Since RFC4880 is now including symmetric ciphers with 256 bit key lengths like TWOFISH and CAMELLIA, is it time to change the limits in gnupg for pubkey sizes? According to some sources (RSA for example) the "equivalent" assymetric key size would be 15360 bits compared to a symmetric cipher using 256-bit key length. Is it really so bad to set the default to something between 2048 and 4096 and the upper limit to 16K?
We know that if the session key is compromised it means one message has been exposed. If the pubkey is factored then all messages encrypted under that key pair are exposed. I know from old posts Werner has been opposed to increasing these limits but am wondering now if he reconsiders based on new chiphers in 4880 and recent events in factoring pubkeys. Sorry if this has already been discussed in the openpgp ietf or elsewhere. I didn't find any hits. Thanks to the gpg dev and user community. _______________________________________________ Gnupg-users mailing list Gnupg-users@gnupg.org http://lists.gnupg.org/mailman/listinfo/gnupg-users
