On Sun, 12 Aug 2012 13:29:57 -0400 Nico Kadel-Garcia <nka...@gmail.com> wrote:
> Such certificates have already been stolen. They're dependent on the > security of the intermediate key owners, and the are demonstrably > unsecure: Check this URL for more details on the release of rogue SSL > signing certificates through a Dutch firm: > > > http://www.computerworld.com/s/article/9219606/Hackers_stole_Google_SSL_certificate_Dutch_firm_admits > So why isn't there a good way for an end user to strictly limit trust in, for example, a "Google Internet Authority" to those domains that are actually owned by Google, and conversely, not to trust any other authority to sign certs for domains owned by Google? A single organization is about as far as I'd ideally want to extend trust at any one time anyway, and only for what I trust that organization for, because this whole SSL/x.509/PKI thing superficially appears to be hierarchical, but (as you've pointed out) in reality there is no hierarchy, because there are no limits to how far the trust is extended once we have these unknown intermediate certs. And just earlier this year, we're informed of a rather interesting "common industry practice": http://www.computerworld.com/s/article/9224082/Trustwave_admits_issuing_man_in_the_middle_digital_certificate_Mozilla_debates_punishment So you can close the barn door all you want with revocation, but that won't bring the cows back home from this particular pasture.
