[Cryptography] BLAKE2: "Harder, Better, Faster, Stronger" Than MD5

James Cloos cloos at jhcloos.com
Wed Mar 26 05:21:23 EDT 2014


Re multiple encryptions, has anyone looked at whether stacked
encryptions in motion lead to any extra weakness?

In particular, things like tls over wifi, ipsec/wifi, tls/ipsec,
tls/ipsec/wifi, pgp or smime over tls (over ...), et cetera?

Often the users have no choice; what may look like tls over clear
channel to them may have multiple encrypted hops.  Some hops may
even have stacked encryption already.

Imagine grabbing a pgp-encrypted file over https over a vpn to the
office over "ethernet" service which is provided as a point-to-point
encrypted vpn routed over mpls, over a "wavelength" service which is
provided via an encrypted piece of a shared wavelength over fibre optic
snake oil.  (I'm sure everyone can guess which oil.)

(Perhaps someone can come up with a deeper potential stack?)

There is the potential that aes might get used at every layer.  Is that
a risk?

Other combinations?

Does such stacking affect traffic analysis?

I've been curious about this for a while, but since there is little one
can do about it I never spent much time researching it.  Has anyone?

-JimC
--
James Cloos <cloos at jhcloos.com>         OpenPGP: 1024D/ED7DAEA6


More information about the cryptography mailing list