[AFS3-std] Second Draft of Standardisation Document

[Jeffrey Hutzelman]

--On Thursday, August 28, 2008 10:12:38 AM -0400 David Boyes 
<dboyes@sinenomine.net> wrote:

> How about the last 10 digits of the National Bureau of Standards atomic
> clock at the time voting is closed divided by the day of the year?
> Public, isolated, independent, and generally non-hackable. Could still
> be gamed, but simple to explain and implement.

And completely not random.

Folks, we are getting ourselves into a rathole here.  We need to specify a 
method that will be used to break the tie, such as using the outcome of a 
random event or giving a casting vote to the individual holding some 
particular role, or some other method.  What we do not need to do is 
specify operational details in what is intended to be a long-term, 
slow-changing policy document.  For example, there's no need to specify a 
_particular_ source of randomness, just that a well-defined future random 
event is used.

-- Jeff