Re: [TLS] [pkix] New version of Multiple OCSP mode of Certificate Status

[Marsh Ray <[e-mail]>]

On 08/04/2010 02:31 PM, Miller, Timothy J. wrote:
>
> First, real browsers already allow mixed content frames with silent
> renegotiation *and* late-binding of request authentication, and
> you're worried about exploiting AIA retrieval?

If it's potentially a problem, then it needs to be worried about. Unlike 
mass-energy, insecurity not conserved. We have an unlimited supply of 
it, so having it in one place doesn't diminish it elsewhere. 
Unfortunately, our capacity to be concerned is finite so we end up 
prioritizing resources over "degrees of brokenness".

Historically, it's shown to be unreliable to expect a small security 
problem to stay small. The network changes dramatically over the years 
while the protocols largely stay the same. What seems obscure or 
unlikely in one context may end up enabling an entirely new vector of 
attack in another.

For example, is this outbound connection willing to authenticate with 
the credentials of the user? Nearly every form of user credentials have 
turned out to be forwardable in one way or another.

> Second, having identified a problem behavior, what's the fix?  Path
> discovery is critically important in bridged PKIs, so if chasing AIAs
> is a Bad Thing, what do we do instead?

Good question.

It may be enough to agree on what the security context of such a request 
ought to be. For example, what should be done with cookie headers that 
might be returned? How does this interact with same-origin policies? 
When are redirects valid? Maybe all this is covered in a document 
already, does every system implement it? Is there a thorough test suite 
or audit tool available?

> Failing is "safe" from a security standpoint, but I'd contend that if
> you put that into a real system it would simply fail to meet any
> practical definition of the term "functional."

At some point we have to accept the fact that the entire purpose of 
security is to make the protected resource harder to access in one 
respect or another. The security/convenience tradeoff isn't going away, 
the best we can do is attempt to optimize it for the most common cases.

- Marsh