I have no idea why you think it's productive to litigate the difference between "block-level encryption" and "full-disk encryption", but if it makes you feel better we can just pretend we switched the terms, because my point applies equally to them --- they're synonyms.
I also have no idea where the "I'm telling the world there shouldn't be a tool like Truecrypt" came from. I think you've misread me.
If you're talking about a security product -- which TrueCrypt is -- the first metric you have to concern yourself with is: does it keep you secure? The user experience and the adoption and the performance and all that other fun stuff is irrelevant if the product doesn't do the one thing that every user unequivocally requires of it.
So yes, it's not just about the crypto...when the crypto works. But when the crypto is insecure, which is what tptacek is saying, then yes, it is ONLY about the crypto.
NB: I'm plenty qualified on UX and general technical matters, but on whether crypto is secure, I defer to the experts.
My only point has been that Thomas, et. al. have been telling us we don't want something like TrueCrypt, despite the fact that we very clearly do. His suggestion of "just use PGP and FS level encryption" is absurd, but NOT from a crypto standpoint (I, like you, defer to Thomas and the other experts on the integrity of the crypto itself). It is, however, absurd from a UX/workflow standpoint.
Horseshit. Round 1 and Round 2 of the audit share technical members. The guy leading the actual crypto review work has been looking at Truecrypt for more than a year. And Matthew Green, who coordinates the whole audit project, just wrote that he and his students have been reviewing Truecrypt's crypto for months.
They did not "only get their act together today". They've thought about Truecrypt far more rigorously than you have, and for far, far longer.
You've been almost completely unable to explain in technical terms what "UX" you want from sector-level crypto that you couldn't get from filesystem crypto. When pressed, you in effect say "yeah, well, name a tool that does that".
The fact that your only options today are [insecure, easy] and [secure, difficult] does not mean that there is no [secure, easy] option possible. But militating in favor of insecure crypto goes a long way towards hiding that possibility from everyone.
This isn't a pedantic point. Ross Ulbricht just got reamed in federal court because a simple physical arrest compromised virtually every secret he had. Why? Because he was relying on sector-level all-or-nothing crypto. By encouraging people to rely on tools like Truecrypt, you are, in a very small but real way, endangering them.
Today was the day Matthew Green released an update on his blog.
I was just reading it, and that's exclusively I was referring to. I look forward to the results and am grateful of the time they're spending. I hope they find nothing.
This isn't a pedantic point. Ross Ulbricht just got reamed in federal court because a simple physical arrest compromised virtually every secret he had. Why? Because he was relying on sector-level all-or-nothing crypto.
There was a bit more to it than just that. He could have used block-level encryption relatively safely if he'd made a series or hierarchy of Truecrypt containers and mounted them only when needed, rather than putting everything on just the one block device.
More importantly, his physical security was lacking, as he hadn't properly considered the threat model. If he'd been working in a secured area (like a locked room) where open laptop snatching was infeasible, that would have given him enough warning to close the lid, and maybe pop the battery out. Albeit still vulnerable to a cold boot attack, if law enforcement have such capacity.
TrueCrypt lets you create fixed sized encrypted volumes, and allows you to decrypt those volumes on any of the three major OS platforms.
There's nothing special about TrueCrypt in how it performs the encryption/decryption (or so we're told), but no tool besides TrueCrypt allows such a flexible approach.
And it's you who refuses to accept that [secure,easy] can exist, because it'd make you irrelevant. It's a completely silly stance to take, but it's yours.
But hey, at least I've wrung your opinion on TrueCrypt out of you:
> By encouraging people to rely on tools like Truecrypt, you are, in a very small but real way, endangering them.
For posterity, in case you edit it away.
Which leads me to the question: Why are you even involved in the TrueCrypt audit, if you think it's a bad idea to use such tools?
P.S. Ulbricht was caught because the FBI owned TOR, and that's about it. Maybe your indignation towards TrueCrypt should consider Snowden's use of TrueCrypt to evade the combined allied world's intelligence community.
Would you like to put money on whether my opinion about Truecrypt is identical to Matthew Green's and Kenn White's, or would you like to include them in your critique?
It's amusing that you feel you've "wrung out" of me something one of the few things I've recently blogged at length about.
I've already answered that question, directly, on this thread.
And no, I blogged "don't use sector-level crypto". In a post literally titled "You Don't Want XTS". Under the subhed "Disks Are The Last Thing You Want To Encrypt". As in, "the last thing in the world".
> This piece is written for software designers, not end-users. If you’re an end-user looking for crypto advice: use Truecrypt, use Filevault, use dm-crypt
It's a great write-up, I read the whole thing. You clearly understand the domain well.
I really just don't get why you'd, in one breath, decry XTS, and then in that same breath, recommend people use TrueCrypt, which is, as you call it, "the best-known implementation of XTS".
Maybe just lead me to the water on this one. It's really the only thing left unresolved in our conversation.
Block-level encryption is a terrible, terrible approach for many reasons (which 'tptacek has referenced a million times). However, Truecrypt is the best such implementation, and it's a required approach in certain cases. You should be doing crypto at the application/filesystem level; if you can't, use Truecrypt. This isn't contradictory advice.
That's not just what he said, he also said, "By encouraging people to rely on tools like Truecrypt, you are, in a very small but real way, endangering them."
You also completely changed the comment I originally replied to. I much prefer your new comment, though my fundamental issue with the fact that you're working on the audit of software you think is dangerous to promote remains.
I don't know what you think you're accomplishing by saying "nuh uh" like this. You've done it a few times, and I don't understand, in any of these cases, why you think anyone would think you'd say otherwise.
If you'd care to elaborate beyond, "nuh uh", I'm sure we'd all be glad to hear it.
The "No you haven't" was in regard to the fact that you haven't answered why you're involved in TrueCrypt at all, if you don't think it should be used.
Well, you have edited a lot of your comments, so perhaps you did include this information in a later edit?
Edit:
Having re-read the thread, you haven't explained why you're involved in the TrueCrypt audit, or why you recommend folks use TrueCrypt if you think XTS is bad.
I'm not being "squirmy". You're playing a semantic game with the word "disk". The technical issue with FDE is that it works at the level of blocks, and so lacks information about message boundaries or the storage flexibility needed to (a) randomize the encryption and (b) store authenticators. Encrypt a physical disk, encrypt a file that pretends to be a mountable volume, same issues.
I get that not everyone understands the technical issues in designing storage encryption, but don't take that out on me.
I also have no idea where the "I'm telling the world there shouldn't be a tool like Truecrypt" came from. I think you've misread me.