4 ms·
Just an idea... should we consider the problem is not with the first visit, but with subsequent visits? If I sign-up on a website (bank, Google, etc), all I wa
by weddpros 12y ago
Just an idea... should we consider the problem is not with the first visit, but with subsequent visits?
If I sign-up on a website (bank, Google, etc), all I want to ensure after that is that when I go back to that site, I will sign-in to the right website and not a phishing one.
CAs don't provide this guarantee, when a simple self signed certificate would!
Certs signed by CAs provide some kind of identity, while a self signed ad-hoc cert would provide strong identity! Let's imagine 3 factor auth: something I know, something I have, something they GAVE me.
In the case of Google, I would gladly install a Google CA cert for all of Google's services. Okay, I just checked and Google is already its own issuing CA :)
When SSL was born, the "problem" was to send your credit card number to an authenticated website (web store).
Nowadays, we're loyal returning customers, and we want to sign-in without any ambiguity.
Maybe we should kill generic CAs? and favor self signed certs after all.
I could trust any self-signed cert, if the only guarantee I need is that I'll always return to the same website where I opened my account.
Remember, that was just an idea, please tell me where I'm totally wrong!
- michaelbuckbee 12y agoCA's don't provide this security, but there's a new security mechanism called HTTP Strict Transport Security - [1] HSTS - which (along with some certificate pinning [2]) would. 1 - http://en.wikipedia.org/wiki/HTTP_Strict_Transport_Security http://en.wikipedia.org/wiki/HTTP_Strict_Transport_Security 2 - http://security.stackexchange.com/questions/29988/what-is-certificate-pinning http://security.stackexchange.com/questions/29988/what-is-ce...
- weddpros 12y agoI wasn't aware of it. Cert pinning is definitely a good solution!