So what, you ask, have I been working on while not posting to this blog?

The story goes like this. I imagine a few of you reading this will be doing so from the comfort of an RSS reader. I have one too, but I keep encountering a problem. Across all my feeds, there are so many items — some interesting, some not — that I can never be bothered to read every single one. Eventually I figured that what I really needed was a way of highlighting which articles in my feeds were more important, so late one night I took a crack at concocting a way to do this.

After some polishing, I dare say the result of my endeavours might be of use to someone other than me. It is (somewhat arbitrarily) named Feedeye, and it’s a Web-based RSS reader that allows you to put feeds into “sets” and will group together items in a set when they talk about the same thing. In doing so, it gives more emphasis to stories the more they’re being posted about in your chosen feeds.

Feedeye is vaguely like Create Your Own Newspaper (the very first Web site I ever got excited about, back in 1996) along with Google News or Memeorandum, all rolled into one very hacked-up beta. Pick some sources you trust and this beast will (try to) tell you what those sources collectively think is important.

Better yet, you can also pick a person you trust, and just read their choice of feeds without even telling me your e-mail address. (That said, Feedeye won’t track what you have or haven’t read unless you log in.)

To demonstrate: here’s a set of news feeds containing world news from mainstream sources. But what if you don’t trust the New York Times or Sydney Morning Herald? Easy, just try this alternate world news set that aggregates the World Socialist Web Site and indymedia.org.

Alternatively, suppose you wanted technology news. Sure you could read CNET, but I’d prefer to focus solely on what’s big in the world of Mac, or perhaps check out the buzz among the Web 2.0 kids.

I’ve polished the thing up to a level where I actually find myself using it to read feeds, rather than just spending all my time on the site making sure it’s working. So what I’d really like is some beta testers to play around with Feedeye and give me some idea of whether or not it actually works for other people.

This is, of course, all very experimental. I make no guarantees that this service will not elope with your RSS-reading Firefox extensions to the Caribbean, only to blithely dump them after one wild night with the funky new search tool from Google.

I also can’t guarantee that it’ll give accurate results, although I’ve been tweaking the algorithm and it seems to mostly work as expected. Do let me know — in the comments, by e-mail, or by carrier pigeon — if and when you find bugs.

A few parting notes:

• Some of you are aware of a project I’ve quietly been working on for the last few months. That’s on hold for the moment (nothing big, just some minor delays), but this new toy of mine has borrowed very large amounts of code from it.
• There are still a variety of bugs, particularly if you’re using Internet Explorer (the fix? switch to Firefox! ). And the colours, design, etc. all still need work.
• Speaking of bugs, there are occasional problems with the automatic updating of feeds (which, for now, should happen every 2 hours). But I have help from Daniel on that front and a fix is coming Real Soon Now.