Luck be a lady, tonight
"With luck, the code will be carried on by others."
The researchers behind the security audit of the TrueCrypt disk-encryption software have completed their work and say they have found no evidence of any deliberate backdoors or serious design flaws in its code. "Based on this audit, Truecrypt appears to be a relatively well-designed piece of crypto software," crypto boffin …
... that the recommendation to use Bitlocker was a red herring, or rather a yellow canary. That particular mention was what made many people quickly jump to the conclusion that the government had somehow "gotten to" the developers. It was deemed so extraordinarily unlikely that they'd advise people to use MS's built-in encryption that the only explanations left were either developer burn-out (not an unreasonable theory), or government pressure. Of course we can't really know for sure, especially since Microsoft's source code is not open to inspection. But then the Truecrypt developers would not have had access to the MS source either, and so would have been making that recommendation based on blind faith in MS -- unless they were actually MS employees or had been MS contractors who got access under an NDA.
Oh damn, my tinfoil hat has overheated. Have to sign off for awhile to let it cool down.
>"TrueCrypt Windows version relying on the Windows Crypto API is "Bad Design", but when Bitlocker also relies on the Windows Crypto API, that isn't bad design?"
No, that's not it; it's the particular way in which TC uses the Crypto API which is bad design: namely, the part where if the call to the Crypto API's initialisation function fails, TC just carries on without using it, when it should warn the user that there isn't enough randomness available from the system for secure key generation.
"Si has it gotten any better?"
The blog you are referring to is by McAfee The Crap Foister - they'd like people to buy *their* Disk Encryption. Please keep that in mind.
The blog starts with two false points - Bitlocker CAN work without TPM.
Smart Cards are also possible. And Bitlocker can be enforced through GPO. Cold Boot/Firewire/Thunderbolt attacks work with all encryption software. Backing up a recovery key or other token is important unless all data is in the cloud already or your backups are always up to date. I'd be vary of storing the recovery data on Onedrive though...
Bitlocker isn't the best solution out there but it works well and is easy to enable. I'm not using it - my laptop has an SED and fingerprint reader. Works seamlessly, instantaneous de-/encryption, doesn't tax CPU. Recommended.
If an encrypted computer is stolen I'm pretty certain no ordinary thief can access any data if the preboot password is well-chosen - no matter which FDE is in use.
Sandbitz I misread your post for a second...I was about to say "But...Bitlocker CAN work without a TPM if you're prepared to carry a USB stick around with you" and then realised that's actually what you were saying. So have an upvote for that and the rest of your post.
McAffee is not saying the contrary. The specific phrase I believe you are referring to states :
"To use Bitlocker without adding additional authentication, you need an enabled, owned TPM1.2+ hardware chip"
That clearly indicates that BitLocker CAN function without a TPM 1.2+ chip, but if you do that, you need additional authentication.
As for Smartcards, could you source your affirmation ?
Well, the 1st greeting you get when login to new Windows 8 tablet is to upload encryption key to Microsoft's server (followed by some cryptic message regarding Bitlocker that I do not recall now). So it's in Windows 8 and surely in 10. The only problem is with key management. You're strongly warned against doing it yourself (paper, file) yet few will trust to pass it to MS. And even then, the proprietary solution comes with possibility of the master key and the whole key backup is just a smokescreen (so they don't have to give the secret away and not so much to help you when you asked for data recovery).
For the professor in the benevolent shadow of the NSA, luck is a relative thing.. for people depending on this tech for their own survival, luck can become an absolute very quickly.
Maybe it's time the crypto community stops using words that belong in the realms of superstition and magic.
"Trust the math" they say..
"... will have to be done by people living in nations other than the US (and the UK, it seems.."
Other than your desire to spread FUD, why? Everything in Truecrypt is completely legal. And it is completely legal to use it. The US might be the only country with decent encryption use protection.
Yes, the PM of the UK would like to ban encryption, but so far no one in the US gov't is getting to chuffed about it.
Well it was largely obsolete with regards to full disk encryption on modern hardware.
Only so much effort any person is prepared to do for nothing.
Probably nothing more sinister than "F**k this! Can't be arsed anymore!"
Sorry bit drunk as been out to the pub and it's my birthday tomorrow...what the hell am I doing here???
Dunno why folks are so against my "Can't be arsed anymore!" theory.
If you look at the timeline there were frequent updates to TC pretty much up till the point UEFI equipped systems started to emerge which TC couldn't support. Then after that it just trailed off for the next 18 month to two years and then the Snowden stuff broke out.
A few months after Snowden we get the whole "TC isn't secure anymore!" alert.
To me it looks like making TC work with the newer hardware was maybe just too big a job and the folks had maybe decided to move onto other (paid) work. Hence the long time slacking off of apparent development.
The only slight conspiracy bit might have been a sponsored effort by someone to put some doubt in the product. If folks are looking to encrypt their laptops all of a sudden and you have a commercial product you'd want to eliminate the best free alternative. Especially one that hadn't been updated for some time. Easy enough to do.
No real need for the NSA to be involved really. I'm sure Symantic etc. have had an upturn in their enterprise encryption products.
Sometimes the simplest theory is the one.
Yawn. Descartes' evil genius: you can't "trust" anything you believe you perceive or conceive. Security is risk assessment; there's no such thing as complete trust. Saying "well, we can't really trust the audit" is sophomoric - obvious from first principles and entirely unhelpful.
Seems clear now that the devs didn't abandon the project because the code was defective or compromised. That leaves the canary theory. Two possibilities: (a) they'd been tracked down and received an actual warrant/national security letter ordering them to turn over their code and credentials for the download site; or (b) there was an effort to recruit them into US military intelligence (a/k/a NSA -- need to keep emphasizing that NSA is NOT a civilian agency, it's part of the DoD, even most reporters don't seem to appreciate that).
Either way, it's a bad sign for the Government. If they had the capabilities they claim in the promotional materials leaked by Snowden (a lot of what we've seen were generated to secure or maintain funding for various programs), then they would not have to go either route. But they clearly don't. If they did in fact approach the Truecrypt devs, that would mean that the Government doesn't have the capability to compromise projects like Truecrypt on their own. That probably boils down to raw computer science talent. For all their recruiting of script kiddies and employment of high-priced consultants, the US Government still doesn't have the talent to overcome existing privacy protection tools.
I'd say that's good news.
>"Seems clear now that the devs didn't abandon the project because the code was defective or compromised. That leaves the canary theory."
It also leaves the "just got sick and tired of it" theory. So there are more than the two possibilities you list.
>"(a) they'd been tracked down and received an actual warrant/national security letter ordering them to turn over their code and credentials for the download site;"
(i) USA NSLs don't apply to citizens of foreign countries, which is what the TC devs are.
(ii) What exactly would be involved in "turning over their code", when it's all been openly published all along?
>"(b) there was an effort to recruit them into US military intelligence (a/k/a NSA"
(iii) I don't think the NSA recruits foreign citizens resident abroad.
>" If they had the capabilities they claim in the promotional materials leaked by Snowden (a lot of what we've seen were generated to secure or maintain funding for various programs), then they would not have to go either route"
(iv) Nothing in the Snowden revelations suggests for one moment that they claimed to be able to break Truecrypt. Quite the opposite: they made clear that TC, along with Tor and Tails, were a serious threat to their abilities.
>" But they clearly don't."
Yes, indeed they do not, as Snowden made clear.
>" If they did in fact approach the Truecrypt devs, that would mean that the Government doesn't have the capability to compromise projects like Truecrypt on their own."
As explained above, there's no reason to suppose that happened. And even if the TC devs had been Americans, and the gov't had approached them, what exactly would the government ask of them? Any backdoor would be liable to be spotted in the source code, and TC doesn't auto-update, so all existing versions would remain fully secure.
"Seems clear now that the devs didn't abandon the project because the code was defective or compromised. That leaves the canary theory. "
No, it leave the actual reason for it: a bunch of self-entitled users bitching about so called "missing features", and then spreading rumors of the developers supposed incompetence. The forums and mailing lists were full of this bitching just prior to the release.
Telling those bunch of freetards to go use Bitlocker was perfect. The developers knew their own code was good, and probably still are developing it themselves.
For another tales of freetards bitching to developers: http://sqlite.1065341.n5.nabble.com/SQLite-and-McAfee-Anti-Virus-td15283.html
Obviously the best way to find if there is actually a back door is to install the software, then write on a twelve year old girls/boys facebook page "I'm a peado! come work for ISIS with us in Syria on our new chemical/biological ordnance project". (a bit like Computer Futures sell speech, but less seedy). That should draw attention. Then you can see if they can break into you HDD.
"Crack Security team"?
The first report released - after a fairly long time - could be produced almost completely automated with some simple and free tools.
The final report was even worse: They missed a bug in the Serpent implementation and a few minor issues too. I wonder if they have ever heard of Valgrind?
They then recommend eliminating cipher cascading as being too complex! Cascading is a way to ensure your data is still safe if an algorithm is broken.
As someone noted above, a bit of injustice here. It would frost your balls that all the money they collected to produce the two very amateur reports was given to some auditors instead of the person who deserved it. Reproducible builds were not even looked at. I found out that someone did that for free earlier. I hope they do not throw good money after bad and pay for that as well. Talk about ingrates!
The final insult was that in the summary, the very issues that really needed to be analyzed were out of the scope of the audit! Now they will say when their report is critiqued oh, we did not analyze that...
Sorry about the rant - a very long day.
The US based security auditors with close ties to the NSA said of TrueCrypt "... they have found no evidence of any deliberate backdoors or serious design flaws in its code". "Truecrypt appears to be a relatively well-designed piece of crypto software."
Ahhh yeah... sounds legit. o_O
Biting the hand that feeds IT © 1998–2019