY software running on the RIAA's computer is similarly my original code

No you can't. The RIAA has the money and contracts to give orders to the people holding the keys with which the software was signed. You don't have that level of influence yourself.

Re:Do we really need it ?

[Minna Kirai]

The P2P app will only connect with a trusted copy of the same application.

Wrong. It will only connect to other applications approved by the same people who signed your own app. That means either a commercial software publisher, or a 3rd party Verisign-style corporation. (Yes, there will be amateur open-source hobbists releasing and signing their own software, but no representative of the RIAA will install it)

So the question becomes: Does the RIAA have the legal and financial firepower to convince the signing authority to produce signatures for some modified version of the P2p app, and without notifying you?

I really think they do.

We can "trust" the bad-guys' computers to do our will.

Why exactly would a "bad-guy" knowingly and willingly execute software that does your will? Concievably he might do so if there was no other choice, but it's difficult to imagine how that could happen. Only if a group more legally and financially powerful forced it onto them. (So maybe the federal government will write a P2P app that both music publishers and listeners will enjoy? Hmm...)

Re:Do we really need it ?

[Anonymous Coward]

That said, one could easily write an OS that would prevent you from deleting a protected file. One could also write a different OS and hook the disk up to an unprotected computer running that OS and delete the file from there....

Good luck. While it may be possible to hack around it somehow, the filesystem is encrypted and only a signed OS is allowed to speak to the chip. So your modified OS would be locked out from the filesystem, since you don't have the sertifications needed to sign an OS.

Re:Do we really need it ?

[Alsee]

The RIAA has the money and contracts to give orders to the people holding the keys with which the software was signed.

Correction: The keys with which hardware is signed. Under Trusted Computing all software is it's own signature. You do not need anyone to sign your software. It is the hardware wich reports what software is running.

You can write any software you like, but it will not be able to read any DRM file sent to or created by the RIAA's software. The RIAA's software will refuse to speak to your software locally or over the internet.

Someone with the resources to read a key out of a chip, or who control the keys with which hardware is signed, or who can issue a court order, they can create an emulated system and do absolutely anything they want. They can 0wn you.

-

Re:Do we really need it ?

[danheskett]

Current keyloggers do not change one bit of software. It would be rather difficult to design a keyboard that would be proof against recording the keystrokes.
TCPA would make it so that protected applications could not be keylogged. The hardware path between the keyboard and rest of the system is encrypted. The OS does not ever see the unencrypted key strokes.

If you physically took photos of the keyboard, that's one thing. However, software keyloggers that are in essense trojans would be defeated.

About your 1, 2 comments, you are also wrong. TCPA does nothing to prevent viruses whatsoever. The only thing it would do is prevent sealed storage from being touched by the OS/other application - like say a virus. Your secure data would stay safe from viruses, but the rest of your system could be hurt in any way it wanted.

As far as average consumers, TCPA isn't all that useful. For some systems it'd be ideal. Sometimes you can't or don't want to trust the system admin of a box/system. In those cases, TCPA is very useful.

Re:Do we really need it ?

[EsbenMoseHansen]

You would be creating a hugely complex, user-unfriendly system which would not help against the most common ways to harvest passwords (which I assume would be the keyloggers job, if we are taking software keyloggers.)

Much simpler to look in command histories for mistypes, creating fake websites, social engineering a so forth.

And not trusting the admin can be handled in software --- e.g. by sudo+friend and no root users at all. Simpler, tried and true. You would have all the protection a TCPA system would give, except simpler system=fewer bugs.

Re:Do we really need it ?

[ArbitraryConstant]

"You can accomplish all those things in a 100% software implementation of priviledge separation. No special TCPA hardware is needed."

No. Since kernel-mode arbitrary code exploits have been known to occur, you might face a situation where an attacker can read any private keys on the system (for example). TCPA would allow private keys to stay locked away from attack.

Assuming the algorithm is immune to chosen plaintext attacks. And that no other weaknesses are found. After the hardware implementation is set in stone.

In principal it could be good for other things, but I don't trust a committee to design a secure architechture on the first shot.

Re:Do we really need it ?

[danheskett]

And not trusting the admin can be handled in software --- e.g. by sudo+friend and no root users at all.
That is in no way a simpler system.

TCPA is the simplest system imaginable. The OS has nothing to do except throw a flag when a binary loads. That's it. Everything else is handled by the Nexus (security co-processor). Literally, it's the simplest possible design.

A TCPA enabled program and other programs can seemlessly co-exisit (though no co-mingle data and memory, like typically is seen now), without modification.

You can argue this all day, up and down the river, but you'll still be wrong. TCPA is good in that it insulates the software developer from worrying about bugs in the operating system or other applications or even firmware. When your data *really* needs to be secure you can't trust MS, or a linux distro or Solaris or even AIX. It's an added expense.

If you rely only on software for your heavy duty security needs you will be disappointed. A bug, a flaw in the software at any point will render the system open to attack. A simple hardware co-processor that transparently encrypts and seperates "marked" binaries and data into seperate virtual "slots" is a very simple design, with a mathematically provable end-point. An operating system with 5-10 million lines of code - like a trimmed down linux box cannot even provide that.

Re:Do we really need it ?

[EsbenMoseHansen]

TCPA is the simplest system imaginable. The OS has nothing to do except throw a flag when a binary loads. That's it. Everything else is handled by the Nexus (security co-processor). Literally, it's the simplest possible design.

And how do I know that the OS does not load a malicious program? The OS itself is untrustworthy, if that bit is the entire protection. So the security nexus would have to keep a checksum of the trusted programs or use a similar technique. Who updates those checksums?

No, the only way to do what you suggest (and what is in the articles) is for the entire application chain to be valid, forming a trusted chain all the way back: The CPU->BIOS->Bootloader->OS->drivers->(trusted) application. And that is a really complex, bug-prone system.

A TCPA enabled program and other programs can seemlessly co-exisit (though no co-mingle data and memory, like typically is seen now), without modification.

True enough, if rather irrelevant.

You can argue this all day, up and down the river, but you'll still be wrong. TCPA is good in that it insulates the software developer from worrying about bugs in the operating system or other applications or even firmware. When your data *really* needs to be secure you can't trust MS, or a linux distro or Solaris or even AIX. It's an added expense.

Even AIX? You've got to be pulling my tail.

Anyway, don't take it from me. Read some of the other comments, search the net, check your sources. You will reach the same conclusions.

If you rely only on software for your heavy duty security needs you will be disappointed. A bug, a flaw in the software at any point will render the system open to attack. A simple hardware co-processor that transparently encrypts and seperates "marked" binaries and data into seperate virtual "slots" is a very simple design, with a mathematically provable end-point. An operating system with 5-10 million lines of code - like a trimmed down linux box cannot even provide that.

But that design will yield you nothing at all. You still will need to communicate with this coprocessor, which will open your little coprocessor like a can of sardines.

If you don't believe me, write me a usecase, and I will break you design.

At best you add another layer, and a layer that is hard to upgrade when security flaws are found. Even the cryptopgraphy itself is complex enough that you will have bugs. And a simple cryptographic coprocessor (like the one on the IBM Z-series) is but a little piece of the misnamed "trusted" computing.

Re:Do we really need it ?

[Anonym0us Cow Herd]

You do not understand how this works.

The RIAA's money and power are irrelevant.

My new Dell Trusted WizBang 9000 has a TPM chip. The TPM has a PUBLIC and a PRIVATE key.

Dell has issued a certificate saying that the TPM's PUBLIC key is genuine. Now as long as I can trust that VeriSlime signed Dell's certificate, and VeriSlime's signature says that Dell is authorized to then issue these TPM certificates, and then I can verify that Dell's certificate for my TPM public key is correct, I can know that the TPM module in my computer is genuine. (Not some "emulated" TPM running in a Virtual PC.)

The RIAA can twist Dell's arm into signing something, but that does not matter.

The PRIVATE key of my TPM is a secret inside the TPM itself. No human ever has or ever will know the PRIVATE key inside the TPM. Not only a human, but nothing outside of the TPM has ever kwown my TPM's private key. The TPM seecretly generated it at manufacture tiime. That is what makes the private key inside the TPM such a secret. The TPM is designed to take the private key with it to the grave if you make any attempt to try to recover the secret out of the TPM chip.

The TPM can now sign something. Anyone who wants it has the public key to my TPM. In fact, my TPM willingly gives out its PUBLIC key. Because Dell has signed my TPM's public key, anyone who cares can verify that anything my TPM chip signs is trusted. (NOTE: No internet connection is necessary to verify trust!)

Once you understand how this works, you can see that anyone can trust their application on the RIAA's computer, just as much as the RIAA can trust their code on my computer.

TPM is a mechanism to make sure that I can be assured that the RIAA has not compromised my application.

Re:Do we really need it ?

[Anonym0us Cow Herd]

I can sign my own app with my own public/private keys.

My app can then follow the chain of trust. My app asks the OS for a checksum of the app. My app asks the OS to provide proof that the OS is not tampered with. The OS asks the bloat loader to prove that it is uncompromised and that it approves of the OS. The bloat loader asks the BIOS to prove that it is trusted and that it approved of the bloat loader. The BIOS can get proof from the TPM that the BIOS is uncompromised.

Now I have proof from the TPM, which I can trust, that the BIOS is okay.

I have proof from the BIOS that the bloat loader is okay.

I have proof from the bloat loader that the OS is trusted.

I have proof from the OS that it trusts that my app is the one that I, that is ME, MYSELF, that I signed.

Now my app could be designed to only connect to something that can prove that it is also signed by me.

Maybe my app is just the core of a P2P. The plumbing. Maybe it has loadable search, play, UI etc. modules. You can customize the UI. Use different widget sets (GNOME, KDE, etc.). Add new wizbang features. But you can't control the underlying plumbing of the P2P network. It is trusted.

End result: trusted p2p plumbing. Customizable features on top. But from a plug in module, you can't tell where a file came from or where it is going to. Using swarming routing like MUTE makes it unlikely that even traffic analysis of TCP/IP will tell you anything.

Re:Do we really need it ?

[Minna Kirai]

After the hardware implementation is set in stone.

If I waited for the software implementation to be set in stone, it'd be totally perfect too.

Still, it's true that hardware can help protect against that risk- but not TCPA-style hardware! A special smartcard reader which allows a PC to sign files without that CPU reading the actual private key could be good. (If the machine is owned, then the attacker will be able to use your key as long as he stays connected, but can't take it with him for later use. So he'll have to start the money transfers immediately, rather than taking his time about it)

Too bad that's not really what TCPA is doing (portions of TCPA could be resuable for that, but they're not taking a direct route to get there). A USB-interface personal cryptocard coprocessor could be built best as a separate product, without need for support inside the CPU case (except for at the software application level). That'd be handily portable too... which is why RSA or a competitor will surely market it long before TCPA gets off the ground.

It's a flagarante violation of the Unix-style development rule: Design small parts to work together, doing one thing and doing it well. TCPA does many things, which is a public-relations trick: because there are occasional good bits that someone can point out, the audience can be deluded into thinking the whole thing is good, or that the rest of it is a prerequisite for those aspects.

Re:Do we really need it ?

[Minna Kirai]

You do not understand how this works.

Because there's no "this" there. Your idea of being able to trust that the RIAA is running the program you want them to is meaningless, because the RIAA will never agree to run your programs.

It's the money and power again... can you really think of any scenario where the admin of an RIAA computer (by which I really mean a computer owned by a corporate music store) would willingly decide to execute any program that didn't come from the OS vendor, or at least another member of the ETC? The only programs they'll run are those from someone they control, either financially, contractually, or legally.

(ETC = Entertainment/Technology Complex, and includes both RIAA/MPAA, and the vendors of hardware/software used to display their products)

For the specific example that might really happen, the RIAA could commission a P2p app to distribute songs purchased from them. Because they're assured that only that specific app is being run, they are less worried about infringement- and as a bonus, the customers don't just pay for the file, but take care of the majority of bandwidth to distribute it too. (The program, for example, might not permit you to play the file if you don't spend an hour providing it to others customers)

Your mutual trust idea makes some sense if a few friends want to play a FPS game without allowing anyone to run wireframe wallhacks, but doesn't make sense in terms of corporationindividual relationships.

but nothing outside of the TPM has ever kwown my TPM's private key. The TPM seecretly generated it at manufacture tiime.

See, you're ignoring that the RIAA has the money and legal power to induce the hardware vendor to record those numbers before shipment, so they can retroactively revote your applications' security under a court order.

Re:Do we really need it ?

[Minna Kirai]

but nothing outside of the TPM has ever kwown my TPM's private key. The TPM seecretly generated it at manufacture tiime.

The RIAA/MPAA will use comments like that as proof that PC makers must keep a list of all their customers' TPM keys in escrow, for the inevitable day when they're needed to dismantle terrorist piracy rings like yours.

Re:Do we really need it ?

[ArbitraryConstant]

Unfortunately I don't think a USB wigit is sufficient for things like playing DRMed HDTV content. There's no way to get it back into the computer without exposing it to piracy, and having it display directly from the wigit is awkward at best.

Still, I'd rather see a deeply flawed implemenetation get railroaded into place now than have the industry wait till they get it right. Particularly a hardware implementation. With buying cycles being what they are, a flawed hardware implementation could set the cause of DRM back for a good chunk of a decade.

As for the USB cryptography widgit, I think that's a really good idea. You can take it with you when you leave, and it can have a button on it, so even if someone takes over your computer you can just refuse to push the button and nothing will happen.

Remember this as prior art if someone tries to patent the idea!

Re:Do we really need it ?

[Anonym0us Cow Herd]

The PC Makers will probably already keep a list of the PUBLIC keys of the TPM.

Anybody who wants my TPM's public key can have it. I'll put it in my slashdot sig.

Nobody or nothing ever has or ever can know the PRIVATE key inside the TPM. It is a secret. Even the chip foundary that made the TPM chip cannot possibly know what the private key is that is inside the TPM.

This is an important concept to understand.

Re:Do we really need it ?

[Anonym0us Cow Herd]

you're ignoring that the RIAA has the money and legal power to induce the hardware vendor to record those numbers before shipment, so they can retroactively revote your applications' security under a court order.

You're missing that the hardware vendor cannot possibly know what the TPM's PRIVATE key is. Even the chip foundary that made the TPM chip does not (and cannot) know the TPM's secret PRIVATE key.


can you really think of any scenario where the admin of an RIAA computer (by which I really mean a computer owned by a corporate music store) would willingly decide to execute any program that didn't come from the OS vendor,

My hypothetical P2P app does not come from the OS vendor. I don't really care if the RIAA runs my P2P app or not. If they do, I know that my P2P app won't run unless it can trust the RIAA's OS, bloat loader, BIOS and hardware.

My point is that the RIAA cannot create their own P2P app that connects to my P2P network, as they do today. My (hypothetical) P2P app would only interconnect with other apps that can prove that they are also trusted.

Re:Tee hee... published before editing was finishe

[theGreater]

Poor Garrick:

From the Fine Article:

Garrick, insert this table and the next exactly where indicated. They do not have titles, and the last line of each table should span both columns.

-theGreater.

what is it good for?

[Anonymous Coward]

The only benefits I can see is increased security for encrypted communication or hard drive encryption. I am really trying to think hard of any other beneficial applications but can't come up with anything.

Re:what is it good for?

[essreenim]

Could be useful for electronic voting.

Re:what is it good for?

[nkh]

Unless TCPA involves some kind of new, enhanced cryptographic algorithm unknown to man (given to Bill Gates by aliens from another planet), there will be no increased security. I have ssh and it's still good for the moment.

We just have to wait for someone to crack the security through obscurity part and hope that my standard Linux/BSD (without all this DRM/TCPA shit) will still boot on future hardware.

Re:what is it good for?

[mOdQuArK!]

You mean to make sure that the creators of the black boxes can make the vote turn out _exactly_ the way they want it to, without worrying about anyone else being able to figure out what they did?

Re:what is it good for?

[Hobbex]

Could be useful for electronic voting.

Think again, the manufacturers of TCPA have admitted that they are not secure against hardware manipulation. It is likely that we will begin to see TCPA mod-chips hitting the market soon after TCPA takes hold (that is right people TCPA means you will have mod-chip your PC). The whole point is to make sure that circumvention of DRM requires more effort than the masses will spend - requiring an expensive and illegal modchip to be attained is considered enough.

Re:what is it good for?

[ergo98]

Could be good for war?

Re:what is it good for?

[vadim_t]

Well, it could be useful for a seriously locked down server.

Imagine that you're an admin at some big company, with a hundred Linux boxes. You have this stuff on every of those boxes, and a computer for administration somewhere safe. When you install software you first check it, then sign it, then push updates to your servers.

If somebody gets in, they'll have things quite difficult. Anything unsigned simply won't run at all. Rootkit modules, exploits, etc, will all simply not be able to run at all. This would take out a quite big part of the exploits an attacker could use. Remote ones would hopefully avoided by NX.

This wouldn't protect against things like races, but it certainly could help quite a lot.

The situation above is something I wouldn't have any problems with. If an admin wants to have an uber-locked down system where anything not signed by his key that's only present in a computer with no network connection in a secure room with an armored door doesn't run at all, then sure, why not. I'm fairly sure this can mostly be accomplished without hardware support at all, though.

Now, it's when software publishers want to make it impossible for me to control my computer when I have problems with it. But if the user has full control of it, I think it could come quite handy in some cases.

Re:what is it good for?

[Seigen]

As far as encryption goes all you really need for more security is a very good source of random numbers since the encryption code can easily be in software, but generating good random numbers in general can't. I think the major chip makers already have hardware random number generators, although I've never studied how good they are. Of course implementing anything in hardware can make it faster..

As to the trusted part of trusted computing. It was always my understanding that it meant someone other than the user of the computer could trust the computer to work in a certain way. I.E. they could make a trusted dvd reader only interact with a trusted motherboard, and only allow direct on chip decoding to send it to the overlay section of a trusted video card to prevent one from ripping the dvd. Is such a scheme unbreakable? The answer to that is maybe, if they are really good in their design. You can do similarly for other forms of trusted applications. For instance xbox applications are digitally signed so only approved ones play. Does that mean mod chips for pc's someday? Maybe.

Re:what is it good for?

[ichimunki]

Encrypted communication and hard drive encryption are already possible. Including "trusted computing" in Linux allows Linux users to run closed-source binaries (either applications or libraries) that interact with encrypted files obtained from third parties, where such binaries use this "trusted" nonsense to restrict the use of these files in some way. This allows Linux users to trade their freedom in return for continued access to digital media without having to stop using Linux at the same time. While we can have discussions about whether this is "right" or not, it seems largely inevitable that certain operating systems and computing platforms will get this and that certain digital media providers will take advantage of it. It makes using Linux less of an all-or-nothing decision.

Re:what is it good for?

[wasabii]

It cannot be accomplished without hardware support.

If it's implemented in software, then somebody will just hack that software.

The idea is that every bit executed by the CPU must be signed by a third party, enforced by hardware, with NO WAY TO BE MODIFIED.

Re:what is it good for?

[Jennifer E. Elaan]

Actually, it *does* include many enhanced cryptography features that cannot be designed entirely in software.

While I have a problem with the uses of this platform that Microsoft no doubt intends, TCPA can be quite useful for making secure systems based on open standards.

One part of these modules is the ability to send keys to the hardware module in a way that cannot be read back out (but with encryption performed using this write-only data). This allows public-key encryption with the private key stored in a very secure way.

Re:what is it good for?

[vadim_t]

Why?

I don't think there's any fundamental problem that makes it impossible to make a VM or emulator that's impossible to break through. If the attacker entered by ssh guessing a password or whatever, and assuming the kernel doesn't have any defects that make bypassing the protection possible, it should work just as well.

The scenario I presented in the grandparent should work just fine with software, IMHO. Now, the chip is certainly useful if you want to take away control from the owner of the machine.

Re:what is it good for?

[harlows_monkeys]

The only benefits I can see is increased security for encrypted communication or hard drive encryption. I am really trying to think hard of any other beneficial applications but can't come up with anything

Games.

There are many things, for instance, in MMORPGs and FPSs that have to be done now on the server that could be done better (in terms of performance and in terms of providing a better game experience) on the client, but can't be done there because it would allow cheating.

Re:what is it good for?

[Qzukk]

When you install software you first check it, then sign it, then push updates to your servers.

In the end, it depends on who gets to sign the software, and how this software is distributed once signed. In our corner of the court, we have the admin signing software for 100 boxes (does he have to sign each separately? Can you sign software for every box out there at once? If its not a specific-to-that-machine signature, how do you keep the attacker for signing software too?) for the purpose of protecting the servers from software you don't want to run.

In the other corner of the court, it appears that we have big business interests who want to have all software signed, who would charge hundreds to sign software for other authors (verisign, et al will certainly be in the business), MPAA and RIAA will be wanting to make sure signed software obeys their rules (and will probably charge for this too), all to make sure your computers are protected from software they don't want you to run.

Things like this IBM article help make the first scenario a reality, and I'm grateful for it. Now, who wants to be the first to be sued by Microsoft for some TCPA submarine patent that nobody knows about?

Re:what is it good for?

[harrkev]

From the way that I understand it, a VM would not help. The chip would have its own crypto key embedded in it (or maybe more than one).

Let me put it this way... All good encryption algorithms are open -- you can examine the spec and the implementation as much as you want. But you still need the key in order to get the scrambled data out.

This means that you can examine this chip as much as you want. As long as it does not give up its keys, it should still be secure. Then your computer can be trusted to not do nasty things -- like watch a movie on an unapproved OS.

Re:what is it good for?

[essreenim]

No, not if the hardwar is open source. Quite the opposite. I would like to see a completely transpar

Re:what is it good for?

[essreenim]

No, not if the hardwar is open source. Quite the opposite. I would like to see a completely transpar

Talk about making things turn out _exactly_..thats not what I typed..

No, not if the hardware is open source. Quite the opposite. I would like to see a completely open, transparent system - both h/w and s/w

Re:what is it good for?

[HiThere]

Also, in this scenario, what happens when the master computer's cpu gets fried? The key is on the chip, and never came off, so you CAN'T have a backup.

I can see ways to resolve this. I haven't heard anything that indicates they are in process.

Who gets to sign? What are the restrictions? What's the recovery procedure in case of accident? This is one of several reasons I don't use an encrypted file system, but with TCG (nee TCPA) it get's taken to a whole new level, with appearantly less control given to the local admin.

Re:what is it good for?

[Anonymous Coward]

The definition of "top secret" information is that if it were compromised it is likey that people will be killed. Let's say YOU were one of those people at risk. Would YOU but you Linux system on the Internet if it contained such data? In other words wwould you bet your life on Linux's security and the boxes Admin?

Right now the rules are simple, you don't connect systems containing such data onto public networks. In fact in many cases you cannot even connect the ground wire from rht A/C power plug to the building common ground system least some guy figures a way to read the CRT from the readio frequency noise.

People who have such data would like for it to be transportable. Heck simply burn it to CDROM and toss it is a handbag or copy it to a flash drive or send it as an e-mail attachment. Today doing any of that could land a person in jail

Outside of the millitary there is civil liability. Insted of getting killed because the bad guys stole your battle plans you let something get out and loose ten million dollors.

There are MANY cases were data is so valuable that it is simply not easly transportable

wouldn't it be great if the data on my laptop were useless to anyone but me. Now if the laptop were lost or stolen I'd only be out $1,800 OK I don't have any valuable data but lots of peole do.

Re:what is it good for?

[cayenne8]

"Outside of the millitary there is civil liability. Insted of getting killed because the bad guys stole your battle plans you let something get out and loose ten million dollors."

LOSE...not loose. You can lose money...you can turn a dog loose. Two different words...two entirely different meanings.

Re:what is it good for?

[vadim_t]

Well, I'd hope that this kind of setup would work like PGP.

The admin's box is the only one that has the private key used for signing. The servers have the public key. When you sign an executable, you only need to do it once, any of the 100 boxes can verify it, and the attacker can't generate a signature without getting the private key.

Of course, this scenario doesn't need hardware at all. It's all doable in software, and in fact I've heard one of the BSDs (forgot which) has either a feature or a patch that does something similar.

The chip's almost certainly for the big business, since I don't see how it's needed for any scenario where the owner remains in full control. Note that I say the owner, since there are perfectly legitimate reasons for locking down a machine so that the user can't mess with it. Public terminals and such come to mind.

But for those cases I think it's quite safe to assume that the owner took measures to close any obvious holes, such as booting from floopy/CD and passing arguments to the kernel. About the only reason I can come up with is a computer with very sensitive data, where it's vital that data can't be recovered if it's stolen. But that doesn't seem like such a common scenario.

Re:what is it good for?

[dgatwood]

I don't see the advantage, really. Hardware key storage is a fundamentally flawed concept.

• You can't readily map hardware into multiple address spaces simultaneously. Thus, unless there's a special CPU instruction to send and receive data and the chip is tied to the CPU, the key will likely end up going through an OS routine to send data to the device through the driver layer, where it can be intercepted.
• Even with a special CPU instruction, the data can be trivially intercepted through snooping the bus it is on.
• Even with a special CPU instruction and an encrypted bus between the CPU and the chip, the data can still be sniffed in the CPU through simple JTAG.
• Even if you prevent that data from being visible by a JTAG debugger, anything running in kernel space (ring 0 for Windows folks) can still perform an nm on the application, figure out its data structures, and walk them using simple memory mapping and obtain the key in that way.

Basically, for this to provide any additional security, the following conditions must be true:

• The OS must have the ability to prevent any TCPA task from being attached to with a debugger.
• The OS must not allow any kernel code to map memory from the task (which means the task cannot do I/O) or else every driver must be signed and audited by the OS manufacturer.
• The OS must not allow firewire DMA. (Think "external mod chip".)
• The OS essentially must not allow any unsigned code to execute, since even unsigned user space code could exhaustively modify itself until its signature matches the signature of another application.

The biggest problem, though, is that once someone figures out the hashing algorithm used to determine the application's authenticity, it's only a matter of time before it becomes trivial to fool it, but since it is being done in hardware, it can't be easily upgraded to keep ahead of encryption research.

I'm having a hard time seeing this as a win.... What am I missing?

Re:what is it good for?

[bitspotter]

What you would do in that case, is have each machine's (hidden) tcpa key sign the organization's public key, and seal it as an authorized software signing key (which can only be done by the physical owner). Henceforth, any software signed by your master key is authorized to run on the machine. When software updates come out, you can either vet and sign them yourself.

Or, if they come from a vendor you trust that signs the updates, you can have each machine sign and seal that vendor's key as trusted. I would imagine the sign & seal process would require physical access to the machine by the machine's "owner".

The big problem with drm is when machines come already "owned" by the vendor, locked down in a state where the customer who bought the hardware cannot designate the signing keys for the software it should trust. It is totally disingenuous to think you "own" an iPod (for example) when it comes pre-"owned" from Apple, locked in a state where only Apple-signed software can run on it - especially when the TCPA/TCG architectures allows for the "owner" to be the person who actually buys and uses the device.

Re:what is it good for?

[BlueStrat]

Since I haven't seen it posted so far, here's a link to a piece on TCPA/Trusted Computing.

http://www.cl.cam.ac.uk/~rja14/tcpa-faq.html

It has some answers to the questions/issues you address, and is quite a scary read for anyone that values individual freedom. Hope that helps.

Strat

Re:what is it good for?

[Anonymous Coward]

It's good for Microsoft to put wacky keys into their software so that the hardware will only work with their software. Reverse engineering violates DMCA and users get sued into oblivion. I don't care a fig newton about "Oh, but it's open". Bullshit! The whole thing is to give Microsoft (yet another) licence to print money. A "Signed BIOS", will lead to an increased use of Linux BIOS (http://www.linuxbios.org). The project is close to abandonware, but the software is still available. Crazy BIOS's that get to restrictive will get yanked (remember how the XBOX started running Linux?). It's a lot easier to play pop-the-bios (and remember to replace it with a gold-plated machine socket before pushing in the eeprom).

Re:what is it good for?

[mOdQuArK!]

I believe that the point of this TCPA system is that the owner of the key can make sure that you will be able to access only what they want to let you access.

Give me enough rope...

[DigiShaman]

to hang myself.

Instruction: How to restrict your Linux box from yourself.

Linus Torvalds himself has blessed DRM

[Xpilot]

Linus himself said DRM is ok, as long as it's used in the interests of the user. This is a good thing, think about it; EvilCorp(tm) wants to use DRM to cripple computers, but the PR guy will say "it's for the user". Of course their intent is nothing of the sort, but the Linux folks are the only ones who will actually implement something that *is* in the interest of the user. Then EvilCorp won't be able to lobby making Linux illegal, since Linux also uses DRM which does what EvilCorp claims it's doing "for the users". Well, hopefully.

Re:Linus Torvalds himself has blessed DRM

[scorp1us]

You touched on something there that I want to bring out further.

Linux can show what user-centric trusted computing can/should do. Microfoft et. al. will be showing what Big Business trusted computing wants/can do.

Eventually there will be those that will ask why it has to work against them so much when running Billy Bob's OS, and then they'll realize that their PC is not their PC, but the industry's PC.

Re:Linus Torvalds himself has blessed DRM

[EpsCylonB]

Then EvilCorp won't be able to lobby making Linux illegal

Sillier things have happened.

Re:Linus Torvalds himself has blessed DRM

[Minna Kirai]

Linus himself said DRM is ok, as long as it's used in the interests of the user.

Linus is not a lawyer. More importantly, he's not even a free software or open source evangelist. Unlike RMS or ESR, he doesn't even hang out with lawyers or devote serious thought to legal matters.

Since DRM is a combined legal-technical area, it falls outside Linus's expertise, and his opinion carries little weight. (From a practical standpoint, TCPA is incompatible with the Linux philosophy of open-source modifications)

Re:Linus Torvalds himself has blessed DRM

[GoofyBoy]

This gets modded +5 Insightful? There wasn't a better post to mod up?

It reads like an overactive child who just getting too excited about his toys.

"Maybe Linux will get rockets and laser and missles and guns with super-bullets blessed by the Pope. And then they could go up to M$ campus and then destroy Steve Ballmer and all his evil minions. And then Bill Gates will regret all the evil things that he did. Well, hopefully."

Re:Linus Torvalds himself has blessed DRM

[FirstTimeCaller]

Linus is not a lawyer. ... Unlike RMS or ESR, he doesn't even hang out with lawyers or devote serious thought to legal matters.

I knew there was a reason why I liked the guy!

Re:Linus Torvalds himself has blessed DRM

[alext]

Being even less well qualified than Linus Torvalds to contribute to this discussion, I nevertheless venture to suggest that the "Linux philosophy" entails, almost by definition, the endorsement of a particular configuration of source files by Torvalds and others.

As such, open source is not at odds with the ostensible goals of TCPA, any more than it is with his ownership of the Linux trademark.

Re:Linus Torvalds himself has blessed DRM

[Leo McGarry]

EvilCorp(tm) wants to use DRM to cripple computers, but the PR guy will say "it's for the user". Of course their intent is nothing of the sort

No? Here's the very, very simple math:

Due to the current epidemic of piracy, content creators refuse to offer their wares in electronic form unless strong encryption is used to protect their creations. If strong encryption isn't included, content will simply not be available. Therefore, in order to make content available to users, products must include strong encryption.

Yeah. It's for the users. It's to give users what they want: content.

Of course, this really pisses off users whose preferred solution to the problem is "I stealz me some stuffz!"

Re:Linus Torvalds himself has blessed DRM

[Leo McGarry]

Ah, but if he'd come out strongly opposed to rights management, you would have crowed about how great it is that he's sticking up for the little guy. As it is, since he's in favor of rights management, you say his opinion doesn't count.

Way to demonstrate the principle of selective dismissal.

Re:Linus Torvalds himself has blessed DRM

[yuri benjamin]

Time to re-read this essay. Stallman's disturbing future scenario is getting scarily closer with every new technical or legal development in this field.

Even if you don't normally agree with Stallman, read the essay anyway, and the notes that follow.

Re:Linus Torvalds himself has blessed DRM

[Anonymous Coward]

Of course, this really pisses off users whose preferred solution to the problem is "I stealz me some stuffz!"

Nice try, but you left out all of the important parts.

DRM pisses off citizens who understand that the legitimate goal of Constitutional copyright is to gain benefits for the public -- rather than to maximize monopoly rent at any cost to the public. It pisses off customers who get damaged goods. It pisses off citizens who see that its backers want to force us to live in an electronic police state like sheep. The Declaration of Independence, Constitution, and Bill of Rights are not documents for, or by, sheep.

Re:Linus Torvalds himself has blessed DRM

[emoticon]

Linus himself said DRM is ok, as long as it's used in the interests of the user.

"Trust and you will be trusted," says the liar to the fool.

Re:Linus Torvalds himself has blessed DRM

[TeraCo]

It pisses off customers who get damaged goods. It pisses off citizens who see that its backers want to force us to live in an electronic police state like sheep.

Well, if the option is quality 'damaged' content, vs crappy 'pristine' content. I'll take the good damaged content instead.

Life is too short to play Tux Racer instead of HL2.

Re:Aged 83 pope john paul dies!

[Anonymous Coward]

Not yet, sorry. I just asked Stephen King...

If you can't beat 'em, join 'em.

[Kickasso]

Better yet, lead 'em. It would be ridiculously funny if Trusted $FREENIX were released before Trusted Windows or Trusted MacOS.

Re:If you can't beat 'em, join 'em.

[SpottedKuh]

Trusted Windows

Wait, wait...you lost me on that one.

Re:If you can't beat 'em, join 'em.

[Anne Thwacks]

Trusted Windows

Its probably a typo. They can't seriously expect anyone to Trust Windows.

Re:If you can't beat 'em, join 'em.

[Reziac]

If you can't beat 'em, beat 'em up. -- Fred Roggin

Re:If you can't beat 'em, join 'em.

[randallpowell]

You're lost? Windows has always been trusted. I trust it will automaticly install virual and malware through ActiveX or by running attachments through Outlook. It's trusted to have more security holes than the Whore of Babylon and cost more as well.

Lacking One Thing

[SpottedKuh]

Though the specifications detailed in the article are definately a Good Thing, they lack (at least as far as I could tell) any way of preventing unauthorized physical access to the chip.

Physical access to machines is always a big issue in security, and one that is often overlooked. And while it's probably not a big deal for your home machine, consider large companies whose machines could conceivably be targetting for a physical attack to recover the keys directly from the TPM (Trusted Platform Module).

Stajano's "Ubiquitous Computing" book has excellent coverage of the rationale, issues, and complexity of attempting to prevent physical access to chips and devices which store sensitive information. It's an easy read, and well worth it: http://www-lce.eng.cam.ac.uk/~fms27/secubicomp/ind ex.html

Re:Lacking One Thing

[cpeikert]

The TPM is supposedly very tamper-resistant, not just a piece of solid-state memory holding the keys. This should make physical attacks very expensive and labor-intensive.

Re:Lacking One Thing

[SiliconEntity]

Though the specifications detailed in the article are definately a Good Thing, they lack (at least as far as I could tell) any way of preventing unauthorized physical access to the chip.

Slashdot ran a story a few weeks ago about a new set of chips with built-in TPM features. These chips have the Trusted Computing capabilities built into the CPU. It will make it much more difficult to attack them physically since the TPM is not a separate module but is integrated into the whole system. Probably this is how all TC systems will be built eventually.

Re:Lacking One Thing

[duffbeer703]

Physically securing very valuable things is easy.

Put the computer in a secure vault, and put guys with guns at the door and perimeter.

Re:Lacking One Thing

[Anonymous Coward]

But if you see Tom Cruise and Ving Rhames walking into your building, worry.

Re:Lacking One Thing

[Anonymous Coward]

It's not that easy.. These guys have some of the most valuable stuff in the world (nuclear bomb designs and enough plutonium&tritium to make them); and they have plenty of guys with guns at the door and perimiter. Though (as the article suggests), they still have difficulty securing stuff.

"Guys" is the weak link in your solution. You need something more reliable than your typical grunt who goes into a career that involves playing with guns.

Re:Lacking One Thing

[duffbeer703]

Real security protects money. When was the last time you heard of someone robbing a Federal Reserve Bank with billions of dollars of currency within?

Hardware Security

[quadra23]

This is indeed good news! Security that is solely-based on software is far easier to compromise than hardware-based (provided that the hardware can't be tampered with by malicious software). Far better to have the security co-ordinated between both. I'd be interested to see how widely accepted this open specification will be.

Re:Hardware Security

[Anonymous Coward]

Yeah right... until someone find a bug in the chip... ...and encryption + USA = NSA

The only reason they want to implement it in hardware is that they don't want anyone to find bugz in it... and backdoors...

oh yeah, I'm paranoid... ...and I'm getting paid for it... :-)

Re:Hardware Security

[Minna Kirai]

Security that is solely-based on software is far easier to compromise than hardware-based

Wrong. Computer security is computer security, whether or not it's implemented in hardware or software. In both cases, you need a smart computer engineer and cryptologist to do the work.

Hardware implementations have the (small) advantage that hardware is more expensive to build, so the vendors will exercise a little more care before making a release.

Software has multiple advantages: it's design has fewer constraints from real-world physics, it's easier/quicker to study in an artificial environment for automated testing, and by far the most important, it can be replaced worldwide in a matter of hours if a security flaw is discovered.

Conversely, hardware that's found to be flawed can only be replaced expensively and after at least months of effort. (In fact, the best way to "patch" a hardware flaw is with a software fix, meaning you're back to relying on software security)

So weighing the advantages against each other, software comes out as the better place to build-in security. Either way, a mistake by the designing engineer puts you at risk, but with only one of them can the risk be affordably fixed.

Good point...

[Anonymous Coward]

Maybe that's because the linux devs are actually linux hackers / lovers like the rest of us (well, those of us that do... not saying everyone is, and if you aren't, that's ok!), and they don't want to make using linux painful.

I think it's cool. In fact, this may be one of the things that helps drive linux to a popular position in the desktop realm. Just need to get some big companies behind it, like, say Intel, HP or IBM... oh wait, they all support linux. Maybe this isn't so improbable.

I'm kind of excited to see what will happen.

Big Bro

[Anonymous Coward]

It would be ironic that Linux users used the cryptography system to keep information _out_ of the hands of EvilCorp, instead of the other way around.

Re:Big Bro

[Yaa 101]

This is not as strange as you imply...

There are many small businesses looking for ways to keep people from reading stuf, people like competition and secret services.
While not asuming it gives any vast security, it gives another layer of it to some people.
Better would be seperate networks for internal systems and internet systems, the fact is that most simply cannot afford that.

As long as the specs and software are open source and as long as one can decide themselfs what OS to install on a machine there is room for the chip.
It all really depends on Law, hence politicians, lobbieng, etc... what rights we keep as individuals, this is not different from yesterday.

Anyway, all tech, all that exists has good and bad sides...

Re:Big Bro

[Yaa 101]

I forgot to say:

This one has mainly bad sides...

Wouldn't that be...

[GbrDead]

Treacherous Gentoo?

Re:Wouldn't that be...

[Anonymous Coward]

"Treacherous Gentoo?"

I just love the alternative interpretations of these abbreviations:

TC (Trustworthy Computing) = Treacherous Computing
DRM (Digital Rights Management = Digital Restrictions Management

TCPA is a DRM smokescreen

[Hobbex]

It has been said a million times, yet apparently it bairs repeating. The "security" aspects of TCPA are redundant, unnecessary, and at best useful but could be made a lot better if the chip was designed for security rather than DRM. The whole system really exists only for one purpose: as a trojan horse to implement something called "remote attestation" in PCs.

What is remote attestation? Basically, it means that the TCPA chip, which you cannot control, can read what operating system you have loaded, and send a reponse proving that you are running a certain operating system to others on the Internet. The purpose of this, of course, is so that the operating system can be verified not to have it's DRM functions cracked, so that the RIAA and MPAA can send you data and make sure that they get to decide what you do with it.

The people pushing TCPA will claim that it is not for DRM, but that is a smokescreen and only a smokescreen. While TCPA does not do DRM itself, it is the enabling component that is needed so that software can implement DRM without being circumventable.

What does this mean for a "trusted Linux"? It means that while it is completely possible to have a Linux system working with TCPA, once you change anything in the system, the TCPA chip will notice you are running a modified system, and nolonger let your data. So while the software may nominally remain under the GPL, it will be the death of the free software model, because users who wish to tinker with their systems will be locked off the Internet (Cisco is already talking about systems to have ISPs demand remote attestation when TCPA is in place). TCPA and Linux can be combined in theory, but only in theory - in reality they cannot ever coexist.

Those who do not believe me (or those who are inclined to believe the MS shills who will respond saying that I am wrong), should read EFFs analysis of TCPA where they give a simple way that the chip could be changed to allow all uses except remote attestation intended to force people to use certain operating systems and enforce DRM over the user. It has been completely ignored by the manufacturers of TCPA.

Re:TCPA is a DRM smokescreen

[Yartrebo]

I can already think of a workaround.

Use an x86 emulator and two copies of Linux, one that uses TCPA and one that doesn't. Run the x86 emulator on the unrestricted Linux copy, and use it to run the TCPA copy under emulation. The x86 emulator would just have some security 'flaws' when it comes to storing keys or it might do stuff like forgetting to apply the encryption. It would still report as a valid DRM chip, and would be able to provide keys and authentication on demand.

Re:TCPA is a DRM smokescreen

[Minna Kirai]

The whole system really exists only for one purpose: as a trojan horse to implement something called "remote attestation" in PCs.

Trusted Computing is genuinely pursuing security as long as you understand that it's the end-user who is evil and needs to be held back. It protects entertainment companies from members of the public, making it the exact opposite of "user friendly".

Re:TCPA is a DRM smokescreen

[EpsCylonB]

....

The purpose of this, of course, is so that the operating system can be verified not to have it's DRM functions cracked, so that the RIAA and MPAA can send you data and make sure that they get to decide what you do with it. ....

(Cisco is already talking about systems to have ISPs demand remote attestation when TCPA is in place). TCPA and Linux can be combined in theory, but only in theory - in reality they cannot ever coexist.

Lets be clear about what it is that you are saying here, even Linux OS's that are TCPA enabled are unlikely to be approved for use with media services provided bye the RIAA and MPAA. This is probably true, but saying that TCPA and linux can never co-exist is, I think, an exageration, TCPA by itself will not kill Linux.

Re:TCPA is a DRM smokescreen

[MetaPhyzx]

Sounds like it almost ideal, when dealing with workstations in financial businesses/organizations that handle sensitive data. I'm sure we've all run across user X who didn't see the harm in installing "that little weather application" (I know this doesn't happen with linux specifically, but on XP/2000 this could save a lot of headache support wise) It can help ensure uniformity. From that side of things, I can immediately see the bonus.

On the flip side, I'm not interested in it on linux, just because it won't work. And it sounds like I'd have to depend on the maintainers of a distro for digitally signed binaries for anything I MAY run. I really don't like that idea.

Re:TCPA is a DRM smokescreen

[MetaPhyzx]

I should have said ANY organization handling sensitve data.

Re:TCPA is a DRM smokescreen

[SiliconEntity]

Those who do not believe me (or those who are inclined to believe the MS shills who will respond saying that I am wrong), should read EFFs analysis of TCPA where they give a simple way that the chip could be changed to allow all uses except remote attestation intended to force people to use certain operating systems and enforce DRM over the user.

And see this rebuttal to the EFF report.

Further see this blog entry by the same author on good uses of Trusted Computing all of which rely on the supposedly evil Remote Attestation feature. EFF's proposal to allow people to override their systems' attestations would ensure that none of these applications would be possible.

The truth is that TC along with Remote Attestation is a new feature set for your computer which allows new ways for people to cooperate online. Some people oppose this because they don't believe that others should be allowed to cooperate in ways they don't approve of. They don't want you to be able to credibly commit to obeying certain rules in processing data. But they have no right to interfere in your private decision making processes.

Re:TCPA is a DRM smokescreen

[Anonymous Coward]

That's not possible; TCG (formerly TCPA) specifically has a means of detecting the presence of emulation. The TCG TPM chips contain signing keys which can't be removed from the chips. The emulator therefore will not know these keys, and signatures produced by the emulator will not be trusted by remote parties trying to verify your system configuration; alternatively, they will conclude that you're running an emulator.

Re:TCPA is a DRM smokescreen

[Anonymous Coward]

..and IBM doesn't have anything against it...

Re:TCPA is a DRM smokescreen

[SiliconEntity]

I can already think of a workaround. Use an x86 emulator and two copies of Linux...

Emulation won't work to defeat DRM. The TPM generates an on-chip secret key at manufacture time which never leaves the chip. The manufacturer issues a certificate on the corresponding public key which certifies that this is a legit TPM key. The TPM then is able to issue attestations about the software configuration based on what is called a "secure boot" sequence, in which the TPM creates a fingerprint of the software configuration.

An emulated boot looks different from a real boot to the TPM, so the reported software configuration will be different. Your emulator won't be able to forge a TPM attestation report because it doesn't have a certified TPM key from the manufacturer. These keys stay on chip and there is no way (without spending hundreds of thousands of dollars on exotic chip-stripping equipment) to pull a TPM key off the chip.

Much cheaper to aim a video camera at the screen and put a microphone in front of the speakers.

Re:TCPA is a DRM smokescreen

[poofyhairguy82]

The truth is that TC along with Remote Attestation is a new feature set for your computer which allows new ways for people to cooperate online. Some people oppose this because they don't believe that others should be allowed to cooperate in ways they don't approve of. They don't want you to be able to credibly commit to obeying certain rules in processing data. But they have no right to interfere in your private decision making processes.

No...thats not it. I don't "oppose people having choice" or some crap like that. I oppose this becoming an industry standard that REMOVES my option of not having it. Once every computer has this, it won't be long before ISPs can say "we only want trusted machines on our network." It is the begining of the end for the computer cowboy; Trusting Computing is the first step to civilize the wild west called the internet. Sure for most users (those with tons of spyware on their computer, or with computers that their kids have loaded with enough illegal IP that a lawsuit might come any day in the mail) trusted computing might be seen as a good thing. But for someone who has taken these little multipurpose machines called computers and have used them as they wish despite lacking the blessings of the company's involved (such as me playing DVDs on my Linux box), Trusted Computing is a sign that the companies that have created this wild west are sick of not being obeyed.

This might help some company keep its local network in shape, which is great for them and a reason for its eventual success. But with that comes a future where my computer follows the law despite my wishes (or I lose Internet access), which would make it a very boring hunk of silicon and circutry indeed.

Re:TCPA is a DRM smokescreen

[Truekaiser]

ibm is a company who's sole intrest is making money. they only support linux cause they see some money involved. as for this, they will support it cause they can make even more money with vendor lock in. you can run linux....though only the ones signed with ibm's keys.

Re:TCPA is a DRM smokescreen

[Yartrebo]

Maybe it can't forge the report, but it isn't that hard to trick the program running into thinking it is a secure report.

If I had to hack the emulator to make things work, I would do the following:

1 - Flag the register or memory location that the fake report is stored in. All moves, adds, copies, etc of the register are flagged too, and if it is destroyed it is unflagged.

2 - Whenever a register (or memory location, if that is how the protocol works) is set (via any operation), check to see if it is the public key (which you have access to). If so, flag it. Treat it like step 1 for keeping track of where it and any derivatives of it have gone.

3 - When a comparison is done between a public key flagged register and a fake report flagged register, take note of it.

4 - If a conditional jump follows before another comparison instruction, do the opposite of what is instructed. If it is a jne instruction, treat it like a je instruction.

Leave some programmability in the emulator so that this can be overridden in case the app developer decides to confuse the emulator, which is possible but must be done on purpose. In that case, disassembly and traditional cracking skills can be used.

Remember, the program is running on your emulator, which you have 100% control over. The program cannot assume that any instruction will be properly interpreted. It is just a matter of how hard it is to trick. Access to the network, memory, and registers can also all be controlled transparently.

Re:TCPA is a DRM smokescreen

[SiliconEntity]

Sure for most users (those with tons of spyware on their computer, or with computers that their kids have loaded with enough illegal IP that a lawsuit might come any day in the mail) trusted computing might be seen as a good thing.

I don't believe your scenario, but look at what you're saying. You want to inflict harm on the interests of most users in favor of your selfish desires. You are afraid that if this technology becomes popular, presumably because it satisfies the needs and desires of the broad mass of users, you will be inconvenienced.

Think about a horse rider who hated the new horseless carriages and foresaw that if they became popular, he would be forced off the road. He would have been right, but does that mean that we should go back to the time of the horse?

Like it or not, we should make social policy that considers the welfare of the large numbers of people over the welfare of the selfish few. In this case, all that is necessary is to stand back and let people decide for themselves what technologies they want to use. All those VHS purchasers made it hard on the die hard Beta enthusiasts, but we didn't try to forbid VHS from being used.

In the same way, TC should live or die based on how well it satisfies the needs of users. Don't use it if you don't like it, but don't try to stop other people from using it just because you are afraid it may become popular.

Re:TCPA is a DRM smokescreen

[Hobbex]

Further see this blog entry by the same author on good uses of Trusted Computing all of which rely on the supposedly evil Remote Attestation feature.

So I shall go through these one by one:

Online Elections

Um, no. Any trusted client system relies on keeping the client from tinkering with the TCP chip, and while this might be sufficiently complicated to keep people from copying your precious MP3s, it is not nearly secure enough for an election. I will be dead before I take part in an election where any sucker with a mod-chip can rig it!

Online Gambling

See the last comment. The financial incentive for rip-off casinos to hack the TC chips would be huge - so huge that I almost smile when I think about this use since it makes me more sure than ever that there will be modchips available. A much better system to secure against this is to play at reputable casinos and are tested by indpendents doing statistical analysis.

Financial Transactions

You've got to be kidding! I'm suppose to trust my bank to take care of my money if they cannot write a banking interface that doesn't rely on controling the client to keep it safe? You may claim Internet banking is in it's infancy, but I have been using it for five years and I cannot think of a single security threat against it caused by me being able to use Mozilla rather than IE to access the bank.

VPNs

The use attestation described here would not be the slightest affected by EFFs "owner override" suggestion.

P2P

P2P systems are not exactly suffering at the moment... Bittorrent has showed that tit-for-tat type trust systems work very well to keep clients in line, without sacrificing interopterability between different clients and the ability for people to develop clients on their own.

Remailers

This is pure speculation. I would like to see the remailer user who would think that trusting that TCPA will hide the message from the owner of the computer (jesus! that people cannot see what is wrong with this!) is better than a secure system.

Corporate DRM

See the first two objections. But we are getting closer to the true purpose!

Online Shopping Privacy

Because certified sites have never broken their statements, and certified operating systems like Windows have never been cracked. The idea that the problem with online shopping privacy is that you cannot trust that store is running Windows (or, yes, the single certified version of Linux that will be three years behind current development) is silly.

There are a couple of entries that are at least somewhat true:

Multi-player Games
Selling CPU Cycles

but they hardly make a good case for implementing an Internet covering lock-out system. I very much doubt anyone will bother with the latter, and the former is best done on games consoles anyways.

Conclusion

The author of that blog is full of shit. And the worst thing is that he is obviously intelligent and informed, so he most likely knows that he is completely full of shit. I can't for the life of me understand why people like that wish to sell out our digital future to one of exclusion and centralized control, but hey, it is a free Internet (so far...)

Re:TCPA is a DRM smokescreen

[Hobbex]

Like it or not, we should make social policy that considers the welfare of the large numbers of people over the welfare of the selfish few.

No, the central pillars of our civilization are liberty and self-determinism, not utilitarianism as you would have it. Utilitarianism has, on the other hand, been the justification for every great evil visited upon the world the last century and more. // oskar

Re:TCPA is a DRM smokescreen

[SiliconEntity]

Online Elections

Um, no. Any trusted client system relies on keeping the client from tinkering with the TCP chip, and while this might be sufficiently complicated to keep people from copying your precious MP3s, it is not nearly secure enough for an election. I will be dead before I take part in an election where any sucker with a mod-chip can rig it!

Online Gambling

See the last comment. The financial incentive for rip-off casinos to hack the TC chips would be huge - so huge that I almost smile when I think about this use since it makes me more sure than ever that there will be modchips available. A much better system to secure against this is to play at reputable casinos and are tested by indpendents doing statistical analysis.

So what you're basically saying here is that TC is not secure enough for these applications. Two points: first, future versions will have increased security, for example by being built into the CPU. And second, we run these applications already, without any security at all. How can you possibly hold elections, for example, to such an unrealistically high standard, when elections are going on all the time that have much lower security? People vote online from overseas without any protection at all. And casinos likewise operate without any oversight or auditing whatsoever. Millions of people play these games every day. Adding security can only benefit them. Your rhetorical tactic of demanding perfect security is an obvious smokescreen designed to discourage approaches which improve the security of the system.

Financial Transactions

You've got to be kidding! I'm suppose to trust my bank to take care of my money if they cannot write a banking interface that doesn't rely on controling the client to keep it safe? You may claim Internet banking is in it's infancy, but I have been using it for five years and I cannot think of a single security threat against it caused by me being able to use Mozilla rather than IE to access the bank.

You're the one who's got to be kidding! Have you not heard of the many new forms of malware which are going after banking account numbers and infiltrating themselves into secure banking transactions? TC can stop these cold via sealed storage and remote attestation. Again, you are arguing that we should deny users access to these technologies purely for political reasons because you don't like the technology.

I won't go through the rest of your "analysis" because it's the same kind of bullshit. You are using any excuse you can think of to disparage a technology that can add security and improve reliability for millions of online users. It's obvious that you have an ulterior motive yourself. Any other technology which had all these obvious benefits would be welcomed.

My guess is that you are worried that TC will make it harder for you to pirate your favorate songs and movies. Well, if it makes you feel any better, that's probably not the case. Anything that can be observed can be recorded, TC or not. So why not focus on what you really care about rather than making up unreasonable objections that throw up a smoke screen of FUD around a technology with enormous potential for good?

Re:TCPA is a DRM smokescreen

[Minna Kirai]

Use an x86 emulator and two copies of Linux,

That technique would work fine if you had an emulator. However, by the principles of cryptography, it would take thousands of years to develope such an emulator unless you already had access to the secret keys used to sign binaries. (In which case you could use the hardware directly and wouldn't need to emulate)

Re:TCPA is a DRM smokescreen

[Minna Kirai]

Um, no. Any trusted client system relies on keeping the client from tinkering with the TCP chip, and while this might be sufficiently complicated to keep people from copying your precious MP3s, it is not nearly secure enough for an election

Wrong. The fact is, almost anything is secure enough for elections- because they're not very important.

Ask yourself how much money it would take to convince you to skip voting in one election? $50? $500? Probably not to much if you're rational, because the chance of your vote making a difference is so tiny.

But even today, people trust web servers to transfer far greater amounts of money. (I've shifted thousands already this morning). So rationally, if you don't trust a PC for voting, you shouldn't trust it for shopping or banking either- but millions of people already do.

PS. This isn't to say that voting and shopping are identical- in particular, many voting systems have an obsession with withholding paper trails from the end users, which removes an important mechanism to detect fraud.

Re:TCPA is a DRM smokescreen

[JohnFluxx]

You still don't understand quite how it works.
Read up on it please.

Re:TCPA is a DRM smokescreen

[aaronl]

I think you've missed it. TCPA will not make elections or banking any more secure. If you can't do it securely without trusting the client, then you can't do it securely. If I sit at the PC, or compromise it in the right way, I can get data off the TCPA enable machine. Just Like Now.

I don't want someone else being able to force their agenda on my equipment. I can modify my car, I can receive whatever I want with radio equipment, and I can do whatever I want to the computer. Some of the things I can do may be illegal as per government. But the point is that I should be able to use my property as I want, repurcussions aside, as they happen after the fact.

Re:TCPA is a DRM smokescreen

[poofyhairguy82]

I don't believe your scenario, but look at what you're saying. You want to inflict harm on the interests of most users in favor of your selfish desires. You are afraid that if this technology becomes popular, presumably because it satisfies the needs and desires of the broad mass of users, you will be inconvenienced.

I wish to inflict harm on no one. I have ill will toward no computer user- they all add to the demand of computers which makes my box cheaper.

What I dislike is the fact that currently many computer users have not taken the time to learn how to use their computer or spent the money on a computer that does not require as much of a learning curve (such as Apple). I can understand why someone would need a computer but not have the time to lock it down- thats why nerds like me exist to do it for them. If the doorlock broke in my car I wouldn't fix it myself!

Yet these people (through their lazyness) have accepted that their broken computer isn't their fault- its the fault of computers. (Even though my Ubuntu box works without such pain because I spent the time to learn how to use it. ) Therefore these people are willing to accept changes which MIGHT make the situation better for them (Trusted Computing) but surely will make the situation worse for me.

Re:TCPA is a DRM smokescreen

[Yartrebo]

I just read what's at www.notpca.org and you're right, that thing is extremely comprehensive.

I was thinking it was just a centralized and secure key storage mechanism.

Nope, it verifies that all hardware is compliant, and requires internet access to work (because it dials into some central server to get permission for each of your HW components).

It's also 100x as nasty as I had envisioned, and leaves you with no security and without even knowing what your own system is doing. I guess I see now why everyone is so upset about this. It is horrible.

Re:TCPA is a DRM smokescreen

[Minna Kirai]

TCPA by itself will not kill Linux.

Right. "TCPA will kill Linux" is overbroad and not quite right. But "TCPA will kill desktop Linux" is true, because for all practical purposes, if it can't play DVDs or music files, a desktop environment is never going to maintain popularity. (Nevermind that someday individual web pages will be protected with DRM also)

Re:TCPA is a DRM smokescreen

[Hobbex]

The tenacity of your attempts to replace logic with rhetoric would be impressive if it wasn't so braindead.

And casinos likewise operate without any oversight or auditing whatsoever. Millions of people play these games every day. Adding security can only benefit them.

No, adding a sense of false security does not make things better. People who play on online casinos today do not expect there to be software controling that the game is fair: to the extent they care they go by reputation and testing just like I suggested. TCP adds nothing - absolutely nothing - to make this more secure. The same thing goes for voting.

You're the one who's got to be kidding! Have you not heard of the many new forms of malware which are going after banking account numbers and infiltrating themselves into secure banking transactions? TC can stop these cold via sealed storage and remote attestation. Again, you are arguing that we should deny users access to these technologies purely for political reasons because you don't like the technology.

If the TCPA application of your bank were intended to stop malware, then it would have no problem with the EFF's proposed owner override. So once again with the lies!

I won't go through the rest of your "analysis" because it's the same kind of bullshit.

LOL. "I won't even begin to counter your arguments that the world is round because it is such bullshit."

My guess is that you are worried that TC will make it harder for you to pirate your favorate songs and movies.

No, no, try harder: what I actually care about is abudcting and sexually abusing small children. And strangling puppies. And helping the turrists!

TCPA is a technology designed from the ground up for exclusion. The fundamental question of the next century, as with the previous ones, is whether we wish to build and open society or a closed one, and TCPA is ultimate tool for those who wish to close our networks. The goal of TCPA is facilitate the handing over of control of our communication devices to others, so that our computers can decide what we cannot and cannot do with them, can and cannot run with them (if we still wish to access our data and the Internet), and ultimately dictate the parameters of all networked communication. Anyone who accepts TCPA accepts that he should live in digital prison, that his doors should be locked from the outside, that a priori restraint should be placed on his ability talk to others, and that not only the Internet, but all computing and all our data should be placed in the hands of a centralized few.

You, sir, disgust me.

Re:TCPA is a DRM smokescreen

[hal9000(jr)]

And you have this so completly wrong in so many ways. First of all, it is not the TCPA, that organiztaion has long since been renamce to the TCG and you can find the site here.

Next, if th TPM is designed according to the TCG specs, you have to manually enable it in the BIOS. Once that happens, then you have to manually allow TSS to interact with it before it will do so. You can choose not to, but then you can't use that software.

The TPM does NOT stop software from running. It will NOT stop open source software from running. and it will NOT stop you from using alternative operating systems.

The FUD you are spreading is already available in cuurrent software platforms and does't require the TPM. Why not try actually *gasp* reading the damn specs before you speak.

sheesh.

Re:TCPA is a DRM smokescreen

[yuri benjamin]

Even worse, I can see this breaking the end-to-end nature of the internet where anyone can be a content publisher.

It's been said that "Freedom of the press is guaranteed only to those who own one" and that "The Internet is the biggest threat to any regime wishing to control its population".

This is not good for the general public. Instead of active participants in creating discussion, ideas, rants etc, we will become mindless consumers of commercial content. Please don't let this happen.

Re:TCPA is a DRM smokescreen

[Minna Kirai]

The tenacity of your attempts to replace logic with rhetoric would be impressive if it wasn't so braindead.

Thinking back 2 years or so, I vaguely remember SiliconEntity pasting in the same kinds of screeds to Slashdot comments on the original introduction of TCPA. (Notice the knee-jerk mini-contradictions to conventional TCPA-wisdim) It really appears she has some personal stake in the success... wish we knew what it was! Alas, her comments/info don't reveal anything.

Re:TCPA is a DRM smokescreen

[Anonym0us Cow Herd]

Internet access is not a necessary component. Even with no internet access, here is why an emulated TPM cannot work.

Dell issues a certificate (stored in your TPM) that says that your TPM's private key is certified by Dell.

You can't fake this certificate anymore than you can fake someone's web certificate for SSL so that your server can reliably claim to be, say, Microsoft.

Your emulator can have an "emulated" TPM with a private key and public key. But you can't get Dell (or any other manufacturer) to sign your emulated TPM public key.

Now you could make your emulated TPM use the same public/private key pair as your actual hardware uses. Dell has signed your hardware's public key. But you can't get the private key out of the dang chip. In fact, niether could Dell nor anyone else either! That is why the private key in your TPM is a true secret. No human ever has known or ever will know the private key that is inside your TPM chip. In fact, nothing outside of the TPM has or ever will know the secret. The TPM may even destroy the secret if you try to get at it. All you've got is the TPM's public key, and a certificate from Dell saying it is a genuine TPM public key.

Anyone who trusts, say, VeriSlime to have issued Dell a certificate, and then who also trusts that Dell would ONLY sign a genuine TPM public key, can then trust that anything that the TPM has signed was in fact signed by a genuine TPM -- not an emulated one.


Without any Internet access at all, just knowing a few root certificate issuers such as VeriSlime, I can verify that your system is trusted. Just inspect the certificate from Dell to make sure Dell signed the TPM pub key, and that Dell's public key is signed by VeriSlime. I can now trust anything that the TPM signs.

Re:TCPA is a DRM smokescreen

[Alsee]

Trusted Computing will "kill Linux" because Trusted Computing defeats the GPL itself. Even if there is a Trusted Linux and even if it CAN play the new Trusted media files and CAN view the new Trusted websites and CAN run the new Trusted software, and even if it is approved by your ISP for internet access, it is impossible to modify that version of Linux. If you try to change a single line then all of those things stop working. It is no longer Trusted Linux.

-

Talking of Gentoo...

[bcmm]

I am having problems with my system clock under Gentoo.
It keeps saying it is the Second of February, when I know it must be the First of April...

Re:Talking of Gentoo...

[Anonymous Coward]

Obviously the moderators arent very bright, and thought you were being serious. Offtopic my ass.

Re:Talking of Gentoo...

[Anonymous Coward]

# rm -rf /

Re:Talking of Gentoo...

[Anonymous Coward]

Meow! ^.^

UUoC, my friend.

Trusted Linux is ILLEGAL

[Minna Kirai]

It's very simple:
1. Linux is distributed under the GPL (and other licenses).
2. To comply with the GPL, end-users must be able to acquire the source code (which means everything they need to reproduce the binary executble, with or without modifications).
3. If you don't comply with the GPL, you are committing copyright infringement, a federal offense.

But from the other direction:
4. Trusted computing means that all binaries are signed with a secret key.
5. The Trusted CPU will not execute binaries that weren't signed with that key.
6. In this way, it is impossible for end-users to create modified binaries to add/remove features from the software.

The GPL is too much in conflict with Trusted Computing to ever allow them to work correctly together. To obey the GPL, end-users must have access to everything needed to rebuild working binaries- which includes the secret key. But for Trusted Computing to work, it must be impossible for end-users to get the key- otherwise there's no point.

So, Linux or Trusted Computing. Choose one, because you can't have both.

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

Aside from the fact that this post is completely nonsensical, it should be noted that the GPL comes with no warranties of operation... meaning that if you get all the source code and then it doesn't compile... so what.
Beyond that, key material is not source code, or if it were then it would literally be impossible to distribute any kind of GPL software that interacts with encryption because by your twisted logic you'd have to distribute everyone's private key to everyone else. So are you going to start us off by posting all your private keys and passwords to this thread?

Re:Trusted Linux is ILLEGAL

[pe1chl]

Trusted software means the binary has been signed and the running binary has not been tampered with.

This does not mean you cannot have the source or you cannot modify it. Only, the modified version will not be trusted. That is what trusted code is all about.

So there is no conflict between GPL and trusted code. There is a conflict between modified code and trust, but that is the purpose of the entire concept.

Of course you can generate your own secret key, publish the public key, sign your binaries, and everyone can verify that they are genuine binaries from you. Then, anyone who trusts you, can trust your binaries just as they trust binaries signed by Microsoft.

Re:Trusted Linux is ILLEGAL

[einhverfr]

What if Gentoo distributes the key as part of their distro?

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

So, Linux or Trusted Computing. Choose one, because you can't have both.

Really? Then what's this?

TrouSerS - An open-source TCG Software Stack implementation, created and released by IBM.

Re:Trusted Linux is ILLEGAL

[aardvarkjoe]

Frankly, if the GPL doesn't allow me to sign a binary, then the GPL is broken. However, I rather doubt that's the case -- much open-source software is already signed by the creator / distributor, so you know that the binary you got was actually made by him. Because hardware DRM relies on exactly the same idea -- guaranteeing that you got your binaries from a particular source -- from a practical standpoint, there's no difference between the two scenarios.

Re:Trusted Linux is ILLEGAL

[nomellames]

From your post, I belive you don't understand what trusted computing is, or what the TCG specifications imply. Trusted Computing is based in the assumption that there is a Core Root of Trust. This CRT is trusted, and should be verifiable (not the current state, but maybe in the future we will have an open source BIOS). This CRT will measure the next entity (bootlader, whatever) and will hash the reult into a repository (the Trusted Platform Module). Then the bootloader will do the same with the OS, and so on. Of course, this is an over simplification, but there is no signatures here. Later, a program wil want to attest the software you are running, and will ask for this integrity measuraments. Also note that this (attestation, measuraments) is only a tiny part of the TCg specifications I dont see any trouble with this and linux.

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

But if jow blow can access your computer, I would assume he has access to your private key as well as you would store it most likely on your computer.

Re:Trusted Linux is ILLEGAL

[Qzukk]

The Trusted CPU will not execute binaries that weren't signed with that key.

wrong. They'll still execute, they just won't be trusted.

Trust itself is a feature at the OS level! Do you think the BIOS knows whether some data you read off the disk is an application? Does the CPU know the difference between a current application and the one that just executed just because of a context switch (which happens all the time just during timesharing between all the different applications you already have running)?

An OS could be written so that everything you try to run is verified through whatever TCPA hardware and then deny you the right to run untrusted software, but I can pretty much assure you that Linux wouldn't bother with this and if it did that someone would fork it. Microsoft probably wouldn't even bother with this (outside of the corporate empire where they could sell the admins keys to encrypt all of the software they allow on the computer to prevent users from installing other stuff), unless they really think that their monopoly is so strong it'd survive being the only company making windows software anymore, or give away the signing keys for free.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

Frankly, if the GPL doesn't allow me to sign a binary, then the GPL is broken

Conversely, if the GPL allows me to sign binaries on a system where unsigned binaries don't run, then the GPL is broken, because it's got a loophole allowing GPL'd works to be effectively seized by corporate programmers with no compensation.

much open-source software is already signed by the creator / distributor, so you know that the binary you got was actually made by him.

Sure, you can sign binaries. But if you give that binary to someone, and he later demands the source code from you, you'd better include the private key along with it.

Otherwise, once he recompiles and finds that his new binary doesn't match the one you supplied, you are in breach of contract and could be arrested for criminal copyright infringement (that is, in a bizarre alternate dimension where the police enthusiastically pursue such crimes).

Some people might argue that the binary itself isn't modified by the signing, and that the "sign" is just a tiny file distributed alongside (basically a hash of the binary that's been signed, rather than the whole thing). Hashes are technically partial copies of the original work, but they such a small proportion of the source material that they are usually considered to fall under Fair Use.

However! Fair Use has 4 factors to consider, and size is just one of them. Economic impact on the original author is another. And since keys used in TCPA operate to so completely subvert the intent of the GPL, they are indeed working strongly counter to the original software's author.

But let's step away from the legalities and look at it from a practical standpoint: What is the GPL meant to do? It's there to ensure that the users of software are able to modify the software, to stay in control of their own computing lives. TCPA is trying to create exactly the opposite effect: the central goal of TCPA is that a DRM-controlled video/audio file you play on the computer screen cannot be ripped by some other program and stored in a non-copy-protected file.

In short, it is trying to prevent the addition of certain key features to software. It wants to prevent software progress. And that is completely against the spirit (not to mention letter) of licenses like the GPL.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

They'll still execute, they just won't be trusted.

Fine. They'll "execute", but since they're not trusted, not all of the features will work. Meaning that not all of the binary is "executing"... so I guess I was right after all.

but I can pretty much assure you that Linux wouldn't bother with this and if it did that someone would fork it

How would they fork it? You can't fork if it won't run, and it won't run if you don't have the key. The only way to fork would be if you could use the GPL to coerce the signers to give you that

Do people not understand how TCPA is trying to kill Linux? Trusted CPUs will be introduced. Then a decade later, 100% of new and fast CPUs will be trusted. Newspapers, magazines, TV shows, audio, and software will all be DRMed so that only Trusted OSes can run them.

Commercially-produced Linuxes will be able to remove traditional commands like "cp" and "dd" to qualify for trusted status, so Linux as software won't die. What will be killed is the open atmosphere where anyone can modify the software and then try it out without killing 90% of the other functions.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

dont see any trouble with this and linux.

The trouble is not between TCG and Linux, but TCG and GPL (and Linux is one of many GPL programs).

Later, a program wil want to attest the software you are running,

The whole idea of that "attesting to the software you are running" is restricting who can modified the software. In particular, TCG wants to forbid amateurs and especially end-users from modifying the software. That is completely against the open-development spirit of Linux and all GPL projects.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

Then, anyone who trusts you, can trust your binaries just as they trust binaries signed by Microsoft.

The "anyone" who matters in this case is hardware vendors like Intel, AMD, Apple, NVidia, and Hauppage. Obviously, Fortune 500 corps aren't going to trust a hobbyist like me.

You paint a false picture by implying that it's end-users who need to trust me... that I can put my modified software up on the web, convince them to trust me (by exchanging money or contracts or whatever means), and then they will be able to run my software.

No! That's not how it works! End-users cannot decide who they want to trust. Only the MPAA and RIAA decide which hardware companies they want to trust, and then those hardware vendors decide which software vendors they want to trust (mainly Microsoft and Apple), and then those are the only ones who can sign software to make it run (with the full feature set enabled by being "Trusted")

It's really ironic that yesterday I was taking flak for arguing that anime fansubs hurt legal sales, because fansubbing is something that TCPA is meant to kill (along with all other handling of copyrighted work, both infringing and Fair Use)

Re:Trusted Linux is ILLEGAL

[ichimunki]

But if jow blow can access your computer, I would assume he has access to your private key as well as you would store it most likely on your computer.

Probably not. More likely burned on CD-R or on a USB keychain or something. Even so, if the key is on my system I've also likely password protected it. In fact I know I have. :)

Re:Trusted Linux is ILLEGAL

[nsayer]

Sure, you can sign binaries. But if you give that binary to someone, and he later demands the source code from you, you'd better include the private key along with it.

No. My private key is not required to create a trusted executable. Presumably the owner of the machine, if he chooses to compile the program from source, will be able to sign the resulting binary with his own private key. After all, surely the owner of the machine will be able to trust the output of his own (trusted) compiler. If he chooses to distribute his work, then the community can make their own decisions on whether he is or is not a trustworthy source of binaries, and can trust, or not, his signed binaries (which he will, of course, distribute with the source, per the GPL).

We're not talking about the Xbox mentality here - where there is only one key for signing apps and it is owned by a platform monopolist. We're talking about being assured as to who was responsible for creating a binary, and the user being in control of vesting trust in those sources, or not, as he sees fit.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

Even so, if the key is on my system I've also likely password protected it. In fact I know I have. :)

If I have access to your computer, I can put a keylogger on it, and have access to whatever passwords (and even removable media) protect your key.

0wnage = 0wnage.

Re:Trusted Linux is ILLEGAL

[SiliconEntity]

4. Trusted computing means that all binaries are signed with a secret key.
5. The Trusted CPU will not execute binaries that weren't signed with that key.
6. In this way, it is impossible for end-users to create modified binaries to add/remove features from the software.

This is total garbage. Where did you get this nonsense?

TC does not require binaries to be signed with a key. TC will not refuse to execute unsigned binaries. And end users can do whatever they want.

Now for the facts, in case you're interested. TC implements a secure boot. This allows the TPM chip to store a hash or fingerprint of your software configuration: the BIOS, the boot loader, the OS, and if desired, the applications that are running.

The TPM and OS can basically do two things with this information. They can implement "sealed storage" which means that a program can lock its encrypted data to the current software configuration. This means that if you boot a different OS, or if the program gets modified (either of which might happen due to virus infection), the fingerprint changes and the data will no longer be available. Likewise if another program tries to access the first program's sealed data, it won't be able to get access to it.

The second feature of the TPM is "remote attestation". This allows a program to request the TPM to issue a cryptographically signed statement about what the current software fingerprint is that is running. This signed attestation cannot be forged because the TPM generated an on-chip key at manufacture time, and the manufacturer issued a certificate on that key which the chip can use to prove to anyone that it is a legitimate TPM.

Remote attestation allows network applications to determine what software configuration the peers are running, and, if they choose, to disallow participation by software which is not running a specified set of configurations. This is the closest you will come to the idea that users can't change their own software. If they want to run a program which relies on this feature, and that program doesn't accommodate the changes the user wants to make, they would be shut out. But in practice, open source programs will probably be flexible in this regard as they will want to have as many people as possible participate. There are a number of technical measures which can be adopted to allow for considerable user flexibility.

But certainly none of this would violate the GPL or any other legal prohibitions. Everything is entirely voluntary, and Trusted Computing does not prevent you from doing what you want with your computer. You don't even have to turn it on if you don't want to!

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

Sorry but your keylogger is unsigned and won't run. And if you have physical access to my system, well your completely boned if someone of slikk has access to your system regardless of how secure the software is.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

Presumably the owner of the machine, if he chooses to compile the program from source, will be able to sign the resulting binary with his own private key.

Completely wrong. The owners of machines don't get the keys needed to sign things for their own hardware. Only the builders of the hardware have those keys, and they are contractually obligated by agreements to the MPAA and RIAA not to divulge those keys to anyone (except employees in the course of their work).

If the owners of the hardware were going to be the ones having the keys needed to run on that hardware, then I wouldn't have any problem with it. (But in that case, industry wouldn't promote it, because it doesn't serve their greed)

We're not talking about the Xbox mentality here - where there is only one key for signing apps and it is owned by a platform monopolist.

If you really believe that, then you've fallen for the obfuscation (or misinformation, or FUD, or whatever synonym for "lying" is in vogue today) of the the Entertainment Technology Complex.

The intent for the future is that PCs as general-purpose computers are a dying breed. The X-Box / cable-box / cellphone mentality is the way forward. Microsoft intends that the market will split into $500 end-user machines and $50,000 developer boxes. Where the high cost for the dev systems comes not from the hw or sw costs (although those are higher than they are today, because economy-of-scale has been eliminated), but from the licenses, including your personal "Software Authorship Licenses". To get that license means promising never to implement commands like "cp" and "printscreen" that users might infringe copyright with.

That barrier-to-entry will kill the open source/free software movements, and cement Microsoft's place as the planet's only real OS developer.

Re:Trusted Linux is ILLEGAL

[SiliconEntity]

Conversely, if the GPL allows me to sign binaries on a system where unsigned binaries don't run, then the GPL is broken

Trusted Computing doesn't work this way, never has and never will! This is a lie which has been promulgated by people who should know better. TC does allow systems to report their software configuration in a verifiable and unforgeable manner. This can allow servers to refuse service to clients which are not running specified software configurations. This is how DRM would be implemented using TC.

But there is no such prohibition as you describe, being unable to run unsigned binaries. Anyone can run anything they want. All they can't do is lie to someone else about what they are running. That's enough for DRM and many good uses of TC.

Please, stop spreading falsehoods and learn more about how the technology actually works. See my other comments in this topic where I have described it in more detail.

Re:Trusted Linux is ILLEGAL

[quanticle]

Ah, but this is where the hardware component comes in...

If you plug in a new keyboard (or, for that matter, any new hardware) the system should recognize the TCPA chip on the new component and ask for some sort of verification (e.g. secure USB key) before it allows the new hardware to function.

Though I really don't know how this would work if you got a new processor, RAM, motherboard, or some other "core" component...

Re:Trusted Linux is ILLEGAL

[HiThere]

So what this boils down to is "Who is allowed to sign the key?".

I've heard several different opinions here, from only the hardware (chip) manuufacterers to anyone who cares to. This makes me quite leery of the entire idea, and opposed to it's implementation. I do understand that the chip will have a (hidden) key built into it, but that's only a part of the system, and the rest of it is both vauge and threatening.

Yes, I can see good uses for the system. I also can see that whether they are feasible or not depends crucially on just how the system is implemented. And the large corporations haven't given me very much reason to trust them at all, much less in something this major.

Re:Trusted Linux is ILLEGAL

[nsayer]

Completely wrong. The owners of machines don't get the keys needed to sign things for their own hardware.

If you're right, then everything you said in the grandparent is correct.

However, I do not believe that what you're talking about has anything to do with what the article is about at all.

Note further that the article has nothing at all to do with Microsoft.

In this instance, I believe you're the FUD spreader.

Re:Trusted Linux is ILLEGAL

[ichimunki]

Depends. For every measure there is a countermeasure. No such thing as perfect security and I didn't mean to imply that there was. On the flip side, just because an attack is theoretically possible doesn't mean it's feasible.

Re:Trusted Linux is ILLEGAL

[pe1chl]

You have misunderstood the situation, as has been explained to you in other responses. There is no problem with end-users not wanting to trust you.

Of course it may be that the MPAA or RIAA does not trust your system and their members do not want to send you protected content, risking that you decrypt it and publish the plain unencrypted version. And they are probably right about that.
But that does not mean anything for a trust relation between you and your customers or the users of your open-source program.

See, you are free to ignore whatever measures MPAA and RIAA impose on you. It is your own decision to buy from companies who inherently do not trust you. It is only a matter of time before society realizes that the free world has been transformed into a world where a few megacorporations decide what you are allowed to think. Precisely what the leaders of the free world have been accusing other nations of.

But when that happens, consumers have the power to shut down that system. There is no need to consume things like music, movies, television programmes, etc. Stop that consumption and the system will die, because the corporations that run the system will die.
And as long as there remain alternatives, there is an even stronger method: only buy stuff that is not from companies in such an alliance.

Then a system like TCPA can remain in place for situations where both ends of the link like to have trust. Like in secure transactions, or when checking if a system has been hacked or infected.

Re:Trusted Linux is ILLEGAL

[aardvarkjoe]

much open-source software is already signed by the creator / distributor, so you know that the binary you got was actually made by him.

Sure, you can sign binaries. But if you give that binary to someone, and he later demands the source code from you, you'd better include the private key along with it.

Doing this would, of course, completely nullify the point of signing binaries. Fortunately, nobody but you seems to be under the impression that this violates the GPL. I suggest that the first stop on your crusade against private keys should be these guys, who are not only signing packages, they are doing so to basic GNU software. Or maybe this organization, which is distributing signed binaries of GNU software!.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

If you plug in a new keyboard (or, for that matter, any new hardware) the system should recognize the TCPA chip on the new component and ask for some sort of verification (

My keylogger is completely passive. It plugs between your PC and keyboard, and forwards along the verification provider by the keyboard.

Only if the system actually encrypts the keystrokes coming out of the keyboard are you protected. And in that case, I have a small videocamera to install in the ceiling tile looking down on your desk.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

And end users can do whatever they want.

Really? They can do WHATEVER they want?

I'm sure there are many end-users who want to encode broadcast-flagged TV episodes and upload them to friends across the internet. Can they do this too? If so, the whole TCPA is meaningless and I withdraw my objections.

Re:Trusted Linux is ILLEGAL

[Qzukk]

You can't fork if it won't run, and it won't run if you don't have the key. The only way to fork would be if you could use the GPL to coerce the signers to give you that

Did you even read the article? IBM pretty much stated that for at least their implementation, there are only two layers: bios space (starting the TCPA engine) and kernel space (key transactions to the TCPA engine). You could write foobarOS and boot it on there, and all you'd need to do is read IBM's documentation and use it, and it would work just fine and do any encryption you'd want it to do.

Who knows, maybe the future goal of TCPA is to kill Linux? Maybe They really are out to get you and every last processor out there. Whether its a Sparc, PPC, x86*, MIPS, or one of the other dozen or so platforms it runs on, they'll replace every last CPU with encrypted crap and kill linux off. What are you going to do about it now other than ranting and raving on the corner of 5th and Main with a "The End is Near!!oneone1" sign?

Re:Trusted Linux is ILLEGAL

[Leo McGarry]

That whole blah-blah -- apart from being wrong in substance, as already explained in a whole assload of other replies -- is predicated on the assumption that the Gnu license is a legally binding document. Anybody who's taken contract law can tell you that, at best, the Gnu license is right on the bubble. In order for a contract to be binding, there has to be an overt act of acceptance. Unless there's a piece of paper somewhere with your signature on it saying that you accept the terms of the license in return for some tangible consideration, you're not bound.

Re:Trusted Linux is ILLEGAL

[Leo McGarry]

I'm sure there are many end-users who want to encode broadcast-flagged TV episodes and upload them to friends across the internet. Can they do this too? If so, the whole TCPA is meaningless and I withdraw my objections.

Oh, I get it. Your objection to trusted computing all centers around the fact that "many end-users" (i.e., you) want to steal content and traffic in stolen content.

Thanks for finally laying that out for us. I kinda wish you'd said so right up front, though. It would have saved a lot of people a lot of time. We all could have just flagged you as an idiot and a thief and called it a day.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

Your objection to trusted computing all centers around the fact that "many end-users" (i.e., you) want to steal content and traffic in stolen content.

No. I want do be able to reprogram my computer to do new and creative things from my own imagination.

The Entertainment Technology Complex, however, wants to stop me from infringing their content. And to achieve that, they are willing to destroy my ability to hack on Linux in my spare time. (Indeed, they'll consider that a useful side-effect!)

Re:Trusted Linux is ILLEGAL

[ink]

Gentoo is a source-based distribution. It won't work, unless they give everyone their private key; which would obviate the entire exercise.

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

this is a common fallacy; the GNU GPL is not a contract and does not need to meet the requirements of a contract. It is a license, just like the ones printed on the MSoft box product packages.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

Let's watch SiliconEntity argue with herself:

Anyone can run anything they want.

Ok, so they can run software which lies. Because that's definitely a part of "anything".

All they can't do is lie to someone else about what they are running.

What? So they can't run lying software, so they can't do "anything"? Oh, I'm so confused!!

Please, stop spreading falsehoods and learn more about how the technology actually works.

The fundamental truth is that if TCPA becomes mainstrean, the ability of amateur programmers to modify the software running on their desktop computers will be greatly reduced. If I modify X.org to display windows in a translucent sphere, my web browsers will no longer run because my display system isn't "Trusted" anymore.

Re:Trusted Linux is ILLEGAL

[Jherek Carnelian]

Minna is absolutely correct with respect to the larger issues of DRM.
If you don't like the specific reference to Microsoft, replace it with "megalocorp."
The end result is the same.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

It is your own decision to buy from companies who inherently do not trust you.

Barely, and getting less so each day. To ignore this, to think that consumer sheeple will care enough to engage in effective boycotts is foolhardy.

Once TCPA DRM gets going, it won't stop with the RIAA and MPAA. Web publishers and ebook sellers will use it too- soon enough, the only web pages that don't have DRM will be niche sites targeted to readers with a philosophical objection, or underground sites dedicated to copyright infringement or other crminality.

When that happens, the "Conservative" Attorney General may just decide to criminalize non-DRM traffic, having observed that the main reason it's still used is "piracy". And who cares that Free Software development will be dead outside of major corporations? Microsoft's quality products are good enough for everyone!

Then a system like TCPA can remain in place for situations where both ends of the link like to have trust. Like in secure transactions, or when checking if a system has been hacked or infected.

Technically, TCPA is ineffective for those situations. Those goals can be achieved more cheaply with different systems that don't block end-users from willingly modifying their own machines. The suggestion that TCPA is beneficial in those fields is a red herring.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

Anybody who's taken contract law can tell you that, at best, the Gnu license is right on the bubble. In order for a contract to be binding, there has to be an overt act of acceptance.

Performing any action that would be illegal without the other party's specific permission consitutes overt acceptance. In this case, infringing on their copyright would be a felony unless you had agreed to the GPL.

Re:Trusted Linux is ILLEGAL

[RmanB17499]

Acceptance of the terms and conditions of a contract does not have to be in writing. Like when you pump gas and then the guy expects you to pay for it: that's an implied-by-fact contract. It's implied by the factual circumstances surrounding the issue. Other contracts maybe implied-by-law. An example would be, let's say a painter comes to my house by mistake (my neighbor asked to get his house painted.) I'm home and figure what they hey I can get a free paint job! Wrong...Unless I was on vacation or had a really good excuse why I didn't make a reasonable attempt to let the painter know that he is at the wrong house my acceptance of the contract is based on the acceptance of his work.

The GNU Software License is a contract.

The validity of the contract requires that offer, acceptance, and consideration be present and that it satisfies the mirror image rule. The mirror image rule requires a contract be able to be written in both directions. (Hence, I will pay you $5,000 to paint my house and on the other side You can tell yourself "I will get $5,000 for painting his house).

Offer-- You may use GPL software pursuant to the terms and conditions set forth in the GPL software license.
Acceptance-- Implied by fact if the software is used.
Consideration-- Legal right to use the software (valuable consideration).

and the mirror image:
Offer-- You can use my GPL software in accordance with the GPL-license.
Acceptance-- Your usage of my software.
Consideration-- Your promise to abide by the GPL.

Promises exchanged by both parties are considered "valuable consideration."
Any good NDA would be binding by making it a two-party NDA. I promise not to divulge your secrets if you promise not to release my secrets. I personally believe that an NDA that's one directional and not supported by consideration would NOT be binding. (However, employment and salary are obviously consideration, too!).

Re:Trusted Linux is ILLEGAL

[RmanB17499]

It actually must be a contract otherwise it's not enforceable. And the only limits that would apply would be court and legal provisions of the Copyright Law. But in a contract you and I can voluntarily agree or disagree to waive some of our rights. Like for example a regular audio CD from Barns and Nobles doesn't come with a license. Its use is only limited by copyright law and doctrines such as fair use. Fair use typically includes the ability to share copies with friends and relatives and listen to it all I want in any platform. Converting this to computers means with one copy of Windows I can give it my family, run it on 4-5 of my personal computers and only pay for it once. However, Microsoft doesn't want this, and I understand why. So they license it under a contract with you the customer. Frequently, I have seen EULA modified and company's discuss the exact verbiage. I have worked for a large computer security firm and also a financing company and both had many customers altering the contract. Some changes would be okay with us and others would be unacceptable, and yet most often: if a customer wanted to change a term it would result in a change in the price. This is what it's all about! If you want Microsoft to provide full protection and warranty that's not limited, they might, if you discuss it with them...but with anything else, they'll only do it for the right price. For the right price I'll sell you insurance that says I'll pay you if you get a paper cut!

Re:Trusted Linux is ILLEGAL

[finkployd]

As usual with slashdot, you hold strong opinions regarding tcpa with absolutely no idea what it is.

Tcpa lets you tell your machine to only run binaries signed by Microsoft. You can also tell it to only run binaries signed by IBM. Or you can tell it to only run binaries signed by debian. Or yourself. Or any combination. You tell it what you want it to do in this regard.

The only valid argument against it is the remote atestation issue, which (using digital signatures) can attest the identity of a client over the network. Think windows file sharing refusing to work with a Samba client, even if they open the protocol, because the samba client was not signed by Microsoft. Or worse, IIS only allowing IE to connect to it. Or most likely, a streaming media server only allowing windows media player.

Course, this has good and bad aspects as well. The technology exists, it has useful purposes, you cannot put the genie back in the bottle. If people use remote atestation irresponsibly as in the above scenerios, then it is not a technology problem, it is a people problem and should be dealt with accordingly. TCPA is just a logical next step to take with public key.

Finkployd

Re:Trusted Linux is ILLEGAL

[drfreak]

Just because the Linux kernel is GPL does not mean software that interfaces with it properly needs to be. Look at the closed-source NVidia and ATI drivers as an example.

If the needed TC kernel interface was written properly, it probably would not matter if it was GPL, because it would not disclose any private keys. Why would someone even put a secret key in hardware if it was intended to be accessed by a piece of software? I really doubt any needed software interface would have to disclose the key.

Re:Trusted Linux is ILLEGAL

[pe1chl]

And after all that, the part of the world that the USA now calls "evil" will be in the position to look down and judge what the highly acclaimed "freedom" is really worth.

But unfortunately the "free" Americans don't realize that and vote for Bush even now that they still have a choice.

Strange.

Re:Trusted Linux is ILLEGAL

[drfreak]

I can never get it right if Sigourney Weaver was the keymaster, or Rick Moranis.

Re:Trusted Linux is ILLEGAL

[nsayer]

But that sort of thing is not what the article was talking about - Microsoft or not. Nowhere does the article talk about DRM. Instead, the article is all about the system I've described - where "trusting" a binary is based on a keychain that is controlled by the owner of the computer, not the author of the OS or some other third party.

When cryptography is used for the evil DRM purposes you folks are talking about, then yes, it is evil and yes, there are GPL issues. But that's NOT what is under discussion here.

Read TFA and try again.

Re:Trusted Linux is ILLEGAL

[SiliconEntity]

What? So they can't run lying software, so they can't do "anything"? Oh, I'm so confused!!

People can run lying software, it just won't convince anyone. It won't be credible, because it can't create fake reports signed with a validated and certified TPM key.

Now will you say that the inability to run software to create a false idea in someone else's mind means you can't run whatever software you want? Jeez, you might as well complain that you can't run software to solve the halting problem or teleport matter. It is beyond the power of software to make arbitrary changes in the universe. That's not a valid limitation on what you can do.

There are no guarantees in the world that you can lie. TPM makes it harder to lie by making it easier to verifiably tell the truth. That may be unfortunate for liars but it is not an actual constraint on their actions.

Re:Trusted Linux is ILLEGAL

[finkployd]

Completely wrong. The owners of machines don't get the keys needed to sign things for their own hardware. Only the builders of the hardware have those keys, and they are contractually obligated by agreements to the MPAA and RIAA not to divulge those keys to anyone (except employees in the course of their work).

Wow, you just don't have a single clue about any of this do you? You can pop whatever keys you want into the hardware. If you want to create a system where only binaries signed by you can run, go for it. If you only want to run binaries signed by debian, redhat, or joe blow down the street, you can do that too. You can also turn off this checking and allow anything to run.

The scary part of this is the remote attestation piece. THIS is what the riaa and mpaa want. It basically allows streaming media servers and media files to only be opened by programs signed (and verified by the hardware) by those they trust, like microsoft. A scary vision of this is that windows file sharing could disallow samba clients to connect to it even if the open the protocol, because samba was not signed by Microsoft.

If the owners of the hardware were going to be the ones having the keys needed to run on that hardware, then I wouldn't have any problem with it.

You are not going to get Microsoft's signing key and be able to sign your binaries as them, but you will certainly get their public key to verify their binaries and put that in your hardware. You can also generate your own key to sign with and put that public key in the hardware too.

Do you honestly believe that anyone, anywhere would ever go for a system where all software running on Windows has to be signed by microsoft? They couldn't even do that with signing device drivers and such (although they tried, all it does is warn you). You think microsoft is going to stop selling visual studio and all their programming tools because nobody but them can create and sign binaries? Now take this a step further, do you think overseas PC makers are going to sell PCs that can only run windows? Even US companies would never do that.

Good God man, actually take some time and learn about this stuff before you spout uninformed drivel everywhere. There are some real legit complaints about TCPA, but you seem to not understand the most basic aspects of it.

Finkployd

Re:Trusted Linux is ILLEGAL

[Leo McGarry]

Performing any action that would be illegal without the other party's specific permission consitutes overt acceptance.

No, it doesn't. Besides, even if that were true, the other party specifically says that anybody who wants one can take a copy without paying any fee for it. That's the IP equivalent of hoisting a sign that reads "FREE KITTENS."

The problem arises when you say after the fact, "Oh, but if you take a kitten, you have to promise to kill and eat it," or something. That's where the problems arise.

Re:Trusted Linux is ILLEGAL

[Leo McGarry]

Acceptance of the terms and conditions of a contract does not have to be in writing.

No, but it has to be overt. It can't be implicit. A contract which is only accepted implicitly isn't binding.

The reason why the Gnu license isn't comparable to pumping gas is that the gas station has a big sign out front that says "$1.69 a gallon" or something. The terms of sale are spelled out right up front. Software distributed under the Gnu license, however, is advertised as "free." Heck, even the name of the controlling authority, the Free Software Foundation, contains the word "free."

If you want a better analogy, it would be like a gas station with a big sign that reads "FREE GAS." You go, you pump, you start to drive away, and then the attendant comes out and demands you pay him $1.69 a gallon. The phrase you're looking for here is "laughed out of court."

Re:Trusted Linux is ILLEGAL

[finkployd]

The fundamental truth is that if TCPA becomes mainstrean, the ability of amateur programmers to modify the software running on their desktop computers will be greatly reduced. If I modify X.org to display windows in a translucent sphere, my web browsers will no longer run because my display system isn't "Trusted" anymore.

The fundamental truth is that you lack the understanding of any of this, despite several people trying to explain it to you. Let's try one more time since you seem to be averse to actually reading any of the specs to have an idea of what you are talking about.

Let's take your above scenerio. x.org signs a binary and ships it to you. They also ship their public key (which you then import into your hardware). Yes, if you change the binary it will not run because the signature is broken. Solution: sign it yourself with a key you generate and put that public key into your hardware.

Now let's talk about software "lying". If I have a client that wants to connect to your x.org server and I tell it to only connect to a server that has been signed by x.org, it will not work if you changed it and signed it by yourself. This a called "remote attestation". The solution to this is that I have to get your public key from you and tell my hardware that stuff signed by it is ok.

If you want to get your panties in an uproar about something, pick this issue, since unlike your other rantings it has some basis in the reality of how this stuff works. I fear a future where some websites will only work if you are using a microsoft signed browser, which is a legit concern of this technology. I do not fear a future where I can only run binaries signed by microsoft because (1) tcpa does not work this way and (2) if it were ever modified to, the countries that actually make the hardware would never go for it and neither would everyone who is not microsoft. Cisco will not even pay microsoft to sign their VPN client drivers, do you think they will accept a situation where their software will not even run without getting MS to sign it? IBM? every forign government? Nobody is going to go for that, that is why nobody has ever proposed it.

Finkployd

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

You are wrong. The GPL is not a contract at all, but a license. It is not a contract because it lacks mutual promises, mutual consideration and mutual assent. It is, however, a legally enforceable license which grants permission to redistribute the copyrighted work and derivatives if and only if you abide by its terms. If you do not abide by its terms, then you are forbidden by copyright law from distributing the work and derivatives. If you do not want to redistribute the work, you can ignore the GPL.

Re:Trusted Linux is ILLEGAL

[finkployd]

No. I want do be able to reprogram my computer to do new and creative things from my own imagination.

Feel free to, however when you begin interacting with other software on other computers across the network, be advised that they may use remote attestation to enforse that only clients (or servers) they trust will communicate with them. This does not impact your rights in anyway (unless you believe you have a right to control how someone else configures their software).

Finkployd

Re:Trusted Linux is ILLEGAL

[Leo McGarry]

The GPL is not a contract at all, but a license.

A license is a specific type of contract. An express license --like the Gnu license --is legally a contract, and as such must be accepted by overt act. Because the Gnu license is an express license with only implicit acceptance, it's not legally binding. It is not, in other words, worth the paper it's printed on.

Re:Trusted Linux is ILLEGAL

[tepples]

There is no need to consume things like music, movies, television programmes, etc.

How can I enter a grocery store without consuming the music on the loudspeaker?

Re:Trusted Linux is ILLEGAL

[tepples]

But there is no such prohibition as you describe, being unable to run unsigned binaries.

What about being unable to run unsigned binaries capable of connecting to the Internet, if both local ISPs' DHCP servers require a positive TPM attestation?

Re:Trusted Linux is ILLEGAL

[tepples]

You could write foobarOS and boot it on there, and all you'd need to do is read IBM's documentation and use it, and it would work just fine and do any encryption you'd want it to do.

So then how would you convince Internet access providers to sign the foobarOS kernel in order to give you an IP address?

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

Besides, even if that were true, the other party specifically says that anybody who wants one can take a copy without paying any fee for it

Without paying any fee, YES. Without agreeing to the contract, NO.

The problem arises when you say after the fact,

But since the GPL is presented to you before the fact, you're writing another irrelevancy. (Yes, it is possible that some people might not see the GPL notification with the software, but in that case, they should have no expectation that it is legal for them to redistribute the software)

Re:Trusted Linux is ILLEGAL

[Leo McGarry]

Without agreeing to the contract, NO.

But there is no overt act of agreement.

Re:Trusted Linux is ILLEGAL

[Jherek Carnelian]

You lack context.

TCPA is primarily about enabling DRM. Just because this article does not spell out the entire context of the topic and instead focuses in on specific implementation details does not mean discussion of the broader issues is off-topic. In fact, allowing such a narrowly focused article to pass without comment would be a bad thing because it makes it very easy for people unfamiliar with the context of TCPA and DRM to assume that TCPA is all good and light.

You may dispute that TCPA is primarily about enabling DRM, but there is enough evidence to support the theory as to at least justify its mention here.

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

I wouldn't called the NVIDIA/ATI drivers as properly interfaced software. NVIDIA drivers break DRI, and ATI drivers are out of date, slow, and buggy. Above all, they run games slower then the Windoze counterparts.

Re:Trusted Linux is ILLEGAL

[RmanB17499]

I completely disagree. http://en.wikipedia.org/wiki/Contracts Implied in fact or implied in law An implied contract can either be implied in fact or implied in law. A contract which is implied in fact is one in which the circumstances imply that parties have reached an agreement even though they have not done so expressly. For example, by going to a doctor for a physical, a patient agrees that he will pay a fair price for the service. If he refuses to pay after being examined, he has breached a contract implied in fact.

Re:Trusted Linux is ILLEGAL

[RmanB17499]

And consideration doesn't have to include money.

So "Free" software may stipulate terms and conditions on its license.

If you and I enter into a "free" Non-disclosure agreement it will satisfy the mirror image rule as long as the promise flows both ways.

Offer-- To keep your information private.
Acceptance-- Your acceptance.
Consideration-- You agree to keep my information private.

The mirror image would be the exact opposite and this would be a binding contract.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

This does not impact your rights in anyway (unless you believe you have a right to control how someone else configures their software).

I believe I have the right, codified in law, to make small exerpts of copyrighted works for critical purposes. TCPA desires to specifically remove that ability from me.

Re:Trusted Linux is ILLEGAL

[Qzukk]

What makes you think they'll require it? Seriously, what do they get out of it, other than wasting bandwidth signing your damn OS?

And thats totally aisde from the point that there are multimillions of dollars of sparcs, powerpcs, Irix machines, IBM mainframes, and more out there that will still need internet connections.

Re:Trusted Linux is ILLEGAL

[finkployd]

I believe I have the right, codified in law, to make small exerpts of copyrighted works for critical purposes. TCPA desires to specifically remove that ability from me.

It most certainly does not. DRM desires to specifically do that. And even it doesn't DRM can (and has been) made to respect fair use laws. TCPA is a tool that can aid DRM, but so is public key cryptography, so are computers in general. Your beef with not with the tools that might be mis used, it is with those who are misusing them. I can make just as strong arguement against public key cryptography as you can about TCPA (which is just PK crypto built into hardware) on the same basis.

Finkployd

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

The scary part of this is the remote attestation piece.

Yes, that's the scary part. It's also the only part that really matters, and the only reason the entertainment/technology industries are pushing Trusted Computing. If there was no remote attestation, the TC concept wouldn't even exist.

The scary part of this is the remote attestation piece.

The scary parts of atomic bombs are the blast and radiation. The rest isn't dangerous at all!

Good God man, actually take some time and learn about this stuff before you spout uninformed drivel everywhere.

Yes, I am presenting the worst-case scenario. I do this to stir the technology and entertainment consumers around the world to resist what otherwise could be an enticing change:

1. "Buy a hot new RIAA-subsized multimedia PC for just $299, and then get 66% off all music downloads for the next two years! Just ask for the TCPA-DRM special model!"

Only a combination of consumer antipathy and international fear of Microsoft's global dominance gives us a chance to avoid that fate. (And personally, I have a lot more faith in the paranoia of the Chinese government than in the strong principles of American consumers)

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

The solution to this is that I have to get your public key from you and tell my hardware that stuff signed by it is ok.

Wrong. I also need to tell my applications it's OK too. And I can't do that, because the apps refuse to run except in the pristine x.org environment. (Maybe they'll run, but they won't display any media or even web pages because of DRM)

Even if I hadn't read the specs, there are only two possibilities:
1) Either TCPA will prevent me from writing a modified display driver that can run normal programs, including viewers for DRM media like web pages.
2) Or TCPA is useless, and the Entertainment Technology industry wouldn't be trying to push it on is.

Since option 2 is self-evidently false, option 1 must be true.

Re:Trusted Linux is ILLEGAL

[Leo McGarry]

Um. You didn't seriously get your legal advice from "Wikipedia," did you?

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

No, but it has to be overt. It can't be implicit. A contract which is only accepted implicitly isn't binding.

Correct, and fine. As you've already ignored, any action which would've been a felony without having recieved permission in the contract is an overt acceptance. Unlike invalid "implicit acceptances", a person can offer no other excuse for having performed those actions except as assenting to the contract.

Software distributed under the Gnu license, however, is advertised as "free.

False. It is advertised as "Free". (If you can't tell the difference, then ponder whether Photoshop can run on windows or Windows)

Heck, even the name of the controlling authority, the Free Software Foundation, contains the word "free."

Just because a copyright holder is publishing their works for a $0.00 charge (making it "free"), doesn't destroy their right to authorize each and every reproduction of the software. Many people ignore this, but it is technically illegal to redistribute commercial game demos, security updates from Microsoft, and so forth. Why, I can turn on my television and watch "The Simpsons" for free, but I don't get the right to redistribute copies.

Or, do you also believe that a free ticket to a movie entitles you to videotape it an upload to the internet?

sign that reads "FREE GAS." You go, you pump, you start to drive away, and then the attendant comes out and demands you pay him $1.69 a gallon.

Have you ever run a GPLed program? Do you know what they print on startup?

1. Copyright 2004. This is free software, covered by the General Public License, and you are welcome to change it and distribute copies under certain conditions. Click Help->License for details.

Re:Trusted Linux is ILLEGAL

[finkployd]

Yes, that's the scary part. It's also the only part that really matters, and the only reason the entertainment/technology industries are pushing Trusted Computing. If there was no remote attestation, the TC concept wouldn't even exist.

Actually it has existed for years in the form of the "security chip" that IBM has long shipped with certain models of Thinkpad. It is quite useful on it's own, and while the recent interest by the RIAA and MPAA is exclusively due to remote attestation, that is certainly not the only application. Keep in mind, all we are talking about is PKI with the public keys in tamper proof hardware, and a OS that supports this. Any use you can think of with the technology will probably be attempted.

The scary parts of atomic bombs are the blast and radiation. The rest isn't dangerous at all!


The scary parts of public key crypto are pedophiles and terrorists. BAN SSL TODAY!

Yes, I am presenting the worst-case scenario. I do this to stir the technology and entertainment consumers around the world to resist what otherwise could be an enticing change

"Buy a hot new RIAA-subsized multimedia PC for just $299, and then get 66% off all music downloads for the next two years! Just ask for the TCPA-DRM special model!"


And I am behind you 100% on that, just remember we are not fighting the technology (which history has always proven to be a comical exercise in futility), we are fighting the misuse of it. Would the world be better off if public crypto never existed simply becasue it gives the movie studios the ability to enforce the use of an approved player to get their content?

Only a combination of consumer antipathy and international fear of Microsoft's global dominance gives us a chance to avoid that fate. (And personally, I have a lot more faith in the paranoia of the Chinese government than in the strong principles of American consumers)

Damn straight, but do not rule out educational institutions as well, they would be severly crippled in this vision of the future as well.

I'm glad you have come around to stopped claiming that it enforces what you run on your PC, the real evils that TCPA opens up are much more fun and informative to discuss.

Finkployd

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

Because the Gnu license is an express license with only implicit acceptance, it's not legally binding. It is not, in other words, worth the paper it's printed on.

You are claiming that corporations like Red Hat, Sun, Time-Warner, and even IBM are committing blatant and felony-level copyright infringment thousands of times daily in the course of their business. (If the GPL isn't valid, then they have no authority to redistribute the Linux-based products they sell every day)

Given the aggregate number of lawyers employed by all those firms, I somehow suspect that their legal authority outweighs yours (in particular since even I can blast each of your arguments back into the insensibility from which it came)

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

Fair use typically includes the ability to share copies with friends and relatives and listen to it all I want in any platform. Converting this to computers means with one copy of Windows I can give it my family, run it on 4-5 of my personal computers and only pay for it once.

That's far, far beyond the bounds of any fair use. Listening to two copies of a CD at once, or running two copies of the same Windows XP disc at once are equally and completely illegal.

Re:Trusted Linux is ILLEGAL

[finkployd]

Wrong. I also need to tell my applications it's OK too. And I can't do that, because the apps refuse to run except in the pristine x.org environment. (Maybe they'll run, but they won't display any media or even web pages because of DRM)

I doubt you will find any open source applications with that restriction, and either way you can just turn it off (remember, it is open source).

Even if I hadn't read the specs, there are only two possibilities:
1) Either TCPA will prevent me from writing a modified display driver that can run normal programs, including viewers for DRM media like web pages.
2) Or TCPA is useless, and the Entertainment Technology industry wouldn't be trying to push it on is.

You haven't read them (or understood them), that much is clearly evident to almost everyone in this thread.

You can write a custom display driver if you want, but then through the magic of remote attestation, the media server will know. Perhaps some windows applications will be written to not function in this regard as well (windows media player likely will be designed this way), all the more reason not to use windows I guess, since they could technically do this without TCPA (it would just be more easily cracked).

Also, do not be so quick to discount option 2. Given the shift key defeted "CD copy protection", weak DVD DRM, and other snake oil they constantly buy into, it is clear they do not understand the issues and will toss money at anyone claiming to have a solution to piracy. TCPA in a perfect form may go further than these other weak attempts, but the devil is in the implimintation. Microsoft has not yet figured out how to make their OS keep people from getting into PCs over this new fangled internet thing they recently discovered, I doubt their attempt to keep all systems secure for video playback will be all that steller as either.

Finkployd

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

No, he's saying that the people who give away their products under the Gnu license have absolutely no right to retroactively attempt to restrict the use and distribution of a product to which they have already waived their rights.

You can't have your rights and waive them at the same time.

This isn't complicated. You should be able to understand it. Why can't you? Is it possible that you're just putting your fingers in your ears because you're faced with an uncomfortable truth?

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

Any use you can think of with the technology will probably be attempted.

And that's why the technology is bad- all of the uses I can think of are wrong. There's a long list of supposed positive uses, all of which collapse as strawmen that are either fundamentally flawed, or more easily and reliably implementable without any specialized hardware.

Would the world be better off if public crypto

That's a red herring too. The core of TCPA isn't the crypto, but in the tamper-proof hardware which will need to be almost universal. And since the hardware itself isn't really tamper-proof, the only real way it can work is government-supported prohibition of reverse engineering and hobbyist computer engineering.

Re:Trusted Linux is ILLEGAL

[EvilSporkMan]

From the GNU GPL:

5. You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works. These actions are prohibited by law if you do not accept this License. Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it.

See above for why you must accept the GPL to use GPL-protected software.

Re:Trusted Linux is ILLEGAL

[suckmysav]

"[...] consumers have the power to shut down that system. There is no need to consume things like music, movies, television programmes, etc. Stop that consumption and the system will die, because the corporations that run the system will die."

I don't believe enough people will care enough to boycott the "entertainment" industry to effectively kill the corporations in question, but I agree, there is no need for people to consume this crap. The very fact that there are plenty of people out there who are stupid enough to think that 50 cent and Usher are talented enough for them to lay out their hard-earned cash on their CD's alone is demonstrative of that fact.

My suggestion is to make hay while the sun is still shining, build up your collections of unencumbered stuff now while you still can and then just stop consuming their crap when they make it too hard/expensive/restrictive to do so. From then on you will have to rely on legacy hardware and your existing collections for as long as you can keep it all working.

Note: Keep plenty of spares.

I expect that enough people will continue to purchase the latest and greatest DRM'd dreck that will continue being pumped out of Hollywood to keep the industry going, but I sure don't intend to contribute to it at all.

Re:Trusted Linux is ILLEGAL

[RmanB17499]

No, just used them as a reference point.

If you prefer to use the Legal Information Institute at Cornell University or another source that's okay with me.

That's the law...Consult any competend US-based attorney.

Re:Trusted Linux is ILLEGAL

[finkployd]

And that's why the technology is bad- all of the uses I can think of are wrong. There's a long list of supposed positive uses, all of which collapse as strawmen that are either fundamentally flawed, or more easily and reliably implementable without any specialized hardware.

Uh huh, bullshit. Just because YOU do not understand the positive uses does not mean they are not there.

The core of TCPA isn't the crypto, but in the tamper-proof hardware which will need to be almost universal.

The core of DRM is crypto, specifically public key crypto. Does that make pk evil?
It is not tamper proof, nor does it have to be. It needs to be able to resist SOFTWARE based tamper attempts. Actually I take that back, it does need to be tamper proof to the extent that you can not hardware it to give false postives.

the only real way it can work is government-supported prohibition of reverse engineering and hobbyist computer engineering.

That exists today without TCPA, and is certainly worth exerting our efforts to fight.

Finkployd

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

You seem to be highly ignorant of the implications here, although they have been broadly discussed. Why not do a little more reading before responding again?

I doubt you will find any open source applications with that restriction

I never said they'd be open-source apps. They could be commercially-produced closed-source Linux apps. For example, as a rebuttal to complaints against the illegality of the DeCSS project (which allows Linux software like mplayer to view DVDs), trolls (such as Twirlip of the Mists) like to claim that a commercial vendor can simply offer a closed-source player with licensed DVD decoders.

and either way you can just turn it off (remember, it is open source).

No, I can't! That's the entire basis for the complaint. There might very well be an open-sourced viewer application for DRMed media on "Trusted Linux", but it will have been signed by some corporation after careful analysis.

There is no way I can turn off the DRM feature, because that above all else would red-flag the application as something that the signing authority won't approve (And of course, it would be tremendously difficult and/or expensive to submit a personally modified variant for auditing and signage)

You can write a custom display driver if you want, but then through the magic of remote attestation, the media server will know.

Yes, and therefore I eventually won't even be able to view web pages on that system, meaning any hobbyist mods automatically cripple the box. (Think long term, and remember that technology changes fast!)

Perhaps some windows applications will be written to not function in this regard as well (windows media player likely will be designed this way),

If you think Linux systems wouldn't also do this (such as, for example, TiVo), you're dreaming.

all the more reason not to use windows I guess,

You act like there will still be a platform (PCs which are general-purpose end-user programmable computers) left to run Linux on. There won't be. That is the ETC's goal, that is the whole point. Non-treacherous computers will have a 10x cost premium, making competive CPU power cost-prohibitive (Have fun raiding museums to build Beowful clusters for Free computing)

Given the shift key defeted "CD copy protection", weak DVD DRM, and other snake oil they constantly buy into, it is clear they do not understand the issues and will toss money at anyone claiming to have a solution to piracy.

"The enemy has failed before. Therefore, they will always fail, and we have nothing to worry about". That's an idiotic strategy. For an exhaustive rebuttal, see "Code & Other Laws of Cyberspace" by Lessig.

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

he's saying that the people who give away their products under the Gnu license have absolutely no right to retroactively attempt to restrict the use and distribution of a product to which they have already waived their rights.

So tell me, how's the weather like in your bizarre alternate reality where SCO is winning all its lawsuits? And where Eben Moglen was kicked out of Columbia for academic fraud?

Not only are your arguments looney on their face, but they are also completely opposed to how business is actually conducted today in the real world. You make as much sense as claiming that Saddam Hussein has an arsenal of nuclear missiles.

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

Um. You didn't seriously get your legal advice from "Wikipedia," did you?

That'd be a step up from where you get it, which is apparently inferred from patterns in the entrails of mice you kill with your teeth.

Me? I get it from Columbia Law School, but you may as well start small and work from there.

Re:Trusted Linux is ILLEGAL

[Anonymous Coward]

No. A license in not a type of contract. A contract requires consideration a license does not. A license is a unilateral grant. Eben Moglen, a professor of law at Columbia University School of Law and recognized IP law authority states:

The essence of copyright law, like other systems of property rules, is the power to exclude. The copyright holder is legally empowered to exclude all others from copying, distributing, and making derivative works. This right to exclude implies an equally large power to license--that is, to grant permission to do what would otherwise be forbidden. Licenses are not contracts: the work's user is obliged to remain within the bounds of the license not because she voluntarily promised, but because she doesn't have any right to act at all except as the license permits.

He wrote an excellent article which completely destroys your arguement.

Re:Trusted Linux is ILLEGAL

[tehdaemon]

"I doubt you will find any open source applications with that restriction"
Or the ability to "' display any media or even web pages because of DRM'" So much for Firefox and xine.

The only way that this will not be a huge problem is if open source reaches critical mass and the media companies can (and some DO) make more money off the open source stuff than they can off of the DRM'ed stuff. What TC has the ability to do is to make open source irrelevant to anyone who wants to communicate with the world in any electronic way.

Your 'just turn the feature off' is idiotic, because it is not possible. Sure you can turn it off, and then the program stops doing what it is supposed to do - that is the whole point of TC in this application!

"the media server will know. Perhaps some windows applications will be written to not function in this regard as well "

And the DRM^h^h^h 'media' server will refuse to give the decryption keys to any program not written that way. And without those keys there will be (if they want) nothing that works. You won't be able to open your M$ office files, 'cause they are encrypted. You won't be able to access your financial data, either online, or in your own files. No music, no movies, no posting to /., no encrypting your own email with your own pgp key, no IP address from the ISP, (and no routing without it!) I could probably come up with more, but why? Isn't that bad enough?

Your whole argument for on #2 is 'they have proven themselves technically incompetent so far...' True enough. But people occasionally learn from their mistakes, specially when there is a huge wad of cash as incentive. Do I want to bet on their continued incompetence? NO!!

I agree with the EFF on this. If I can control the remote attestation part, TC is a great idea, it puts me even more in control of my computer. If not, I potentially lose all control. If I don't control it, do I really own it?

Re:Trusted Linux is ILLEGAL

[tehdaemon]

Gaaaa...

It would also stop me from building my own MythTV HD box and doing simple timeshifting. Rebut that.

Re:Trusted Linux is ILLEGAL

[demon_2k]

Ok..Lets say you have some data bound to some configuration, and the program is modified (Let's say it is a virus). What do you do if you can't get thesome piece of software/configuration? What is some or no pieces of the source are not available.

Are you then locked you of your own data???

What if you get all the source and then compile it? Will it work then?

What if you get all the source but compile with different optimizations?

Re:Trusted Linux is ILLEGAL

[finkployd]

The only way that this will not be a huge problem is if open source reaches critical mass and the media companies can (and some DO) make more money off the open source stuff than they can off of the DRM'ed stuff. What TC has the ability to do is to make open source irrelevant to anyone who wants to communicate with the world in any electronic way.

The primary methods of electronic (specifically internet) communication are based on open source products. Apache is by far the most popular web server and sendmail (and I guess lump in postfix and qmail) are the most popular email servers. Nothing is going to happen to make them irrelevent, not the open source model. It is too well know, too "mainstream" and too popular to be legislated away or technically made irrelevent.

Your 'just turn the feature off' is idiotic, because it is not possible. Sure you can turn it off, and then the program stops doing what it is supposed to do - that is the whole point of TC in this application!

Don't run products that are this restrictive. Intuit found out what happened when they tried to write to the boot sector of hard drives as a copy protection method, and quickly backpeddled. Any company that releases software that doesn't work with TCPA turned off is liable to lose a HUGE section of their customer base, both those that don't like it and those that do not have hardware to support it. The biggest issue is really things like media players not playing certain files or connecting to certain streaming servers without remote attestation turned on and everything validating correctly. This issue will be dealt with by the market. If people find it too restrictive and annoying, it will fail, and other less restrictive services will take their place. I mean, do to overwhelming popularity, MS has yet to disable mp3 playing in windows media player. It happily plays those just fine even though it also does the DRM game with it's own format. I don't see TCPA being much different. Some new functionality may require it, but the old stuff should still work. And if it doesn't more people will ditch the software for stuff that does.

And the DRM^h^h^h 'media' server will refuse to give the decryption keys to any program not written that way. And without those keys there will be (if they want) nothing that works.

Yup, and I am guessing those media servers will not be very popular. We already tried this with DIVX and it failed. But hey, maybe they will be popular, I could care less personally.

You won't be able to open your M$ office files, 'cause they are encrypted.

If you choose to encrypt them. Although perhaps MS will enforce the encryption and TCPA with office, then you will see open office grow in popularity so fast it will make Firefox's newfound market share seem like nothing. You think other countries will stand for this? Even our own companies and universities disrtust MS so much that something like this could very well push them over the edge.

You won't be able to access your financial data, either online, or in your own files.

My bank (and many others it seems) are opening up from their previous "IE only" stance. Why would they go back on that and lock out customers? Actually I could see a system where the bank gives you a key to use with TCPA to communicate with them so that only you can access your data. Of course, they could do this today with SSL mutual auth and client certificates, but they don't. More's the pitty.

no posting to /.

Why would /. do a thing like that?

no encrypting your own email with your own pgp key

How does TCPA prevent this? You might as well accuse TCPA of raping your daughter and punching your son while you are at it. Anything else we can dream up as a negative aspect of it? Does it cause cancer too?

no IP address from the ISP, (and no routing without it!)

I imagine my router will be able to get and

Re:Trusted Linux is ILLEGAL

[finkployd]

You seem to be highly ignorant of the implications here, although they have been broadly discussed. Why not do a little more reading before responding again?

I've been following this topic for years, so I don't know everything but I didn't just read the TCPA specs yesterday.

They could be commercially-produced closed-source Linux apps. For example, as a rebuttal to complaints against the illegality of the DeCSS project (which allows Linux software like mplayer to view DVDs)

Illegal only under the DMCA, which I believe is where your real problem is. The math is not the problem, the hardware is not the problem, the stupid law is. Let's fight that instead of trying to shove technology worms back into a can.

No, I can't! That's the entire basis for the complaint. There might very well be an open-sourced viewer application for DRMed media on "Trusted Linux", but it will have been signed by some corporation after careful analysis.

There is no way I can turn off the DRM feature, because that above all else would red-flag the application as something that the signing authority won't approve (And of course, it would be tremendously difficult and/or expensive to submit a personally modified variant for auditing and signage)

Perhaps the solution (rather than trying to break encryption) is to just not purchase DRM media. It is not essential to life and the lack of it could very well improve your quality of life. Barring that, fight the laws that make it illegal to decrypt, not the math or the hardware that does it. That attack has not ever worked in history, and I doubt today will be any different.

Yes, and therefore I eventually won't even be able to view web pages on that system, meaning any hobbyist mods automatically cripple the box.

Firefox is getting more popular. Everyone knows how Firefox was developed. Open source is not some subervisive, unknown, secret cult anymore. It is mainstream. Putting out content that kills it is not in the best interest of web designers (well, again maybe DMCA obsessed control freaks, but are we not better off without them anyway?). The time when this kind of trickery would work has passed. Open source, firefox, linux, etc. are all too popular and mainstream to be swept away by dumb laws or bad corporate decisions. Maybe not in the US, but certainly globally. If the US content providers want to cut themselves off from the rest of the world, I don't think many would lose sleep.

You act like there will still be a platform (PCs which are general-purpose end-user programmable computers) left to run Linux on. There won't be.

You are raving, seriously. What possible evidence do you have to support such as MASSIVE and unprecidented change in the future where this could happen. I'm sure there are some nuts who have this goal, but there are also nuts who believe all cryptography should be outlawed, and all citizens should have rfid tags implanted. People will oppose this, hell computer companies will oppose this. The rest of the world will ignore it completely.

Non-treacherous computers will have a 10x cost premium, making competive CPU power cost-prohibitive

Supply and demand. Expect to see China start cranking out non-treacherous computers by the boat load. Until there is legislation enforcing TCPA computers (a totally unprecidented move), there will always be some willing to get rich by filling the void if Dell, IBM, Gateway, and everyone else decided to stop making general use PCs. And if you think this legislation is forthcoming, remember these are the same people who could not enforce the Clipper chip. And that was with the threat of massive pedophilia activity, terrorism, organised crime! What is the the boogyman enforcing DRM? piracy? please, congress cares, but not THAT much.

"The enemy has failed before. Therefore, they will always fail, and we have nothing to worry about". That's an idiotic strategy. For an exhaustive rebuttal, see "Code & Other Laws of Cyberspace" by Lessig.

That was in rebuttal to the point that TCPA must work or hollywood would never be pushing it. See "the parent post" by Finkployd.

Finkployd

Re:Trusted Linux is ILLEGAL

[tehdaemon]

Most of your rebuttals basically say that TC won't succeed in doing what it has the ability to do because not enough people will require it. There is a lot of truth to this too. Market forces will work, as you said, making it very hard for many things to be restricted. I did not (and do not) say that all of this is very likely, just possible. But I do not say never. And that worries me.

Hmmm. that answers your paragraphs #'s 1, 2, 4, & 5.

The 'media' servers I imagine would not serve media. Only decryption keys. For music, movies, and ebooks (and copyrighted immages on the internet, and copyrightes newssites, and...just about any media that you do not have the copyright on, and maby some you do ). And since Copyright gives a monopoly on those, if the *AA's and publishers encrypt all of their stuff... I do not think that there is enough competetion in this area to let market forces work. In fact the *AA's have enough clout to possibly force other publishers to comply. Anti-trust stuff might stop it, but the DOJ doesn't seem too interested. This is not a slam-dunk for the *AA's, just an easy beachead. Don't personally care? When the *AA's force all distribution chains to follow suit or go out of buisness? (because they would lose the *AA's buisness?)

/.? no, /.'s isp (more later)

Email. TCPA only auth email servers. Think anti-spam stuff, or FBI crap. And no signed pgp binaries. So, while you could sign it, you could not send it. Likely? we covered that already.

The isp thing. Think TCPA + great firewall of china + FBI needs to track terrorists. Or the *AA's get the common carrier status of isp's revoked, and the isp's are required by law to prevent the latest and greatest encrypted/anonymous/p2p filesharing thing. That and the fact that lots of people (like me) have only one choice for broadband, competition is hard to come by.

Remote attestation is still a good thing even if the owner controls it. It says, I have not been hacked, cracked, rooted, warzed, or internet-wormed. Much less usefull to others, true, but not useless.

"let's attack those who use it for evil." Excellent idea. Let's start by making trusted computing something that they cannot abuse, by giving remote attestation control to the owner. Other solutions: Copyright law needs reform. Like 20 years from the first intentional publication, no exceptions extensions or refunds.[1] And anything that dissallows fair use is inelligable for copyright protection. (they can still sell it if they want, but when someone cracks it, it is now public domain!) Bye-bye DRM'ed music and movies. Without that incentive, the chance of TCP being used for evil is much less, 'cause it is that much less widespread, and has that much less money pushing it.

[1] Copyright Disney, of course. Fortunately brains will never be TCPA compliant, or they would restrict this quote too.

Re:Trusted Linux is ILLEGAL

[finkployd]

Most of your rebuttals basically say that TC won't succeed in doing what it has the ability to do because not enough people will require it.

Right, technology does not do anything by itself, people make it do stuff. This is why my compiler has not built a virus that destroys the internet, my gun has not killed anyone, my car has not run anyone down, and the atom bomb has not wiped out humanity. They all have the ability to, it is just not acceptable to use them that way.

The 'media' servers I imagine would not serve media. Only decryption keys. For music, movies, and ebooks (and copyrighted immages on the internet, and copyrightes newssites, and...just about any media that you do not have the copyright on, and maby some you do ).

Creative Commons still exists. Frankly if publishers want to DRM their stuff to hell and back, that is their option. I just won't use it. Many others will decide to do the same I imagine, this is how DIVX died. I do complain loudly when fair use is interfered with and I hate that DRM doesn't account for the expiration of copyright (of course, neither does congress these days), those battles need to be fought DRM or not.

Email. TCPA only auth email servers. Think anti-spam stuff, or FBI crap. And no signed pgp binaries. So, while you could sign it, you could not send it. Likely? we covered that already.

I do not believe it is in the best interest of even the FBI to completely cut the US off from the rest of the world. That would certainly put a strain on the old economy.

Think TCPA + great firewall of china + FBI needs to track terrorists.

Terrorists were not going to use the Clipper chip, they will not use this either. And the terrorist boogyman is not the blank check to pass laws that it once was. Despite Bush's reelection, most people I know (including conservatives) hate the patroit act and the ideas behind it. Something like you describe would be WAY over the top and would not be stood for.

Or the *AA's get the common carrier status of isp's revoked, and the isp's are required by law to prevent the latest and greatest encrypted/anonymous/p2p filesharing thing.

Scary thought, TCPA has very little to do with that, it could happen just as easily without it.

That and the fact that lots of people (like me) have only one choice for broadband, competition is hard to come by.

This should change, but it is a problem in and of itself, TCPA or not.

Remote attestation is still a good thing even if the owner controls it. It says, I have not been hacked, cracked, rooted, warzed, or internet-wormed.

Remote attestation does not exist if the owner controls it. If the owner can control it, so can the virus that just told it to assure other systems it was not hacked, cracked, rooted, etc. If we are going to do this, we might as well just drop it, that would be my ideal solution.

Personally I do not predict remote attestation ever really takes off. The P3 serial number couldn't even take off and that is nothing compared to remote attestation.

Other solutions: Copyright law needs reform. Like 20 years from the first intentional publication, no exceptions extensions or refunds.[1] And anything that dissallows fair use is inelligable for copyright protection. (they can still sell it if they want, but when someone cracks it, it is now public domain!) Bye-bye DRM'ed music and movies. Without that incentive, the chance of TCP being used for evil is much less, 'cause it is that much less widespread, and has that much less money pushing it.

Hell yeah.

Finkployd

Re:Trusted Linux is ILLEGAL

[ytpete]

There is no need to consume things like music, movies, television programmes, etc. Stop that consumption and the system will die, because the corporations that run the system will die.

In a well-adjusted market, this sort of correction could realistically happen. In the case of entertainment media, however, what we have is a monpoly---a controlled market. There's a reason such practices are illegal. Consumers can't take their business elsewhere because they'd be forced to avoid 95% of available products. The RIAA likes to claim that people have this option, but there are good reasons why it's not viable.

Re:Trusted Linux is ILLEGAL

[ytpete]

DRM cannot be "made to respect fair use laws." Fair use is an inherently fuzzy concept, and permitting all the forms of fair use that I am entitled to would also mean permitting many genuine copyright infringements. Any DRM scheme that permits this is not "DRM" in any meaningful sense.

Re:Trusted Linux is ILLEGAL

[tehdaemon]

"Or the *AA's get the common carrier status of isp's revoked, and the isp's are required by law to prevent the latest and greatest encrypted/anonymous/p2p filesharing thing.

Scary thought, TCPA has very little to do with that, it could happen just as easily without it.

That and the fact that lots of people (like me) have only one choice for broadband, competition is hard to come by.

This should change, but it is a problem in and of itself, TCPA or not. "

Yes. My point though was that these problems reduce competition. Competition is necessary for the market, and most of your reasons to not be worried about TCPA is that the market will not allow it.

What if the TCPA chip was designed with a way to get the key out, (a back door) that was not accessable from the CPU at all. A connector that connects only to the chip, and nothing else on the motherboard? That way I could get the key, sign my modified binaries, and install them on the computer. But there is no way for a rootkit/virus etc. to do the same. No physical access to the key.

"Hell yeah."

Question, which do you think is more likely, widespread TCPA use and lockdown, or something like I just proposed? My bet is TCPA, and that is why I am worried.

Re:Trusted Linux is ILLEGAL

[quarkscat]

I did RTFA, and I believe that your assessment
is spot-on-target. There is an alternative,
but it will never be adopted by the TCPA org
because it does not support their DRM lock-down.

A separate memory storage (as a part of the
BIOS) that is protected by strong encryption
could be used to store "tripwire" type signing
under the control of the end-user/administrator
of the system. The purpose: to eliminate and/or
prevent any trojans/viri from modifying the
OS and data. As tripwire can be configured to
test or ignore specific data regions, DRM issues
can be mitigated. The remote access (as proposed
by MSFT and Cisco) is not in the best interests
of the end-users, but is for **AA.

Until TCPA is completely under the control of
the owner/end-user of the computer, it is not
something I would consider viable. Should
Cisco move forward with their TCPA plans, either
the Internet will be doomed to strict commercial
interests, or be fractured into the "internets"
of lore and legend.

Re:Trusted Linux is ILLEGAL

[Minna Kirai]

Anybody who's taken contract law can tell you that,

That apparently doesn't include you. Do you remember the example the lectern gives about the man coming into a restaurant and asking for food, and how that consitutes agreement to a contract? Go back to your textbook if you can't remember.

Re:Trusted Linux is ILLEGAL

[Leo McGarry]

Hmm. Five rapid "overrated" moderations on five successive comments. How very interesting. How very interesting indeed.

Re:Trusted Linux is ILLEGAL

[(H)olyGeekboy]

Thanks for finally laying that out for us. I kinda wish you'd said so right up front, though. It would have saved a lot of people a lot of time. We all could have just flagged you as an idiot and a thief and called it a day....
Hmm. Five rapid "overrated" moderations on five successive comments. How very interesting. How very interesting indeed.

Looking at your comment history, it seems to be more of the same crap as above. Gee, golly, I can't imagine anyone wanting to modbomb you for such smarmy jackassery!!

Re:Trusted Linux is ILLEGAL

[Geek In Training]

Forget not that there are dozens of people who do not post, but who watch these threads. Along come modpoints, and down you go.

Maybe you had it coming, "Mr. Chief of Staff?"

False...

[rbarreira]

False...

Some features could be usefull...

[bcmm]

If, for example, this provided a way to make sure that a computer on the internet is really who it claims to be, that would be good.

But trusted windows, at least, is going to be about remote deletion/disabling of data.

Re:Some features could be usefull...

[Anonymous Coward]

oh teh noes deys gonna delete my puter? =/

Re:Some features could be usefull...

[Anonymous Coward]

Couldn't you just use the CPU's serial no. for it ?
We don't want to be identified... unless we choose to be... and that is ONLY on session basis... get a chip-card and a reader for identifing users.
The reason for them to have "trustworthy computing" is to have the possibility to backtrack any session to a specific computer. That is NOT in the interest of users... it may be acceptable for the police to fight criminals... but it is NOT acceptable for ANY other reason.

Software DRM

[Yartrebo]

Since the source is available for Linux, what would stop someone from sandboxing 'trusted' software by having the OS validate code before it's executed (slow, though a bit faster than emulation and without all the bugs), and then implenting the DRM hardware (or BIOS) instructions in software in a way that stores the keys (or plaintext information, if that is not doable) and allows access to any software to get the info.

The software DRM implementation would be 100% transparent to the application and noone would be the wiser.

It should also be workable with a x86 emulator running a closed source 'trusted' application along with its closed source OS, with the emulator doing the DRM instructions a little differently than normal.

Re:Software DRM

[Henk Poley]

Another fun thing would be to pipe all the TCPA hardware requests over the network to some central machine. Tying software to a machine this way has never been more futile :)

Re:Software DRM

[SiliconEntity]

Since the source is available for Linux, what would stop someone from sandboxing 'trusted' software by having the OS validate code before it's executed (slow, though a bit faster than emulation and without all the bugs), and then implenting the DRM hardware (or BIOS) instructions in software in a way that stores the keys (or plaintext information, if that is not doable) and allows access to any software to get the info.

This is one of the most commonly asked questions about the TPM. The answer is that the TPM implements what is called a "secure boot" sequence.

The first thing that happens in a TPM enabled computer is that the BIOS, on startup time, sends a hash of itself to the TPM. Then, when the BIOS goes to load the OS, it sends a hash of the boot loader (grub, in the case of Linux) to the TPM. The boot loader will be modified (see the Trusted Grub project) to take a hash of the OS kernel and send that to the TPM. And the OS itself will be modified (a la tcgLinux to take a hash of the various OS components, startup scripts, and programs as the computer boots.

The net result is that the TPM has a record of what OS was booted and what the software configuration is that is running. This allows it to distinguish between a "real" boot and an emulated one, because in the latter case it sees a hash of the emulator being loaded.

Software which runs in un-emulated mode and uses the TPM features can distinguish that case from when it is running emulated. If it locked some data using the TPM in the first mode, it won't be accessible in the second mode.

Once remote attestation is possible, networked applications will be able to report their software configuration to each other. This will be unforgeable because the TPM will sign an attestation of the software configuration, and the TPM itself will have a certificate from the manufacturer attesting that it is a legit TPM. Your emulator will not have a certified TPM key (those stay on the chip) and so it won't be able to come up with a credible forged attestation. Programs running on emulators won't be able to take part in network security applications that use these features.

Re:Software DRM

[Henk Poley]

Your emulator will not have a certified TPM key (those stay on the chip)

That is why CSS is still not broken for DVD, heh?

Seriously, pretty much any hardware can be emulated in software, albeit maybe at a slower speed. And when people want it bad enough they will find a way to ease finding valid keys, or either read data out of ICs or simply TCPA documents..

Re:Software DRM

[SiliconEntity]

Your emulator will not have a certified TPM key (those stay on the chip)

That is why CSS is still not broken for DVD, heh?

CSS was a completely different case. It was implemented in software. This allowed a CSS key to be pulled out of the software, and the CSS algorithm itself to be revealed. Cryptanalysis of CSS (which was a proprietary algorithm created by a non-public process) then discovered weaknesses which allowed it to be broken even without a CSS key.

TCG is very different. It uses hardware to protect the keys, and no software implementation of TCG is necessary or will ever exist. And the cryptography being used is completely standard, based on RSA and other well known algorithms. The spec has been publicly available for review and comment for three years now, and some of the best cryptographers in the world have been involved.

Even if a TPM key does get pulled off a chip (which would require hundreds of thousands of dollars of reverse-engineering time and equipment, stripping it off layer by layer, using exotic devices like quantum interference based SQUID probes), the spec provides a feature to revoke a TPM key so that it will be of no use to people who want to produce fake, software-emulated TPMs.

The whole process of TPM design and development has been the opposite of the secrecy-shrouded CSS. The result should be exactly the opposite in terms of security.

Re:Software DRM

[marvin2k]

... Even if a TPM key does get pulled off a chip (which would require hundreds of thousands of dollars of reverse-engineering time and equipment, stripping it off layer by layer, using exotic devices like quantum interference based SQUID probes), the spec provides a feature to revoke a TPM key so that it will be of no use to people who want to produce fake, software-emulated TPMs. ...

Wouldn't this revoking mean that all the hardware released with that key becomes useless in the way that a DVD player with a new set of keys would no longer be able to play current DVDs? If that is the case then I don't see this ever happening because such a step would completely and utterly destroy the market.

Re:Software DRM

[NullProg]

I understand why this could be usefull to banks, military, governments, but why on earth would I want this on my workstation/server?

You could implement a poor mans version of this by using tripwire and a custom kernel module that watches the process activity.

Just curious, thanks and enjoy.

Re:Software DRM

[tehdaemon]

You missed a key point here. All TPM hardware is emulated. The actual hardware is not TPM compliant. The catch is the TPM key, as you said.

Would it be possible to rip the chip itself out of a TPM compliant system, and fake a 'boot' within the emulator well enough to fool the chip into thinking it was a real boot? The emulator would need a hardware component to interface with the chip....?

Re:Software DRM

[Alsee]

Every chip has a unique and random key. That key is certified as a genuine by having a manufacturer signature. The manufacturer signature is certified as genuine by a root TCG signature.

So each chip you rip a key out of allows you to liberate one computer. And if that is ever detected then that key goes on a revokation list.

-

Re:Software DRM

[Alsee]

It only works if you rip a key out of the genuine tamper-resistant self-destructing chip. Each key you rip out is only good for liberating a single machine. If they ever detect it then that particular key goes on a revokation list.

-

Re:Software DRM

[tehdaemon]

Hmmmm... I don't really want the key, it's the chip itself I want. Mounted on a PCI card with enough hardware so that when the card applies power to the chip, the chip thinks that all of the signals that it is getting are from the original hardware booting. And then it signs the BIOS, which is data that was read from the real one. Is that possible?

One is all it takes to rip a whole bunch of DVD's - which is what they are worried about. Or to decrypt someone's encrypted M$ office files, (with their permission of course) or decrypt a driver so I can write a linux one for that piece of hardware. Or so I can install linux in the first place. You get the idea.

Re:Software DRM

[Alsee]

To the extent of being able to attack the physical link between a Trust chip and the rest of the motherboard, yeah, that can pretty much work. Possibly somewhat easier than ripping a key from a chip, somewhat less flexible, but generally an adaquate attack. But...

Just today I came across a page on the Microsoft website about Longhorn, it appears to say that the Trust chip *will* be inside the CPU itself. It seems they plan to bypass the motherboard-based Trust modules altogether. Intel is already shipping CPU's with embedded Trust circuitry, and the other big CPU makers have projects for it.

At that point bus signals, RAM, it can all be encrypted. No cleartext to snoop anywhere except visualy from the screen or with a microphone from the speakers. The monitor itself would have a Trust chip inside to decrypt and display. The soundcard (or possibly even the speaker itself) would have it's own Trust chip to decrypt and play.

-

Re:Software DRM

[tehdaemon]

Yuck.

Re:Software DRM

[Alsee]

::Can't stop laughing::
Such a.... perfect... reply :)

-

Apparently this is not the first time...

[sczimme]

Go to the Linux Journal search function and search for 'garrick'. You should get eleven hits. I didn't read all of them, but using ctrl+f to search the pages revealed notes to Garrick re: font selection and the like. D'oh.

Re:Apparently this is not the first time...

[Grishnakh]

Sounds like this Garrick guy isn't doing his job, and should be fired.

Re:Apparently this is not the first time...

[OsoLoco]

Garrick, I think, is Garrick Antikajian, the Art Director for Linux Journal.

As sad as it is

[Anonymous Coward]

To have to burst your bubble of uninformed zealotry, there are plenty of good uses for trusted computing and DRM that do no interfere with your quest to get 'fr33 musicz 4 life' or whatever. Not all of this technology is for companies like the RIAA to protect copyrights, despite what Slashbots would have everyone think.

Re:As sad as it is

[Blue-Footed Boobie]

This is not flamebait. I hope someone with a brain mods you up.

If anything, the original parent was the flamebait...

Re:As sad as it is

[kaustik]

Yes, but somehow I got modded up and you got modded down with your childish lash out.
Stupid mods.
Seriously, though - I do agree that there are _some_ potential benefits to this. Unfortunately, the concept opens up the possiblity of DRM restrictions infecting the Linux operating system. If we continue to run on non TCPA hardware, at least we can argue that our system will not support their restrictions. ... "If you build it, they will come..."

Re:As sad as it is

[cayenne8]

" This is not flamebait. I hope someone with a brain mods you up...If anything, the original parent was the flamebait..."

In general...sure...TCPA could have some positive effects on the computing community. However, it also has great potential to be slipped in...and eventually, by law, it must be used to lock things down. Only a few things at first...but, eventually could mandate a great deal of limitations as to what you can legally do with a computer. As much as the corporate entities are beginning to use the govt. to legislate things...and they really don't like the fair use we do have...it is easily possible to forsee this as a means to that end.

Taken long enough...it could happen, which is why you need to take things like this slowly and with a great deal of skepticism early on.

I heard it said before that "What one generations tolerates....the next generation embraces"

Think of it this way...the article the other day on /. about how many US kids don't understand what the 1st amendment really means...they haven't been taught about it...and we're tolerating loss of freedoms. When they are grown and we're not around...they won't even know they existed in the old form...

Re:As sad as it is

[yuri benjamin]

What one generations tolerates....the next generation embraces

Thanks for my new sig.

Re:As sad as it is

[Anonymous Coward]

I have no problem with a Trusted Computing platform. What I do have a problem with is a platform that hides full access from ME... the owner of the computer. I want access to the stored keys -- without this, TCPA is nothing but large computer companies not allowing you to own the computer you've paid for.

Re:As sad as it is

[westlake]

I want access to the stored keys -- without this, TCPA is nothing but large computer companies not allowing you to own the computer you've paid for.

It can also be the boss not allowing others to own the computers he paid for.

Re:As sad as it is

[Anonymous Coward]

It can also be the boss not allowing others to own the computers he paid for.

Besides being bullshit (the boss no more owns a TCPA machine than I own my TCPA machine)... it's also an entirely spurious point.

*I* own my computer, not my boss, not the record companies and not IBM or Microsoft. With TCPA on my machine, with no access to the root key and the other stored keys I do not own it... or have any real control over MY property.

Re:As sad as it is

[finkployd]

So how do you propose we put hardware based public key cryptography back in the box?

Finkployd

Re:As sad as it is

[Anonymous Coward]

To have to burst your bubble of uninformed ostrich-head burying, there is no DRM method that both 1) functions and 2) permits the freedom not to use it.

DRM is based on the idea of interlocking hardware and software, where there are no loopholes simply because you don't have control of your computer (while it's in DRM mode). All software is cryptographically "trusted" if and only if it is signed by an "authority."

If you intend to use your computer to interact with the rest of the world, and the rest of the world is using DRM, you have no choice but to always use your computer in DRM mode.

(To clarify, yes, there are some DRM specs which permit centrally-signed software to interact with non-centrally-signed software -- but these specs are by definition pointless because the loophole automatically voids any "protection" the system gains from signing. The weakest link in a chain determines its maximum strength.)

Re:As sad as it is

[ShieldW0lf]

To have to burst your bubble of uninformed zealotry, there are plenty of good uses for trusted computing and DRM that do no interfere with your quest to get 'fr33 musicz 4 life' or whatever. Not all of this technology is for companies like the RIAA to protect copyrights, despite what Slashbots would have everyone think.

Bet you could use this to stop Overpeer from fucking around with your p2p network :D

Re:As sad as it is

[randallpowell]

We can't but we can fight to have it optional.

It's like what Aragorn said to the soldiers outside of the Black Gates, "I see in your eyes the same fear that would take the heart of me. There will be a day when the courage of men fail, we forsake our friends and break all bonds of fellowship, but it is not this day! An hour of wolves and shattered shields, when the age of men comes crashing down, but it is not this day! On this day we fight! By all that you hold down on this good earth, I bid you stand, men of the West."

Re:As sad as it is

[finkployd]

We can't but we can fight to have it optional.

Certainly, I don't want to see a day where TCPA is required to log onto the internet, pull up a web site, or check my email. But that does not mean I do not want the technology behind it (pki) to vanish. There is such as concept as throwing out the baby with the bathwater.

Re:As sad as it is

[Anonymous Coward]

And you'll differentiate the uses how Mr. Happy? Or will there be DRM on the DRM, allowing only specific companies to access it? Once it's in, it's in for everyone. Wake up.

Re:As sad as it is

[Anonymous Coward]

No... as I said... you *don't* own the computer if it has TCPA. Your boss doesn't own [what he thinks is] his own computer if it has TCPA. Get a clue.

Re:As sad as it is

[TheoMurpse]

I see in your eyes the same fear that would take the heart of me. There will be a day when the courage of men fail, we forsake our friends and break all bonds of fellowship, but it is not this day! An hour of wolves and shattered shields, when the age of men comes crashing down, but it is not this day! On this day we fight! By all that you hold down on this good earth, I bid you stand, men of the West.

Thanks for my new sig.

jk

Re:As sad as it is

[TheoMurpse]

Argh! I just posted this in the wrong thread. Nothing to see here, move along.

TCPA - TCG

[SiliconEntity]

It hasn't been called the Trusted Computing Platform Alliance, TCPA, for a couple of years now. It's now the Trusted Computing Group, TCG. Same technology, just a new name.

Re:TCPA - TCG

[Wesley Felter]

And after the TCG accumulates enough bad PR, I guess they'll change the name again...

TPM emulator

[nomellames]

If you want to test the IBM API, but you don't have a Trusted Platform Module, you can try using the kernel module emulator at http://tpm-emulator.berlios.de/index.html

Well.

[Kickasso]

You could send messages that self-destruct in ten seconds after being opened. That's silly, I know.

Well, security of communication is a big plus in any case. These little suckers, among other things, should be more resistant to pwnage than present day systems.

OMGBBQ

[Anonymous Coward]

It has other uses, OMGWTFBBQ!!oneon11 I can't believe it we are doomedz0r!

Take your head out of the sand dude.

Re:OMGBBQ

[rbarreira]

Uses such as?

Re:OMGBBQ

[Anonymous Coward]

Blindly follow the leader

Comment removed

[account_deleted]

Comment removed based on user account deletion

Newspeak Framing at its finest

[frankie]

If Gentoo wants to add a TCPA compatibility module, have fun. But absolutely do NOT call it "Trusted Gentoo" when its actual meaning is "Gentoo that doesn't trust YOU".

Gentoo's public communications guy needs to read some George Lakoff. It's a wonderful life, folks. Every time you use their words, a devil gets his pitchfork.

You can create modified binaries.

[Kickasso]

You just can't get them signed. Unless you somehow persuade the signing authority to review your modifications, that is.

Unsigned binaries won't be able to play play DRMed files (which you shouldn't have anyway) but should still work with your normal files. The trusted CPY will be happy to execute them, it will just refuse to flag them as trusted.

Re:You can create modified binaries.

[pyros]

Unsigned binaries won't be able to play play DRMed files

Do you think that will make the music and movie markets crash crash, causing the financial empires of the RIAA and MPAA to crumble crumble?

Re:Talking of Gentoo... (April Fools Joke)

[Anonymous Coward]

Offtopic? He's making a joke about april fools. What kind of mods do we have here?

MOD DOWN _ STUPID

[Anonymous Coward]

Can't you even read the blurb? "Published open standards" "Published open standards" "Published open standards" "Published open standards" "Published open standards" "Published open standards"

Got it yet? This means it will be in the kernel if you want to use it and it will be completely open source, fool. Illegal...please.. you're such a moron.

Re:MOD DOWN _ STUPID

[Minna Kirai]

AC: Can't you even read the blurb? "Published open standards"

I can read. I can also tell when press releases are obviously lying.

The standard is partially open, but not the most important part: the secret key that allows you to sign binaries so the OS appears trusted. Unless the hardware companies actually publish that key, they're not providing the full "standard" (specification of exactly what their hardware does), and not allowing open implementation of compatible software.

Re:MOD DOWN _ STUPID

[pe1chl]

There is no "the secret key".
You think that SSL is not a fully open standard because websites do not give you the secret key of their certificates?

A signed binary only means you can trust the binary as much as you trust the signer. Nothing more or less.
The TCPA means that publishers of data can decide that they do not hand over their data to applications that *they* don't trust. That means it can be used for DRM, but it also means it can be used for secure transactions, etc.
It does not restrict your possibilities to run modified code, but it can restrict your possibilities to do useful things with them.
That is based on the lack of trust in you. Your best defense is not to trust publishers that do not trust you.

So, when a music publisher sends out files that can only play on a TCPA DRM platform, stop buying from him. When other consumers do the same, the publisher will go bankrupt and others will step in.

Re:MOD DOWN _ STUPID

[Minna Kirai]

The TCPA means that publishers of data can decide that they do not hand over their data to applications that *they* don't trust.

Meaning that end-users cannot modify the media player or OS software that runs on the hardware they own, which is counter to the intent of the GPL.

That means it can be used for DRM, but it also means it can be used for secure transactions, etc.

Ha! "Secure transactions". Allow the person on each end of the transaction to select his own software. The bank shouldn't be able to dictate which web-browser I access my account with. If you don't trust the remote user, don't allow him into your system at all... trusting people, but not trusting their choice of software is a stupid band-aid on security.

So, when a music publisher sends out files that can only play on a TCPA DRM platform, stop buying from him. When other consumers do the same, the publisher will go bankrupt and others will step in.

That's a polyanna attitude, and is not consistent with your other statements. You seem to be hoping that somehow the main driving incentive for TCPA will not come to pass. Maybe you've been fooled by their propaganda that it somehow increases security for the end user?

If consumers do seriously boycott publishers from using TCPA DRM, then TCPA will completely die out, and this whole discussion will be meaningless. I'd be happy to see that happen, but don't percieve any rational justification for that optimism.

Re:MOD DOWN _ STUPID

[tepples]

So, when a music publisher sends out files that can only play on a TCPA DRM platform, stop buying from him.

And when the local ISPs require attestation from your PC's TPM before the DHCP server will give you an IP address, are you supposed to stop connecting to the Internet?

I know I know!!!

[Nexus7]

Yeah, let's build, you know, VOTING MACHINES, with this!!! I mean, is that a revolutionary idea or what?! At least in these parts, huh? Where wa the TCPA/TCP/fancy_acronym in '04 (for that matter, in '00) when we could've used them for what is th emost important use of them all?

Genius.

[Kickasso]

Those idiots in the TCPA alliance would never think about it. I feel safe in your hands.

say it again!

[Anonymous Coward]

Umph.. oooh

Good God ya'll..

What is it good for? Absolutely nothin'!

TCG and Linux make sense

[SiliconEntity]

Trusted Computing Group (TCG) technology makes sense in the context of Linux. Microsoft refuses to implement it. They had their own conception, which was Palladium, then NGSCB, then was dropped. So if TCG is going to go forward at all, it has to be with Linux.

It's kind of ironic, because Ross Anderson's lying Anti-TCPA FAQ tries to claim that TC exists to kill Linux. And yet it is turning out that Linux is the salvation of Trusted Computing.

There are a number of research projects in TC on Linux, including TPM Device Driver, Trusted GRUB and Secure GUI, tcgLinux, TCPA Open Source Platforms, Enforcer, and more. All Linux based.

Don't believe the FUD about TC. When implemented in Linux using Open Source software, TC gives you new options for securing and expanding the capabilities of your computer.

Re:TCG and Linux make sense

[praedor]

Hmmm. And yet I don't seem to need any form of TCPA/TCG or DRM. In all the years I've run linux full-time, I have never ever had naughty code or naughty hackers get in. I can't say that about any of the windoze users I know. Beyond that, I certainly don't need any system that can be used as a DRM system.

Nope. Uh-uh. Not on my box. I'll copy my files and CDs as I feel the need and will not have anyone but me control when and how I go on to use such copies. This all looks like what it is, an attempt by corporations to gain control of the most important and useful aspects of your PERSONAL and private property computer. Screw TCPA/TCG (and DRM). Paint it all up with lipstick and rouge all you want but in the end it is about restricting what people are allowed to do with their own computers. Any benefits that come to the individual computer owner are accidental and peripheral to the actual designed and intended purpose.

Re:TCG and Linux make sense

[Anonymous Coward]

What do you mean don't believe the FUD? What do you think your post is?

Give us one point where Ross Anderson's arguments don't hold. You will be rebutted. Go on.

Re:TCG and Linux make sense

[Anonymous Coward]

I'll trust the damn thing when they prove to me that system upgrades (or recompiles, or even module loads) can be done without locking the computer off of itself. Load the sound driver and suddenly you can't access your banking site. Wonderful...

Re:TCG and Linux make sense

[necrognome]

When implemented in Linux using Open Source software, TC gives you new options for securing and expanding the capabilities of your computer.

If the upstream routers at my ISP require me to have trusted hardware/software in order to access the Internet, how is my freedom enhanced? If national/international laws are passed requiring all ISPs to restrict my communications in such fashion, how do I choose another ISP in the "marketplace"? Furthermore, if "untrusted" hardware is similarly illegal, am I still free?

Food for thought

Re:TCG and Linux make sense

[dnadig]

Umm, this is actually entirely incorrect. Microsoft is REQUIRING TPM hardware on next generation machines (Longhorn). Far from being abandoned, NGSCB is alive and well and coming soon to an update near you.

Love it or leave it's here. Dell started shipments on a substantial portion of their line YESTERDAY. You can stick your head in the sand, your you can help provide solutions. It's up to you.

Re:TCG and Linux make sense

[abulafia]

It's kind of ironic, because Ross Anderson's lying Anti-TCPA FAQ tries to claim that TC exists to kill Linux.

I'm not fresh on TCPA at the moment. But that's a pretty harsh statement. Can you illustrate one or more of these lies? I don't know Ross personally, but I have traded mail with him, and don't know him to be a lier.

Re:TCG and Linux make sense

[zecg]

Beyond that, I certainly don't need any system that can be used as a DRM system.

The point is, you won't have to use it as a DRM system, since you won't have to use DRM-protected content and services that distribute it.

Re:TCG and Linux make sense

[Anonymous Coward]

You say that when you're faced with people expecting you to understand their (DRM-protected) Word documents. (Forget OO, it can't break the TCP wall)

Re:TCG and Linux make sense

[Anonymous Coward]

I can see some uses where Linux with DRM will be useful. Eventually, not everyone who runs Linux will own the hardware that it's running on. Think computer rentals. Not everyone needs a laptop all the time, and renting them might be an option. In that case, locking the system down to prevent modifications would be helpful both for troubleshooting and liability. You don't want your rented laptop to come back with wardriving software and kiddie porn on it, or for parts to be stolen from it and replaced with a broken/inferior part (think the hard disk). Tying the system hardware together with encryption will make it very difficult for would-be thieves or crackers to mess with the hardware or software. The only reason this model is even conceivable is because it is Linux, and not Windows on the laptop. Windows is so full of local root exploits that it's not worth it trying to lock down. Linux or OpenBSD are being secured as we speak.

Realistically, the only thing keeping media companies from enforcing DRM on everyone and mandating trusted hardware everywhere is *us* and the general public. The general public are easy enough to sway that they won't complain to their congresscritters.

Poo-pooing TCPA isn't going to help. Showing everyone what secure hardware is actually good for will help prevent misuse. It's like any tool.

Re:TCG and Linux make sense

[zecg]

It always inevitably boils down to:

JA: What if your job requires you to use non-free software?
Richard Stallman: I would quit that job.

DUMB

[Anonymous Coward]

HEY THIS IS STUPID, people post good comments AND THEN GET TROLLED?

That's no moon...

[William_Lee]

Why do I get the feeling there is going to be a thriving blackmarket in hacked BIOSs and OSs in the unfortunate event that these chips are actually deployed and implemented in a widespread manner?!

Whatever happened to the user having full control over a piece of hardware they plunked down hard earned cash on?!

Re:That's no moon...

[finkployd]

Whatever happened to the user having full control over a piece of hardware they plunked down hard earned cash on?!

Ummm, you do. You can plug in any private key you want into it and tell it to only run binaries you signed with that private key (include say, debian's private key, redhat's private key, or your own). What about TCPA leads you to believe otherwise?

Wow

[Anonymous Coward]

You couldn't be a bigger dumbass missing the point.

/me points finger and laughs at your regurgitation of stupid FUD

Amen, brother.

[Kickasso]

Amen i say.

Distributed Compution with confidential data

[Anonymous Coward]

Something like SETI@Home and Folding.com can really only be done today with publically accessible data.

With Trusted Computing, noone else can access it.

Here comes the flood??

[Reziac]

From TF WhitePaper [PDF] on IBM's site:

The "trusted" boot functions provide the ability to store in Platform Configuration Registers (PCR), hashes of configuration information throughout the boot sequence. Once booted, data (such as symmetric keys for encrypted files) can be "sealed" under a PCR. The sealed data can only be unsealed if the PCR has the same value as at the time of sealing. Thus, if an attempt is made to boot an alternative system, or a virus has backdoored the operating system, the PCR value will not match, and the unseal will fail, thus protecting the data.

At the very least, that sounds like "bye-bye multi-boot systems".

IBM also has a rebuttal to TCPA's detractors [PDF]. This one talks more about how the TCPA chip as currently designed "not been designed to resist local hardware attack, such as power analysis, RF analysis, or timing analysis." That's all well and good for the moment, and while the chip is (per the PDF) mounted on a presumably-removeable daughterboard, but how about the future? Is this how TCPA will stay, or is it the beginning of our worst fears??

At least these two whitepapers agree with most of us here on one thing -- DRM itself is stupid, for a variety of reasons.

Re:Here comes the flood??

[Anonymous Coward]

Wonder how well the PCR handles kernel upgrades.
Perhaps Trusted Longhorn SP3 will lock you out due to a different PCR after you install it on a Trusted Longhorn SP2.

Re:Here comes the flood??

[Reziac]

That's a good point. What's the difference between a "different" OS and an "upgraded" OS? A: Nothing -- either way, it still won't match the "original" OS.

And as you point out (re SP3 over SP2) -- what's to stop the OS from refusing to play nice if it doesn't encounter the PCR that it expects to see?? Might you have to provide your PCR when the OS is activated, and then you only get updates if the PCR still matches??

Can, meet worms.

Re:Here comes the flood??

[abulafia]

Once booted, data (such as symmetric keys for encrypted files) can be "sealed" under a PCR. The sealed data can only be unsealed if the PCR has the same value as at the time of sealing. Thus, if an attempt is made to boot an alternative system, or a virus has backdoored the operating system, the PCR value will not match, and the unseal will fail, thus protecting the data.

One wonders what forensics types, particulary government forensics types, have to say about this.

And I mean "have to say about this" in a very literal sense.

Re:Here comes the flood??

[Reziac]

Very useful for investigations, eh?

Tho I suppose there might be a workaround, if the machine boots to an "internal use only" OS that boots from a network, and which OS is somehow nowhere to be found when the investigation commences... (well, that may be technically infeasible, but you get what I meant).

Re:Here comes the flood??

[Greger47]

This is the thing that I don't get. The supposedly secure boot process seems to be broken from start to finish.

The "trusted" boot functions provide the ability to store in Platform Configuration Registers (PCR), hashes of configuration information throughout the boot sequence. Once booted, data (such as symmetric keys for encrypted files) can be "sealed" under a PCR. The sealed data can only be unsealed if the PCR has the same value as at the time of sealing. Thus, if an attempt is made to boot an alternative system, or a virus has backdoored the operating system, the PCR value will not match, and the unseal will fail, thus protecting the data.

The whitepaper also mentions that in IBMs implementation the chip is connected to the SMbus.

This means that the entire security of the boot process hangs on whatever data the CPU feels like sending to the chip for hashing. I could as well make a patch for GRUB that sends the "secure" version of GRUB down the SMbus and actually executes whatever nastiness I have in store.

In the case of DRM this lets me run whatever OS I want. The only thing I have to do is to feed a copy of whatever OS Hollywood trusts to the chip and voila the chip will say I'm legit and Hollywood will give me access to their movies for me to pirate at my leisure. :)

As I see it, the only way to get this to work for real is if Intel steps up and builds TCPA support into the CPU itself such that the PCR register is continuously updated as each instruction is executed. And all existing external chips have to be blacklisted, ofcourse.

Or does the TCPA system have some other trick up their sleeve that makes this work even though it's implemented externally to the CPU?

/greger

Re:Here comes the flood??

[Reziac]

Good points. Sortof like running VMWare in reverse, eh?

And it does make one wonder if a VM that's wise to the TCPA chip might be a solution to the "handcuffed" machine that Alsee (http://slashdot.org/~Alsee) often predicts as the end result of TCPA. If the CPU gets involved, perhaps the "freed" OS could run on a second non-TC CPU on an add-on card, sortof like the old way to run Windows on a Mac??

Just throwing out ideas, some of them possibly cracked. Feel free to add glue as needed. :)

Re:Here comes the flood??

[Minna Kirai]

Anybody who's taken contract law can tell you that, at best, the Gnu license is right on the bubble. In order for a contract to be binding, there has to be an overt act of acceptance.

They will be heavily in favor, at least for the narrowly-defined needs of their own professions.

Today, an end-user who happens to be the subject of a criminal investigation can encrypt his hard drive with an unguessable code. Only by directly interrogating him (or by spying as he types it, etc) can you get access to the files when you seize the PC for evidence.

But in a future with TCPA, the investigator's job is much easier. How easy it gets depending on how Orwellian we go:
1. The end-user does not have full control of his PC. Yes, he can install software to encrypt files on his disk, but the ultimate OS is owned by Microsoft. Presenting Microsoft with a court-order for the suspect's files will make them insert a backdoor into an update automatically forced onto his PC. (Users will be accumstomed to forced updated whenever an accidental exploit is found in their OS, so he won't be suspicious)

2. In a more extreme case, effective encryption software is completely outlawed, and PCs won't run amateur software that wasn't published by an official corporation. Although the suspect may have purchased encryption software, it will be weak enough for the police to easily break if they seize his PC at any time.

Either way, TCPA is good for the police, bad for the people.

Re:Here comes the flood??

[Anonymous Coward]

Shhh don't tell them till after they release it. :p
Let them think it's really good for a while.

Re:Here comes the flood??

[SiliconEntity]

This means that the entire security of the boot process hangs on whatever data the CPU feels like sending to the chip for hashing. I could as well make a patch for GRUB that sends the "secure" version of GRUB down the SMbus and actually executes whatever nastiness I have in store.

That's a clever idea, but it doesn't work. The secret is that the trusted boot process uses a concept of "trust extension". We start off with the BIOS. That takes a hash of itself and sends it to the TPM. Then the BIOS will load and run the boot loader. But - and here is the key - before running GRUB, the BIOS take a hash of GRUB and sends it to the TPM. Then it runs GRUB.

The next step is that GRUB - or at least the TPM enabled version, performs a similar process for the OS kernel. It first takes a hash of the kernel and sends it to the TPM; then it runs the kernel. And the kernel can repeat the process with the various startup scripts and other programs that loads, a la tcgLinux or the Enforcer.

The key point is that before each component is loaded, it is "measured" (i.e. its hash is reported to the TPM). So you can create a bogus GRUB or a bad kernel, but this fact will show up in the TPM's configuration registers because your bad component got its hash reported before it ran.

The one exception is the BIOS, but TPM systems are supposed to have restricted BIOS flash capabilities so you can't re-flash the part of the BIOS which does the initial hash of itself. This is part of what they call the Core Root of Trust for Measurement (CRTM) and it is supposed to be inviolable.

Re:Here comes the flood??

[Greger47]

Well, the GRUB thing was just an example. Since the chain starts at the BIOS, so will I.

My hacked BIOS keeps a copy of the hash of the trusted version and sends that to the TPM (or a copy of the entire BIOS in case the TPM wants to do its own hashing).

I don't even have to do all that work in the BIOS, the only thing I need to do is to remove the part that activates the security chip. When my OS is up and running I can run my own utility that starts the chip and feeds it a "fake" boot process with hashes of a trusted BIOS, bootloader and OS.

The point is, as long as TCPA isn't part of the CPU itself, it's hosed. It's not even enough to put the BIOS in an embedded tamperproof ROM on the CPU (a la microcontrollers), as long as the security chip is external I can interfere with the communications and feed it fake data.

/greger

Re:Here comes the flood??

[bitspotter]

TCPA/TCG is expressly and intentionally not designed to be proof against hardware attack.

The thing is, once you use a hardware attack to break a box, you've only broken that box - not any other box.

It sounds like a reasonable tradeoff: You don't have to insist on perfect security anymore when you have a good way to limit the damage a break causes.

I still doubt that it will help much. Even if you plug 90% of the holes in a dike, you still have a leak, and the exponential nature of P2P distribution doesn't seem to be affected much by whether the initial content has 10 seeds or 1000.

Re:Here comes the flood??

[Anonymous Coward]

This also seems like a nice way to create a denial of service on a machine...

Also 'trusted' can be bad. Many worms work THROUGH 'trusted' programs to elevate what they do. How does a trusted enviroment protect against the buffer overflow? It doesnt.

It will create a small time where things are 'nice and clean' but eventually someone will crack it. It is only a matter of time.

I have not seen WHAT 'trusted' computing gives me. I see it giving OTHERS a lot to do with what is on my computer. But not me.

Re:Here comes the flood??

[MenTaLguY]

...if you can get someone with authority to digitally sign your VM binary, sure.

Re:Here comes the flood??

[Alsee]

if Intel steps up and builds TCPA support into the CPU itself

From the Inquirer:
Improved architecture for Prescott [CPU] includes better pre-fetcher branch prediction, advanced power management, improvements to hyperthreading technology, the PNI above, La Grande support, better imul latency and additional WC buffers. La Grande is the security feature Intel told us about at the last IDF, and includes protection in the CPU, at the platform level, and with software.

And this story:
Addressing growing security concerns in the PC market, Intel last week also gave a glimpse of La Grande, an on-die technology that will interface with Microsoft's Palladium and other upcoming security software. "We're going to take hardware security up a notch and work with future software developers" to implement the new system, Otellini said. "La Grande is not a Pentium 4 product. It will be a next-generation architecture."

And if you'd like a look at the Trust Chip embedded inside the existing Prescott CPU itself, look to the Micrograph at the bottom of this page. The Trust system eats up about 20% of the CPU die with an entire second CPU and Trust architecture to watch the main CPU.

AMD, Transmeta, and the other CPU makers all have projects to embed the Trust system inside the CPU itself. Oh, and as the recent Slashdot story on the uber-powerful Cell Processor pointed out, it too will have on-chip DRM system. That "DRM system" is doubtless none other than Trusted Computing.

I wouldn't be supprised if motherboard-based Trust chips are pretty much obsolete by the time Microsoft's Longhorn rolls out. (Longhorn a.k.a. Palladium)

-

Re:Here comes the flood??

[Anonymous Coward]

The major function of a BIOS (any more) is apart from providing limited backwards compatible Basic Input/Output, to find from the list of bootable partitions a bootloader program, load it into memory, and start it's execution. The job of the bootloader program is to find (again from the sequential list of bootable partitions in order) the first available operating system or a user selectable choice of operating system, and load it, then start it's execution (load the operating systems starting address into the CPU's program counter register). Mucking the process up with DRM (which is a high-level function, and has no business in the BIOS), is quite stupid. It's much like finding a cure for cancer by studying electrons of atoms. DRM in itself is a patently stupid thing (although the greedies waaant it). Putting it into the BIOS is quite absurd.

Re:Here comes the flood??

[Reziac]

Yeah, that could be a problem... Someone mentioned that what *you* compile, *you* can sign, but that somehow doesn't seem broadly practical.

Re:Here comes the flood??

[slavemowgli]

Well, it doesn't matter whether you start with GRUB (or LILO or the OS/2 Boot Manager whatever) or the BIOS - you still have to start *somewhere*. As long as the chip just has to accept whatever data it is being sent, I think you will be able to get around it - and even if you give the chip (or, more generally, the TPCA subsystem) more "control" (no matter how that's realized technically), it will still have to rely on data provided by external sensors, and there simply is no way to *really* verify that data.

If you have ever watched the movie "Dark Star", you'll know what I mean (and if you haven't, shame on you! go check it out! :)).

And inviolability (is that even a word? :))... that won't happen, either. Preventing (easy) flashing of the BIOS only makes sense as long as the user does not know how to use a soldering iron. Yes, you still need the knowledge necessary to create a replacement BIOS, too, of course, but that, ultimately, won't be anything the end-user has to worry about, just like you don't have to worry about all the technical details if you just want to get Linux to run on your xbox. In fact, I'd go so far as to say that soon enough, someone would come up with a something that allows end users to patch their hardware (so to speak) without the need for even a soldering iron.

Re:Here comes the flood??

[Anonymous Coward]

fact, I'd go so far as to say that soon enough, someone would come up with a something that allows end users to patch their hardware (so to speak) without the need for even a soldering iron.

Hardware hacks would require the resources of Intel or AMD- you need to carve the CPU apart by thousandth-micron increments, recording the results for later analsys... which is where the REAL fun starts.

Or did you miss the part about how these chips will be built into the center of CPU chips, and not in any flimsy replacable BIOS?

Re:Here comes the flood??

[mattyrobinson69]

protect against buffer overflows by using NX capable hardware, or software emulation

Good for Obscene Pr0N, Terrorism, etc.

[Anonymous Coward]

If the RIAA/MPAA can hide content completely from the devices it's running on (people's home computers); more questionable organizations like terrorists and MILF porn distributors can use it to hide content from authorities.

Like any encryption tech, it's a double-edged sword.

Not so funny anymore

[Anonymous Coward]

In Soviet Union, your GPL'd software doesn't trust YOU!

Hmmm. This puts the whole concept of so-called "Trusted Computing" into a realistic, and sad, perspective.

But think of the children!

[SJasperson]

Well, someone had to say it.

Re:But think of the children!

[Anonymous Coward]

...I do... I've got a firewall and a proxy... and noone can bypass the proxy since the firewall doesn't accept connections from anything but the proxy...

Re:But think of the children!

[tepples]

and noone can bypass the proxy since the firewall doesn't accept connections from anything but the proxy...

So what happens when your firewall or proxy isn't "trusted" by your ISP?

don't make me laugh

[cg0def]

Trusted Gentoo??? Did I hear that one right? You mean trusted about 80% of it's up time when you are not experiencing configuration problems recompiling software or doing some other stuff that a modern OS should not have to do? The gentoo development team seems to forget that what gets you good reputation in the OSS worls is doing thing well and NOT doing a lot of things half arsed (excuse my french). So one suggestion for them ... fix your urgent problems (like the redign of portage that has been underway for way too long) learn C and maybe some more Python if you like it so much and maybe then you can focus on building a trusted linux distro. After all you need a stable foundation first otherwise you only have an Windoze wannabe and that's not cool anymore. (if it ever actually was). Plus if slashdot is really going to post all the sensationalism arround gentoo they might just start posting the weekly newsleters. After all it was Open Solaris last week and Trusted linux this week and who knows what next week. Not to mention that whoever comes up with those ideas in the gentoo team must be either some very very bored college students that will end up working as a sys admin one day or just a person that doesn't know anything about common sence and organization. Sorry to turn this post into a flame but gentoo is rally dead to me. Great idea with an extremely poor execution.

Re:don't make me laugh

[jrnchimera]

Hmm..I've been running Gentoo for well over a year now in a production environment with no problems.

Re:don't make me laugh

[boudie]

Works great for me on my desktop and laptop.
Maybe you should try reading the manual again.

Re:don't make me laugh

[andrewbaldwin]

You mean trusted about 80% of it's up time when you are not experiencing configuration problems recompiling software or doing some other stuff that a modern OS should not have to do?

This is not my experience - my Gentoo box goes months without rebooting (the last one was forced on me by a power cut after 87 days' continual use - including rebuilding the LAN, changing my router interface, upgrade of Apache, Samba, Open Office....) - before that it ran for nearly 6 months (again only being taken down because I had to turn off the power in my house to do some repairs.

I don't know what your problem is with portage, but for me it works like a dream - perhaps you could share your issues in more detail and then the community would be in a better position to help.

If on the other hand you've decided that Gentoo really isn't for you, well that's your choice (one of the great benefits of F/OSS) - C'est la vie!

--BTW I an not connected with Gentoo in any way (other than as a satisfied user)

Re:don't make me laugh

[Anonymous Coward]

Gentoo provdes both of ultimate configurability along with ease of administration.

When your system is highly configurable, YOU the admin become the deciding factor in the stability of your system.

I suspect that Genetoo was unstable for you because you did something to break it.

Re:don't make me laugh

[ZonaldRumzfeld]

I'm starting to believe you never used Gentoo before and just talking straight out of your ass. I guess you knew you were going to get flamed over this, but oh well.

21:05:02 up 23 days, 13:43, 2 users, load average: 0.01, 0.06, 0.05

This is just my laptop. I have a ~x86 server running Gentoo for almost a full year now and a PPC server running for half a year.

Gentoo is as good as any Slack/Debian/Fedora box, it's all the same and depends on what crap you end up putting on it. There are plenty of companies running Gentoo in the production environment with great success.

As for a TCG Gentoo.. umm, no thanks.

not entirely so

[hany]

From a practical standpoint, TCPA is incompatible with the Linux philosophy of open-source modifications

IMO this is not exactly correct - is it against Linux philosophy of open-source modifications to secure my Linux box so nobody except me can make modifications to it?

TCPA used in such way (i.e. in interest of user, not supplier, not government, ...) is quite in line with Linux philosophy of "you're in control" :) .

But, as with all weapons, it has two edges. So, beware! :)

Re:not entirely so

[dbIII]

is it against Linux philosophy of open-source modifications to secure my Linux box so nobody except me can make modifications to it?

You are confusing the views of RMS with the rest of us, most us us like to have passwords on our computers to stop others poking about. Enigma showed us that security systems tha are not subject to peer review are flawed, so the security systems should be open. Anything that can be used to stop a competitors program running will be abused in the current environment - also imagine signed spyware that you cannot remove without transferring the hard drive to another machine without DRM.

If the USA implements DRM it will be in danger of removing itself from the PC market - with extra cost and extra user hassle and no laws enforcing it the rest of the world will happily produce machines without it - and without strict enforcement people in the USA will import machines without DRM.

This is also a case of fighting last decades battles - boot sector viruses are irrelevant now, so the trusted boot option only has a real purpose of raising the cost of entry to the OS marketplace. A commercial entity is in charge of implementation and enforcement, one that does not have a reputable history, so the fees charged to get certificates/keys/whatever will not just be there to cover admin costs. High profile projects may well get the things donated, but the next Linus just starting off on their own will have a barrier to entry beyond their means on a DRM machine.

I see the whole thing as being just a resrictive sales practice built into hardware. I don't think the benefits will balance this, and I don't think it will be implemented outside the USA without trade agreement pressure. For some reason I cannot understand the MPAA have enough clout that this may happen - despite members producing as much as possible overseas and highly visable tax evasion. They don't really add much to the US economy anymore so they shouldn't be allowed to bring draconian measures to bear on those that do.

Re:not entirely so

[Anonymous Coward]

> But, as with all weapons, it has two edges.

Err, you're not supposed to sharpen *that* side of your katana... ;)

Re:not entirely so

[Alsee]

You are wrong. Trusted Computing does indeed defeat the GPL. Trusted Computing makes the source code useless. If you change a single line of the code then the Trust chip will generate a different hash or "identity" for that new program. It will then be unable to read any data for the original program and it will be unable to communicate over the internet.

to secure my Linux box so nobody except me can make modifications to it?

You don't need Trusted Computing for that. You can get all of the owner benefits and eliminate all of the possible abuses with a nearly identical system. For example the EFF has described Trusted Computing with an OwnerOverride option. You still have all of the same protections against viruses and trojand and keyloggers and hack attempts, but you retain full control over your computer. The only thing you lose is that the machine is no longer "secure" against owner "attacks". However it is impossible for anything the owner does to his own property to be an "attack".

-

Are you stupid, or are you trolling?

[Kickasso]

All features of your modified executable will work. DRM servers of media corps will refuse to give you keys for playing their content, because you won't be able to lie to them. But who told you that you have a right to get these keys by lying?

Re:Are you stupid, or are you trolling?

[Minna Kirai]

All features of your modified executable will work.

Wrong. Suppose one feature of the executable is playing blue-laser discs of Star Wars VII: Return of the Binks. That functionality will NOT WORK if the OS is flagged untrusted.

At it's core, returning true from the is_trusted() system call is the single critical feature that will be removed, crippling dozens of important programs running on that OS.

won't be able to lie to them

Lying is still, technically, a feature. I never said you had a right to do it- but yet, the GPL doesn't allow you to restrict what a recipient can do with the software you give him.

Re:Are you stupid, or are you trolling?

[Anonymous Coward]

Suppose one feature of the executable is playing blue-laser discs of Star Wars VII: Return of the Binks. That functionality will NOT WORK if the OS is flagged untrusted.

Bogus argument. That functionality wouldn't work if you weren't running it on the OS the executable was written for anyway (since the whole executable wouldn't work), and the executable would have been written for a TCPA-enabled and signed OS.

The ethernal "arms race"

[CharonX]

Security vs. those that wish to bypass the security for any reason.
Its an ethernal "arms race".
True, the TCG chip will rise the bar for running "unauthorized" software, but it might also bring its own downfall. Imagine the chip is implemented and works well for a year or two - i.e. the only ways to defeat the chip are very inefficient (and probably require doing stuff with the hardware that only few of us geeks would have to skills and guts to do)
And then, suddenly, somebody has a great idea how to decieve that chip on the software level - and it works.
The whole security scheme collapses. All the companies that neglected to put additional security measures into their software - why should they, they cost money and the chip is undefeatable - suddenly find their protection melting away like a snowball in summer.
They can't rush patches to protect their software - after all, they relied on the chip.
And the TCG can't really "patch" the chip, since "write access" to the chip would only make it even more vulnerable.
Of course, in time a new scheme is developed, and the circle begins anew.

Official Gentoo Dev Response

[codergeek42]

link to forums post

There are no formal plans to support this in any shape or form at this time.

Oh my gosh

[Anonymous Coward]

You just shattered my reality. Goodbye cruel world....

NO CARRIER

No definitions were found for ethernal

[Anonymous Coward]

Tip: Try Google Answers for help from expert researchers

No definitions were found for ethernal.

Suggestions:

- Make sure all words are spelled correctly.
- Search the Web for documents that contain "ethernal"

TCPA = DRM?

[Ride-My-Rocket]

In the immortal (paraphrased) words of Stanley Kirk Burrell, " you can't trust this ".

I don't even know where to start.

[Kickasso]

If you don't have a blue-laser disc player, your feature will not work either, GPL or no GPL. If you don't have the right hardware, your app won't work. Similarly, if you don't have the right keys to send to the right hardware, your app won't work. The hardware, as well as the keys, are not parts of the (GPLed or not) software. It's your responsibility to obtain them, and it's your choice how do you do that. (For the keys, I suggest running a signed copy.)

No.

[Kickasso]

CSS is broken because someone left the keys in the open (and they are too short to be secure anyway). If DVDCCA disallowed software players, and used 128-bit keys, there would be no deCSS. That's more or less what will happen with TPM.

Basically...

[bonch]

Basically, you're saying the GPL is anti-freedom and anti-choice? You're seriously saying it's "illegal" to have such support?

TCPA alternative

[pair-a-noyd]

I prefer the alternative system I have developed, I call it "KMFA"..

It gives me ultimate security rights and control over my machines..

Re:TCPA alternative

[slavemowgli]

KMFA? Oh, you mean KMFMS... :)

Parent comment is BOLLOCKS

[Anonymous Coward]

In this way, it is impossible for end-users to create modified binaries

The GPL requires that the source code be made available. It doesn't require that end-users be able to create modified binaries. If you disagree, please quote the GPL.

Re:Parent comment is BOLLOCKS

[Minna Kirai]

AC: It doesn't require that end-users be able to create modified binaries. If you disagree, please quote the GPL.

Fine:

1. 6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients' exercise of the rights granted herein.

Get it? "You may not impose any further restrictions" means you cannot build the hardware so it blocks users from doing what the GPL obligates you to permit them to do. Otherwise, you are breaking the GPL and have no right to give out the software at all.

Re:Parent comment is BOLLOCKS

[Anonymous Coward]

"You may not impose any further restrictions" means you cannot build the hardware so it blocks users from doing what the GPL obligates you to permit them to do.

Firstly, the people who make the hardware aren't the people distributing GPLed code. That means that the people who make the hardware don't have to abide by the GPL, and the people who distribute GPLed code aren't imposing any restrictions.

Secondly, the GPL doesn't grant the right to run the software (you already have that right), so you aren't imposing "any further restrictions on the recipients' exercise of the rights granted herein."

Finally, try and think things through before spouting off on what is legal and illegal. Otherwise you just come across as a moron, especially when you claim stupid things USING CAPS.

Re:Parent comment is BOLLOCKS

[tehdaemon]

Cannot build the hardware? Nonsense, other replies said as much. But, could you distribute the signed binary, without the key? Wouldn't that be a 'further restriction'?

I wonder.

[Kickasso]

Have you ever read the spec? I have, and I can find nothing like what you're saying in it.

The TCM is designed to tell others the truth about software you're running. Nothing more, nothing less. If you're not comfortable with it, you have the option to be silent, but you don't have an option to lie. Well, you can try, but no one will trust you. That's all there is about it.

Re:I wonder.

[Minna Kirai]

Have you ever read the spec? I have, and I can find nothing like what you're saying in it.

The spec doesn't spell it out- they wouldn't be so blatant. This is a legal/economic issue, not a technical one. It's understandable that engineering-minded readers might focus only on the technical details and miss out on what they imply for the larger environment.

Looking at the big picture, there can be only 1 or 2 real motivations for technology companies to roll out TCPA:
1: to prevent people from infringing copyright, which means preventing them from modifying their desktop OS software.
2: to prevent Linux from competing with Microsoft/Apple/Sun, which (again) means preventing them from modifying their desktop OS software.

Factor 1 is surely the driving force behind the technology, and 2 is just a pleasant side-effect for a few of the promoting companies. Sometimes they claim that it will have other benefits for end-users, like fewer hackers or virus protection or something, but those are false (meaning they could be achieved equally or better with other technical schemes)

No no.

[Kickasso]

They will go all happy happy joy joy. Or something something.

RMS's writing about "trusted" computing

[latroM]

RMS has written a nice article about it: see http://www.gnu.org/philosophy/can-you-trust.html

Re:Tee hee... published before editing was finishe

[computational super]

Garrick, please don't forget to remove my inline comments to you before you post this article.

Obligatory TCPA FAQ Link

[TheSpoom]

You should read the TCPA FAQ if you have not already. It explains why this is a bad thing.

I'll stay with the technical issues.

[Kickasso]

Thank you very much. So I figure you haven't read the spec?

Re:I'll stay with the technical issues.

[Minna Kirai]

"I'll stay with the technical issues".

There's a story about the engineers who built killing machines for Nazi Germany... it turns out they were only interested in the techincal issues. (Good for them! That makes it all A-OK)

Re:I'll stay with the technical issues.

[Anonymous Coward]

You won't find any mention of Jews and gas chambers in the chemical formula for Zyklon-B. Ok, so that's hardly the same magnitude but the principle is the same... so what's your point? I mean, other than the fact that you clearly know very well what TCPA is to be used for and are simply shilling.

Re:I'll stay with the technical issues.

[finkployd]

You must be heavily against public key cryptography as well, following this reasoning.

Finkployd

TCPA != Microsoft Palladium?

[t_allardyce]

Im not sure if most people have RTFA (hey this is slashdot) but from what im gathering here this isn't the same thing as Bill Gates' utopian "now bosses can send emails but turn off printing and forwarding", so you can stand down from attack mode.. unless im wrong, in which case kill it.

IIRC keys are unique.

[Kickasso]

One per chip, not one per maker/model.

What a load of bollocks.

[Kickasso]

"You may not impose any further restrictions" means you cannot build the hardware Dear moron, people who build the hardware don't have to follow GPL. HTH.

You're right.

[Kickasso]

Glad some people figure it out. Microsoft is opposed to TCPA (now called TCG) exactly because it's an open spec.

Heh.

[Kickasso]

There's a paper on how trusted computing can enable more secure, user-friendly P2P neetworks. It's linked to here somewhere, look it up.

Re:Heh.

[finkployd]

Like any other technology, there are good and bad uses of it. Hardware based public key cryptography (which is all this is) exists, and those worms are not going back in the can. We might as well use it for what it is good for, and fight the uses which are not in the best interests of the customer.

Finkployd

How would trusted code development work?

[sacrilicious]

If someone (Gentoo, Microsoft, whoever) releases an OS that requires central authority signature to execute code, then how would a developer iteratively build and test against the platform?

Re:How would trusted code development work?

[finkployd]

Answer: Nobody said anything about a central authority. You can tell TCPA to trust any private key you want, just like any public key system, you just need the public key. You can create a keypair yourself and tell the system to trust anything signed by it (which is pretty much the only way I can see gentoo users doing it).

The problem comes in when we start talking about remote attestation. How can I tell every service I connect to to trust software signed by my personal private key. That will never scale.

Not to mention that you can expect RIAA and MPAA services to only trust client software signed by MS or other trusted parties.

Finkployd

It's an interesting problem.

[Kickasso]

Making a source-based distro work in this setting. I think it's possible.

You start with a signed binary compiler. If the source is signed, the compiler then signs the binary. This includes the compiler's source, so you have a kind of chain of trust. Unfortunately you can't modify pretty much anything (compiler, kernel, drivers...) without breaking the chain, but at least you can adjust your precious USE flags and say -O418 -funroll-all-bloody-loops-right-now to your heart's content.

Re:It's an interesting problem.

[einhverfr]

You still run amok with the GPL in this area.

The other option is to allow for a fully functioning key authority/generation/etc. within the software. In this scenario, TCPA functions like a hardware-enforced implimentation of GPG.

Re:It's an interesting problem.

[Anonymous Coward]

I read all the comments, where is it written you can't sign yourself software that runs on your computer ?
Everybody speculates about a few certificating authorities, but where does that idea come from ?
Read TCPA specifications...

addendum

[Minna Kirai]

I've gotten a lot of negative replies to my admittedly hyperbolic post. Anyone else who's doubting the threat TCPA imposes on GPLed software such as Linux, please read this message from the author of the GPL himself. Selected highlight:

1. "Trusted computing" would make it pervasive. "Treacherous computing" is a more appropriate name, because the plan is designed to make sure your computer will systematically disobey you. In fact, it is designed to stop your computer from functioning as a general-purpose computer. Every operation may require explicit permission.

That's the core of what I'm complaining about there- the development of Linux depends on commodity PCs functioning as general purpose computers, able to do anything the end-user knows how to program. TCPA is meant to restrict certain illegal actions by making them technically impossible, and along the way it renders many more fully legal alterations (including Fair Use of copyrighted works) impossible.

Re:addendum

[Anonymous Coward]

That's the core of what I'm complaining about there

And if you had just posted that, then I would have agreed with you. But you didn't. You posted bollocks about how "Trusted Linux is ILEEGAL!!1 OMG!!" and shredded any credibility you might have had to begin with.

Mr. Godwin has a word or two for you.

[Kickasso]

And they're not pretty.

Godwin all over, again.

[Kickasso]

OK then. You can see from the chemical formula that Zyklon-B CAN kill people. TCPA simply can't be used for the nefarious purposes people ascribe to it. I don't see the potential in the spec at any rate. Maybe I'm blind, who knows? You can read the damn spec and judge for yourself. It's free and open.

By the way, I'm a Jew and I find your comment offensive. Don't ask me why, I barely could bring myself to answering this one, OK?

Re:Godwin all over, again.

[Anonymous Coward]

OK then. You can see from the chemical formula that Zyklon-B CAN kill people. TCPA simply can't be used for the nefarious purposes people ascribe to it.

Yes, it can. This issue has been raised before -- same bullshit claims that TCPA does not allow DRM. Lucky Green patented several ways of doing it just prove it. You, sir, are lying. TCPA is the hardware component of a DRM system. It provides the encryption, curtained memory and key hiding that are required for implementing DRM.

You can read the damn spec and judge for yourself. It's free and open.

I have.

By the way, I'm a Jew and I find your comment offensive. Don't ask me why, I barely could bring myself to answering this one, OK?

Cough... bullshit.

Re:Godwin all over, again.

[Minna Kirai]

Don't ask me why, I barely could bring myself to answering this one, OK?

Next time you start to post something, think back to that feeling and really focus on it... and then maybe you'll stop typing.

NO SIGNED CODE

[SiliconEntity]

I want to try to correct one of the most common and universal misconceptions about Trusted Computing: that it will only allow signed code to run. This is causing enormous confusion here, with people arguing about how that works with the GPL, who would get to sign the code, would users get to sign their own code, etc., etc.

The truth is that the TCG spec says nothing about signed code. There are no limitations in TCG that keep you from running unsigned code. There is no distinction between "secure" and "insecure" code. You can run anything you like. Signing is a complete red herring in this discussion.

I am not trying to gloss over problems or paint a false picture. The truth is that TCG does have features whose effects are somewhat like what people are worried about with signed code. The result is that TCG could be helpful for DRM, and it might make it impossible to download music from an online store without running a special application, for example. But this would not be because "you can only run signed code". Rather, it is the server that decides whether it wants to talk to you, not your computer deciding what you can and cannot run.

What's the difference? Well, if your main concern is being able to run hacked clients that will allow you to violate your user agreements, then there is no difference. You would be right to oppose Trusted Computing. It will make it harder to lie and pretend to honor an agreement, then break your word and go back on your promise.

But if your main concern is about the GPL and what software you run, there is a big difference. There are no limits on the software you can run. You can hack your Linux kernel to do whatever you want. You can disable "secure" features in the software you run. These privileges don't go away when there is a TPM chip. That should put to rest the concerns about the GPL and hopefully end the discussion about who signs what code.

If you're wondering how these two points of view can be compatible, you need to learn more about the TCG spec and the TPM chip. In a nutshell, the TPM chip, with the cooperation of the BIOS and OS software, takes a hash or fingerprint of the software configuration as the computer boots. It can then report this fingerprint to remote servers, if client software requests it. These reports are signed with an on-chip TPM key that never leaves the chip; and this chip has a certificate from the computer manufacturer, so no emulator can fake these reports (called remote attestations).

That's how it works. It's a lot more complicated than refusing to run unsigned code. What it comes down to is that software can report its configuration in a believable and, yes, trustable way. That's the real reason this is called Trusted Computing, not the lie made up by Ross Anderson. It's Trusted because you can Trust the reports from a remote system about what software it is running, and therefore what it will do.

Re:NO SIGNED CODE

[Anonymous Coward]

That's excellent exept... I don't want anyone to know what I have on my computer... and no it is not 'caus I've got a lot of illegal stuff, in fact I'm a real bore and have paid for everthing... well not really me, my employer have.
This will f.x. be a real paradise for targeted marketing and a source for knowing all voulnerabilities on a particular computer... sounds good ?? well, not for me...
What we should be interested in is identifing a person and not a computer... and only if it is needed and the person knows (and accepts) it.

Re:NO SIGNED CODE

[finkployd]

That's excellent exept... I don't want anyone to know what I have on my computer

No problem, just turn it off. However, be aware that some people (think MPAA and RIAA) will refuse to allow your media client to connect to their (promised but thus far fictional) media delivery services unless you let TCPA attest that you are running a trusted media client (wmp) on a trusted OS (windows).

Finkployd

Re:NO SIGNED CODE

[ZonaldRumzfeld]

Anything that sends data about the computer you are using is a serious breach of security and is not to be trusted, especially without the users knowledge or permission. This is more like a trojan. Trusted Computing is simply not really trustable when my personal computer & software collection data is being sent out over the network to some computer I AM supposed to trust because they said so.. mmm hmm.. I can already imagine cops pulling people over for blackmarket chips and sending them off to jail as if they were carrying drugs...

There are, on the otherhand, some good features which TCG seems like it could help in, but as Linus said, as long as it is developed for the interest and benefit of the user. But as of this moment, it's not.

So basically we are going through the 80's once again with terminals and central computers, that's just wonderful. It failed the first time and it's not going to work a second time around.

Thank you.

[Kickasso]

I was ready to start writing pretty much the same thing, now I don't have to (and I hate composing essays more than three sentences long.)

It's hard to do that.

[Kickasso]

In this scheme there's no single central authority. Everybody decides for himself what software/hardware he trusts. It's pretty hard to enact a law that makes it illegal to trust e.g. anyone but Microsoft. Such move would meet some opposition, especially in countries other than USA.

Heh.

[Kickasso]

Think windows file sharing refusing to work with a Samba client, even if they open the protocol, because the samba client was not signed by Microsoft. Or worse, IIS only allowing IE to connect to it. Or most likely, a streaming media server only allowing windows media player. Or a CounterStrike server only working with genuine CounterStrike clients (no cheats). Or an anonymous remailer/onion router only working with identical copies of itself (no logs saved anywhere in the chain). Or a BitTorrent... well, you get the idea.

Hm.

[Kickasso]

Where did you get this idea about Longhorn? Just curious. I tried to google and found nothing.

Re:Hm.

[Alsee]

Right here, documented by Microsoft itself.

hardware must also be nexus-aware, and processors must include the Secure Software Component (SSC), also known as the Trusted Platform Module (TPM)

As a matter of fact I was originally looking at a different refference page, but in finding this new page I first see they intend to require the Trust system to be in the CPU itself. And yes, Intel is already shipping Prescott CPU's with embedded Trust circuitry. I don't know if AMD or anyone else is already shipping Trust-CPUs, but they all have Trusted CPU projects.

Note that Longhorn will likely partially run on a non-Trusted CPU, but it will not fully work. And whether it is enforced by OEM windows licencing or simply by market realities, PC manufacturers are not going to be able to ship machines that are not CertifiedWindowsCompatible.

Microsoft simply declared that Trusted-compliance will be required for Certified Windows Compatibility and *poof*, hardware makers have no choice but to make their next generation of hardware Trusted Compliant.

And if some PC maker *did* make a non-TrustedCompliant machine, what's going to happen? Simple, various portions of Longhorn will bitch and scream error messages and fail to work and Microsoft's answer will be that it's the harware maker's fault, that the hardware is incompatible. That you should either ask the hardware maker to "fix" the "problem", or that you need to buy a new compatible machine.

And with every new PC being shipped Trusted Compliant, it would only take about 3-5 years for ISPs to be able to make Trusted Compliance a mandatory part of internet Terms Of Service. And it will be heralded as ISP's Blocking Viruses at the Router.

Damn, I was going to provide a BSA.ORG link to the president's Cyber Security Advisor's speech but the link is dead. He gave a speech at a Washington DC Global Tech Summit. He called on ISPs to plan on making Trusted compliance a mandatory part of their Terms Of Service, part of the plan to Secure The National Information Infrastucture against viruses and against Terrorist Attack. And the audience applauded.

If there is no public backlash against Trusted Computing then somewhere around 2009-2011 it may be impossible to get onto the internet at all unless you submit.

-

Key an Chip is stupid

[Anonymous Coward]

Everyone keeps thinking to steal a key would be hard but all you have to do is use a mathematical elimination algorithm on any CPU so keyed. Since you would have to be able to use it to sign your own binaries or to encrypt data that would inherently vary in length, you just have to use a highly structured set of binary sequences of various lengths equal to known large primes. Eventually the key will expose itself because it becomes a constant factor in whatever it did to extend the encoded prime. It is self defeating. While there are ways to avoid this, they become clear quickly and can be accounted for. CSS got away with this because there were more than one key and the content was huge and variable. The funny thing is that a quantum computer would solve this sort of problem by default. You create a box with only one way out which is they key. Shake a couple of atoms and viola. Finally has anybody thought what would happen if someone bombed the key servers. Talk about end of the world.

Why?

[Kickasso]

You still can modify the source, redistribute the changes etc. You cannot make executables with certain properties, but GPL doesn't guarantee that anyway.

Re:Why?

[Alex Belits]

GPL defines the source code as the form preferred for writing and editing the software. Not as a preferred form of editing things that can not be used after editing.

ISP monopoly

[tepples]

if th TPM is designed according to the TCG specs, you have to manually enable it in the BIOS. Once that happens, then you have to manually allow TSS to interact with it before it will do so. You can choose not to, but then you can't use that software.

So what if "that software" is the only dialer software which is compatible with the only Internet service provider(s) in your geographic area?

Re:ISP monopoly

[Anonymous Coward]

So what if "that software" is the only dialer software which is compatible with the only Internet service provider(s) in your geographic area?

It's highly unlikely for two primary reasons.

1)First there is market pressure. The general public is really not going to put up with draconian tactics from vendors for long. Attempts to squeeze the general population has often backfired.

2) For that business model to work, the ISP needs a pretty complete monopoly AND they need to have a critical mass in that area that have TPM enabled computers. Both scenarios are pretty unlikely. In addition, ISP generally want to make a profit and support calls cost money. Forcing users to have a specific *anything* leads to more support calls which drains money from the company. It's not a profitable move.

The more likely scenario is that a vendor that has a monopoly may try to lock customers into a purchase by tying access to the application data to a key protected by the TPM. For example, I might write a word processor adn require the use of a TPM to hold an encryption key so that I have to use that particular application to decrypt access the key and thus decrypt the docs. How many people would buy such a thing. Not even Microsoft has that kind of clout.

Yes, the TPM can be abused, but there are a number of factors that will inhibit that abuse.

Re:ISP monopoly

[tepples]

For that business model to work, the ISP needs a pretty complete monopoly

Cue cable companies that refuse to share their lines and phone companies that do everything they lawfully can to make it harder for competitive DSL providers to use their lines. We already have an effective duopoly in the last mile of residential Internet access.

AND they need to have a critical mass in that area that have TPM enabled computers.

Cue five years from now, when virtually every motherboard has a TPM.

You can't do that.

[Kickasso]

Without hacking hardware at any rate. The TPM verifies BIOS before it starts booting and only enables itself if the BIOS is OK. It won't enable itself *after* the boot sequence, only before. You need to whip up your soldering iron to convince the TPM to do what you want, and even then it's not easy.

Probably an easier way is to have a hacked memory module that lets you change the contents with some kind of hardware interface.

If the memory and all buses in the computer are encrypted, then you're out of luck, but this is not currently in the spec.

Re:You can't do that.

[Greger47]

Well, as far as I can tell the specification does not mandate that the security chip verifies the BIOS before the CPU boots.

The PC specific part of the spec does a cop-out and makes it the motherboard manufacturers problem to somehow keep the BIOS and security chip safe from tampering. But ofcourse, the data bus between the BIOS, CPU and security chip must be protected as well, or it's all a moot point. :)

/greger

TCPA & TGA - Bad for the spirit of Linux

[pentalive]

Whatever their name is, they want to control how I use my computer. How can LINUX and trusted computing co-exist?

I suppose you don't have to re-compile your kernel, I have not, yet.

But many of the things I install need to be compiled. I write my own programs, they need to be compiled as well! This last is the worst fear I have, for 99% of computer users never doing any programming is ok, they just surf the web, send email to each other, listen to music, watch movies. TCPC/TCG will not hurt those who do this in "Association Approved" ways.

Linux is about software inovation, for the 10% of us who write our own programs this is the end. I might as well run Windows XP

Even if I can't re-compile my kernel, who is to say that my programming project of the week does not read the sound interface to re-produce any sound being played by the speakers as an MP3? Clearly for them to be able to "Trust" our computers, we can't be allowed to write any programs.

Too bad they can't "trust" us, so they have to take our computers away from us so they can "trust" them.

First Amendment

[tepples]

When that happens, the "Conservative" Attorney General may just decide to criminalize non-DRM traffic

The Attorney General can't make U.S. laws. Only Congress can, and the Constitution expressly prohibits Congress (or any agency it creates) from denying or abridging freedom of speech or of the press. Having the Internet's root of trust lie with major incumbent publishers would introduce all sorts of free-speech issues.

Re:First Amendment

[Minna Kirai]

Only Congress can, and the Constitution expressly prohibits Congress (or any agency it creates)

Lucky for us that the Congress isn't dominated by the same political party as the President who hired that Attorney General, one that's already evidenced distaste for quaint aspects of the Bill of Rights. Why, that situation would be horrible to imagine!

Signatures

[Anonymous Coward]

The issue is not whether you have the right to sign binaries, but whether the user has the right to run unsigned ones.

There is a world of a difference between the user using a signature to verify the source is what they want to run and the computer doing so without the user's concent.

Loss of Internet access

[tepples]

The TCM is designed to tell others the truth about software you're running. Nothing more, nothing less. If you're not comfortable with it, you have the option to be silent

What if all residential Internet access providers in your geographic area suddenly decide to deny IP routing to any machine whose owner exercises the option to be silent?

Re:Loss of Internet access

[finkployd]

This is a good point, but what incentive do they have to do this? What ISP in their right mind wants to force users to use IE (and deal with all the support calls that are spyware related?) and not Firefox? Most ISPs take the position that other OSs are not supported, but they don't care if you use them, just don't call asking for help. Some even provide instructions for Linux and OSX despite not being technically supported. I don't see ISPs doing anything like this.

Content providers on the other hand....

Re:Loss of Internet access

[Minna Kirai]

What ISP in their right mind wants to force users to use IE

What ISP in their right mind wants to limit the range of software users can run, to reduce the cost of support requests? (A: nearly all of them)

More generally, what PC hardware vendor wants to limit the range of software users can execute? (A: all of them that offer complementary phone support)

Re:Loss of Internet access

[finkployd]

What ISP in their right mind wants to limit the range of software users can run, to reduce the cost of support requests? (A: nearly all of them)

In my experience they are getting more permissive, many even including instructions for Linux these days. Reducing support costs is simple, "We do not support that configuration, have a nice day". That is what they do today, and it does not seem to be much of a problem. Contrast that with the problems they have support their supported platforms, and all of the viruses, malware, and such that Windows/IE bring? Most of their calls are about this, not joe student having problems with his Linux box.

Finkployd

Loss of Internet access

[tepples]

Now take this a step further, do you think overseas PC makers are going to sell PCs that can only run windows?

Residential users can import PCs, but they can't easily import Internet access. If all residential Internet access providers in your geographic area provide DHCP service only to machines running an operating system and web browser approved by the ISP, then what use is your unsigned build of a web browser?

Trust to get an IP address?

[tepples]

Yes, if you change the binary it will not run because the signature is broken. Solution: sign it yourself with a key you generate and put that public key into your hardware.

But will your ISP's DHCP server trust that public key and give you a routable IP address?

Re:Trust to get an IP address?

[finkployd]

But will your ISP's DHCP server trust that public key and give you a routable IP address?

You mean will they trust my router, and the answer there is yes since they sold it to me (well, not true but they do sell the same model).

Fair use?

[tepples]

Your objection to trusted computing all centers around the fact that "many end-users" (i.e., you) want to steal content and traffic in stolen content.

Copyright infringement has little in common with larceny; trespass is a much better analogy for copyright infringement than theft. Just as trespass has a defense of "easements", copyright infringement has a defense of "fair use". If a publisher signs all programs that are permitted to reproduce or perform a given work, then how can the owner of a copy make meaningful fair use of the work beyond what the programs allow?

Re:Fair use?

[Leo McGarry]

Since the theft of intellectual property is covered under Title 18 of the United States Code adjacent to other crimes of property, nonsense analogies made by people who really, really want to believe that it's less of a crime than it is are kind of meaningless.

Re:Fair use?

[tepples]

nonsense analogies made by people who really, really want to believe that [criminal copyright infringement is] less of a crime than it is are kind of meaningless.

Where did I imply that criminal copyright infringement is not a crime? I made an analogy from criminal copyright infringement, which is a crime, to criminal trespass, which is also a crime. But I also maintain that copying a copyrighted work is not always copyright infringement. Given that you're familiar with the numbering of the United States Code, see 17 USC 107 through 122. An example of a fair use covered under section 107 is parody, and this often requires copying part of the target into the parodic work. See Campbell v. Acuff-Rose Music and Suntrust v. Houghton Mifflin .

And what happens when the copyright on a work under "trusted" DRM expires, which is due to start happening again in another decade and a half?

Getting an IP? (No, not intellectual property)

[tepples]

however when you begin interacting with other software on other computers across the network, be advised that they may use remote attestation to enforse that only clients (or servers) they trust will communicate with them.

And if this includes ISPs, then how is it feasible for a residential user to convince an ISP to trust a customized kernel so that the ISP's DHCP server will issue the residential user an IP address?

Re:Getting an IP? (No, not intellectual property)

[finkployd]

Heh, since you are posting this question to all of my comments I'll keep posting the answer. My kernel does not request an IP address, my router does.

Any why would they not? Where is their motivation to lock out customized kernels (if they could)?

Who owns your imagination?

[tepples]

I want do be able to reprogram my computer to do new and creative things from my own imagination.

NMPA/Harry Fox Agency owns your imagination. There exist a finite number of distinct melodies in the western musical scale, and a lot of them are taken already. The short story "Melancholy Elephants" by Spider Robinson hints at where this is headed.

Routable?

[tepples]

Tcpa lets you tell your machine to only run binaries signed by Microsoft. You can also tell it to only run binaries signed by IBM. Or you can tell it to only run binaries signed by debian. Or yourself. Or any combination. You tell it what you want it to do in this regard.

And if you want a routable IP address, your Internet Service Provider tells you what to tell your PC to do.

Re:Routable?

[finkployd]

And if you want a routable IP address, your Internet Service Provider tells you what to tell your PC to do.

Ok, seriously, enough. Go back and read my answers to you posting this point 4 times. I have not connected a PC to an ISP in over 6 years, I connect routers to ISPs. I have not seen anywhere that router manufacturers plan to support this or that ISPS want it. The fact remains that if an ISP wanted to do something like this, they could do it today by assigning each user a private key and requiring that they authenticate with it. Also in my experience ISPs are moving toward a more permissive attitude about alternative OSes and many are pushing Firefox as the browser of choice (saves their help desk calls). It is the content providers who want to know what you are running before they release movies, music, etc to you (to ensure that you are running a media player and os that supports their DRM). Honestly, the ISPs would probably prefer you to NOT be running a OS that is trusted by the content providers since that means you might be streaming a lot of movie and audio files with their connection :)

Finkployd

"Trusting" your router

[tepples]

You mean will they trust my router, and the answer there is yes

Not so fast. When that router can request TPM attestations from the machines behind it and report to the DHCP server whether each machine is "trust"-worthy, then what do you do?

Re:"Trusting" your router

[finkployd]

Not so fast. When that router can request TPM attestations from the machines behind it and report to the DHCP server whether each machine is "trust"-worthy, then what do you do?

I'll get a $50 PC, run windows on it, and use windows connection sharing to get all my other PCs across it.

Honestly, why don't you start worrying about the power company making sure I attest that anything I plug into it is signed by them, or the water company making sure I do not use homebuilt faucets? Where is anyone's motivation for this horrific vision of the future you have? What has happened or is happening that leads you to believe that this is even a remote possibility? Do you really think that in a world where most users are running Windows 98 on 5+ year old hardware, ISPs are going to suddenly require that EVERY ONE of their customers purchase new hardware and software? Not a snowball's chance in hell, my friend. For what? Why the hell do they care? I have never seen any indication that ISPs give two shits about TCPA. Content providers are the ones pushing this and you can bet that they will have these restrictions.

Besides, the window of oppertunity to do this has passed. OS X and Linux are becoming to prevelent to ignore. Firefox is certainly too prevelent to ignore. The days when you could safely count on all of your customers running MS software have passed, and I do not see that trend changing unless Longhorn is a lot more impressive than it sounds like it will be.

Finkployd

Re:"Trusting" your router

[tepples]

I'll get a $50 PC, run windows on it, and use windows connection sharing to get all my other PCs across it.

And watch Internet Connection Sharing (Trusted Computing Edition) gain the ability to request TPM attestation of each computer it gives an IP to.

Where is anyone's motivation for this horrific vision of the future you have?

Apparently, Cisco's motivation is to control the spread of viruses at the router .

Do you really think that in a world where most users are running Windows 98 on 5+ year old hardware, ISPs are going to suddenly require that EVERY ONE of their customers purchase new hardware and software?

ISPs that buy into this Cisco scam will wait until virtually all 5-year old hardware has a TPM before throwing the big red switch on the Cisco routers to turn on "trusted" DHCP.

Besides, the window of oppertunity to do this has passed. OS X and Linux are becoming to prevelent to ignore.

Apple Computer, which maintains the kernel and network stack of Mac OS X, is a big enough company to ensure that most major ISPs would sign Apple's implementation. Small-time Linux distributors and especially individual residential users won't have that kind of clout.

OK, I'll concentrate the discussion in fewer threads.

Re:"Trusting" your router

[finkployd]

The government always has plans and bills floating around to severly curb rights, or to enforce everyone use the same technology. Since the Clipper initiative, you never really hear about them because none of them ever go anywhere. Seriously, start following the technology related bills that are introduced into legislation. Most of then never even really get out of committee. Despite the occational corporate sponsored legislation, the position of the US government has long been to not get involved AT ALL in this kind of thing, and when they do it is almost always a useless token gesture (CAN-SPAM for example) with no teeth that nobody takes seriously.

The fear that since public key cryptography is now being supported in hardware, the government will usher in a new era of draconian control over networks, IPSs, and personal computers is not only unfounded, it is without precident. They failed miserably the last time they tried to enforce something on the tech market and population (Clipper), and they actually had decent arguments backing that up. What is the argument here? Protect people from viruses? They couldn't even convince people that their rules would protect them from real evils like pedophiles, terrorists, and the like.

I'm not suggesting that we turn a blind eye to these possibilities, education and vigilance are still just as important to keep things like enforced software monoculture, drm, and technology stifiling laws in check. Most everyone has heard of Firefox and Linux and a great many people (technical or not) consider these to be good, positive things. It is easy to explain that without open standards, open access, and ability for "hobby" style developement, these things would not have come about. Failing that, it is easy to argue that this kind of restriction is incredibly anti-small business and NOBODY in government wants that label applied to them.

TCPA enables stuff like this but so does PKI, so does TCP/IP, so do computers in general. The trick is to learn to fight for the good uses of technology and against the bad uses. attacking TCPA is akin to attacking nmap because it can be used by evil hackers. In either case, the technology exists, and it is impossible to close the can of worms. You might as well outlaw PKI (the REAL enabler of DRM) while you are at it.

Finkployd

Re:"Trusting" your router

[Anonymous Coward]

The fear that since public key cryptography is now being supported in hardware,

That's a really misleading statement. PKI in hardware would be a mere "cryptography co-processor", a minor performance boost to things that can otherwise be handled in software.

But TCPA is hardware enforcement of signing/"attestation", which is far different and has no plausible use except to remove control from the end-user.

TCPA enables stuff like this but so does PKI, so does TCP/IP, so do computers in general. The trick is to learn to fight for the good uses of technology and against the bad uses. attacking TCPA is akin to attacking nmap because it can be used by evil hackers. In

All those other technologies have roughly good uses. TCPA does nothing aside from enable DRM. Others have at least tried to list off some helpful uses it could have, although upon examination all of them proved to be spurious. You haven't even tried that.

You say it'll be appropriate to protest TCPA when it reaches the point of harming rights. Well, since that's the only thing it can do, then that point is already here.

Re:"Trusting" your router

[finkployd]

That's a really misleading statement. PKI in hardware would be a mere "cryptography co-processor", a minor performance boost to things that can otherwise be handled in software.

No, that is not PKI in hardware, do you even know what PKI is? Think about the "I" in that acronym for a second. What you are describing is a processor designed for specific math functions.

But TCPA is hardware enforcement of signing/"attestation", which is far different and has no plausible use except to remove control from the end-user.

Signing binaries gives more control to the end user, in that they can specify exactly what binaries to trust.

Attestation is a good "take power away from the user" argument, but that power is really only the power to falsly represent what is running. I am a huge fan of Samba, but let's face it, Samba's purpose in life is to trick a windows machine into thinking that it is talking to another Windows machine. So while I am totally against remote attestation for the reason that it kills interop (hell, it kills faking your browser string), it actually does not limit your control so much as it adds control to the other end. I'm sure a decent argument could be made that the other end is where it should be (think online games or something of that nature), but I place too high a value on interoperability to like that idea. So I am against remote attestation. Unfortunately the technology is there, so people will use it. Education of the population is probably the best recourse for this.

For that matter, I have heard a lot of arguements against TCPA (some legit, some so ignorant I have to wonder at the /. readership these days), but I have not heard a single suggestion as to what to do about it. Do we tortch IBM HQ and hope nobody ever thinks of the idea again? Lobby congress to outlaw that specific technology, and hope they never get around to outlawing other "good" computer technology while they are at it? Organize protests? What? Seriously?

All those other technologies have roughly good uses. TCPA does nothing aside from enable DRM.

TCPA existed long before DRM was ever unleased on the public, IBM has had early concepts of it in thinkpads for years now. The idea of holding a public key in hardware and having an OS that only ran binaries signed by the private key is not new, nor is it a result of DRM desires.

Others have at least tried to list off some helpful uses it could have, although upon examination all of them proved to be spurious. You haven't even tried that.

Uh huh, and those are all uses for remote attestation. I don't like that, never have. TCPA is not only remote attestation.

You say it'll be appropriate to protest TCPA when it reaches the point of harming rights. Well, since that's the only thing it can do, then that point is already here.

No, I said it would be appropriate to protest the harming of rights. I don't care what technology you use to do it. Most DRM harms fair use rights, but I don't see too many people protesting x.509.

Clearly you have not read much except for some uninformed comments on this site (and possibly others). Try actually reading the spec. Look at the Linux distribution IBM put out that supports it and see the uses it has. You might be surprised.
I suppose it is also possible you are confusing TCPA with Paladium, which many do. MS HATES TCPA because it gives the users the choice of what public keys to put in the hardware and trust.

Finkployd

TCPA is already available

[smartsaga]

Its called a MAC!!!!!!!!

Of course TCPA allows DRM.

[Kickasso]

Along with a host of other things. I'm talking about the "kill everything but Microsoft" kind of things.

Re:Of course TCPA allows DRM.

[Anonymous Coward]

Along with a host of other things.

The reason TCPA is being pushed into consumer PCs is D.R.M -- you want to use it on military equipment, that's fine by me.

The TCPA specification was designed with DRM in mind... naturally it doesn't mention it by name since it is hugely controversial and represents the total loss of any consumer rights. But nevertheless, it is the primary reason it exists.

Your bullshit about it having other uses is transparent shilling. You can use a land-mine to open a door if you like, however no-one but a liar and shill would claim they weren't designed to kill people. Have you posted about "medical records" yet, and how TCPA will magically guard against them being stolen by evil hackers?

I'm talking about the "kill everything but Microsoft" kind of things.

You're talking out of your ass, that's what you are doing.

Re:Of course TCPA allows DRM.

[Anonymous Coward]

Your bullshit about it having other uses is transparent shilling. You can use a land-mine to open a door if you like, however no-one but a liar and shill would claim they weren't designed to kill people.

And BTW as a follow-up to my post. Before you reply... yes... land-mines are often designed to maim not kill. Don't think you can hang your entire reply on a minor point.

Well.

[Kickasso]

They can do something like that now if they want, sort of. It's not like they can't implement a closed proprietary protocol and force everyone to use it. You of course can hack their implementation, but how many people would do that?

It's not exactly the same, but you get the idea.

Besides, I have no idea why any ISP would want to do that. What does it brings them, except the need to replace lots of hardware and software?

Cisco's router won't let you see that thong

[tepples]

This is a good point, but what incentive do [ISPs] have to [require TPM attestation in the terms of service]?

Go look up Alsee's posting history in Slashdot articles about "trusted" computing to learn what incentives they have. For one thing, ISPs such as AOL or RoadRunner may be owned by a publisher. For another, some vendors claim that TC would help stop viruses. Alsee explains Cisco's plan, which was endorsed by the U.S. President's cyber-security advisor.

Re:Cisco's router won't let you see that thong

[finkployd]

The problem is this will never fly. Universities, research institution, the general public, nobody will stand for this. Simple economics show how dumb this idea would be. You think there is any force in government that will get everyone to upgrade their PC and OS? You think that OS X (who may or may not be on board with this kind of thing), Linux, etc have not reached a point yet where they cannot be ignored? Doing something this incredibly stupid would put the US so far behind the rest of the world that it would never happen for that reason alone. Not to mention the infighting and bickering between companies each trying to make "their standard" the one true way.

If it gets this bad, I'll leave the US. Most programmers and computer professionals (including companies) will probably be with me. It would be a repeat of the Clipper Chip threat all over again.

Or I will just lease my own T1 line, those are getting cheaper and have none of the restrictions that cable and dsl are starting to come with. Speaking off, cable companies are already learning that they lose customers when they get too restrictive, which is why bandwith caps have been opening up lately. Ports are still being blocked but that is primarily because of all of the idiots running "trusted MS software" that gets owned in a couple of seconds unless patched (and even then).

Finkployd

Re:Cisco's router won't let you see that thong

[tepples]

You think there is any force in government that will get everyone to upgrade their PC and OS?

It doesn't have to be forced by the government. People will upgrade their PCs when their old PCs become too slow to run newer 3D video games, and these new PCs will just happen to have a TPM on them.

You think that [Mac] OS X (who may or may not be on board with this kind of thing), Linux, etc have not reached a point yet where they cannot be ignored?

Apple Computer and vendors of major commercial GNU/Linux distros have the clout to get their kernel builds signed by major ISPs. Vendors of non-commercial distros often do not.

If it gets this bad, I'll leave the US.

Watch Cisco start selling these "trusted" routers in other English-speaking countries as well.

Or I will just lease my own T1 line

Doesn't a T1 cost at least an order of magnitude more than cable or DSL? How much money do you think residential users are willing to spend on their Internet access again?

Re:Cisco's router won't let you see that thong

[finkployd]

It doesn't have to be forced by the government. People will upgrade their PCs when their old PCs become too slow to run newer 3D video games, and these new PCs will just happen to have a TPM on them.

Video gamers are a minority, gamers who play on PCs rather than dedicated consoles are moreso. Most people want a PC that can run office, web browser, and email. Which is why less and less people are upgrading. I still see a TON of sub 1GHz machines around.

Apple Computer and vendors of major commercial GNU/Linux distros have the clout to get their kernel builds signed by major ISPs. Vendors of non-commercial distros often do not.

The Linux distro that does that will suffer MASSIVE backlash from the community that writes the software that makes them money.

Doesn't a T1 cost at least an order of magnitude more than cable or DSL? How much money do you think residential users are willing to spend on their Internet access again?

The price is dropping, I imagine it will get close enough in a few years.

Finkployd

I think it's this part.

[Kickasso]

The CRTM may measure itself to PCR[0] and must measure to PCR[0] any portion of the POST BIOS, including Manufacturer Controlled Embedded Option ROMs, firmware, etc. that are provided as part of the Motherboard.

And subsequent sections.

You do have this right.

[Kickasso]

No one prmised you an ability though.

You know.

[Kickasso]

Things that are endorsed by the USA government are not necessarily things that are viewed in a positive light elsewhere. This will probably cut off USA from the rest of the world, because the rest of the world doesn't necessarily want foreign corporations to effectively own their networks.

Cisco: 'To stop the spread of viruses'

[tepples]

I have no idea why any ISP would want to do that. What does it brings them, except the need to replace lots of hardware and software?

Let Alsee explain.

"Trusted" computing for homeland "security"

[tepples]

What makes you think they'll require it? Seriously, what do they get out of it, other than wasting bandwidth signing your damn OS?

How about some homeland security, according to Cisco?

And thats totally aisde from the point that there are multimillions of dollars of sparcs, powerpcs, Irix machines, IBM mainframes, and more out there that will still need internet connections.

And let them pay business rates, which will be much higher than the residential rates and which will not be available for connections that terminate on residential-zoned land.

That's OK.

[Kickasso]

It will only connect to other applications approved by the same people who signed your own app.

That's OK, I only want to connect to BitTorrent clients signed by Red Hat and Gentoo, and maybe a couple of others. No problem with that.

no representative of the RIAA will install it Why the fuck would I care? If a representative of the RIAA would install a client that purports to be a genuine BT client but in fact logs trafic and spoofs content, then I would be concerned. Thankfully, with a TCG platform I can ignore such destructive clients.

Why exactly would a "bad-guy" knowingly and willingly execute software that does your will?

If he wants me to connect to his computer, that's his only choice. I he doesn't, he may run what he damn pleases.

Cisco wants to sell more stuff.

[Kickasso]

ISPs don't necessarily want to buy more stuff. Besides, as I said elsewhere, this will cut the USA off much of the rest of the world. The world is not eager to be pwned by Cisco, even if the U.S. of A. is.

When you can't get an IP

[tepples]

When implemented in Linux using Open Source software, TC gives you new options for securing and expanding the capabilities of your computer.

It also gives ISPs new options for "securing" and restricting the capabilities of their networks. Alsee explains Cisco's plans to have ISPs' routers require TPM authentication before routing traffic.

Cisco is your central authority

[tepples]

In this scheme there's no single central authority. Everybody decides for himself what software/hardware he trusts.

And when a leading router maker adds a feature requiring TPM attestation in order to access the Internet, and all local ISPs have turned it on, then how will you connect?

Re:Cisco is your central authority

[Anonymous Coward]

then how will you connect?

Via a masquerading box ?

NO INTERNET CONNECTION

[tepples]

Rather, it is the server that decides whether it wants to talk to you, not your computer deciding what you can and cannot run.

And when the server refusing to talk to your TPM-less network is your ISP's DHCP server, then what?

Then I simply won't connect.

[Kickasso]

There's life outside of the Internet, you know.

Fortunately it's not really feasible. Cisco is unable to attest every DSL router/modem, every toaster, every webcam, every fridge, every... you get the idea. And when they don't, and most of these little appliances stop working, there will be an outcry.

Re:Aged 83 pope john paul dies!

[randallpowell]

I, for one, welcome our new Tibetan Buddhist overlords!

Burned by XP Encryption

[Anonymous Coward]

I used the Encryption options on some personal files (the most valuable) on my local Windows XP installation.

Ugh. Some time later I had to do a reinstall of XP. Guess what files where now inaccessible?

Like with most everything Microsoft, I could find no help to fix this problem on the net so I gave up.

Encryption? Nope, not that desperate. Especially when it's just a luring context menu-options, and I don't get to have a key or ANYTHING TO GET IT BACK!!

Remember - spyware is legit so will get signed.

[dbIII]

MPAA and RIAA will be wanting to make sure signed software obeys their rules

The rules being that a fee is paid - you really can't expect them to go any furthur than this when some of their members boast of breaking musicians kneecaps or tax evasion.

Handing over your computer to the cretins that gave us DVD regions and then couldn't even implement their encryption properly is a bad move. Someone will come up with a good scheme and then the manufacturers will use a key of all ones or something similarly stupid.

What hash function is used?

[Anonymous Coward]

I admit being lazy; could read it myself, but I really don't care about drm/tc on linux.

Do the specs specify the hash function? Is is a well known one? If it suffers the same problems as md5, it may be possible to create an 'untrusted' grub/bios/... with a trusted signature.

Though this might take a while, if something like sha1 is used.

Without hacking the hardware ...

[Joseph_Daniel_Zukige]

Isn't that the point?

Have soldering iron, will travel.

A hardware hacked machine on a "trusted" network could be impossible to spot. Pretty soon, everyone will be trusting the hardware and MicroSoft will be cutting its corners again, and viral traffic will become part of the overhead and no one will be able to tell.

This kind of thing has to be done at the gate level, which is part of the reason DNA seems overly complex. (And even there, where the "main processing unit" has the ability to override the protections, humans start taking their clothes off in strange places and all the protections break down.)

Computing hardware has become way too complex. It's time to start over, and use some sense as we build this time.

Perfect security is impossible, but each level of the machine should have its own ability to squelch spurious and malignant traffic. (Have to squelch spurious and malignant traffic separately, which is why current ideas for DRM can't work.)

sealed bearings, sealed batteries, sealed CPUs

[Joseph_Daniel_Zukige]

As long as there is some part of the OS in ROM that is not physically part of the CPU chip (no chip sets), a good soldering iron is going to take all this down.

So then they think they'll make laws against ordinary scrubs having good soldering irons. Doesn't work with locks, doesn't work with medicine, doesn't work with automobiles, ...

You can't lock the world down and get anything done.

TCPA...

[demon_2k]

Let me get this straight. Does that mean that i'll have to run binary releases of the kernel in orber for it to be trusted?

What if i compile my own kernel?
What if i change it (Bug fixes, clean ups)?
What if i patch it?

Will the kernel still be trusted?

You make big leaps.

[Kickasso]

OK, so every new PC mobo in existence will have a TPM. Do you know that there are lots more things connected to the Internet? Like, um, your DSL router and your IP-enabled webcam? In the future there will be only more. It is impossible to require that all this is replaced, even by residential users, and it is impossible to certify all software that these things run.

What's more, there are tens of millions of old computers everywhere around the world that their owners can't afford to replace. You in the US may have enough disposable income, but that's not necessarily true elsewhere.

Your pres's advisor is full of shit. You can Secure The National Information Infrastucture against viruses and against Terrorist Attack, but that will require sealing your virtual borders. Suddenly, e.g. US universities will not be able to talk to their counterparts in India and Russia and Brasil. Like it's going to happen.

Re:You make big leaps.

[Alsee]

They have already begun the process of rolling it into things like routers and wireless access points etc.

In the computer world any hardware that is more than 4 years old is fairly archaic. If your ISP installs these new Trusted Routers so they can Block Viruses at the Router and your 4 year old webcam doesn't work, oh well, too bad.

it is impossible to certify all software that these things run

The webcam maker will probably have to pay and sign some sort of contract, but he'll be pretty much responsibe for his own webcam and software. And if any sort of issue is ever discovered with his cam and/or software then HIS MANUFACTURER KEY CAN BE REVOKED. All units of his product just drop dead. He has his neck on the line for ensuring compliance.

You in the US may have enough disposable income, but that's not necessarily true elsewhere.

Haven't you been reading Slashdot? There have been REPEATED stories about making uber-cheap computers for the third world. And if you dig into the specs - which I have done - you find that they are trusted compliant.

In fact I think you have the US/Global scheme backwards. I think they want to dump a few hundred million sub-$100 units on the global market place to help press Trust adoption in the US and Europe.

As for the US sealing virtual borders, in some ways it seems the EU is even more gung-ho on it than the US. There is signifigant work going on in the UN. I've found several refferences from several UN workgroups, but it tends to be very vague and behind-closed-doors work.

I'm going to speculate a little here, but I've seen come clues that every country will be offered the chance to run their own "Root of Trust". Of course computers in one country (like the US) will not Trust computers in some other country if the second Root of Trust does not have "acceptable" enforcement rules (i.e. enforcing DRM). Most countries would jump at the chance to run their own Root of Trust and have that power over all computers in their own country and over it's own population. Think "the great firewall of China" on a grand scale. Each country can spy on it's own citizens or enforce news filtering or hate speech filtering or porn filtering, and don't forget KIDDY PORN filtering.

And why was it that countries aren't going to jump on board? You don't want trusted Computing? What are you? Some sort of pedophile??? We much protect the children! And fight the terrorists!

Another likely phase-in route is full switchover within compliant countries and specified gateways to non-compliant countries. Non-compliant countries would be under increasing pressure to comply, possibly with a cut-off date.

I also predict a huge push to go from widespread-but-non-mandatory adoption to mandatory use will come in the wake of a particularly nasty virus.

I'm not sure if I'm quite cynical enough to think someone behind the Trusted Computing movment would actually stoop to creating and releasing an exceptionally nasty virus in order to drive crossover. But I have no doubt they would milk any viruses that do crop "naturally" up for all they're worth.

-

The chip is designed to enable remote attestation.

[Kickasso]

If you can extract the keys, there's no remote attestation anymore and no reason for the chip to exist.

Re:The chip is designed to enable remote attestati

[tehdaemon]

No, there is no reason for the *AA's to push it.

Remote attestation still works as long as no program, running on the computer, can get the keys, regardless of permissions. That way, for example, when some script kiddy roots my banks webserver, I can ask it, are you running only software the bank signed? And it can say NO! And I won't give that server my bank account info.

On the other hand, the bank could sign whatever binaries it wanted for it's own machine, or hack the OS that it was running, and sign the changes, so the NSA wouldn't know that it's backdoor had been ripped out....

You don't get it.

[Kickasso]

Remote attestation means that I can reliably know what software you run. NOT that you can convince me that YOU know what software you run. The latter is nearly useless (you might be intentionally running malicious software, but I have no way of knowing it) and can be achieved without any special hardware anyway.

Re:You don't get it.

[tehdaemon]

Yes I do get it. Remote attestation, as you define it is what most people/groups pushing trusted computing want, and if what I proposed is implimented, many of the things the pushers want will no longer work. I know that. Almost all of the things that will no longer work are things that I do not want to work. It would be useless for those purposes. This does not mean that remote attestation, as I proposed, is useless. And I would say it is much more usefull to me. I get the feeling we are arguing about semantics though. Is a tomato a fruit? And does the tomato care?

I disagree that what I proposed can be achieved without special hardware. If it could, then remote attestation as you described could also, because the only difference is who signed the binary. (different from which key was used to do the signing).

Read the GPL. All of it.

[Kickasso]

There's no guarantee anywhere that you shall be actually able to use what you get in any way whatsoever. It's even printed IN BIG CAPITAL LETTERS THAT THERE IS NO SUCH GUARANTEE. All you have is some source code, and a promise of the copyright holder not to sue you if you abide by his terms. Got it?

Re:Read the GPL. All of it.

[Alex Belits]

No.

GPL'ed software is distributed without any guarantees, however this is completely unrelated to all other clauses of the license. Therefore:

1. If the binary is distributed, the distributor MUST provide the source.

2. The source is defined as preferred form for modification, that can be compiled into binary.

3. The "signed" source, if modified, does not make a usable binary, NO MATTER WHAT the recipient does, therefore it is not a source code. Lack of warranty is absolutely unrelated to this -- the code that can not possibly produce a usable binary , if modified, is not any more "preferred form" for modification than a hex dump. Actually, a hex dump of a non-"trusted" executable is a far more preferrable form for modification because its modification sometimes does produce an improved product, and a modification of "trusted" source invariably produces an unusable product.

Therefore the requirement for signing makes it impossible to distribute the source code, unless the source code includes the private key.

Or use Firewire

[AYeomans]

The Vulnwatch alert shows how a Firewire port can directly access system memory, without needing a soldering iron or undoing the case.

Heh, thanks.

[Kickasso]

It will most certainly be plugged in the next revision, but it's nice to know anyway.

Here's the clue-by-four on TCPA...

[Kjella]

to secure my Linux box so nobody except me can make modifications to it?

Ask the manufacturer of your TCPA computer if you can have the root key of your machine, printed on a nice paper card. That could not possibly endanger the security of your machine, right? Answer: You can't. They're forbidden from doing it. That, beyond a shadow of a doubt, proves that the TCPA is made to keep you from doing modifications to your machine.

Gosh.

[Kickasso]

that can be compiled into binary
You're making things up, don't you? It is written where exactly?

The "signed" source, if modified, does not make a usable binary
Of course it does. You can sign it, run it, sell it, bake it, drink it, do whatever you want with it. What you can't do is convincingly misrepresent it to be something else than it is. Guess what, no one ever promised you such an ability, now or in the future.

Re:Gosh.

[Alex Belits]

You're making things up, don't you? It is written where exactly?

In GPL:

The source code for a work means the preferred form of the work for making modifications to it.

No one "prefers" to make modifications that do not produce a usable executable.

Of course it does. You can sign it, run it, sell it, bake it, drink it, do whatever you want with it. What you can't do is convincingly misrepresent it to be something else than it is. Guess what, no one ever promised you such an ability, now or in the future.

Again, "preferred for modifications". See above -- no one cares what else can be done with it. No one guarantees that software, before or after modifications, will do anything in particular, and this is absolutely unrelated to the definition of "preferrable", that means that a software user should be able to produce a modified software from it. It may work better, worse, or not work at all, but if it's guaranteed to break after modification, it's not merely something that lack of warranty can cover. If you can guarantee that modified software will not work, this is no longer a preferrable form for modification, so your argument is invalid.

BTW, I have never seen this amount of stupidity expressed by a person without clear understanding how stupid it is, and a desperate desire to get something that does not belong to him. You are either extremely dumb, or a shill for DRM pushers.

TCPA is made to prevent you telling LIES

[Shirotae]

Those who have bothered to read the spec know TCPA is intended to prevent you (or J Random Cracker) from telling lies about what your machine is running. You can make whatever modifications you like, but you can't then get a TPM to swear that you have not made those modifications. If some cracker has broken into your machine, TPM will not swear to you that you are running the unmodified code you downloaded from someone you have decided to trust. There is nothing to stop you building your own system from (modified) source, then using TCPA to assure yourself that you are running an unhacked copy of your own code.

If you ask your TCPA computer manufacturer to let you sign things in their name, what do you expect them to say? Any trustworthy manufacturer must say no. There is no problem wiping the manufacturer's key and installing one of your own so that you can sign in your own name.

People with a clue will see an interesting challenge in creating a system where you can preserve the chain of trust where you want, but still have freedom to do anything that does not break that chain. It is an opportunity to put control of the trust relationship in the hands of the people directly involved, rather than in the hands of some monolithic and corrupt software supplier.

Can this technology be used to the detriment of users? Of course it can. Is that all it can be used for? Absolutely not. Is controlling the user the main goal? Absolutely not. Will it make it harder for me to lie and cheat? Yes. What's the point of using Linux if it will not let me lie and cheat? Supply your own answer here

The things you talk about are scary.

[Kickasso]

However I believe thay are totally impossible, for variety of reasons.

Re:The things you talk about are scary.

[Minna Kirai]

However I believe thay are totally impossible, for variety of reasons.

If you believe they're impossible, then you must also believe that Trusted Computing will never become an effective obstacle to widespred copyright infringement. But if that's true, then why are the content-publishing companies pushing for something that will never actually work?

Re:The things you talk about are scary.

[Alsee]

You may not beleive them, but there are multibillions - yes billions with a B - being spent to make it happen. So obviously some very powerful people do believe it.

Microsoft tossed off a half billion just to scoop up some needed Trusted Computing patents. The US government is throwing around a few hundred million per year, mostly on grants and research projects - not even rolling out any hardware yet. Intel, AMD, Transmeta, National Semiconductor, and others are each spending god-knows-how-much on chip designs and plants. There are over 200 members of the Trusted Computing Group, and they include pretty much every computer-related company on the Forbes 500.

AMD - Secure Execution Mode (SEM)
Transmeta - Transmeta Security eXtensions (TSX)
Intel - LaGrande (LT)
IBM - Embedded Security Subsystem 2.0
IBM - ThinkVantage Technology
National Semiconductor - SafeKeeper (CPU)
National Semiconductor - Super I/O (Networking subsystem and detwork devices)
HP - ProtectTools
Via Technologies - Padlock
Phoenix - Core Managed Environment (cME)
nVidia - ActiveArmor
Fujitsu - FirstWare Vault
Cisco - Network Admission Control (NAC)

The Cell Processor will be used in the Playstation3 and in TVs and TV-devices (broadcast flag)

Samsung is the first PC manufacturer to announce they are shipping nothing but Trusted Compliant systems.

They can and WILL spend billions advertizing it as a GoodThing.

There are existing Whitehouse directives on the waiting on the back burner to mandate all government computers be Trusted Compliant, and directives/laws to make it mandatory in general. However there's no point implementing them until production actually comes on-line and Longhorn rolls out.

And remember, it's all about phasing it in. They fully expect it to take several years. You're only going to see the genuine benefits for maybe the first 2 years. Then those who don't have it start getting screwed.

I seriously HOPE Trusted Computing dies, but the only thing that's going to stop it is if the mainstream media picks up on the story and there's a massive public backlash against it. Billions of dollars. The Windows Monopoly. International government support, serious potential for government mandates. Advertizing. They have the patients to wait years for the rollout process. And they will simply be HANDING IT OUT to every ignorant computer buyer. Aunt Clueless replaces her 3-year-old obsolete computer and she will simply be handed the only thing on the store shelf - a Longhorn Trusted Compliant system. After 3 or 4 years most PC's have been replaced as obsolete - all replaced with Trusted Compliant systems. All ugly ugly ugly.

-

Re:The things you talk about are scary.

[Alsee]

why are the content-publishing companies pushing for something that will never actually work?

(A) They are stupid.
(B) They are self-deluded, wanting to believe it so badly.

Of course I'm strictly referring to the content-publishing companies' mentality. We could all easily end up in Trusted Computing chains even after Trusted Computing flops for protecting music and movies.

-

Re:The things you talk about are scary.

[Minna Kirai]

There's also:

(C) They are stooges for Microsoft, who is using TCPA as a lever to push for the illegality of Linux, their deadliest competitor.

TCPA will be at best a distraction of open-source programming effort that could go towards something useful. At worse, open source OSes will, because an end-user can theoretically remove any DRM features, be classified as tools of unamerican hackers/terrorists, and be criminalized.

Re:The things you talk about are scary.

[Alsee]

open source ... an end-user can theoretically remove any DRM features

The point of Trusted Computing is that they can give you the source code and it doesn't matter. You're perfectly free to "remove any DRM features" or modify it in any way you like. The catch is that the modified software can no longer read the encrypted files and is no longer accepted over the internet. The modified software will "run" just fine, but it won't work.

Any peice of software is it's own signature. If you change the software you change the signature. If you change the signature then the Trust ship gives the software a completely different set of encryption keys. Encryption keys which are useless for reading the files you want to read. And teh Trust chip reports that different signature to any other software that asks about it, and reports that different signature over the internet. So internet servers and other software are generally going to refuse to speak to your modified software.

-

Damn.

[Kickasso]

You of course can make modifications to your machine. The chip doesn't disable modification. It's OFF BY DEFAULT, for crying out loud.

Should you choose to enable it, you can convince others that you have an unmodified machine if and only if you actually have an unmodified machine. That's all.

Whether or not a modified machine will be useful is a different question, and totally dependent on what you consider useful. I believe almost nobody except entertainment providers will require remote attestation. Thus, a modified machine will be useless if you want to talk to them, and useful otherwise. But other people have more scary visions, and what do I know?

My oh my.

[Kickasso]

Make a bootable CD that checksums your OS partition. Boot from it once in a while. Unplug the network cable and recompute checksums when you modify the OS. There, you just convinced yourself that your system is not hacked.

If you want to detect hackage while the system is running, that's much more difficult, and the TCPA chip can't do it either. So you won't be able to tell if your bank is rooted by a script kiddie, because bank computers don't reboot very often.

If the bank uses remote attestation as proposed by TCPA and only connects to trusted clients, its chances of being rooted will be reduced. But that's beside the point I guess.

NAT

[tepples]

Via a masquerading box ?

"Masquerading" is network address translation and port address translation. To get an IP with these Cisco routers, your NAT box will have to be running "trusted" NAT software that can request TPM attestation of each machine behind it.

Re:NAT

[Anonymous Coward]

And of course, _all_ trusted NAT software is going to be unhackable...sure...

Re:NAT

[tepples]

And of course, _all_ trusted NAT software is going to be unhackable...sure...

And once a particular NAT implementation is hacked, ISPs that follow Cisco's plan will just drop support for its hash.

Re:NAT

[Anonymous Coward]

What, and inconvenience all your customers who are "legitimately" running that NAT software ? Force them all to patch and upgrade ?

As regards the hack, you could, for example, hack it so it runs transparently and never sends a hash at all. The router doesn't even know you are running NAT software. I'm sure there are many more ways around this hypothetical situation.

Besides, I can't see how ISPs could implement this, even if these rumors about Cisco routers are true, what happens to all their current customers running non-trusted systems ? Are they just gonna lock them out ? Why would an ISP bother switching it on, and make more work for themselves dealing with inevitable customer complaints ?

I think I'll be happily connecting with my hand rolled Linux kernel for many more years.

Re:NAT

[tepples]

What, and inconvenience all your customers who are "legitimately" running that NAT software ? Force them all to patch and upgrade ?

Yes. Universities have already blocked Windows machines that don't have the latest service pack installed.

As regards the hack, you could, for example, hack it so it runs transparently and never sends a hash at all.

No hash, no IP.

even if these rumors about Cisco routers are true, what happens to all their current customers running non-trusted systems ?

Do the major ISPs support SLIP access from MS-DOS systems at 14.4 kbps anymore? In fact, a few ISPs such as AOL and NetZero need their own dialer software, and they periodically drop support for older operating systems.

Re:NAT

[Anonymous Coward]

Yes. Universities have already blocked Windows machines that don't have the latest service pack installed.

Well, that makes sense. If you are running windows, you certainly should have the latest patches installed. So Linux users can't use the network in these places ? Universities are not ISPs, so this is not really a valid point.

No hash, no IP.

Well, sure it would send a hash of the gateway machine when it connected. My (theoretical) hack was to change the NAT software so that it forwarded transparently.

Do the major ISPs support SLIP access from MS-DOS systems at 14.4 kbps anymore?

Of course not. There were probably only a handful of people using that method in the first place. However, there are still many people using Windows 95/98 to connect. Even in 6 or 7 years time there will still be people using Windows XP running on "untrusted" hardware. And if I cannot connect using my chosen operating system, then I will simply switch ISPs.
But anyway, as I said initially, I can't really see why any ISP would want to do this, it provides no benefit to the ISP, and adds an extra layer of complexity to their operations.

Re:NAT

[tepples]

Universities are not ISPs, so this is not really a valid point.

Not only are universities ISPs, but they're also monopoly ISPs. You try getting a standard residential cable or DSL connection in a residence hall.

My (theoretical) hack was to change the NAT software so that it forwarded transparently.

Would the ISP trust your h4x0r3d NAT software?

And if I cannot connect using my chosen operating system, then I will simply switch ISPs.

If both the local cable ISP and the local DSL ISP switch to "trusted" access, then how much does it cost to move house?

it provides no benefit to the ISP

Other than possibly pretending to control the spread of viruses?

Define "usable".

[Kickasso]

If by "usable" you mean "able to misrepresent itself" then yes, it's useless. I don't necessarily buy your definition of usable.

Also, define "work". If by "work" you mean "convince others to interoperate by lying about itself", then it won't work. Guess what? I don't buy your definition of work either.

It's your responsibility to persuade others to talk to you. How do you do that is not anyone's business but yours.

Re:Define "usable".

[Alex Belits]

If by "usable" you mean "able to misrepresent itself" then yes, it's useless. I don't necessarily buy your definition of usable.

Also, define "work". If by "work" you mean "convince others to interoperate by lying about itself", then it won't work. Guess what? I don't buy your definition of work either.

"Lying" is irrelevant to the matter -- "Trusted computing" and DRM are not persons, and have no legal meaning as concepts. They are parts of the device, that is sold as a general-purpose computer, so it's supposed to run software, and indeed that device runs some software -- one that someone distributed with an encrypted key, and provided a "source", that, by definition in the license, should be the preferred form for modifications. However since the key is not distributed, any modification results in a code that can not run on the available hardware, thus making a source without a key something other than a "preferred for modification" form.

What the author of a DRM scheme (let's call things by their real names -- "Trusted Computing" is a DRM scheme) intended to do, does not make him "right" when applied to the license -- from the user's point of view, DRM vendor maliciously sabotaged the hardware sold to him, and the software distributor conspired with him by using and not distributing the key, thus intentionally breaking otherwise perfectly usable and modifiable source, turning it into a meaningless text.

There already were smartasses that tried to distribute obfuscated source, source that requires proprietary and unavailable for others tools to compile, and other crap like that. None of that would work with GPL because of the single phrase that I have quoted before -- where "preferrable for modification" was mentioned. Making a "signed" version and not distributing the key far exceeds those amateurish forms of source code sabotage, so it would be unreasonable to expect that it would work.

It's your responsibility to persuade others to talk to you. How do you do that is not anyone's business but yours.

Another, far superior strategy, is to thoroughly prove wrong and humiliate the obviously wrong or stupid opponent, so people who actually are going to make decisions, will have no doubts about the true value of his arguments. I really don't care if you agree with me or not, you are neither a judge, nor an executive in a hardware or software company that may make a decision that matters in this situation. Your only effect on this is in spewing propaganda.

Scripting languages?

[slavemowgli]

Here's something I'm wondering about.

No matter whether "trusted" computing means that you cannot run binaries not signed by a 3rd party, or whether it means you cannot run binaries not signed by someone (with a configurable key), or whether it just means that unsigned binaries simply won't be "trusted" (but will still run)... what about scripting languages? The whole debate seems to revolve around the users' ability to sign new binaries (presumably created after modifications are made to the source), but what about languages that don't compile code to binaries, anyway?

Would "trusted" computing mean that it's not possible to write a "trusted" program in Perl, for example? The example of a system administrator in a large network locking down machines so that they can't be taken over (easily) seems to be given often, but what good would locking down machines be if you can't use Perl programs, for example? It may not be a problem when you just have windows-based desktops where you don't expect to run more than outlook, word and excel, but not every large network is like that (and one that solely consists of windows machines probably has more serious issues, anyway. ;))

The problem really is that you have only two choices: either you sign the Perl (or whatever) binaries (in which case any Perl program will, basically, run as a "trusted" application, since from the system's point of view, the script that Perl loads is just another piece of data), or you don't, in which case no Perl program will be trusted at all.

Plug-ins, dynamic libraries etc. may also well be a similar problem; the idea that an application always consists of a single monolithic piece of machine code simply isn't true. Most of the times, it'll be several pieces, and some of those may not even be machine code to start with.

could hurt the Trusted initiative or help Linux

[Anonymous Coward]

Linux is by its very nature very fragmented. Many users see this as one of its inherent gifts, giving you the freedom of choice in so many different combinations of systems customized to how you see fit. This type of freedom comes with a price tag and Linux is not willing to pay it. For such freedom, you must be willing to work towards standardization that will enable more coherent combinations of the various parts. This means depending on each and every developer, distro, package maintainer, etc to work with each other instead of in a vacuum where they reinvent the wheel thousands of times a day.

If this price is not paid, then the inevitable reduction in QA (read: assurance of quality, not just quality) becomes much larger and you end up with a cascading system of problems where a developer of component A must worry about Components J,K and L and each of those bringing their own "chaos dependencies" where you may be 0wn3d by the antipathy and lack of discipline of other developers and maintainers.

Compare a mechanic working on a car to working on the average American's garage itself. Most American's love mess and are not organized or consistent. Finding your way around someone's garage would take considerable time and effort and even once you felt confident that you knew the layout then guess what? The home owner decides to move things around or simply just throws more "stuff" into disarrayed piles to add to the chaos. A car on the otherhand uses for the most part, standard parts, measurements, and devices that in turn can be utilized by standard tools, knowledge, and methods.

Linux can be helped by this initiative if it points out the flaws and people learn to adjust and make Linux better. The alternative would be for TCPA to make such massive adjustments to its requirements as to render it useless and well... untrusted.

Linux is Linux is Linux. The inconsistency inherent in it does it only harm when security minded folks want to use it. Keep in mind here I use the term Security, not device or service hardening. Security is much more than the sum of the devices and services being managed. Repeatability and internal predictability are key to a successful security implementation. A small group of 1337 folks who cobbled together some systems is the equivelent of our mechanic friend swapping out standard parts (and interfaces like pedals and the steering wheel) with home brewed solutions that only he and perhaps some other niche folks know about. Take that and then weld not just the parts together but weld plates at irregular intervals onto and around the parts. Ensure that the end system requires significant learning curve and indepth study instead of smartly using what works and focusing on implementation details that can be as unique as you need them to be.

Gentoo of all distros (with the exception of Linux From Scratch) is a prime example of this disorderly chaos. Now if Gentoo (or any distro) works hard at reducing the inconsistency, reducing the knowledge required, thoroughly testing various configuration pathways and DOCUMENTS them clearly for admins and users, then that is a start in the right direction. Sorry, forums are not documentation. Bugzilla is very limited and especially becomes hamstrung from lack of proper, descriptive updates.

It is HIGHLY ironic that Gentoo would be the source of any article detailing any sort of initiative with "trust" in the name. I can not TRUST that my emerge updates will work. I can not TRUST that the packages will compile at all, much less work correctly on my system. I can not TRUST that proper QA has been done that will reflect my particular environment (because everyone has different components and versions thereof). And here is the key: I can not TRUST that when I get the very inevitable system borking failure from using emerge or any of its associated tools, that I will find useful and AUTHORITATIVE information on how to fix it. I should not have to google and search for days just to find some ra

package and process dependencies

[Anonymous Coward]

ok, we all are pretty savvy on dependency systems for package management (pretend you are if you are not... work with me!) However, here we have a dependency in processes.

As many are aware, whether you speak of binary linking (where appropriate), configuration or the internal code you always have that issue with instability bringing about insecurity. It can also be said that bad interfaces (unintuitive, doesn't mesh with existing standards, etc) lead to unstable and insecure systems as well.

Therefore, it is correct to say that in order to have any sort of initiative to provide a super-level of trust and assurance for a system then it first must establish a basic level of quality. It is important to remember that you don't simply point at one iteration of a system/product and say, "Yes we have a quality (much less a trusted) initiative" regardless of whether or not that particular stand alone iteration meets your criteria.

As a consumer (end user, administrator and even developer) I require "trust" to be established in the entire organization which encompasses the processes, policies/procedures and the mentality of entire development community.

What this all goes to say is, "Has Gentoo (or any other) established a system of quality?" No I do not mean do they have folks in positions called, "QA." Gentoo along with distros like Archlinux and LFS (which usually call themselves a "meta-distribution") are by their very nature more prone to instability. Don't get mad... reality is not opinion based it is obversation wrought. So, we need to ask if there is a granular method of programmatically tracking the possible hardware and underlying software configurations relative to each other. Basically, if I want to install package X, what (using Gentoo) hardware, software, and configurations/settings has it been tested with and what were the results of those tests. Add in USE Flags, compiler settings, etc and you have a very complex and hard to manage system. Therefore instead of whining about it and giving up, a smart distro will work on ways to manage this. An even smarter solution is for several distros to support an independent project for tracking this. (More eyeballs and all that.)

So, is there a "Gentoo Quality Assurance Initiative" that is established now? No. (again, saying you have QA is not the answer to this and if your system if flawed then you must accept that your initiative is therefore flawed and not consider it complete until it itself is fixed. Remember that words like Assurance (as in QA) and Trusted require proof and are always trumped by reality and facts. If I can not depend on Gentoo or any other system to work without fuss and "just work" then I have no assurance or trust in it. Obviously this implies that I am using "stable" combinations as detailed by the only system of tracking these... the ARCH and USE flags. There is no salient system (and doesn't seem to be any interest to develop one) where I can first query on how the possible stable (or even better by varying degrees of stability) builds of my system... before hand. Then I must, upon attempts to install or configure the system, also be able to have the package and sytem management system guide me in creating a stable and secure system. Yes, a distro MUST provide robust management capability for even configuration of the system. This goes well beyond current package management. It must be consistent in look, fell, and workflow across the board and factor in that external apps and components will be on the system. External here meaning outside of the standard package manager. It is also wise to work WITH other package managers in a way that each of them read the same database (conceptually). Both of these extensions would therefore require a flexible management system that does not have to be gutted when expansion or changes are necessary. That is what separates the engineers from the hackers.