I saw an announcement about Fever, which is a self-hosted $30 RSS reader. It's a graphically beautiful application which has a nice hook: it'll scan your feeds and create a hot list of links which reoccur in your reading list.

I've been using Bloglines since forever, and I've been getting more disappointed with its performance as of late, so I shelled out $30 and gave it a whirl.

Let me first say the installation process is near perfect. What you do is you download a "Can my server run this application?" type helper. After it successfully says you can install it; you get this screen: 

After you input an activation key (which I purchased), it seamlessly installs it (part of the default helper is chmod-ing your directory to 777 - security hole!) ... so you go right the application. I really wish I captured a video on it - it is an amazing process which is just very smooth.

About the application itself? It's really pretty, but when I spend some time trying to read actual feeds, it didn't do a good job. For me, a deal breaker in any RSS aggregator is interweaving feeds based on time. I don't know why Google does this, I don't know why every other RSS application does this except Bloglines. It really boggles me - I don't see how it's beneficial to have to context switch between articles!

I have roughly 300 feeds I subscribe to daily, and each has a very different writing style, and require context to read. Scanning things is much faster when they are grouped by feed - I can skip whole feeds if I know it's the same old information (I usually skip the Apple feeds around big announcements, as there's a lot of noise). When feeds interweave, it's nearly impossible to avoid the noise. Financial crisis? I want to focus on my grouped feeds of economist blogs. And when I got Keynesian-slanted feeds mixed with Friedman-slanted feeds, it really is hard to read.

Graphically:

Bloglines scanning style:

1. Title of feed (I get context)
2. Title of posts (I can scan headlines, based on context)
3. If the title catches, I read the post. I can skip whole sections if I'm not interested (Hot Deals Club, not so much lately since those are usually impulse buys)
   

Fever's scanning style:

1. Title of post - ok, maybe this is interesting. But... what about context?
   
2. Scan up for feed name - Oh, this is coming from Daring Fireball - an Apple-type feed. It must be really good. If that title came from Instapundit (conservative blog), I'd skip it.
   

My eyes get really tired scanning Fever feeds.

And one downside I hadn't anticipated for utilizing a server-side app refreshing 300 feeds takes a long frickin' time. (Fever has a cron job feature to alleviate this)

I've used Google Reader, I've used Fever, and I've used various Firefox plug-ins (memory fail) and I tried Thunderbird (didn't like it mixing with work emails) ... but in terms of being able to read tons of feeds quickly, Bloglines is still the winner. That really is a sad state of affairs given the fact that Bloglines is incredibly old - Google Reader (for which I was never a huge fan of) seems to have influenced the other crop of applications in adopting the worst components (at least Fever is able to "group" feeds so I can click once - a feature that seemed to be lacking when Reader launched).

If you're OK with the interweaving of feeds, I'd check out Fever. If you're not bothered by that reading style, then I'm not sure what Fever offers you more than Google Reader - I have a pretty large set of feeds, and the "tempereature' feature didn't seem to push up anything interesting.