Refactoring Yahoo! Pipes
I’ve been griping on Twitter about the fact that the official copies of most IEEE and ACM papers are hidden behind paywalls, which is a great way to ensure that they don’t reach people in industry who might otherwise, you know, read them, learn something, and possibly even adopt the tools, practices, or ideas they contain. Luckily, most researchers routinely break those copyright agreements and post their work on their own sites, so I’m actually able to read some of the papers that were presented at ICSE 2011 this year. The best so far is Kathryn Stolee and Sebastian Elbaum’s “Refactoring Pipe-like Mashups for End-User Programmers”, in which they look at the dataflows people are constructing using Yahoo! Pipes  and categorize ways of reorganizing them to eliminate redundancy, improve efficiency, and so on. It’s solid, practical work, and the paper is a pleasure to read.
 Pipes is the best approximation I’ve seen to date of something that puts the power of the mashable web in the hands of everyday people. If you haven’t seen it, give it a try, and then round up some friends and create an open source clone using HTML5, processing.js, and similar tools.