[Enigmail] Good news, everyone!

[Robert J. Hansen]

Michael Kane wrote:
> I've recently become very interested in email encryption, and I'd like
> to talk to like-minded people about how to make it easier for
> grandmothers to use.

Lost cause, unfortunately.  I don't say that lightly.  For almost two 
decades I waved the banner, but I finally have to admit defeat.

Email encryption is _hard_.  It's hard to think about, it's hard to talk 
about, and for newbies it's often genuinely _impossible_ to 
differentiate people who know what they're talking about from people who 
should be ignored at all costs.  E.g.:

	"In the beginning nobody used long keys because they took so
	 long to encrypt and sign messages.  Now we have much faster
	 computers, and there's no reason not to use the best keys
	 possible: 4096-bit RSA."

	"Bit for bit, Elgamal and its family offer better strength
	 than RSA."

	"Remember that all algorithms can be brute forced--it's only
	 a question of whether you can afford the computers required."

	"Sign every email that you send, regardless of to whom you're
	 sending it.  That way if anyone ever forges your message,
	 you can say 'I sign all my messages.  If it wasn't signed,
	 then it wasn't me.'"

If you were to ask people who considered themselves savvy on the subject 
of email encryption, pretty much everyone would agree all four of these 
statements are pretty much true.  The problem is they're all false.  All 
of them.  True responses would be:

	"The absence of reasons not do something is not, itself, a
	 reason to do it."

	"This is wild conjecture."

	"This is ignorant of even the basics of information-theoretic
          security."

	"No information is conveyed by the absence of a good signature."

But ask a newbie to evaluate each false claim and compare it against the 
true claim, and very few of them would be able to figure out who's 
telling the truth and who's lying.  This is how I think the conversation 
with the newbie would turn out:

"He says Elgamal is bit-for-bit better, you say that's wild 
conjecture--he says he read it on alt.security.pgp, and you're talking 
the Springer-Verlag Lecture Notes in Computer Science series--what the 
hell are these things?!  I just want my email to be safe and secure!"

There is so much misinformation out there about email crypto that it is 
utterly implausible it could be an CIA plot.  The CIA is good at 
muddying the waters, but this is the Ganges River during monsoon season.

There's good news, of course--analytical thinking and logic will help 
clear the waters, leaving only some of the deep depths enshrouded in 
murk.  But that requires the user to know the rules of logic and be 
willing to apply them to analytical thought.  In our modern society, 
most people would rather drink hemlock than think critically and 
rationally about a subject.

I doubt that it is possible to teach email encryption to a general 
public that has at best an eighth-grade understanding of mathematics. 
And for that reason, I'm no longer an advocate of making email 
encryption simple enough for people's grandmothers.  Won't happen.  Lost 
cause.  Pure quixotry.

I'm an advocate of making email encryption simple enough for someone 
who's taken two semesters of college math.

Believe it or not, if we could do that it would be a massive, /massive/ 
step forward in the usability of email encryption.  And that's exactly 
what I hope we can do with Enigmail.