Can you go beyond basic algorithm testing and test AES and SHA algorithm used in SAFE with the NIST Cryptographic Algorithm Validation Program (CAVP)?
I thought NIST was known not to be trusted, given the rot introduced in the past. They’d likely recommend ROT13 though prefer you just use ROT26.
Still it’s a good thought to have all and sundry test openly in the wild.
NIST CAVP provides test vectors for various crypto algorithms. It just proves that AES implementation is giving for known inputs and keys expected results.
Beyond me but would that not be based on the Rust implementations… SAFE is not writing their own basic functions afaik.
It would be good to know which cryptographic algorithm implementations (and versions) are being used and if they are tested. I’m sure some implementations out there have backdoors built in, or they will fail if thoroughly tested.
It’s a fair point and I think falls to marketing perhaps to describe the many and varied ways that SAFE is necessarily secure. Something that is convincing to the majority of us and then with the detail there for the expert eye too. Over time I’d expect necessarily the case will need to be made as to why SAFE can be trusted… but while the code is in flux perhaps it’s too early to expect that to exist.
I would not expect the module(s) that handle encryption/decryption or other crypto operation to change to often. In fact if they do change I would seriously question security it provides and would suspect backdoor being intentionally or unintentionally.