Automating Web Access

Over the past few weeks I managed to resume reading about programming in general, and Python in particular. I've watched a great presentation by the BDFL himself, talking about Python 3.0. Great stuff is coming. Another thing that caught my interest is a library for automated web browsing, named "twill". Seems great for web application testing, and even for implementing something like "web pipes" or "mashups" over existing content (that is not accessible as XML, or SOAP, or whatever is the current acronym for such thing).