On 18/08/2013 02:26, Warren D Smith wrote:

Any recommended secret-key cryptosystems?

My desiderata are simple+short algorithm and high security level. Speed is not very important.

I suggest being more explicit about all those terms. (A description of what it's actually for might clarify them all at once.) Does "simple+short" really mean "must fit in so many bytes on such-and-such an architecture", or "source code must be simple enough that some third party can verify the implementation with extremely high reliability", or "must be suitable for teaching students", or ... ? Does it need to be secure against the full might of the NSA, or against a single smart cryptographer? Do you need it to be in some sense provably secure, or to have been thoroughly examined by experts and had no big holes found, or to look to a few amateur cryptographers like it should be very secure, or ... ? Is speed unimportant because it's purely paedagogical and will never be run, or because it's going to be used once a day in some protocol, or because it's going to be run on super-fast hardware, or ... ? -- g