I have posted a corrected version of my ISIT paper on randomized coding for channels whose state can depend non-causally on the transmitted codeword. It turns out the result as presented there was incorrect. Perhaps if I get a little more time I’ll write something expository on my little nook of information theory.