I have never understood why RSS, in addition to being a document format, doesn't include a specification for a really simple API that supports requesting the full archives for a feed. Every feed could be responsible for providing its own canonical archive. It would be a trivial addition to the implementation of blogging software, and would be immensely useful.