[cryptography] ICIJ's project - comment on cryptography & tools

ianG iang at iang.org
Mon Apr 8 08:38:38 EDT 2013


On 8/04/13 04:06 AM, Peter Gutmann wrote:
> "Kevin W. Wall" <kevin.w.wall at gmail.com> writes:
>
>> I think you're giving the NSA way too much credit on why security sucks. Even
>> if we were to restrict 'security' to the scope of cryptography, even there, I
>> think the NSA has much less to do with dumbing down crypto security than
>> other factors.
>
> Exactly.  If the NSA didn't exist at all the only difference we'd notice is
> that there'd be less of this weird obsession with ECDSA (via pressure to adopt
> Suite B).  Computer security as a whole wouldn't suck any less.


I think we all suffer a fair amount of cognitive dissonance on this one.

We all know stories.  DES is now revealed as interfered with, yet for 
decades we told each other it was just parity bits.  The same process 
happened to GSM -- MiBs specified the 40 bit key, but because it was a 
secret design, they didn't need to create a legend to hide the 16 bits 
of zeroes.  Add in export control regs, add in the war against PRZ.

If someone where to do a longitudinal study of the public knowledge of 
the interference, I think it would mount up.  Individually, we can 
ignore those stories as conspiracy theory, but in aggregate, much harder.

>> IMO, the biggest factor is that 95% or more of developers are completely
>> ignorant of best practices in cryptography.
>
> At the other end of the scale, 99.9% of developers who do know security have
> no idea how to create *usable* security.  At the moment there are exactly two
> crypto-using products I can think of that I'd feel confident a random member
> of the public could walk up and use, those being Skype and iMessage.


This is the good news.  I think the message has finally got through that 
usability is more important than classical CIA, etc.

> (Unfortunately to the crypto-purists they're not good enough because they're
> MITM-able.  You should be tunnelling SIP over OpenVPN, it's really easy,
> here's a pointer to a list of links to 100-page discussion threads on web
> boards for ways of doing this that may work sometimes).


Yeah.  This is a mystery to me, where did this crap come from?  Although 
it aligns perfectly with the geek mentality, other specialties in CS 
tend to create a greater resistance to the guild mentality.  I can't pin 
the causality on it as yet.


> Incidentally, the NSA is, from all the reports I've seen, even worse than we
> are at making security usable.  My favourite publication on security
> usability, Laura Heath's "An Analysis of the System Security Weaknesses of the
> US Navy Fleet Broadcasting System, 1967-1974, as exploited by CWO John
> Walker?", goes into this in more detail.


A great read!

An interference attack can be extremely high-leverage.  Being good at it 
can do a lot of damage.  This however doesn't mean that one is any good 
at defence.


> Peter.
>



iang


More information about the cryptography mailing list