k-Collector was aggregating many feeds and building a browsable content database indexed by a community created taxonomy based on facets such as who, what, where, and when. Yes, each tag had its own RSS feed (I believe we also supported feeds for combinations of tags, it was definitely on the roadmap) but we were also connecting together related tags, forming tag groups, supporting filters. We were also starting to think about how to address the problems in shared taxonomy. It was a very fertile period.