Re: unconfined domains and Dan Walsh's article