i realise that this doesn't affect the main point of the article, but the example on non-random symmetric keys in https is wrong, isn't it? the key is chosen by the client, not the server. http://stackoverflow.com/questions/3936071/how-does-browser-...