Slashdot Mirror
Questions for a Lecture on Microsoft's Palladium?
An anonymous reader asks: "Microsoft is going to be giving a lecture on Palladium
for my Computer and Network Security class at MIT this Thursday. We're told that it's going to be the most technically detailed lecture publically given to date, and that we should be armed with questions as a result. Any suggestions from the Slashdot crowd? What technical details have you been dying to know about Palladium?" It would be interesting to hear back from someone who is planning on attending this. For those who wish they were, but can't for one reason or another, what would you have asked by proxy?
More of a basic business question, but didn't anyone learn from Intel's ill-fated processor serial number "feature" in the Pentium III, or the Div-X movie fiasco? Why would consumers want this at all, and why will they choose it over other alternatives?
The biggest question in my mind on Palladium is how it's supposed to help users. Why we're supposed to use it, instead of just keeping on using our old Palladium-free computers.
TANSTAAFI: There Ain't No Such Thing As A Free iPod.
We don't need palladium for viruses...this just seems like a system for pervasive DRM. Why do we need this?
And how does "trust" have anything to do with Palladium. Palladium is a system of control, not of trust.
What will Palladium do to those of us who release independent content? (As in, independent of major corporations.)
The only way I can see it possible to effectively implement DRM is to require computers to not play any digital content that does not have a valid encrypted signature, as provided by the various media companies, and/or Microsoft and Intel.
My main concern, is that independent producers/composers/moviemakers will be locked out of distributing digital content, because the companies involved in Palladium, and other DRM schemes, can choose to withhold issuing these encrypted signatures to them, therefore rendering their content unplayable on Palladium-enabled systems.
I feel, as a copyright owner, and musician, that this infringes upon my rights to distribute my work signature-free, for anyone to be able to play. I do not want a special tag on my releases telling people this is official. I would just like to see my stuff "out there". Therefore, this infringes upon my right to the "pursuit of happiness", as ordained by the constitution.
Anyone else have thoughts?
-----
"You spilled my egg... I needed that egg."
Trolls and humor aside, I would like to know how they are expecting to fix problems with Palladium should they arise. The only way they can fix X-Box "security" problems right now is to release X-Box 1.1, and if they have to re-release computers to fix security problems, how would they do it? and who gets the bill? (maybe I shouldn't ask that last question...) And what is to stop people from mod-chipping computers? At any rate, I believe like many of my fellow /.'ers that X-Box is a Palladium Preview... or Rhodium (the element before Pd, get it?)
Hmmm.. On that note, maybe Palladium is a preview to Microsoft Silver?
I'm the Devil the Windows users warned you about.
Can open source software and Palladium coexist?
Go even more general than this, so you don't even have to bring up competition:
How can user written software run on a 'trusted' system?
jello.
aka aron.
Ask questions that will make the lecturer either reveal how evil it is, or make his evasions obvious. Possibilities:
1. If you turn it off - as MS claims they're going to allow - will the system then appear to apps, content & the network as "a Palladium PC with Palladium turned off" or as a non-Palladium PC? (Hint: it's the former.)
2. Will I still be able to flash my BIOS? *All* of it? replace it completely? (Assuming TCPA hardware, they're lying if they say 'yes'.)
3. Why would I want to buy this, if I'm not interested in Hollywood movies but do want complete control over my computer?
I see alot of questions here that refrence things from the open source movement. I would use more ambiguous words in their place because as soon as the folks from MS realize that your into open souce they're going to give you the run-around. IE, don't say open source projects, say personal software projects. in place of Ogg Vorbis, say alternitive audio codecs.
There was a MS representative at the career fair here at UVA and as soon as I mentioned the word linux, the conversation pretty much ended.
my other penis is a vagina
/ONLY playing devil's advocate; DON'T get on my case as this is not how I really feel/
Their answer will be:
"Providing adequate protection for digital content helps ensure that the quality of that content is protected, and maintaining the rights of the content producer will help maintain the quality of their work, which helps us all."
Again, I don't agree with this nor do I think it is a compelling reason, but if I were a Microsoft Market-bot-3000, that would be my standard output.
El Karma: excelente(principalmente la suma de moderación hecha a los comentarios de los usuarios)
In fact, stay away from the obvious questions in general. Answers will have been prepared and you will waste your time.
If you want to make them squirm, you need to come up with some direct and highly pointed questions that will be very difficult to avoid answering directly without making it very obvious they are so avoiding it. (You can't prevent avoidance, but you can try to make it obvious that that is what they are doing.)
If I could ask a question, I'd try something like the following:
- What kind of data recovery plans will exist if I buy $1000 dollars worth of digital music that is tied to my processor, only to have my processor get fried in a power surge? Will there be any way to recover my investment, or is it lost? If so, what's to prevent hackers from using that recovery mechanism? If not, how can this be a benefit to customers?
The meta-point: Perfect protection implies no recoverability. Recoverability implies imperfect protection. You can not have it both ways.It's pointed, and it will be very difficult to avoid giving an answer, or making it obvious there isn't one. Either there is a recovery procedure, or the customer is SOL... it's pretty binary. If there is a recovery procedure, hackers might exploit it. (Or do we have to dial home to Master Microsoft first?) If there is no recovery procedure, then how can they honestly claim this is a benefit to the customer?
Me, I'd lay money on a handwaving answer... but it should be obvious, if you do it right.
There is no reason you couldn't write an open source browser or office suite and have it run on a palladium system. The reason why there have been murmurs of a possible palladium/OSS conflict only apply to a certain type of program, specifically that which uses palladium/tcpa's "security" features.
;) )
Picture an open source media player. As it stands, xmms could be run on a palladium system and the oss model would work fine. It would play oggs ripped from your own personal cd collection and any company that takes the source, modifies it, and distributes a binary would have to release the source back to the community. No problem.
Now let's say a company takes the xmms source, adds support for drm-infested media, and releases a binary that's been digitally signed by MS, meaning that MS has examined the source and seen that it will not ever expose unencrypted, drm'd data to user access. It still plays oggs (they haven't removed that feature yet), but here's what happens when you try to connect to Disney's server to upload your credit card data and download Mickey Mouse 2010 (subtitile: Yes, we still have the copyright):
1. Disney queries your machine for it's unique ID (yes, all PCs must have them for the system to work).
2. Upon verification that the unique ID is a valid one from the central unique ID database, it asks your system for a signed, timestamped, digitally signed (by the TPM [trusted platform module) message saying that your system is running a drm-compliant OS.
3. If it gets an affirmative answer back, it queries the OS as to whether the app is digitally signed by MS. I'm not familiar with the system that will be used in this case, but I think identd would be an accurate model (i.e. "Is the app connecting from port xxxx on your machine to port yyyy on my machine digitally signed?").
4. If it gets an affirmative answer back, the server will then send content encrypted with the platform's public key (not the "unique ID" key, that one is a single purpose sign-only).
5. xmms, upon receipt of the data, plays it back according to the drm rules.
Now, imagine you want to modify the new xmms sources (that include drm support) to play a new audio format or to add a media manager function (or whatever). You still have free access to the sources, but once you modify and compile them, you get an unsigned binary out of your compiler. It still plays oggs, but when you try to buy a movie from Disney, the OS responds (in step 4 above) with a negative answer.
"No, the binary making that connection is NOT signed."
The result is that Disney will not send data to that app. I'll get the obvious question answered right now:
Q: What if you modify your OS to respond to all step 3-4 "is xyz app signed?" questions with a "yes" answer? Couldn't you break the system that way?
A: No. The authentication process would fail on step #2 above because your recompiled kernel wouldn't be signed so the TPM on your motherboard would refuse to vouch for it.
What does this mean for OSS? Well, not much. Open-source, non-pd/tcpa software won't be affected at all. OSS that does "handle" secure content as one of its main functions would be affected - you wouldn't be able to fork it unless you wanted to pay MS for a digital signature on every release to you want the pd/tcpa portions to keep working. In a nutshell, only the portions of OSS that normally depend on pd/tcpa would be nonfunctional.
So why is palladium/tcpa still a big problem? Well, a couple of reasons, but first, more Q&A.
Q: What if I were to physically crack open my trusted platform module and extract its private encryption and sign-only authentication keys.
A: You would have broken palladium/tcpa security.
Q: What if I were to replace my core root of trust for measurement (CRTM, aka my BIOS) with one that always reports the system is booting in a "secure state" to the TPM?
A: You would have broken palladium/tcpa security.
Q: What if I find a buffer overflow or other bug in a signed application (e.g. windows media player) that allows me to execute arbitrary code as that process?
A: You would have broken palladium/tcpa security.
Q: What if I find a buffer overflow or other bug in the OS or a signed driver that allows me to execute arbitrary code as the OS kernel?
A: You would have broken palladium/tcpa security.
I don't mean to make it sound easy - tcpa is designed to place these activities beyond the means of the average script kiddie. However, they are all very real valid security problems that palladium/tcpa _will never be able to solve_, specifically because of the nature of cryptography, mass-produced hardware, and information itself. I guess you could say that information really does "want to be free".
(Note to grammar nazis: Yes. I'm aware I put the period outside the quotation marks. I did this because I believe it enhances the readability of printed english. Putting the terminating semicolon from a line of C code inside the quotes around a quoted string just doesn't make logical sense. However, any its/it's, there/their/they're, or other stupid mistakes that detract from my ability to communicate clearly are fair game.
So why is it such a bad idea? Because people think it will work. The latest issue of PC World (November [?] 2002) features an ad from IBM touting the advantages of the latest Intel Pentium 4 processor's LaGrand Technology. If I could find it I'd post the page number, but if you look through the issue it's on the left side somewhere in the middle-ish section. It promises freedom from viruses and a more secure operating system. I think it promises completely secure e-commerce as well. The average PC World readers are going to read this and their eyes are going to pop out of their heads. "Really? No more viruses? No more trojans? Secure e-commerce? How wonderful!" When online companies start pushing "secure" online movie rentals (broadband only, some restrictions may apply, void where prohibited, etc...) the ones surviving heart failure will scramble to buy new pcs with this LaGrand Technology (or amd's equivalent). After all, who wouldn't want a virus-free secure PC that does new and exciting things?
Nevermind that the reason 99.999% of the computer-using public have to even think about viruses is because outlook is so incredibly insecure. Nevermind that the only things stopping global availability of secure online shopping are the certificate authorities' greed and US crypto export laws*. Nevermind that online movie rentals will most definitely not take off soon considering how much bandwidth is available to home users even with broadband. (Yes, you may have 2mbit cable, but what's going to happen when a large enough percentage of friday night movie watchers decide to download and cable companies are overselling their last mile _and_ backbone bandwidth at a ratio of 50 to 1?) Nevermind that LaGrande Technology is designed to be the cpu-side hardware support for tcpa/palladium which is already flawed. I'm not saying that IBM won't be able to make good on their promises of perfect security and a virus-free environment (that's a separate debate) - I'm saying that they're pushing a unique PC ID and Digital Restrictions Mechanisms into every home in trying to do it.
(* Yes, I'm aware that you can get strong ssl encryption in linux outside the US. Here I'm referring to windows, a product from a commercial entity that has at least a slight interest in pretending they obey US law.)
So that's how it's going to get into homes and businesses. What harm is it going to do once it gets there? Well, just because it's going to be hopelessly inadequate when it comes to serving its intended purpose of stopping online piracy of digital media doesn't mean that it won't restrict fair use rights. Sure, anyone can use a cracked pd/tcpa box to download a film from disney and then distribute it online, but if Joe user can't rip his legally purchased CD and send it to his car stereo because of draconian DRM code, that's a problem. And that's only the copyright/fair use side of the issue. What about security? What happens when a certain OS vendor, with complete confidence in its supremely planned but critically flawed transition element, starts getting lax on security and starts depending on pd/tcpa keep everything together? Even worse security holes than we've seen before due to inattention to important detail and (at least) internal code review.
I hope you see what I'm talking about now. The worst possible outcome is not that palladium/tcpa will progress as planned (which violates the "possible" part). It's that it will approach an uneducated public and fail miserably.
Are you a paying member of the eff yet?
* Only DRM/"Trusted" systems will be able to play content from the Music industry or Hollywood.
* For an operating system to be trusted it needs to be vetted and signed for use with DRM. i.e. it needs to be a "known quantity".
* An OS where the user can modify it at will is not a "known quantity" or signed, and even if it was, as soon as you recompile it you would break the signature. Basically, an OS where you are allowed to modify it, can not be trusted. (Allowing modifications being a large part of the "Freedom" involved in Free Software. You can't have it both ways).
The result being a world where only non-Free operating systems can play the entertainment industry's content, by design.
If you thought playing Windows Media files on Linux was tough now, wait until Palladium.
--
Simon