Re: rand() - implementation ideas [Q]

[z007400b@bcfreenet.seflin.lib.fl.us (Ralph Silverman)]

Kent Budge ("kgbudge,"@,sandia,.gov) wrote:
: Ralph Silverman wrote:
: ...
: > 
: >         systems based on such predictable behavior
: >         can hardly be viewed as 'random' enough
: >         for serious (e.g. tactical) purposes...

: Depends on the application.  For Monte Carlo integration, the 
: predictability is irrelevant (and may actually be advantageous, 
: for reproducibility.)

: For cryptography, your conclusions is most definitely correct.

: >         actually,
: >         there are serious random generators
: >         in the
: >                 hardware realm
: >         which depend,  alternatively,  on
: >         principles of
: >                 diode noise
: >         and
: >                 radiation source and detector
: >         .
: ...

: If it's encryption you're after, these gadgets are a very nice way
: to generate a one-time pad, which is not breakable even in theory.
: (You still have to worry about someone Xeroxing your one-time pad, or
: intercepting its transmission to users, but that's a whole different
: ball of wax.)

: -- 
: Kent G. Budge
: kgbudge@sandia.gov
: (usual disclaimer)
: Return address hacked to foil junk mail; edit before replying.

--
****************begin r.s. response******************

	re.
		"Depends on the application..."
		(posting cited above);
	really?

	seems reasonable that top
	hardware randomization is
	preferable in all applications
	where a stream of unpredictable
	(i.e. unbreakable) randoms is
	wanted...
	tactical applications would seem
	to fit this,
	yes?

****************end r.s. response********************
Ralph Silverman
z007400b@bcfreenet.seflin.lib.fl.us




--

Ralph Silverman
z007400b@bcfreenet.seflin.lib.fl.us