Seb is writing about an algebra that works on feeds. Being an ex query optimization guy, I can't hear the word algebra without thinking of query languages. Which in this space, lands you squarely in XQuery land. According to conversations I've had with Dare, the .NET folks will have this functionality in the platform in the next release. On Java, there's Michael Kay's SAXON. For C (and thereby Python, Perl, Ruby), there's Galax. There are also countless commercial products, but most of these seem to be aimed at large persistent datasets. There's room for XQuery implementations that are going to work on "small" in memory XML documents.

We need more projects like this. I applaud the effort to do something off the beaten path.