The conference was overbooked (there were easily over a thousand people attending), and generally had the feeling of the beginning of something great. The presentations were uneven, but the number of interesting ideas presented greatly overshadowed the number of bad implementations.
The conference attendees were a varied bunch. There seemed to be roughly equal numbers of industry and academic types. The industry types ranged from people from large, established companies looking to make use of the World Wide Web, to people from miniscule startups with a few good ideas and lots of attitude. The academic types were from across the globe, from all sorts of areas, but areas that involve enormous amounts of data (e.g., astronomy and social sciences) seemed the most common.
I'll divide my specific comments into the following areas: