To be clear, that tutorial was made in 2007, so is a bit dated. Also, it shows how to set up FF to use the proxy, but the idea of a tunnel is not FF-specific, nor is this vulnerability.

One big issue with SSH-tunnel as a solution is that anything not set up to use the proxy still works, it's just quietly vulnerable.

Any suggestions on making TCP traffic which doesn't go through the proxy totally fail?

I love SSH tunnels, but in regards to this particular problem, it really just pushes the problem off to wherever you ssh tunnel terminates. Do you trust you server operator? ISP? This is addressed in our presentation, here (VPN's are essentially doing the same thing): http://codebutler.github.com/firesheep/tc12/#20

Totally agree!

But right now I'm more worried about a co-worker or stranger in a Starbucks taking over my personal Facebook or Gmail account than my server operator trying to spy on me.

I trust my server operator much more than anyone squatting on attwifi, yes.

Thanks for the link!

I'm going to be traveling for a while pretty soon and using a lot of internet cafes and other free wi-fi spots so I should probably get this set up - I'm worried someone will be able to grab my password while logging in to check mail.