discussion: LLL and Security

Johan Van Schalkwyk jvs@iconz.co.nz
Fri, 18 Nov 1994 00:36:54 +1300 (NZDT)


Dear Tunespeople,

On Wed, 16 Nov 1994, Francois-Rene Rideau wrote:

>    There is a big problem with a LLL: Security. [...]
> be ok may be more difficult to prove as computing the argument; i.e.
> requiring that the argument be a prime number or any such thing).

Erm [1] I don't think its difficult, I think it's _impossible_ to be 
certain (Analogy, it's not always possible to determine whether a 
particular program will even _terminate_, let alone whether it will screw 
the system). You could also spice this with a bit of Godel, if you wish, 
and sorry about the neglected umlaut!

>    All such things can't be expressed in a *low-level* language. A low-level
> [...]
> * external object specifications in a high-level specification language are
>  given with the low-level equivalents

Erm [2] If the hi-lang has corresponding lo-lang specs, for whom/what are 
we providing the hi-level specs ?

Or am I being a complete twit?

JVS.

PS. Erm [3] Shouldn't we be talking turkey, not the first metatarsal of 
the third leg of the fourth Demodex mite in the left nostril of the turkey?