This one’s a bit more advanced than the others. Not only does it merge a couple of feeds (Metafilter and Ask Metafilter), make sure there are no dupes, and filter based on keyword (it’s a permit policy, so you have to add more keywords to let more posts through), but it also does some basic string handling to get the posting domain (i.e. Ask or Main) into the post title on the feed, so you can tell what general feel the feed article will have (hopefully).

I quite enjoy thinking in these ways. They’re not precisely new ways for me as I do it in programming when I get to program, but you have to factor in the entire set of data and then set up your pipe to deal sanely with all possible/probable values. It’s an expansive way of thinking that makes me feel smart.

As you all know, I’ve occasionally written comments/feedback to big blogs I read about their overall content, but since I certainly can’t expect them to respond only to my own feedback and change their tunes (esp. with respect to diet/health/exercise topics which I often find noisome), I’ve started filtering their feeds to my Google Reader myself.

How? With Yahoo Pipes, which is a drag/drop type programming utility specifically for aggregating and filtering feeds and other data on the web. It’s pretty neat and usually pretty easy to use, as long as you’re careful and you test as you go.

I already built a pipe for Lifehacker that filters out posts to do with the “diet” category and the “health” category (and may need to change it to add “cooking”). I also built a pipe for BoingBoing that for now only filters out non-unique posts (by title) but may in the future implement a similar filtering mechanism for certain categories.