<p dir="ltr"><br>
On Jun 27, 2015 11:51 AM, "Paul Rubin" <no.email@nospam.invalid> wrote:<br>
><br>
> Michael Torrie <<a href="mailto:torriem@gmail.com">torriem@gmail.com</a>> writes:<br>
> > Furthermore you cannot prove a negative, which is what proving<br>
> > security is for anything but the trivial case. Are you saying this is<br>
> > untrue?<br>
><br>
> I've always thought that there are no two even numbers that when you add<br>
> them together, give you an odd number. Are you saying that statement<br>
> can't be proven?<br>
><br>
> > But how does one prove a system is secure except by enumerating attack<br>
> > vectors<br>
><br>
> In the case of encryption, you do a reduction proof to a recognized<br>
> primitive like AES. That is, you show that if your system is breakable,<br>
> you can transform the break into a break against AES itself. That's the<br>
> best you can do at the moment, because the open status of the P!=NP<br>
> problem means that no one knows how to prove that any primitive (such as<br>
> AES) is secure. The reduction proof means that the evidence for AES's<br>
> security also applies to your system.<br>
><br>
> Of course that's just for the cipher itself. For the entire surrounding<br>
> software/hardware/process system which is mostly not mathematical,<br>
> you're right, there's no way to (mathematically) prove security or even<br>
> to define it.</p>
<p dir="ltr">Ahh okay. So what he's referring to must be such reductions and proofs of these provable aspects, though he spoke very broadly.</p>