I've been working hard during the last years to spread the word about Objects, Patterns, Frameworks, OO Middleware, and Agile Methods in Brazil and South America. My research focuses on reflective middleware, frameworks, dynamic configuration, distributed objects, and computer music (that doesn't sound very focused does it? :)
OOPSLA is the place I go every year to get some OO oxygen into my academic blood.
What I've been noticing is that the main technical treck is getting less and less interesting over the last years while the panels, BoFs, workshops, and keynotes are getting more interesting...
It would be nice to find a way to introduce some freshness and audacity into the technical treck. It seems that papers that bring really new ideas cannot compete with papers that provide very small but consistent technical improvements in the state of the art. That's a pitty!