Feeds

back to article Crowdfunded audit of 'NSA-proof' encryption suite TrueCrypt is GO

A fundraising effort to pay for an independent, professional security audit of TrueCrypt, the popular disk encryption utility, has raised enough money to pay for an arguably long overdue audit of the security software. TrueCrypt is a widely used utility that encrypts and decrypts entire drives, partitions or files within a …

COMMENTS

This topic is closed for new posts.

Page:

Great, fine...

Give the source a code review.

But don't neglect to review the sources of the libraries and compiler every whit as closely.

This was shown to be necessary long ago. And even then, don't trust too much:

http://cm.bell-labs.com/who/ken/trust.html

5
1

Re: Great, fine...

Examining the source code of the compiler won't help you - if it has been bugged in the way Ken Thompson describes in the article you linked to, it will replicate a trojan horse even if a clean copy of the compiler source code is compiled with a bugged compiler. An attacker doesn't need to ship the bugged source code - a bugged binary and clean source will suffice just nicely.

This is the point so many open source nuts simply don't get: If it were as easy as simply recompiling the code, it wouldn't be an issue. If you download a Linux distro and recompile everything yourself with the compiler in the distro, you are trusting the distro not to have embedded a backdoor in the compiler they have shipped. It matters not one jot whether the compiler is open source or not.

That TrueCrypt is open source simply makes it easier to target with a compiler Trojan, because the target is a known quantity. How many open source enthusiasts - in all seriousness - thoroughly check the compiler that shipped with their system?

1
4
Silver badge

Re: Great, fine...

@corestore, @Oliver Jones

It's not a major problem given the number of c compilers out there. Or write your own (in jscript/haskell/lua/pascal/elisp whatever) and bootstrap from that. If supposed professionals can't deal with that, they aren't professionals. Jeez.

> This is the point so many open source nuts simply don't get

We don't get it because we're stupid but you're clever enough to point it out to us. Thanks...

3
3
Silver badge
Headmaster

Re: Great, fine...

This is the point so many open source nuts simply don't get

Yeah, about that, Mister invective-hurling Superiour Intelligence ... you may not be as superior as you think:

Countering "Trusting Trust"

"Now, this technique only detects when the binary doesn't match the source, so someone still needs to examine the compiler source code. But now you only have to examine the source code (a much easier task), not the binary"

2
1
Bronze badge

Re: Great, fine...

To be honest, it's even money that the NSA funded the effort.

They've already reviewed the compiler and libraries long ago and kept up.

Or did you think that SELINUX was only for show?

0
0
Bronze badge
Big Brother

Audit

But who's doing the audit?!!

1
0
Coat

Re: Audit

The NSA.

2
1

Re: Audit

Nope, a company called 'TNNSA'...

(Totally Not the NSA)

3
0
Bronze badge
Facepalm

Re: Audit

From TFA:

"the name of the firm who submitted the tender [to perform the audit] is being withheld pending evaluation of its offer and those from potential rival bids"

Reading comprehension is a useful skill.

2
0
Bronze badge

Re: Audit

Who's doind the audit?

Quiscus Todiet, of course.

0
0
Big Brother

Still not "secure"

Why was the electronic computer invented, to crack encryption systems!

Encryption is a time and effort based method of making it tough for the other side to read, it is not a truly secure system.

Ignore the end of the universe time quotes from people like the FBI in how long it takes to crack an encryption, they are usually asking for bigger budgets when they make those statements.

In real life, the crack time for an RSA-155 (512 bit) message was 5.5 months, using a couple of hundred PC, and a server with a lot of memory, for 1999, i.e. 4gb.

(a) PC's have got a lot faster since then

(b) the more compute cycles (resources) you throw at it, the quicker it breaks

(c) the NSA has a LOT of compute cycles to throw at things, even the average 3rd world banana republic has more compute power than a 1999 research team.

So when wanting to secure something, think how long the data is sensitive, and plan accordingly.

4
0
Silver badge

Re: Still not "secure"

"So when wanting to secure something, think how long the data is sensitive, and plan accordingly."

And also remember that lots of locks are more secure than just one lock. Encrypt the contents of the files, and store them in a differently encrypted container on a differently encrypted disk. And make sure it's physically secured too. And don't use the same password for all them all.

Layers. Security is substantially stronger if you do it in layers. Like Ogres, and onions. Onions have layers, Ogres have layers.

3
2
Anonymous Coward

Re: Still not "secure"

http://security.stackexchange.com/questions/18087/is-multiple-encryption-a-good-idea

Basically multiple encryption isn't any mroe secure than a single good layer of encryption.

0
3

Re: Still not "secure"

Also, consider this little bit...

Even information that in itself is no longer sensitive may have meta information that IS sensitive.

The cleartext of a previously encrypted message may be used to help cracking the cipher used. Sucks if other messages was also sent using that cipher...

(This is why one-time pads should never be reused... )

A document may contain verbatim text from a classified or encrypted message which will reveal the source of the intercept. And if you know the source, you can probably figure out which documents he had access to, and which may have been compromised.

Documents that always have the same beginning really sucks...

German Enigma encrypts had a pair of characters that was repeated in the very beginning of each message. Ordnung über... whatever... That is what the Polish military intelligence realised and which was the key to cracking the messages.

(The characters were 'offset codes' to be used to modify the 'daily code' into a one-time code.)

The Enigma system was in use until well into the 60s...

2
0
Silver badge

Re: Still not "secure" @AC

"Basically multiple encryption isn't any mroe secure than a single good layer of encryption"

I read the page your link took me to, and it didn't say that at all. Read harder.

1
0
Bronze badge

Re: Still not "secure"

Increasing the length of the key by one bit, will double the time needed to perform a brute-force attack.

So, in your example a 513 bit key would require 11 months to crack.

A 514 bit key would require 22 months to crack and so on, a 522 bit key would require 5632 months (over 450 years).

I'll leave it up to you to work out how much more secure a 1024 bit key is, assuming that the cypher can only be attacked using brute force, but one starts hitting 'end-of-the-universe' type timespans quite quickly.

And remember, I'm not the FBI, and I'm not asking you for a budget.

0
0
Silver badge

Re: Still not "secure" @AC

Taken from that post, first paragraph of the first answer.

Personally, I would avoid multiple encryption protocols most of the time. It adds significant extra implementation complexity without making your data any more secure in the real world, unless the encryption protocol you are using is ultimately broken or becomes computationally feasible at a later date to break.

0
0
Happy

Re: Still not "secure"

Pairs on enigma was a possible reason for the allies not shuting down the german weather (wetter) station on spitzbergen.

As nice reliable germans sent the wetter report first thing in the morning.

Two E's and two T's in the opening of the message. A predicatable known plain text decrypt.

Given how predicatable XML files are in their start, how many people are transmtting encrypted XML?

PS

the UK government was quite happy to supply second hand enigma after the war, with out mentioning we had cracked it. So NSA backdoors are not really a new idea.

1
0
Happy

Re: Still not "secure"

"A 514 bit key would require 22 months to crack and so on, a 522 bit key would require 5632 months (over 450 years)."

The RC4 encryption system relies on pairs of primary numbers, which do not necesisarily double as th key length doubles.

You also need to factor in Moore's Law, the RSA-155 crack used single threaded CPU's at 200mhz - 300mhz, a modern cheap pc is quad cored at 3ghz+, and today we would not use a CPU, we would use a graphics card which is roughly an order of magnituded better at crunching numbers, and also follows moore's law.

The process used in brute force of a key also tends to use a lot of hard disk space, so the move to SSD's, would also make a significant dent on time (i.e. typical HDD 10ms access, SSD 0.1ms access)

The paper on the crack of 768 bit encyption can be found here eprint.iacr.org/2010/006.pdf, from back in 2009

the relevant bit of the summary is...

"Our computation required more than10^20 operations. With the equivalent of almost 2000 years of computing on a single core 2.2GHz AMD Opteron, on the order of 2^67 instructions were carried out. The overall effort is sufficiently low that even for short-term protection of data of little value, 768-bit RSA moduli can no longer be recommended. This conclusion is the opposite of the one arrived at on [39], which is based on a hypothetical factoring effort of six months on 100 000 workstations, i.e., about two orders of magnitude more than we spent"

PS.

note not GPU compute cycles, which is the way to go for this type of crack.

1
1

Re: Still not "secure" @ despairingcitizen

This is why messages should be compressed before they are encrypted: That way, you can safely remove speculation about the content of the file, because many of the characteristics will have been mangled by the compression algorithm prior to encryption - making use of cribs much harder.

0
1

Re: Still not "secure" @ despairingcitizen

and the compression alogrythm is different to a second layer of encryption how?

compression is predictable, hence how you decompress.

If you are looking to make known format files more difficult to decrypt, then you need to have a random length of salt at the front and back, as a starting point, and work out from there.

You also need to factor on that the data in system to system comms is inherently predictable in most cases, for example the claims number for an insurance claim follows a set pattern, there are some business operations where looking at a file reference number, at any one of a couple of hundred operators, I can identify exactly which software they are using for administration.

0
0
Silver badge

Re: Still not "secure"

Moores law will soon run into the ground and that statement will be true for CPU and GPU processors alike. So brute-force cracking of encryption may get very hard in the near future. The big question is whether human ingenuity can come up with fundamentally new ideas on improving computers beyond Moores law.

1
0
Bronze badge

Re: Still not "secure"

Nobody should have been using RSA-1024 (or less) any time recently, and I seem to recall that RSA-2048 becomes standard at the end of this year. RSA-768 was reported broken in 2009 - in about 2.5 years - and the authors projected that RSA-1024 might be solvable by their methods around 2020. RSA-2048 or larger should be good as long as I care about keeping my secrets.

RSA-1024 was reported cracked by a server attack based on creation of controlled power faults. Feasibility of a brute force attack is doubtful.

2
0
Silver badge
Facepalm

Re: tom dial Re: Still not "secure"

"....and I seem to recall that RSA-2048 becomes standard at the end of this year....Feasibility of a brute force attack is doubtful." Yeah, but decryption by waterboarding works in minutes! Oh, sorry, did that make the sheeple's paranoia spike?

If you are of serious interest to the authorities then they will arrest you and then play at decrypting your stuff later. In the UK you will be formally asked to decrypt it - if you don't you get a separate charge and conviction and your unwillingness can be pointed to in court regarding the original suspicions. If you are unlucky to be the subject of a rendition to a "friendly" despotic regime rather than arrest, then they'll probably go to more "aggressive" methods after about a week of your not being co-operative. At that point you'll probably tell them everything you had encrypted plus how you used to steal cookies when you were a kid.

But the simple fact is the vast majority of you sheeple simply are of zero interest to anybody, let alone the secret squirrel types, so encrypting anything is simply pointless beyond how it helps you cope with your paranoid delusions.

1
1

Re: Still not "secure"..new ideas on improving computers beyond Moores law.

Interesting who much money is being put into quantum computing by organisation with ties to the likes of GCHQ and the NSA

Also the introduction of 3D IC's will make some dent on the processing power.

We also have the pure maths mob finding better ways of cracking keys.

I would observe that GCHQ sat on Public Key Cryptography for over a decade before it was "discovered" by RSA, hence one could wonder at what else they are sat on.

1
0
Silver badge

Re: tom dial Still not "secure" @Plump & Bleaty

Oh we agree here plumps, Farmer has all the power so us sheeple had best suck it up, yes? And you're all behind him.

Strange but plumps just said in another post

"Capitalism and democracy are pretty poor systems but they allow advancement and development and at least attempt at equal opportunities, BECAUSE it takes into account human nature, making them the best compromise yet."

So, plumpywumpy, are you in favour of democracy where we the people hold the government to account with the prerequisite that we know what they're actually doing at all, or not?

> Yeah, but decryption by waterboarding works in minutes!

You endorse torture as well?

> the vast majority of you sheeple

Awww, plumpy, we both agree that you're the biggest, plumpest, bleatiest, obsequious-est sheep around here.

0
1
Silver badge
FAIL

Re: Loser Re: tom dial Still not "secure" @Plump & Bleaty

"...... are you in favour of democracy where we the people hold the government to account with the prerequisite that we know what they're actually doing at all, or not?....." I'm not, simply because there is nothing in the democratic idea that says everyone has to know everything. The democratic process is about how we elect our leaders, not about constraining how they govern to suit your petty paranoias. Should the idea of how they are governing, be it their foreign or economic policies or lack of transparency, upset you then the democratic process allows you to try and convince the majority to support the idea of complete transparency, it does not make it mandatory. And until you do manage to convince the majority you're just urinating into the wind. But you may want to bear in mind that the majority seem to have bigger problems to worry about than your pet delusional peeve.

".....You endorse torture as well?....." Most definitely. No, seriously, they could have torn out Khalid Sheikh Mohammed's fingernails, wired his testicles to the mains, and I wouldn't be worried. The guy planned the mass murder of civilians in 9-11, let alone the misery he inflicted on his own people, and his "torture" saved others, so I have zero problems with it. Your spoonfed opinion no doubt differs, but that's your right. You enjoy your moral hobbyhorse, I prefer to deal with realities, so please do realise that the hysterical, ego-stroking lecture you have in mind is, IMO, simply pointless moral posturing and more self-delusion.

0
3
Silver badge

Re: Loser tom dial Still not "secure" @Plump & Bleaty

@plump 'n bleaty

> Should the idea of how they are governing, be it their foreign or economic policies or lack of transparency...

mmm, but my point is that without the relevant information (deriving from substantial transparency, not necessarily total transparency but a good in-principle idea of what's occurring), we tautologically cannot know how they are governing and therefore cannot cast an informed vote. Your idea of democracy is not something I recognise nor find attractive.

Let's turn it around, plumps, suppose the government withheld all information except what it wanted us to know, would that be acceptable to you? Where is the line drawn, between total transparency and total control of info released to us?

> ".....You endorse torture as well?....." Most definitely. No, seriously [...]

Christ.

Ok, I always feel that if one endorses an action then one should be morally willing to perform that action. Do you agree, if so do you accept that you would personally pull a person's fingernails out and subject them to electrocution designed to cause intolerable pain?

> please do realise that the hysterical, ego-stroking lecture you have in mind ...

More interested in comprehending how far down this path you've gone. I don't expect to change your mind.

0
0
Silver badge
FAIL

Re: Loser Re: Loser tom dial Still not "secure" @Plump & Bleaty

"....but my point is that without the relevant information (deriving from substantial transparency, not necessarily total transparency but a good in-principle idea of what's occurring), we tautologically cannot know how they are governing and therefore cannot cast an informed vote....." Complete male bovine manure. You can see the RESULTS or lack of from policies. A simple example is Treasury spending - I do not need to know the exact sums used by the Treasury to calculate the NHS budget to know there are waiting lists for treatments. Even with a higher education and some understanding of budgets I could not take Treasury analysis and do the calculations myself because I am not a trained economist or medical practitioner or surgeon or hospital equipment buyer. I can see when it is working to an acceptable level when I can walk into my local GP's surgery and get an appointment to see him. Similarly, I do not need to know exactly what the SAS, MI5, MI6 or the GCHQ (and by extension, the CIA and NSA) are doing to be happy they are defeating terror attempts because we are not seeing the daily bombings and other attacks rampant across many other areas of the World where the terrorists can act more freely. Please do try and pretend AQ and their chums would not like to be exploding a bomb at least weekly in London and Washington if they could. Please also pretend it's not like they have plenty of supporters and sympathisers in both the US and UK that would help them if they thought they would not get caught.

".....Your idea of democracy is not something I recognise nor find attractive....." That is because you do not understand democracy and instead conflate your spoonfed baaaah-liefs on transparency with democracy. You are attempting to insist that democracy cannot be real democracy unless there is complete transparency, which is obvious horse manure. Voters are quite capable of making electoral choices on the results they see rather than having to read and understand every single minute action, and there are endless numbers of experts and analysts that can provide summaries for the public which can be openly challenged and debated. Your problem is you cannot accept that the majority are quite happy not knowing the minutae of each and every political action, or each and every military action or each, and each and every clandestine action, because they understand that revealing everything (especially the clandestine actions) would actually be damaging to us.

".....Let's turn it around, plumps, suppose the government withheld all information except what it wanted us to know, would that be acceptable to you? Where is the line drawn, between total transparency and total control of info released to us?...." Except that is exactly what they do. In the case of the UK, civil servants working for HMG get to decide what should be released and what should be classified. They operate to set rules, it is not done on a whim, and unsanctioned release of classified material is punished under the Official Secrets Act. Again, if the MAJORITY of voters had a real problem with that then a political party would adopt its revocation as the central plank of their electoral policy to ensure their electoral victory. But you were obviously too busy sticking your head up your rectum to notice the electorate have a lot more burning issues they want politicians to work on. I know you want to pretend that UK politicians are simply not listening to the public but the fact is the three main parties spend a lot of money and time trying to gauge public opinion through workshops, polls and the like.

Once again, you are whining because your minority views are not being accepted by the majority, and since you cannot provide a rational argument as to why they should accept your views you instead insist that the majority don't understand (because you like to think we're too stupid and need our thinking done for us) and attempt to redefine such simple terms as democracy to incorporate your views instead. Busted!

"..... I always feel that if one endorses an action then one should be morally willing to perform that action....." Really? So if you have such a hankering for transparency you should be more than happy for everyone to know everything about you such as how you vote, your bank account details, where you work, what you earn, which donkey pr0n sites you no doubt peruse, how you plan to spend your money in the coming year, etc. But then I'm betting the whole transparency thing suddenly stops then, right?

0
4
Silver badge

Re: Loser Loser tom dial Still not "secure" @Plump & Bleaty

@Plump & Bleaty

Remarkably long post, plumps, derailed by your misrepresentation of my clearly stated position, and failure to address my question.

So,

> A simple example is Treasury spending - I do not need to know the exact sums used by the Treasury to calculate

Your words: "the exact sums". Indeed, exact is unnecessary, an approximation is necessary as you tacitly admit (by using the word 'exact'). So you do need information.

> I do not need to know exactly what the SAS, MI5, MI6 or the GCHQ (and by extension, the CIA and NSA) are doing

Again, the word 'exact', implying that you do need enough information to know.

> because we are not seeing the daily bombings and other attacks rampant across many other areas of the World where the terrorists can act more freely.

Prove that without such tapping these events would be daily, and that they are being prevented by mass undeclared surveillance. Oh you can't? Because you don't have the necessary info? So you take it on trust Farmer is doing a good job!

Extra salt lick for Bryant tonight! You've earned it plumps.

> You are attempting to insist that democracy cannot be real democracy unless there is complete transparency, which is obvious horse manure.

Oh, agreed about the horse manure! because I never claimed complete transparency was desirous. You even quoted my exact words. Let me quote your exact quote of my words back at you: "....but my point is that without the relevant information (deriving from substantial transparency, not necessarily total transparency but a good in-principle idea of what's occurring), we tautologically cannot know how they are governing and therefore cannot cast an informed vote.....

Read it again pillock.

I said

> suppose the government withheld all information except what it wanted us to know, would that be acceptable to you?

It would appear that your answer is yes, because they already do ("Except that is exactly what they do."), so you admit that we are not being given the information we need and you're happy with that. OK.

Right, now the question you dodged, the one about whether if you endorsed torture you'd be willing to do it yourself. I'll repeat it, broken down into two for ease of comprehension:

* Do you agree that if a person is willing to sanction an action that they (if they are physically capable of it) should be morally willing to perform that action personally?

* If the above is true do you therefore accept that you would be willing to personally torture another person?

It's a simple and important question, please answer it directly without spewing clouds of chaff and attempted misdirection.

1
1
Silver badge
FAIL

Re: Loser Loser tom dial Still not "secure" @Plump & Bleaty

"....Indeed, exact is unnecessary, an approximation is necessary as you tacitly admit (by using the word 'exact'). So you do need information....." Blimey, you really are obtuse! You score zero on reading and comprehension - I stated I looked at THE RESULTS, not that I tried to do or asked for even rough sums. So that must mean you think you yourself are a complete pillock. Probably the only correct thing you've ever thought.

"....Again, the word 'exact', implying that you do need enough information to know....." You are so determined to ignore the facts you simply want to read whatever you want into what people post. I do not need the government to disclose anything about what the relevant services are getting up to as, again, I can see the results in the lack of attacks in the UK compared to other countries. That is the result that matters to me.

"....Prove that without such tapping these events would be daily...." So you do want to pretend that AQ and co would not like to attack us. I suppose it was past your bedtime when they had all the news with Bin Laden's threats against the West, since repeated many times by Al Zwahiri and the Talibints. You really are just too stupid for words, so convinced of your own "righteousness" that you have stuffed you head so far up your rectum you probably can see your own tonsils. So, do you think the London Tube bombings were the work of aliens? Just to help you fill in the yawning blanks in your knowledge, here's just a list of the Islamic terror attacks in the last thirty days (http://www.thereligionofpeace.com/index.html#Attacks), please note that whilst the majority of attacks are Muslim on Muslim violence in Iraq and Pakistan, countries such as Nigeria, Kenya, Russia, China and Turkey are all on the list. Please do explain what exactly you think it is that makes us so much more "clean" than those other countries so that no-one would be thinking to attack us?

"....I never claimed complete transparency was desirous...." I know exactly what your idea of "substantial transparency" would be and you also know it is never going to happen because it will never gain the support of the majority. It's called the secret service for a reason, and that's beacuse it requires secrecy to operate, it's just you want to stop it operating so you can feel stupidly morally superior.

"....so you admit that we are not being given the information we need...." Nope, I'm saying people like you don't need to know anything about what the security services get up to because (a) your posts expose the fact that you are simply too stupid to understand what they do or why, and (b) that you would simply cripple our ability to defend ourselves so you could feel morally superior, and (c) because your views are of the tiny majority of morons, and (d) because you may think you're important but the reality is you are just a whining cretin I wouldn't trust with scissors, let alone make actual suggestions on security policy.

"....now the question you dodged....It's a simple and important question......" You mean your attempt to drag the thread off into an area you feel your vacuous dribblings give you some form of moral superiority? OK, if it keeps you happy.

"....Do you agree that if a person is willing to sanction an action that they (if they are physically capable of it) should be morally willing to perform that action personally?...." Stupid question. I know many people that would be happy for a surgeon to perform a life-saving operation on their child but be too squeamish or too worried about making a mistake to do the same themselves, are you suggesting their child doesn't deserve the surgeon's efforts? That's probably an unfair response seeing as it will take you at least a year to work it out in your head. If you are asking would I have tortured Khalid Sheikh Mohammed, knowing what he had already done and knowing his friends intended more attacks, and that by torturing the piece of scum it could save hundreds if not thousands of lives, then I wouldn't enjoy it, but I would do it.

Now, let's try a different tack to see just how much you value your morals. A question for you - do you seriously think your "shining moral fortitude" would have made a shred of difference if you had been on one of the 9-11 planes or on one of the Tube trains on 7th July 2005? Because the truth is the people that want to attack us won't give two hoots about whether you're a handwringing moron with your head up your rectum, you're just as much a target as everyone else. Enjoy!

0
5
Silver badge

Re: Loser Loser tom dial Still not "secure" @Plump & Bleaty

Hi plumpo,

> I stated I looked at THE RESULTS

No, what you actually said was "I do not need to know the exact sums used by the Treasury to calculate...". Had you said "I do not need to know anything about the sums used by the Treasury to calculate..." your case might have been stronger.

It happens I foresaw the recent crash (that occurred in 2008) and was warning people somehting economically rough was going to happen - did you? Sometimes the results come long after the initial unwise decision.

> I can see the results [of mass covert electronic interception] in the lack of attacks in the UK compared to other countries.

Maybe. Or maybe not. That one follows another is your assumption, not a proof.

> Please do explain what exactly you think it is that makes us so much more "clean" than those other countries so that no-one would be thinking to attack us?

Actually, I have long had my concerns about terrorism and bio- and nuclear terrorism (bio is more likely). I don't dispute the rather ugly role some branches of islam are performing or the danger they can be. My question, which you don't understand, is, is bugging everyone's e-transactions the most efficacious way of dealing with it? You choose to trust government that it is, and they'll do it right. I and others want the right to decide for ourselves. And you call us sheep...

Incidentally, roll back a couple of hundred years and what what catholics and protestants did to each other. There's nothing special about what islam is currently going through, just that it's happening now. And to repeat, mass covert interception does not strike me as necessarily the most effective thing to counter it.

> I know exactly what your idea of "substantial transparency" would be

Oh? Go on then smarty pants.

> Nope, I'm saying people like you

entire para of MBZCC

> are you suggesting their child doesn't deserve the surgeon's efforts?

Wow, you confuse life-saving surgery surgery and torture. Niiiiiiice one plumpy. At least you answered the question, albeit 2nd time round.

Just a small point, you talk about human nature but you have little understanding of it and seemingly no insight into your own.

0
2

This post has been deleted by its author

Silver badge
Happy

Re: Boring Loser Re: Loser Loser tom dial Still not "secure" @Plump & Bleaty

"....No, what you actually said was "I do not need to know the exact sums used by the Treasury to calculate...". Had you said "I do not need to know anything about the sums used by the Treasury to calculate..." your case might have been stronger....." And now you're just desperately trying to split hairs because you lost the argument again. Again, yet again. The NHS is a very good example as the results are very transparent, it is a service most of us will use during our lifetimes and one which is often analysed in the news. Other simple examples are bin collections, or road maintenance. Please do pretend you know anything about budgeting for the NHS or a transport. I'll. accept that you are an expert on rubbish and recycling as your posts are full of it.

".....Maybe. Or maybe not. That one follows another is your assumption, not a proof....." And there we have the biggest of the gaping holes in what you have been told is The Truth - you cannot provide an alternative reason. You dance around it and spout male bovine manure by the ton, but you simply cannot account for the lack of attacks by anything other than the surveillance, by the effectiveness of the very actions you wring your hands and whine about. Carry on frothing and dancing, but your pointless drivel has been shown up for the vacuous ranting it is.

"....Do you agree that if a person is willing to sanction an action that they (if they are physically capable of it) should be morally willing to perform that action personally?...." Hmmm, looks like you insisted that a person had to be willing to commit an action themselves in order to allow it to happen. No surgery for you! Oh, sorry, did you just lose again, yet again, again? Why, yes you did! Dance around some more, but everyone knows your silly prattle got shot down. Again. Maybe you should try stopping to think for a minute between rants? I know it would be a novel strategy for you, just think of it as growing up just a bit. You might want to take a minute to consider if you'd be willing to spend a day emptying bins, or maybe slaughtering animals for the supermarket you buy your food at, or even just putting in a day on the farm so you can eat vegetables. Oh dear, you might talk yourself into starvation!

"..... I and others want the right to decide for ourselves...." You and a minority, the majority are fine with the decisions they make for themselves as part of the democratic proces. You know it and that's what really gets your goat - the majority have real problems to deal with, not your paranoid and delusional fantasies of being important enough to be of interest to anyone. Get over yourself. And deep down you know, come election time, your minority ranting will have zero impact. Enjoy!

"....and seemingly no insight into your own." Well I won't need to ask a vet what's wrong with you and the rest of the sheeple - paranoid conspiracy theorists trying to cope with the pettiness of their lives have been around for decades. You're so desperate for someone to be reading your emails and tracking you online because - face it - no-one else is bothered to. Oh dear, you've been busted for the boring attention-seeker you are! ROFLMAO!

0
2

Not an expert on this, but it seems to me that in order to have any credibility, the identity of the auditors must be known. On the other hand, once they are identified, the NSA / other such body can get at them and threaten all types of nastiness unless they get the result they want...

Seems to me then that we should be trying to keep the auditors identity a secret until the very moment the report is published...

3
0
Silver badge

To take it slighlty one step further, it would be even more credible if two(or more) seperate , independant and unknown to each other, teams should be given the task.

Each of the teams could then analyse and compare the results of the others eventually creating an aggregated result.

Then releasing the test scenarios in order that the greater public can attempt to reproduce the results and make their own decisions..

4
0
Anonymous Coward

That's why they had crowd-funding, to pay auditors, because obviously nobody was doing it for free.

I'm just glad this panned out and they managed to collect enough funds.

2
0
Silver badge

The free part is the biggest flaw in the perceived security superiority of OSS. Granted, the fact it can be independently analyzed is a positive thing, but if no one is doing the audits then having that option is meaningless.

The unaudited nature of a popular security product is stupefying. If nobody is doing line by line audits of security related code what makes anyone think that other, not security specific products, are being audited? Having 'more people looking at it' obviously isn't a valid point, nobody looks at the actual security parts... It's like storing all your valuables in a vault but not bothering to make sure the door locks: Why bother building the vault?

Expecting people to turn out extremely high quality work for free has always been a weakness in the system. Actually paying people is obviously going to be necessary as the community isn't going to do it for free. They just expect someone else to do it for them and they'll accept whatever you hand them. It will be very interesting to see if funding audits is sustainable, if it isn't then we're all back at IE6.

1
1

Use two rival teams

I think you need to give the job to two teams who are in competition with each other. Ideally, to two security researchers with big egos and reputations to defend.

That way, neither has any incentive to overlook something at the request of the NSA. If they do that, and their rival doesn't and spots the backdoor, then they will be shown up as either incompetent or corrupt.

It's the only way to be confident they do a proper job.

5
0
Bronze badge
Meh

audit? paid for that already...

Didn't we already pay for that with the billions we funnelled to the spook agencies?

Probably the biggest fallout from the whole SpyingSaga is that it has muddied the waters as to what is secure...

P.

0
0
Bronze badge
Pint

Once upon a time...

Once upon a time, I wrote a BASIC program to stuff (POKE) "^H" (backspace) into BASIC source code, overwriting a placeholder character. The backspaces would overwrite the actual code when LISTed, so the LIST would display whatever one wanted (fake code embedded in the REM section of each line).

With that technique, one could do anything (assorted proof of concept demos):

A program that looked as it it would print "No No No", but would print "Yes Yes Yes".

A program that appeared to LIST when RUN, and appeared to RUN when LISTed.

A program that would clearly demonstrate that 2 + 2 = 197 (or whatever).

All in BASIC. Around 1980.

Search the source code for "^H".

And think about similar techniques of source code obfuscation, matured by 30+ years of new concepts such as Unicode.

Very (very, very, ...) unlikely, but necessary for completeness.

1
2

Re: Once upon a time...

Nice idea, but most text editors aren't so cooperative when it comes to actioning ANSI control codes: They just print the code you were trying to hide, plus a load of rubbish on the screen, followed by the code you were hoping to obfuscate it with.

You're also assuming (and making a very big assumption, I might add) that the compiler wouldn't choke on those non-standard characters when it came to lexical analysis.

0
0
Silver badge
Devil

Classic bait and switch coming up....

It wouldn't take more than a few lines of code to back door the whole lot, so if there is one all that needs to be hidden is a few lines of code. At that point it becomes a simple maths problem, and the NSa employs more maths grads than anyone else. Calculate the MD5 hash value for what you need to get your doctored version to come to with the backdoor code removed but so it looks like the authorized and backdoored download, add in a few lines of comments to get to the same hash, and give the auditors the safe version. They check it, give it a clean bill of health, and then the backdoored version goes back on the download sites with the auditors seal of approval and a supposedly safe MD5sum value. Ah, I do so love the sound of paranoid numpties' heads exploding!

0
2

Re: Classic bait and switch coming up....

You do know what the properties of a cryptographic one-way hash are? Admittedly MD5 sum is insecure, and has this has been theoretically known since 1996 and proven in 2006.

Now, it depends on your level of paranoia about the NSA developed SHA-2 and the NIST sponsored SHA-3 hashes...

0
0
Silver badge
Pirate

Re: Classic bait and switch coming up....

"You do know what the properties of a cryptographic one-way hash are?...." Yes, thanks. An MD5 hash is a calculation derived from the contents of a file, but there are in theory an infinite number of different files that could give the same MD5 hash value. Whilst it should be very hard to generate two files which are almost identical and have them calculate to the same MD5 hash value it is not impossible by any stretch. MD5 has been declared "broken" for years but is still by far the most common "safeguard" seen on download sites.

0
0

Re: Classic bait and switch coming up....

But it only takes one person to compare the files instead of the checksums and the game's up.

1
0
Silver badge
Facepalm

Re: Werner McGoole Re: Classic bait and switch coming up....

"But it only takes one person to compare the files instead of the checksums and the game's up." Not when one file is a binary, which is what the download is, and the other is the doctored source code file.....

0
0
Anonymous Coward

The key isn't the code, it's the algorithm and process used to encrypt the data. Choosing an algorithm that is NOT on the US Dept. of Commerce approved list is the best approach.

1
1
Silver badge

Is it really the best approach? I'm not in any way saying the NSA Compliant Certified algorithms are the best, but somewhere, several places actually, in the chain there must be implicit trust. Trust not only in the technical abilities of those implementing the algorithms, but trust in their ethical capacity as well.

Whether it's a nation, a building or code, security is only 50% technical, the other 50% is reliant on Humans. If you don't trust the Humans involved you don't have security, you have a resource intensive, micromanaged system that can be manipulated. Snowden is a great example of this. When everyone is constantly under scrutiny the guys watching are nearly always looking the wrong direction and you know what they're looking at; its an unavoidable side effect of any highly structured system, it can only do what you tell it and that's a significant weakness.

Anyway, at present we don't have an independent body that is both technically and ethically capable that is worthy of the public trust. The higher up the food chain you go, the more shadowy everything becomes. There are mystery parties involved in all of this and that's never going to be secure. Until/if such a body is extant, there's a strong argument to be made that using one of the 'government approved' algorithms is the best option. At least with those you can take appropriate measures to ensure data you might not want them to have doesn't travel online. With the mystery algorithms you don't know the enemy, or even if there is one. That isn't security either.

0
0
Anonymous Coward

Pretty irrelevant

Seeing as TrueCrypt still can't handle guid partition tables, and almost every new laptop now has the disk formatted this way. Therefore a significant slice of the market (windows laptops) cannot do full disk encryption with TrueCrypt without reformatting the hard disk and starting again. Telling people who turn up with BYOD laptops this doesn't go down well and we have given up.

TrueCrypt used to be the protection of choice for all our laptops but with an increasing number of BYOD laptops we are having to tell people to upgrade to Win 7 Enterprise or Win 8 Pro and use BitLocker. Shame.

1
0

Page:

This topic is closed for new posts.