🚀 go-pugleaf

In article <21772@rpp386.lonestar.org> jfh@rpp386.cactus.org (John F. Haugh II) writes:

> In article <BwuLv6.L10@acsu.buffalo.edu> owens@acsu.buffalo.edu (Bill Owens) writes:
> >Except that our passwords are specifically designed to be immune to
> >dictionary attacks. The passwd program checks each new password against
> >a rather large dictionary (assembled from several sources), and after
> >it was first installed we ran the usual dictionary attacks against the
> >entire password file and forced users to change weak passwords.
> 
> It doesn't matter if your passwords are "designed to be immune" from
> nuclear explosions.  If you have ten accounts with ten different
> passwords, the odds that you can guess - by sheer luck alone - one of
> them is ten times greater than if you have one account with one password.
> It's just the way that probability works.

Aren't you assuming that you can check each dictionary word against
the list of 10 encrypted passwords in the same amount of time as you can
check it against 1 encrypted password?  This isn't the case with Unix,
since the `salt' used in the encryption process means that the encryption
will (in the vast majority of cases) have to be performed 10 times.

Admittedly, there is a *slight* increase in risk from this approach,
simply because there is 10 times more chance that you have an easy
password in, say, the most popular 10000 words, and this is higher
then the chance that 1 person's password is in the first 100000 words,
but this is a relatively small increase, and is probably offset in
terms of overall security of the scheme by the removal of the need to share
passwords.

Regards,

Granville

 
Granville Moore                                  gvm@nemesys.demon.co.uk