5 ms·

Perhaps I'm missing something but I don't see too much benefit from OTPs. If the machine in front of me is compromised (e.g. Firefox bug, Flash bug, Adobe Reade

by nas 17y ago

Perhaps I'm missing something but I don't see too much benefit from OTPs. If the machine in front of me is compromised (e.g. Firefox bug, Flash bug, Adobe Reader bug, Linux kernel bug, Windows bug, etc etc), a OTP doesn't stop a man-in-the-middle attack. Once the attacker is on the remote system they can open a backdoor. Game over. Sand boxing seems to be a good idea (e.g .http://theinvisiblethings.blogspot.com/2008/09/three-approaches-to-computer-security.html http://theinvisiblethings.blogspot.com/2008/09/three-approac...). KVM and Xen seem pretty heavyweight. I was looking into using Smack to do lighter weight sandboxing but didn't get too far yet.

gnosis 17y ago
"a OTP doesn't stop a man-in-the-middle attack" This is true. " Once the attacker is on the remote system they can open a backdoor." If you're using OTP's, then there's really only one way an attacker could do that: hijack your session after you use the OTP to log-in, but before you have logged out. OTP's won't protect you from such a session hijack. However, they will protect you from a simple keylogger on the untrusted machine you're logging-in from, since recording the password you used to log-in will be useless for the attacker (as far as future logins are concerned). This is why OTP's are very, very important and useful for access from untrusted systems... particularly for uses such as reading your email through webmail interfaces while away at on a business trip or on vacation, where your only access to your email might be through an untrusted computer at an internet cafe or somesuch. "Sand boxing seems to be a good idea" I'm all for it. But that is a solution to a completely different problem: how to keep the rest of your system from being compromised further once someone has already hacked in to a part of your system. OTP's are a way to (hopefully) prevent that from even happening (by neutralizing the usefulness of a password that has been compromised on an untrusted machine).

by nas 17y ago

gnosis 17y ago

"a OTP doesn't stop a man-in-the-middle attack" This is true. " Once the attacker is on the remote system they can open a backdoor." If you're using OTP's, then there's really only one way an attacker could do that: hijack your session after you use the OTP to log-in, but before you have logged out. OTP's won't protect you from such a session hijack. However, they will protect you from a simple keylogger on the untrusted machine you're logging-in from, since recording the password you used to log-in will be useless for the attacker (as far as future logins are concerned). This is why OTP's are very, very important and useful for access from untrusted systems... particularly for uses such as reading your email through webmail interfaces while away at on a business trip or on vacation, where your only access to your email might be through an untrusted computer at an internet cafe or somesuch. "Sand boxing seems to be a good idea" I'm all for it. But that is a solution to a completely different problem: how to keep the rest of your system from being compromised further once someone has already hacked in to a part of your system. OTP's are a way to (hopefully) prevent that from even happening (by neutralizing the usefulness of a password that has been compromised on an untrusted machine).