Re: Paper & pencil password algorithm

I'm not convinced about the usefulness of what James Taylor
is going to do.
Really?!

Yes. The problem is the time and effort spent doing the hashing and
the error rate.
It's very interesting, but I'm afraid I'm too lazy to use it.

The hash function is everything. Maaartin, I really like the idea you
have had of passing up and down a list of number performing a simple
operation. Even my granny could manage that (for addition and
multiplication modulo 10). My worry is whether she would understand
modulo 11.

Sure she would. Just don't tell her what she's gonna do. I computed it
always like
5*5 % 11 = 25 - 22 = 3
8*8 % 11 = 64 - 55 = 9
not thinking about any modulus, just subtracting one of the "nice"
numbers 11, 22, etc.

Unfortunatelly, multiplication mod 10 is quite useless because of zero
divisors (i.e., things like 2*5=0).
You'd have to add a product to something else or alike.

I was also thinking about performing the whole computation modulo 11.
The ring Z/11 is a field and you get some nice things as easy to
compute MDS functions for free.
But it's a bit harder to do and maybe more error-prone, I didn't
really tried.

Did you tried to perform the rounds I proposed lately?
This is the point - if you can't do it fast and comfortable enough,
it's lost.
Showing how good it works is important, too, but currently I have no
time for it.
.

Relevant Pages
Quantcast