I’ve mentioned in the past how much I like theFeedDemon RSS reader from Bradsoft. One of the reasons I am so fond of it is the ‘watches’ feature it has – you can set watches on a collection of terms and phrases (say ‘New York Giants, NY Giants, Giants Football, Giants NFL), and it will automatically build you an RSS feed with links to any items it finds in all your RSS feeds that match on those terms. It’s extremely useful and makes it possible to much more quickly refine your RSS feeds for specific topics you’re following.

Unfortunately FeedDemon is Windows only and so far I haven’t encountered a mac or Linux RSS client that has a similar feature. If anyone knows of one please comment. Anyway, enter Newstrove.com. The site is a search engine that tries to one-up news.google.com with a massive index of news from around the world. The site also has a very handy feature similar to FeedDemon though – look to the right side of the page near the top for the build your own topics link. Build your custom RSS feed, then copy it into your favorite RSS reader. You’re good to go with the same sort of news ‘watches’ that FeedDemon provides.

The downside is that this is a server-based system. Personally I’d prefer it be client based for performance reasons, but on the positive it does expose a lot more feeds to your search terms than the client-based one does.

Google needs to get on top of this stuff. They could do a masterful job with this. And while they’re at it they should add beysian analysis to the feeds, so that instead of seeing 120 links to posts covering whatever the media buzz dujour is (8 dead in iraq!) I see 1 aggregate feed for that topic. This is the next step in the evolution of browsing information space in my opinion, and it’s actually critical – even with RSS the situation is a total overload in terms of the voulme of content I would like to be able to look through compared to what I can reasonably manage.

I’ll return to this subject again, as I want to touch on email and usenet as other components in this information space, and I want to return to the visualization systems liketouchgraph and some of the tool prototypes like MIT’s Haystack and the OSAF’s Chandler. None of these is hitting all the bases yet but folks are working on this stuff. I feel like a drowning man waiting for a life preserver, and they haven’t cut a hole in the middle yet 😉 But at least as I said, there’s progress.