Robert Weir wrote:
>AES is not a panacea either. I know that some foreign governments are
>suspicious of anything that the NSA says is OK for use. So best to
allow
>a choice of several strong algorithms.
Mmm, at least AES is a standard, so like I suggested before, let's go
for at least Blowfish and AES. If you don't trust AES, you can use
Blowfish and if you need a NSA-approved cipher, go with AES.
If that isn't enough, I would suggest using the algorithms mentioned in
OpenPGP / RFC4880 (TripleDES, AES-128, CAST5, IDEA, Blowfish and
Twofish).
Those algorithms have already been implemented in freely available
crypto libraries (like Bouncy Castle in Java/C# and Botan in C++), so
most developers don't have to write these routines themselves...
Best regards
Bart
