I had heard good things about BlogLines, so I thought I would give it a spin. I didn't plan on switching, but I did: that's how good it is.
What sold me was the integration with the way I use Mozilla. With SharpReader, I would skim titles, dip into those that seemed interesting. Of those that merited further investigation, I would open in a separate window, and come back to once I had completed my initial scan.
With Bloglines, I skim articles, and open the ones that merit further investigation in a separate tab. All in all, I'm finding that I can read more in less time.
Other benefits: by being server based, I don't have to worry about keeping machines synchronized. Also by fetching feeds once and then serving them multiple times, Bloglines acts as a collector/distributor, which will make feeds even more scalable.
Now for two niglets:
- Bloglines appears to understand content:encoded, xhtml:body, and atom:content. While this is good, I'd like an option to prefer description and atom:summary instead. This is based on my usage pattern: I would prefer to see a summary and then open in the background the full text for later perusal.
- The interface isn't RESTful. HTTP GETs are not idempotent. What this means in practical terms is that once the server has provided the entries, the assumption is that I have read them. Even if, for some reason I close the window.
All things considered, Bloglines has become my new default aggregator.