4 ms·
If the private keys are very close primes than square root quickly breaks the encryption.
by mc4ndr3 5y ago
If the private keys are very close primes than square root quickly breaks the encryption.
- olliej 5y agoYes, but if they're too far apart they also become easier to break. Yay RSA \o/ Seriously, don't use RSA for anything, the problems are unending in both the maths and the implementation.
- rurban 5y agononsense. there is enough room for secure pairs, just don't be too stupid. RSA is secure, only government agencies are argumenting against it
- olliej 5y agoNo, RSA is worse than ECC in every metric. Every cryptography expert recommends against RSA so I'm not sure where you are getting "only governments" arguing against it. Problems with RSA include the core algorithm being extremely difficult to ensure constant timing for (even using Montgomery ladders); padding being super easy to do wrong, which breaks the security; Different keys have drastically different security properties (see this article for example, but numerous others exist); The required key size for a secure connection is at least 2kb, preferably 4 etc, etc ECC cryptography does not have the padding problems; despite being technically slower, constant-ish time RSA rapidly matches or exceeds the ECC time; You don't have the same "bad keys" problem; A secure key is only 256 bits. No new protocol should be using RSA.
- rurban 5y agoyou forgot to mention that ECC curve parameters came out of thin air from the NSA, the FIPS ones, whilst only the djb ones can be trusted. RSA params are randomly created, and only a few are bad. Unlike ECC.