| Who | When |
Messages | |
|
|
|
|
|
Matthew Hamrick
|
15
|
 |
|
08-01-2003 12:37 AM ET (US)
|
|
|
Peter R. Morris
|
14
|
 |
|
07-31-2003 03:04 AM ET (US)
|
|
This will not keep RIAA away from your kazaa or any other files.
they catalogue what you have by spidering the kazaa network just as any other user does when searching for a specific file.
In order for you to use the hdd you have to have unlocked it. When you are sharing using Kazaa or any other network, it MUST be accessible from the network.
It is only at the stage of a raid when this is useful and then only if you have the time to turn off the computer.
In addition, what method is used to present the authentication key? if it has to be typed in, it must be of a relatively short length and so will be susceptible to brute force attach, if it is on some "tag", a USB device or floppy, then you have to be able to destroy its contents quickly when/if a problem happens. and to re-create it should you need to after a false alarm...
this is not a surefire fix to the problem.
Just as it is not possible to generate a truely secure DRM, it is not possible to generate a system which will allow you access to your data while maintaining its security.
the point is always that at some point within your computer the information has to be converted into a form that is useful. What technology can obscure, other technology can make clear.
|
gfm
|
13
|
 |
|
07-30-2003 09:05 PM ET (US)
|
|
This has been around for a long time, it's part of the Secure ATA protocol. I think a lot of motherboard manufacturures have supported this for a while. This is same tech that's in the xbox and the hp digital entertainment center.
This won't do jack to protect your hard drive if they have your motherboard.
Can anyone say: RIAA Honeypot? I think some groups need to start setting these up. Create hundreds of fake accounts in Kazza that appear to be sharing thousands of songs from popular artists, but they're really all the same file--one that isn't under RIAA jurisdiction.
|
Fencepost
|
12
|
 |
|
07-30-2003 05:54 PM ET (US)
|
|
Correct on the IBM disk passwords - those have been around for quite a while, though I think that ThinkPads are the only notebooks that actually use the drive password. Those are more than adequate for most people's security needs, and would also serve my desire to keep my files private and the laptop useless if it's stolen. In that way it's much like zip file encryption - fine to keep out the casually curious, inadequate against the determined attacker.
I'm not really the target market for anyone doing a secure laptop though - governments and businesses are, particularly anyone who deals with sensitive information (investigations, 5-year plans, hardware schematics, whatever). For those folks who might be specifically targeted, a drive with all the data encrypted would be very desirable. If it's something you do with any regularity, it wouldn't be any problem at all for someone to take the platters out of a notebook drive and put them in another drive of the same type. A company that planned to do that could even be prepared with on-hand drives of all the sorts used by their targets - how many different drive manufacturers does Dell for example use in their lightweight laptops, the kind given to tech-savvy CEOs?
Some of it may be urban legends, but I have absolutely no doubt that there's a significant amount of laptop theft and the like going on out there. Encryption that was maybe a little bit of a headache (a fob on a keychain) but that couldn't be skipped because it was too much trouble to enter a password every time would go a long way to improving things.
|
Wes Felter
|
11
|
 |
|
07-30-2003 04:41 PM ET (US)
|
|
Cory, I only vow not to write blogs at work; reading them is part of my job. Fencepost: My ThinkPad requires a password before the hard disk can be accessed; removing the disk and putting in a different system (supposedly) can't defeat the password. The data probably isn't encrypted, so if you took out the platters you could read them, but this system is probably semi-useful for business users. (I believe the Xbox uses a similar system, but the geniuses at MS gave every disk the same password...) WiebeTech's FireWire Encrypt is similar to this Abit product, but FW instead of IDE and it's based on a password instead of a dongle.
|
Avi Bar-Zeev
|
10
|
 |
|
07-30-2003 02:35 PM ET (US)
|
|
Edited by author 07-30-2003 02:36 PM
Possibly showing my ignorance of deeper cryptography... My first thought is that keys are weak if you have both the full cyphertext (the encrypted disk) and part of the cleartext (the known portions of the disk) at hand.
The part of the cleartext your know are the standard headers written to every drive for each OS and filesystem. You can assume certain volume contents and even guess your way into the root directory structures. So producing the key wouldn't be as hard as, say, a single file on the disk that had just its [presumably unguessable] contents encrypted. My guess is that some of those MS-Word "password recovery" programs use this principle since word files have known structure too (and weak keys, I'm sure).
On the fob approach, if they can get the laptop or harddrive, they can probably get the fob too. The case where it works is where someone steals your laptop when you're not looking, not that this doesn't happen often enough. My old CEO once left his handspring on the curb...
|
Eli the Bearded
|
9
|
 |
|
07-30-2003 02:16 PM ET (US)
|
|
Interesting, but probably not useful for the stated purpose of keeping "the RIAA away from your Kazaa files". If the RIAA knows you have Kazaa files it will likely be because they saw your IP download them.
|
Fencepost
|
8
|
 |
|
07-30-2003 02:14 PM ET (US)
|
|
I've been hoping to see something like that for laptops for years - you get a laptop with a code and 2-3 fobs; the fob has to be in for the system to be powered on or used. When the system starts the key value is copied from the fob to an internal register that's used in encrypting/decrypting all HD traffic. Removing the fob blanks the screen and locks the keyboard, etc. (or suspends the system) until it's put back in.
Allowing the fob to be removed without requiring a shutdown makes it much more convenient and practical (no shutdown needed for bathroom breaks), and while it means that someone could in theory open the laptop and try to read unencrypted disk traffic, they'd have to do it by stealing the laptop, opening it while it was running, soldering connections to running chips, etc.
Comercial purchasers could presumably get a system that used multiple keys (allowing a master) or one where they got copies of all fobs for the systems they'd purchased. Extra cost to include it in a system would probably be less than $10, selling price would probably be $100 higher, and I'd buy one if so.
I suspect that the UK intelligence services would be waiting in line for these given their fairly well-publicised past problems with lost systems.
|
CNO
|
7
|
 |
|
07-30-2003 02:08 PM ET (US)
|
|
|
spiffy wiffy
|
6
|
 |
|
07-30-2003 01:13 PM ET (US)
|
|
As Bruce Schneier says: "Snake Oil."
The devil is in the details and there are no details here. <ol> Any cryptosystem which can only keep a supercomputer at bay for a few weeks is pretty crappy to begin with. There are the practical problems of key management. Where does the key(s) reside? On the motherboard in NVRam? On a removable fob? On the Drive's NVRAM? In the head of the end-user? There are problems of scope. Encrypting a drive at the level of the controller might be a good idea. But the controller only knows about raw disk blocks. It knows nothing of the filesystem built on top of it and, therefore, can do nothing to selectively control access to individual files. Even "Kazaa" files. There are problems of trust. I can't see/fix/modify the code because its burned into the IDE controller. How can I trust it? For all I know, the manufacturer has had to cut a deal with various governments to hand over master keys in order to get an import license.
Conclusion: You're better off securing your computer using technologies you can see and software which is open to scrutiny.
|
Cory Doctorow
|
5
|
 |
|
07-30-2003 01:01 PM ET (US)
|
|
But I should point out that you're breaking your vow not to read blogs during working hours! I'm tellin' IBM!
|
Cory Doctorow
|
4
|
 |
|
07-30-2003 01:00 PM ET (US)
|
|
I'll take your word for it, Wes.
|
Wes Felter
|
3
|
 |
|
07-30-2003 12:57 PM ET (US)
|
|
This doesn't have anything to do with trusted computing.
|
Suppafly
|
2
|
 |
|
07-30-2003 12:27 PM ET (US)
|
|
who wants to bet there is a secret key to unlocks all of them.
|
wlodekf
|
1
|
 |
|
07-30-2003 11:52 AM ET (US)
|
|
what a bulshit.. if nobody can access you hard drive, how are you gonna share you files with others?
|