/dev/random question
Ted Mittelstaedt
tedm at toybox.placo.com
Mon Sep 17 02:19:20 PDT 2007
> -----Original Message-----
> From: owner-freebsd-questions at freebsd.org
> [mailto:owner-freebsd-questions at freebsd.org]On Behalf Of RW
> Sent: Sunday, September 16, 2007 1:56 PM
> To: freebsd-questions at freebsd.org
> Subject: Re: /dev/random question
>
>
> That's a poor analogy because they haven't improved /dev/random so it
> doesn't block, they've taken a /dev/urandom implementation and renamed
> it. In terms of your analogy they've blocked off the road, diverted
> everyone onto the highway, and renamed it to main street.
>
> Using Yarrow for /dev/random is not an intrinsically bad idea, but it
> is controversial.
I really don't see what the issue is here. If you really want a
/dev/urandom on your system then fine - symlink /dev/random to /dev/urandom
and be done with it.
Historically on UNIXes, /dev/urandom has been LESS random than /dev/random.
In short, it forced the application developer to make a tradeoff - if they
needed a lot of random numbers quickly, they had to be content with the
stream of numbers being less random.
Where you often saw this is in game programming - and less-random inputs
to games made them predictable, and thus, not as fun to play. With crypto
apps, they just made those apps run -slower- as the app waited for the
random device to give it randomness.
if you really want a source that is kind of random but really isn't then
use the rand library call which doesen't use yarrow.
And as for using Yarrow for /dev/random, well the /dev/random device uses
the hardware random generator on the VIA C3 Nehemiah (stepping 3 or greater)
CPU, not Yarrow, if such a CPU is present. For those that think Yarrow is
controversal then fine - add in support for the hardware random generator
of your favorite processor (most have one now) and quit complaining.
Ted
More information about the freebsd-questions
mailing list