Fever is a new self-hosted RSS web application developed by Shaun Inman, creator of the popular Mint web analytics application. I have been a long time user of Netvibes but I find that it’s been pretty slow of late, so this gave me an opportunity to give Fever a go. Unfortunately no trial version is provided by Shaun but a quick check on Google turned up quite positive reviews of this application.
Setting up Fever on my host was very easy. I think Shaun has established a best practice in terms of PHP application installation. Shaun provides a small server compatibility application that you install and run to determine if your hosting provider has all the necessary supporting applications to run Fever. This is important since Fever costs USD30 and being a PHP application, the source code is completely exposed, hence no refunds are offered once purchased. By providing a compatibility check, it is a good move on Shaun’s part to stop any disgruntled customers complaining about the application not working.
Once you pass the compatibility check, you are taken back to the Fever site to obtain an activation key. The Fever application is up and running pretty much immediately after entering the activation key and MySQL database information.
Fever handled an OPML file import without any issues which I generated out of Netvibes. All the categories that I had created in Netvibes were retained in Fever. After getting all my feeds set up in Fever, I went through a process of classifying certain feeds as “Kindle” and “Sparks”. Herein lies one difference in approach that Fever brings to the RSS game.
Fever uses a temperature metaphor to identify “hot” topics in my collection of feeds. It does this by looking for common topics that multiple feeds are referencing. “Sparks” are feeds which I don’t necessarily want to read on a regular basis but they provide additional context or reinforcement to my “Kindle” feeds. This approach is very similar to how Google Pagerank works, the more feeds that mention a certain topic, the hotter that topic becomes.
In the week or so that I’ve been using Fever, I’ve found that this “Hot” feature of Fever to be quite a hit and miss affair. In the screenshot above, it has done a good job of identify the hot topics but I’ve noticed many other instances where it hasn’t been as successful. On those occasions, the topics appear to be just common phrases or words used in different feeds eg “announcement”, or “wikipedia”. The screenshot below is a good example. In this case, the topic was actually correctly identified but mislabelled as just “Gizmodo”. In other cases, it can be a random collection of unrelated articles.
Given it’s unreliability, I’ve found that I’m not using this feature as much as I would like. I still am going through my feeds manually so that I don’t miss anything. I think the concept is a good one but definitely needs refinement and polish.
Another nuisance that I’ve found is that Fever doesn’t seem to update the unread count in each feed properly. It works sometimes but not always. It does have a very fluid Ajax interface that feels quite responsive. Articles in the feed can be toggled between full and excerpt view by clicking on the article body.
There is also some rudimentary integration with social media services such as Facebook and Twitter, allowing you to post articles directly into those services. It would be nicer though if the Twitter integration was tighter including URL shortening. Currently, it is posting the entire URL into Twitter which is not very practical.
Overall, I’d give Fever a 6.5 out of 10. Being a new product, I’m expecting a lot rapid enhancements to this application. It has promise but it’s not there yet.
