[Cryptography] Heartbleed and fundamental crypto programming practices

[ianG]

On 27/04/2014 23:31 pm, Phillip Hallam-Baker wrote:

> But I know what the pushback is going to be; MAKE IT SIMPLE! Because
> the last problem that arose in the USA was privacy that is the only
> problem we are going to allow to be discussed. Any proposals to do any
> other security will be rejected as OUT OF SCOPE. Because stopping what
> the NSA might do is much more important than stopping what we know is
> going on in Iran etc.


I just spent the last year in Kenya.  There, they have little clue what
western style privacy is.  But they do have greater needs in security
than the west, very much greater and more pervasive.  In order to secure
themselves, they do what the rich world eschews, they do things like
share their financial information amongst each other ... for security.

It takes a while to unravel the western internalisations;  It's no
surprise that these committees get it wrong, given their starting
positions and their assumptions.



iang



ps; apropos crypto, we give them rsa/sha1hmac/aes over the wire and
chach20/poly1305 for the static data, self-authenticated using their
social groups with CAcert-style assurance and photos.  Look ma, no CA!
Probably over the top for transactions <$100 but it is far easier to do
the job properly once, then move on to something else, than as you say,
start adding fixes in when the real needs start slapping you in the face.