Florian Weimer, <Florian(_dot_)Weimer(_at_)RUS(_dot_)Uni-Stuttgart(_dot_)DE>,
writes:
Hmm. What about introducing a new secret key packet with a hash
instead of a checksum? This approach seems to be much cleaner, and it
doesn't cause surprising problems for the end user.
I think that is basically what I was suggesting, except using an HMAC
(which is a keyed hash). This would be a new format for secret key
packets. Then we'd have to work out the details in terms of what we do
with version numbers and such.
Then there is the issue that if we are changing the packet format, maybe
we should make other changes as well. Then, if we're changing the secret
key packet format, should the public key packet be changed as well,
which introduces interoperability and backwards-compatiblity problems.
It kind of opens a can of worms if we go this way. On the other hand,
given that any new key format won't be backwards-compatible, if there are
other secret-key-specific changes this might be a good time to make them.
Hal
________________________________________________________________________
Total Internet Privacy -- get your Freedom Nym at http://www.freedom.net