J. Wren Hunt wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Ian G wrote:
| <snip!>
|
| Do you really think MF should purport to make the
| decision that the user should trust 20 different CAs
| without a choice?
|
| Yes, users can remember the brands needed. Huge
| numbers of branding studies have shown the user
| has a capability to deal with brands. The entire
| western commerce system runs on it, and relies
| on it to get bread to your door, petrol in your car,
| your car itself, and beer at the end of the car
| journey. Quick, how many beer brands do you
| know and recognised?
|
There is a difference between brands of products that you use and have
experience with, vs. those that you may simply have no basis for knowing.
Yes. If you are dealing with a brand you have
no basis for, ... Take Care!
With your analogy I know that millions enjoy Guinness; me, I can't get
the stuff down! ;-) That's not to say it's bad, which it seems like
where your argument is going: recognized brand = good thing(tm).
Un-recognized brand = bad thing.
mmm... Close! Recognised brand - know whether it is
a good thing or bad thing. Un-recognised? Well, take
more care.
In the Schmoo thing, we both saw the cert details and knew rather
quickly that something was afoot because it was *not* Verisign.
Right. But that was only the techies who know how
to do that. Firstly it was obviously not Paypal, because
it said Meoow, and we were all told in advance it was
a bug. Secondly, we all knew that we had to go looking
for the cert (which took me minutes to find).
Yet, our average user does not know that. She has no
tools to know anything about Verisign. If she is in the
top 10% of average users she might look at the lock.
But I
see no way whatsoever that joe6pack on the street (joe6pack being
defined as a non-crypto geek) would have the faintest clue or even know
why he should care?
This is how brand works - ask your mates. If
you are asked about which phone to buy, you
ask your mates. If you ask about what mp3
player to buy, you ask your mates. You don't
need to know the brands, or even have the
faintest clue about any of them, you just need
to have plenty of mates.
It's like me taking my car in for service. When
the mechanic says I need brake pads, I merely nod and say yes; we don't
even bother going through which brand he slaps on there because we both
know it would be meaningless to me. The only thing that's likely to get
my attention is when he asks "Do you want the $10 or $100 pads?". But
when I drive off the lot and my car successfully brakes, I quickly
forget the entire brake pad converation, not just the brand. For me it's
just not important. Just like most people haven't a clue what that damn
padlock is for anyway. Not on their radar. Un-unh. Nope. Not gonna
happen.
Yes, users go to Mozilla like you go to your mechanic.
Now your mechanic, he knows his reputation is
on the line if he puts in duff brake pads. In a
sense, so is Mozilla's. It's worse for the mechanic
because he can be sued and people will flee from
his shop. Mozilla doesn't have that to worry about.
But, get this: When the dodgy brake pads of the
Shmoo cert came out recently, few said it was
Mozilla's fault. Some people said it was the
registrar's fault. Others (surprisingly few) said
it was the CA's fault. Some say it is ICANN's
fault. I forgot what I said....
So you say the user has no capability to decide
these issues. I think we've seen here this week
that it's not as if anyone here has the answers.
Giving the user more control over their CA -
whether to vote them up or down - can't then
do more damage than has already been done!
We the crypto-geek community have to ensure that we give consumers the
necessary tools but that doesn't mean we need or should attempt to make
them experts. For no matter what clever hack we code, RFC we write, or
GUI we display, it boils down to Abe Lincoln's "You can fool all of the
people some of the time, some of the people all of the time, but you
can't fool all of the people all of the time". We can only address
parts of the argument IMO, not the whole thing.
Having an opinion on Verisign doesn't make
anyone an expert. It's like going to a football
game. You don't have to understand the rules
to form an impression.
To double up on your Abe, it all boils down to
what Ludwig van Mises said:
"centralised planning making might work
in a small community like a village. But, for a
community and economy like a big country, it
is impossible to build the computer big enough
to analyse all the information. **"
iang
PS: That's a paraphrase, I don't know the original
quote, but it's the one that destroyed communism
(centralised planning) as a viable model. It's called
the von Mises calculation argument.
--
News and views on what matters in finance+crypto:
http://financialcryptography.com/
_______________________________________________
mozilla-crypto mailing list
[email protected]
http://mail.mozilla.org/listinfo/mozilla-crypto
