Re: [SAGE] Security tokens

[Steve Willoughby <steve@ichips.intel.com>]

> 1) OTP users can (and should) choose long secrets (and not limit
> themselves to the old 8 character DES maximum)

This will make the attack exponentially more difficult, maybe unreasonable
to perform.  However, my experience is that if you open the capability for
your users, 99% of the time they will pick really dumb passphrases which may
be trivial to crack.
 
If you can manage to enforce good passphrase selection, OTP is much more
secure.

> Frankly, I'd be more worried about session hijacking using OTP over
> a clear text link than I would be a brute force attack on my OTP secret.

As would I.  I've just seen to many people get into trouble by picking a
solution that sounds good on paper, implementing it and naively assuming
that they're protected.  

And I've run into people who assume that OTP is completely secure with
any passphrase because it's hashed n times.

Not that anyone on *this* list would use 'aaa123' as an OTP passphrase,
of course, but I'm assuming that most of us have a general user population
to support, so the caution flag is always good to raise.
-- 
Steve Willoughby             | Let thy software go forth amid the hostile
Intel DPG Eng. Computing     | input, accepting liberally and crashing not,
Engineering Apps Development | yet sending forth only harmless output.
<steve@ichips.intel.com>     |     -- Matt 10:16 (programmer's translation)