[userbinator]

This is interesting because it's one of those cases where insecurity can turn out to be a good thing - had those cybercriminals been more careful with their systems and made them more secure, this may have never been possible; but then again, the malware might not have been able to do this in the first place if the users' systems were more secure. How that could be accomplished is also worth considering - there is a school of thought that suggests taking control away from the users and disallowing them from doing anything that some entity (corporate or government) does not approve of on the assumption that users will always make mistakes (e.g. Trusted Computing), but this also means loss of freedom - as the saying goes, "freedom is not worth having if it does not include the freedom to make mistakes."

However, if on the other hand we allow the users freedom, and thus assume that mistakes (such as being infected with malware like this) will happen, then it makes sense that a means of recovery should be available, which is not something that "perfect" security allows. To use an analogy, people who have lost their keys or had them stolen should still be able to gain access to their house. In the physical world, perfect security is nearly impossible, but with digital data, it's not. Locking an item in a safe means it can still be retrieved if the key is lost by, in the worst possible circumstance, cutting open the safe, no matter how physically strong it is. Encrypting data with a long-enough key and sufficiently strong algorithm means it's truly practically destroyed without the key. I think this point - that encryption can be really, really, really unrecoverably strong - needs to be made more aware as we continue to use more of it.

It would be particularly ironic if this recovery was made possible through exploiting the malware servers with something like Heartbleed...

[kijin]

> it's one of those cases where insecurity can turn out to be a good thing

Well, it's usually a good thing when the bad guys make a mistake, isn't it? "Oh, I wanted to blow up this building, but I set my timer to the wrong time zone." Oops, now the police has an extra hour to evacuate the building and dismantle your bomb.

What matters is: Good for whom? Obviously, insecure tools are not good for the person who relies on it for mission-critical tasks. But what is good for that particular person and that particular task might not be good for other people and other tasks.

Since "good" is relative, "perfect security" is also relative. Perfect security for whom? And what do we mean by "security", anyway? Let's say we think of security as the ability of a system to resist interference from anybody other than its legitimate user(s). But then the question becomes, who are the legitimate users?

If Apple is the sole legitimate user of a device, it makes sense for that device to resist your attempts to interfere with its Apple-approved functions. That's perfect security for Apple, perfect security for Steve Jobs's posthumous ego.

If you are the sole legitimate user, on the other hand, the device should resist Apple's attempts to tell you what you can or can't do with it. That's perfect security for you, but it comes at the expense of perfect security from the point of view of Apple designers.

As for CryptoLocker, the whole purpose of that program is grossly immoral, so does it even have a legitimate user?

Unfortunately, it is becoming increasingly clear that perfect security for one party does not always align with perfect security for some other party.

[derefr]

You're listing the clear, black-and-white cases.

The interesting case is: if I am the sole legitimate user of the device, should my device resist my attempts to run cat_pictures_infected_with_cryptolocker.jpg.exe?

[kijin]

> if I am the sole legitimate user of the device...

On the flip side, if your device resists your attempts to run cat_pictures_infected_with_cryptolocker.jpg.exe, it is clear that somebody else has some degree of control over the behavior of your device, and this somebody does not consider you to be the sole legitimate controller of your device.

Who is this somebody, and what right does he/she/they have to retain partial control of your property? That sounds like a more interesting question to me. Because unless you're like RMS and only use free software on open-source hardware, you're never the sole legitimate controller of any device these days.

[Kliment]

How about if said somebody is a less tired, more sober, more attentive and less stressed me?

[kijin]

If you wrote a program to restrain your own careless self, or installed someone else's program with the expectation that it will restrain you, then I suppose everything's kosher. We're all consenting adults, right?

Things get more complicated when a stranger decides to tell you what you can or can't do, without your consent, against your expectations, and sometimes even in spite of your loud complaints.