Re: Conspicously absent: social engineering and cross-domain problems

James A. Donald wrote:
>
> Amir Herzberg wrote:
> > we can easily turn one password into many
> > site-specific keys. Plus, we can try to force users to
> > use different passwords (which, of course, is not as
> > good, but easier to do - see problems below).
> >
> > There are `only` two problems: 1. This requires the
> > password manager to set or change the user's password.
> > This _can_ be done, but since no standard exists for
> > this, this is problematic. A standard may help. 2.
> > What happens when the user moves to a new machine,
> > etc.?
>
> User has master password, which he gives to the program
> that manages all his logins and site specific passwords
sure
>
> Login program uses master password, plus a large random
> number locally stored on the computer, to generate
> account specific passwords for each account.
this already requires us to change passwords (problem 1). This is 
tricky, I wish we could standardize this. Or, any tips on how we do 
this? We actually are implementing , you know this makes life harder..
>
> Oops, we are on a new computer?  Random number is not
> there?  Then do an SRP login to the server of the
> company issuing the login program, and get a copy of the
> large random number.  This means that the company
> issuing the login program can launch a dictionary attack
> on your master password, as can anyone who has access to
> one of your logins and access to a computer on which you
> used the login program, but no one else can launch a
> dictionary attack.
Do you mean to authenticate to the `login helper trusted party (LHTP)` 
using as a shared key the hash of your master password, and they'll send 
the user's `random number` ?
That does seem a reasonable solution.

Tks, Amir Herzberg

Received on Tuesday, 13 June 2006 15:57:48 UTC