[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Target audience and security (was: Re: [Demexp-dev] Thoughts on voti

From: Brian Hurt
Subject: Re: Target audience and security (was: Re: [Demexp-dev] Thoughts on voting machines)
Date: Tue, 21 Sep 2004 17:10:40 -0500 (CDT)

On Tue, 21 Sep 2004, David MENTRE wrote:

> We are *not* targeting an electrinic voting system for "classic"
> elections. Other people and free software projects are doing it.

Thanks.  That's the decision I was looking for.

One neat idea I had, to keep in our back pocket:  There are about 30,000
words in common use in the english language.  Make it a nice, round
number- call it 32,768 common words.  We might be able to get it up to
65,536 words, if we include common first names, place names, etc.

There will probably be times when we need to have humans deal with large
numbers.  We've all had the hell of reading 20-digit alphanumeric strings
over the phone, knowing that one wrong digit means it won't work.  
Instead of encoding our large numbers as strings of common words.  Each
word represents 15 or 16 bits of the number.  So that 256-bit hash of the
public key is represented to humans as a string of 16 common words, as 
opposed to a string of 64 hexadecimal or 43 base-64 characters.  

The compression is less (we're talking about 80+ characters at least for a
256-bit hash key, vr.s 43 or 64 characters), but the user friendliness is
much higher.  It's much easier to read "robot cat herd enrage violet ideal
..." over the phone.  Much more likely for the human to get it right, too.  
Note that we can also take advantage of any spell checkers we can glom
onto.  Note that capitalization is irrelevent.

Internationalization is easy.  If word 12,345 in English is 'tomato', and
in German it's 'schreib' (sorry, my german is awful- what's german for
'write'?), then all we need is a translator program that when it see's
'tomato' spits out 'schreib'.  Each language can have it's own list of 64K
most common words.  If the language doesn't have 64K words (Klingon,
Esperanto?), it can use multiple words.  The translation is simple, 
mechanical, and precise- because it doesn't care what the words are, only 
what number they are.

"Usenet is like a herd of performing elephants with diarrhea -- massive,
difficult to redirect, awe-inspiring, entertaining, and a source of
mind-boggling amounts of excrement when you least expect it."
                                - Gene Spafford 

reply via email to

[Prev in Thread] Current Thread [Next in Thread]