Slashdot Mirror
TiVo Awarded Patent For Password You Can't Hack
Davis Freeberg writes "TiVo has always been known for thinking outside of the box, but this week they were awarded an unusual patent related to locking down content on their hard drives. According to the patent, they've invented a way to create password security that is so tough, it would take you longer than the life of a hard drive in order to figure it out. They could be using this technology to prevent the sharing of content or it could be related to their advertising or guide data, but if their encryption technology is really that good, it's an interesting solution for solving the problem of securing networks."
Reminds me of a trick I pulled on an old HP-UX box. I somehow managed to put a backspace keystroke in my password I could log in on the console (which treated backspace as a normal keystroke) but not over ssh or ftp (since there is no obvious way to type a backspace into one of those clients).
I suppose if I ever figured out how to put a newline into my password I would have one heck of a time logging on.
If it exceeds the life of the drive theres an easy way to just clone the drive or remove the platters and put them into another hard drive (yeah very sensitive operation likely requiring the conditions of a clean room).
Its hard to make something undefeatable and if you claim such it is only going to attract people as a challenge. Maybe that is what they want?
Of course if someone proves that it isnt 'impossible' then does that void the patent?
09F911029D74E35BD84156C5635688C0
+2 Troll is Slashdot's way of saying groupthink is confused
This has nothing to do with networks at all. The patent is about making sure a hard disk can only talk to a certain host.
Its just another attempt to prevent people form using their own hardware how they want to.
I am no expert, but couldn't you create a device that reads the input + output of the hard disk, then grab the challenge + response and by doing so improve your chances of cracking the key?
Or maybe the password is just "Iceberg" -- "Even if they hit that key, it won't cause a crack."
On the dangers of assuming keyspace => security:
from ''Computer Security and Cryptography'', Alan G. Konheim.
Belief is the currency of delusion.
I know that I'm probably not their target audience, but the one reason that I have two subscribed tivos is that I can hack them and disable the DRM and generally they've been pretty cool about it. But the day they lock me out of my one boxes is the day that I cancel my subscriptions and either continue with the hardware on my own or switch to MythTV.
The difference between Canada and the USA is that in Canada healthcare is a right and gun ownership is a privilege.
It's basically just a DRM-machination with the cryptography on chip. Basically, the same that AACS has on HD-DVD, and the patent specifies that guessing the password woud take longer than the lifetime of a drive. Euhm, I guess even guessing 56-bits encryption would be enough.
The problem is still, the user has HIS content, he can do whatever he wants with it as long as he can see it. Unless you encrypt the lightwaves that reach our eyes and plant a DRM chip in our brain, we're going to be able to copy your precious content.
Custom electronics and digital signage for your business: www.evcircuits.com
I've already canceled my TiVo service due to their rapidly-decaying "rights" issues. For the obscene price lately on TiVos, plus service, it's cheaper to buy a few components and build a MythTV or similar box.
Quickly, before Cringely ruins it with bad math, I need to point out some very obvious weaknesses in making this work correctly:
Okay, you all can go back to your regularly scheduled cheap shots.
I have no idea how the process of reverse-engineering of a microchip works, but does it depend on inspection of features revealed only by light? I.e., is there a reason it can't theoretically be performed in a darkroom? Failing that, if some part of the process does depend on visual information, would a high-resolution camera with a high shutter speed be able to capture that information sufficiently well before the chip was destroyed?
What I'm trying to ask is: does photosensitivity make it practically impossible to examine the guts of the chip, or does it merely make it harder?
FTA: According to the patent, they've invented a way to create password security that is so tough, it would take you longer than the life of a hard drive in order to figure it out.
So it's security is that a brute-force/birthday attack is just so improbable that the drive will wear out before i can test enough possibilities to have a measurable chance of getting it? Besides, twofish, blowfish, AES, any virtually any other standard encryption algorithm could boast the same thing. Tell me if I'm wrong, but couldn't i make a bunch of 1:1 copies of the disk and use those to crack it?
If i had one dollar for every brain you dont have, i would have $1.
but I do know this nifty card trick:
Give your friend a deck of cards. Turn around and have them shuffle it, select a card at random, memorize the card and put it back in the deck. Have them shuffle it some more (without you looking at it). Take the deck from them and take a card from it and say 'this was your card'.
In the long run, you'll be right about 1 in 52 times. If you happen to be right the first time with a particular friend, and never do the trick again, they will be scratching their head for a long time trying to figure out how you did it.
So, the point I'm trying to make is that it could take longer than the life of a hard drive to crack the super secret code, or you get get it right on the first guess (or the second one, or the third one...). So it seems rather silly to claim that it is uncrackable.
There are a lot of reasons:
As for restrictions, the box itself doesn't do much of anything to restrict me as far as I've read. And, of course, for what I'd be using it for (a DAAP client), it's really an ideal solution (lack of S-Video and composite outputs notwithstanding). It's easy to use, can connect to a DAAP share on the MythTV backend box easily (it looks like an iTunes share), etc. Output formats fr older TVs notwithstanding, it's a plug-in-and-go solution that can easily integrate with the MythTV setup, but is still tweakable under the hood if I feel the need to do so at some point in the future.
That's what I look for in technology products---products that (as much as possible) just work when you plug them in, but are still sufficiently easy to mod to add features if/when I outgrow their functionality. The AppleTV gives me a lot more room to grow than an Xbox. That means that I'll be able to keep using an AppleTV long after I'd need to replace the Xbox with something else. That long-term viability is worth an extra hundred dollars to me.
Check out my sci-fi/humor trilogy at PatriotsBooks.
When I read this I though "Okay, so you have to steal the box to get the content or do a lot of work to get the data off of the drive using the chip in the machine.. no big deal right?"
Then it occurred to me, maybe the host computer isn't the local Tivo box, maybe it is Tivo's system (remote) that they're calling the host. What does that mean? Now you can't get data off of the drive unless the Tivo calls home, swaps keys, and stores a decryption key/algorithm in RAM. This means that if Tivo says no, you can't get at data on the device you now own. So... well if you can hack the OS then you can just have the keys stored after/during exchange or you could read out of RAM, but maybe the OS is built off of a network boot scenario with the initial sending of the system happening only after the handshake. Tricky.
If (big if) that is the case then the way to beat it will have to be capturing the data in RAM from a running system. It sounds tough but I suspect you could do it by setting up a virtual machine that intercepts the call coming from the box, and on return sends all output from the chip normally destined for real RAM into virtual RAM (which is really filesystem based, heck make it a ram drive so it is as fast as RAM but readable as a file.) Copy the virtual ram file, and you've got an unencrypted OS. Hack your unencrypted OS to store the keys, and now you have your drive decryption key, your "call home" key and a hackable OS. Want to do something Tivo doesn't like? Make your OS think the commands came from Tivo, not too difficult now. Maybe they have a changing algorithm where the chip uses a new key (in predicted order) for each call home, incrementing after each successful exchange. Maybe then you have to talk to the chip every time with your Virtual Machine, but it still accomplishes the goal of having complete access and control.
Okay, what I think they really have is a scheme to make sure that a chip and drive are tied together so you can't get at the drive without the chip, thus no Tivo drive swapping and they really don't care right now anyway and just wanted to get the patent because they think their method might be marketable some day. I wonder if I'm giving them ideas.. nah, they'll never read this post, right?
B) Eliminate all the stupid users. This is frowned upon by society.
Why not encrypt the HDD at the level of the drive electronics? That way a user would have to physically remove the platter to read any useful data. That process would cost more than most data one could recover from an average user's tivo.
On the other hand, yes, this does appear to be a simple patent on tying a hard drive to an electronics unit. Viable attack vectors are already obvious.
The ______ Agenda
I think you all should just stop watching T.V. I haven't watched T.V. regularly in over 5 years now and it feels great. Just think a moment about how obsurd cable T.V. is . . . you're paying money to be advertised to. It should be the other way around. THEY should pay YOU to watch thier crap.
Think about how much head space you will be saving yourself. Hell, I still have commercials floating around in my head from the late 80's. I certinatly don't need any more of that filth polluting my thoughts.
In conclusion, T.V. sucks. Stop watching T.V.
That is a dreadful patent, and it would be ridiculous to see it issued; hardware challenge-response dates back to at least the first IFF machines in the second world war, they're not even mentioning having a deliberately slow password-hashing algorithm, which is itself at least as old as UNIX, and the technique is vulnerable to bump-in-the-ATA-cable extraction of the data from the disc in the first place, and probably also to an attack where you swap the drive controller board for one from a drive of similar model without Special Tivo Sauce.
In general, you seem to be correct. You can patent just about anything. But there is an exception. Since 1911, the words "Perpetual Motion" have been the kiss of death for a patent application. In order to patent your perpetual motion machine, you have to obsfucate its nature -- for exmple by claiming it is an anti-gravity machine. No, I'm not making this up. Wish I were. See http://209.85.165.104/search?q=cache:SzVmVt9_BIwJ: news.nationalgeographic.com/news/2005/11/1111_0511 11_junk_patent.html+patent+perpetual+motion&hl=en& ct=clnk&cd=2&gl=us&ie=UTF-8
You can't see ANYTHING from a car, You've got to get out of the goddamned contraption and walk...Edward Abbey
Wait what? I have an enterprise WD drive installed in my home PC with a 5-year warranty. As far as reliability, In the aftermath of Hurricane Katrina, I was working at a college whose campus was across the street from the gulf of mexico. One of the professor's computers which was recovered had a WD Caviar Drive in it. Due to location the thing was under sea water for 10 hours. Circutiry on the underside was corroded, it shook salt when you tapped it, and smelled like dead fish. After a lot of sad grinding sounds, Symantec Ghost had made a working clone of it in 20 minutes. No file loss. In normal operation, many of our (past warranty) WD drives worked like a champ as well. I will also admit our newer seagates never had a single issue, but the older models were less reliable than the aptly named Quantum Fireballs.
The password might not be cracked. Well, at least not cracked in a meaningful or useful way. I can think of several ways this could be accomplished. Tying the drive to the mainboard with a kill switch that burns out the firmware controler could be one. This could mean all ads and all content is useless outside the tivo and the drive is borked if tried outside it too.
But if this patents is invalidated, it is meaningful in several ways. First is other devices might be forced into using it by the media companies or something and this will raise the costs of consumer electronics. The next thing is, suppose someone discovers this as a way to keep usable information out of anyone's hands who don't have permission to use it. There is another royalty that needs to be payed and it will come out of our pockets too. But most importantly, A patent takes an entire piece of software off the market for most. Imagine if the word processor was patented when it originally was developed. Whatever the first word processor was and anyone willing to pay the royalties to them are the only word processors we would have. Openoffice.org wouldn't be here, Microsoft could have bought the patent and stopped everyone from using it other then them, so on and so on.
So what happens when computers are fast enough that to be somewhat reasonable secure, you need this patent. If it is still valid, again, everyone pays TIVO to use it. But if it was copy written instead of patented, then many other players could attempt to do similar things and hopefully competition would make things better and all. But if we are stuck with this one implementation and it turns out not to work, any working implementations from other companies will have a payment to TIVO associated with any costs.