This is somewhat related to the question if a website in general
should be allowed to know if the user uses non-approved certificate
that is installed on her device. This is important if one considers a
possibility of using social engineering to trick the user to install a
new certificate authority that allows to perform MTM attacks. I know
banks worry about that.

However, allowing such knowledge to websites has a privacy
implications. Besides social engineering arguments are weak as a user
can be just as well tricked to install a special browser that skips
any ssl checks. So the proposed text change is good.

On 27 May 2014 20:28, Chris Palmer <[email protected]> wrote:
> On Thu, May 22, 2014 at 12:49 AM, Yoav Nir <[email protected]> wrote:
>
>> Interesting question. IMO (no hats) the answer should be no. If the UA has 
>> disabled pin validation (as section 2.6 says it may) then it should not send 
>> reports either.
>
> Thanks for raising the question, Igor. I am inclined to agree with
> Yoav. Anyone else have thoughts?
>
> Is section 2.6 a good place to put a note about this issue? Proposed text:
>
> """
> If Pin Validation is not in effect (e.g. because the user has elected
> to disable it, or because a presented certificate chain chains up to a
> locally-installed anchor), and if the server has set a report-uri in a
> PKP or PKP-RO header, the UA SHOULD NOT send any reports to the
> report-uri for the given certificate chain.
> """

_______________________________________________
websec mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/websec

Reply via email to