ietf-openpgp
[Top] [All Lists]

Re: [openpgp] rfc3880bis - hard expiration time

2015-04-27 07:09:08
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1



On 04/25/2015 11:02 PM, Nicholas Cole wrote:
I don't see the benefit of a hard expiration time.

Full ACK. I also don't see the need for hard expiration times.

For me soft expiration is a way that prevents the usage of keys after
a certain amount of time, e.g. one year, when the secret key has been
accidentally deleted or the passphrase has been forgotten. So - and
this is important - these use cases are not really attack scenarios
they only provide convenience for key distribution.


More users should be encouraged to use expiration times, because
that would help to limit the problem of losing control of keys
through forgetting a passphrase (probably the most frequent AQ on
the gnupg users list).  It seems to me possible that so few do so
because they do not realize that expiration times can be changed.

Introducing a hard expiration time would introduce complexity
because there would then need to be two kinds -- hard and soft.

What are the use cases for a hard expiration time?

1. Perhaps an organization wishes to be sure that employee keys
are not used beyond a certain date.  If so, the answer already
exists: refuse to renew certifications of the UIDs on that key and
make sure that all certifications have an expiration date.

ACK


2. Enforcing key rotation.  But if this is important to individual 
users, the answer is simply: set an expiration time on your key
and don't extend it.

ACK


3. Preventing an attacker who has gained complete control of a
private key, and who can prevent the dissemination of a revocation 
certificate, from extending the life-time of a key, assuming that
that same attacker is not in a position to forge or coerce the
creation of a replacement key.

If one consider the average time that a key is still valid (5 years
expiration on key generation in Enigmail) so 2,5 years on average
valid from the time it got stolen by an attacker, the attacker has
enough time to do her "thing", e.g., forge signatures. The scenario
that an attacker needs to use the key after 2,5 years is totally
unrealistic. Thus, hard exp solves nothing.

Regards
Dominik
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)

iQEcBAEBAgAGBQJVPibRAAoJEHGMBwEAASKCzIcH/3vZ/YGwSLGQuavi8KT9WwZZ
kW9oBp7oZaGhokCh1El/+XJuMxPIbPHy/sLhw75zF417fxh0BqsnEQ6/zdgv0yTO
nT7JZ79U9O/pwpqF/e206P0PDg4dLT+Hfr2BIoqbDMXssz2osdigvjnMBeT1lAku
QiKSrfj0PWN0+J7Lkst7F5KGHAnDIHMs2gdpTdTr5rAThOuDGummaNRAQr1ezyBS
NFBVGSeMMsRJNRHbfiqyfndYmyBzW2babZu/rmw3e6FB1y70vZn6GiDmysXFMLpH
5lXyChP0WBsaTwVncCkCzur6QfI9os09v9jhIClu3Y16vgDWP02RzaKnhmLE4E8=
=n38i
-----END PGP SIGNATURE-----

_______________________________________________
openpgp mailing list
openpgp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/openpgp

<Prev in Thread] Current Thread [Next in Thread>