I'm glad I'm not the only one yawning. Wake me up when it can talk to the full web and when it works with XPath/regex/webified-awk (oh please, somebody make a webified awk).

I really like the idea of viewing web resources as stdin & stdout but here's what make pipes on unix work: authors of unix tools know that they have to spit out regular data. Unix apps that spit out the equivalent of tag soup are not very usable as stdin for another app. So you shove it through awk/sed/perl/od/you name it and hope it's usable enough. Usually it's not.

The problem with the web, even RSS, is that nobody thinks about their output the same way unix geeks think about their app's output. Pipes are completely ingrained into the unix landscape in a way that they aren't in the web. Hopefully Pipes changes that.