Julien Pierre wrote:

>> Back to Ian's comments on market value... if all you risk is $50 a year
>> in credit card fees, $40 is well worth the money, $900 isn't...
> 
> That's a false premise, though. Even though it's one of the most popular
> applications, SSL is used for much more than online credit card
> transactions, but also for many more transactions which do not have the
> same legal protections as credit cards do in the US, including stock
> trading, electronic check transactions, wire transfers, etc. For many of
> these transactions, the authentication is crucial, as you may not have
> any recourse if something goes wrong with your confidential data.


Julien,

Let's see if we can tease out that false premise.

I'd say the false premise is that the protection
in HTTPS be set for everyone at the same level, as
required to protect the highest value of transaction
for some small set of people.

Firstly, we do not know what the users are doing.
  We don't know which user is doing what value
  transaction, or when, or how it is related to any
  other activity.

Secondly, we do not know which transaction to target
  and thus protect.  We don't know its value, the attacks
  facing it, costs of those attacks, whether identity
  matters or not, what regulation or rules applies...

Thirdly, we do not know whether we can stick with
  one transaction set, or whether we are trying to
  cover all transaction sets.

  I.e., it is a fact that nominally, SSL was
  required/invented in order to protect credit
  card transactions.  Yet, it is also a fact that
  the SSL model is singularly unsuited to that
  role (c.f., SET), and that what was built instead
  was a general purpose connection security protocol
  for client-server channels.  E.g., within the life
  of the design of the protocol, not to mention its
  implementation, the original transaction requirement
  was disposed of, and something different was built.
  (Whether this was good or bad is not debated here,
  but the point is that SSL is not really tuned to
  the task of protecting credit card transactions,
  something that many people in the credit card
  world will be happy to discuss at length.....)

In short, we do not know anything about what the
transactional usage that we might be trying to protect.

So, setting the protection at the highest level MUST
BE A MISTAKE, because we cannot decide on what that
highest level might be.  In fact, we have no tools
whatsoever to decide *any* level, nor practically
anything about the transactions.

Which is how it should be.  The users should decide
what protection they need and what they can afford
(by this I mean server users and browser users).
Funnily enough, they know much much more than any
security developer about their own risks.

The only problem is that HTTPS *denies* them the
capability to set their own protection.  Or, as I
recently ranted, they have a binary choice:  Servers
can do only high end purchased CA certs, or nothing,
and browser users can only do whatever the server
has decided to do, "nothing" or "everything."



Under these conditions, it is unreasonable to suggest
that we are trying to protect "high value."  It's
also wrong, in point of fact, to suggest that HTTPS
does protect the value of serious transactions.

For example, hard or high value, say, RTGS payments
systems, those with no recourse (gold payments, or
smart cards), have to spend a lot of money protecting
their transactions with add-ons to the security model.
The have to do this because the security that comes
with browsing is only worthwhile up to maybe $100 or
so and maybe 1000 users.  And that's being kind.

Many institutions expect to deal with transfers of higher
than that.  If they have the unfortunate characteristic
of *relying* on those transactions, (many banks in the
US and other places have to do rollbacks of value when
they get hit, which means they are not included here,
they are not relying totally on the security of the
browser) then they need to do things like send out
challenge/response hardware, add in fraud detection
software, write turing tests, send two part transaction
confirms, etc etc (every DGC does something different).

In fact, those sites that really really need protection
would be happy with 40bit crypto and a self-signed cert
if they could just get the users to click on a browser
button to create their own client cert.  Now, *that*
would allow them to develop a relationship that could
be protected by a client-side authentication, instead
of dealing with this whole phishing and password
crunching and merchant hacking stuff that they have to
deal with.

That would be protecting users!  Instead, they have to
pay money to other places and for other techniques to
protect their users.


iang


PS:  Oh, well, I suppose there is one thing that we
know:  the users tend to do their transactions
with a site that they are familiar with (banks,
transfer agents (paypal, goldmoney), auction houses,
etc etc).  Funny thing, one of the requirements for
SSL was that it protect transactions where the user
had no prior relationship, which happens to be an
unusual transaction - I mean, who just clicks on a
URL and starts filling out a form with their credit
card details without doing some checking??? - and
not the common transaction that we normally see
needing protecting.  It is not explained why this is
for SSL, just assumed.  Just another false premise.
_______________________________________________
mozilla-crypto mailing list
[EMAIL PROTECTED]
http://mail.mozilla.org/listinfo/mozilla-crypto

Reply via email to