Ian Goldberg (iang@CS.Berkeley.EDU)
6 Apr 1999 13:51:02 GMT
In article <19990406091756.B22236@progressive-systems.com>,
Ge' Weijers <ge@Progressive-Systems.Com> wrote:
>On Mon, Apr 05, 1999 at 11:23:01PM +0000, Ian Goldberg wrote:
>> In article <Pine.LNX.4.10.9904051648130.31211-100000@deeped.gloop.org>,
>> John S. Bucy <bucy@gloop.org> wrote:
>> >
>> >For ElGamal and DSA, when we say that the key length is n bits, which of
>> >the numbers of the key is n bits long. x?
>>
>> p. But that's not the key.
>>
>> - Ian
>
>So the question is: what is 'the keylength' when you use a generator g
>of order q, such that (e.g.) |q| = 160 bits, and a modulus |p| =
>1024. The private key x has a length of 160 bits, but the public key
>g^x has a length of 1024 bits. My vote goes to the length of g^x, as
>that seems to determine the work factor for an attacker as long as q
>is in a suitable range.
Yes, that's what is usually said; the _public_ key is 1024 bits long
in DSA, even though the private key is shorter.
- Ian
The following archive was created by hippie-mail 7.98617-22 on Thu May 27 1999 - 23:44:21