Due to the amount of content people are receiving on FriendFeed, several applications have been developed to make our lives easier. There are some tools that allow for the "basic" uses of FriendFeed, like AlertThingy, Twhirl and MySocial247. These client applications allow a user to read, like and comment on stories just like the Web interface of FriendFeed. You can also view the public feed, or filter for a specific user or service like Twitter.
As useful as these applications are, they do not solve the fundamental issue with FriendFeed: there is too much information. As many people say, it is like drinking from a firehose. There have been many discussions on how filtering is needed, or some other way to view information. But not many applications have appeared to fill that need. However, three applications try to make your FriendFeed usage a little easier or just more interesting.
FriendFeedMachine
FriendFeedMachine is the simplest offering of the three. It allows you to view activity for everyone just like FriendFeed. You can also mark items as read, which is a big feature missing in FriendFeed. However, the biggest addition it gives you is a "good friends" list. You can add someone as a good friend by clicking on their avatar and checking the good friend checkbox. This gives you a list of users that you want to read separate from your normal friends' activity. This is a helpful feature, but having only one list available and limited customization options restricts what you can filter.
NoiseRiver
Your interests are specified by keywords which can also be imported from Delicious and you specify how much you like or hate a topic using the slider. You specify your neighborhood in the same way, just supplying the nickname of the FriendFeed user. NoiseRiver then uses this information to display how much you should like a shared item. This allows you to quickly browse your activity based on your interests and the recommendations generated by NoiseRiver. The interestingness measures become really useful if you use the search feature. Searching on FriendFeed results in a lot of information, getting the green and red bars within the search results is immensely helpful.
Another feature that will be coming soon is something called "Smart Social Connexions". The bigger the intersection (red zone in the pic) of two persons' sets is, the more likely they are to be real friends in a social context. This intersection is based on the interests and neighborhoods of the two users. Having a better friend recommendation will also give you a better list of recommended items. However, the extremely cool features only help with the scanning of items of FriendFeed, and early adopters are notoriously fickle so their interests may change rapidly. You could still miss shared items that could be of interest because of the proliferation of activity.
mioNews
The topic folders are really useful because a lot of good content can slip down the activity stream before you get a chance to view it. If you select a user's activity, the right side of the page has two panels which can be display top & bottom or left & right. The first panel lists the items in the activity stream. The second panel shows the detail of the item with likes and comments, just like it would be shown on the FriendFeed pages.
Obviously, the advantage of using mioNews is the fact that you can keep items as unread and view a particular user's activity. You can also create folders to group friends in order to get the same "good friends" list as FriendFeedMachine, or even a group of friends that are "must reads". This folder navigation is probably very familiar to many people, but it does not really solve the information overload problem.
Conclusion
Given that these applications are fairly new, it is logical that they have focused on a particular feature that they felt was needed. However, these applications need to mature before they can become the ideal FriendFeed client. The interestingness measurements of NoiseRiver could really streamline your daily perusal of FriendFeed, but you could miss things from people you really want to read. MioNews takes the folder organization approach but tends to miss the recommendation side of NoiseRiver. However, the author of mioNews, Patrick Lightbody, did reveal a little magic in the background:
mioNews asks you to like/hate individual articles. Then, using some autotagging secret sauce, the topics and people are tuned behind the scenes.
I have not been using mioNews long enough to figure out how this autotagging and tuning affect the normal experience, but there does not seem to be any way to customize the settings.