The problem with feed readers

Everybody’s made their point about how great RSS (or Atom) feeds are, and how they help folks syndicate content and keep in touch with a lot of websites in a very efficient fashion – problem is that everybody seems to not like to think about how to manage the syndicated content they read. The cool thing about feeds is supposed to be that you get easy access to a large amount of information, which you then should be able to process and digest sooner than you could have by just visiting each individual website and scanning or reading all the articles! With current feedreaders, the usage flow is like that of browsing – just that instead of clicking on, say, a bunch of bookmarks in your browser, you click a bunch of feeds in one panel that then open up a list of articles from the feed(blog) in another panel! It saves you some time, but not too much.

Alright, so you get the feeds from a lot of websites, but right now, things are getting out of control – at least for me they are. I cannot read updates from 200+ websites within an hour. Also, I don’t want to read everything written by everyone I find interesting. When such a large set of data is involved, I would expect to be able to manage and trim the data down according to my personal preferences.

So, basically, can we please have a feedreader with some intelligence, please? RSSOwl is neat in that uses Amphetarate to “rate” articles, but you have to manually rate what you read, and that is such a pain. There is a dropdown menu for the ratings, so it takes a few additional clicks and some moving around of the mouse to rate articles. I dislike how the UI is designed as well, and how one has to double-click a feed to read the contents. Needlessly complicated. Findory.com is neat too, but then it is not easy to use it to read a bunch of blogs. Searching for the blogs you want to read is so difficult in Findory! Adding a set of blogs you want to subscribe to is tough in findory – not intuitive in any case. I would love to see a desktop (or online) feed reader that learns from my behavior – what do I like reading?, what articles do I spend the most time reading?, without having me do anything extra besides just reading the articles in the various feeds. This is the only way I can keep on adding feeds of interesting sites, and hope to catch everything interesting in a reasonable amount of time. I suppose computer science is advanced enough to have thought about problems of organizing large amounts of data in an intelligible fashion, and that there must be folks who work on analyzing and utilizing usage patterns. This will be the next killer internet app – a feedreader that is intelligent, and magically presents you with the most interesting articles published the last day or two – by understanding who you are and what you like.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

6 Responses to The problem with feed readers

  1. Podz says:

    But finding only what you are interested in automatically reduces the chances of going off in an interesting tangent does it not ? Although the feeds / sites you will read may be interesting, they could be blinkered so although it appears you are getting a lot of information, you are actually getting a finely focussed set of data from which you are further filtering. Or something like that.

    I have:
    – custom RSS feeds from the blog indexing sites
    – custom RSS news feeds
    – FeedDemon watching for certain terms in all the other feeds I take.
    That way I can concentrate on what I really do want to read but also read many others.

    I do think we save time though – scan 20 feeds or open 20 browser tabs ?

    And Big Brother …. wouldn’t you find it just a bit spooky that an algorithm somewhere could select just what you wanted ? By the same token, it could do more.

  2. Carthik says:

    Podz,

    Why, maybe it can work the other way then, and remove stuff that I totally would not be interested in reading – say like I never read about American Football, or Genomics. I agree that some interesting tid-bits could be left out, so to speak, but that is would be a beginning.

    I wouldn’t think of an intelligent search and rank algorithm as a “big-brother” sort of thing – it would just be a smart version of a Bayesian filter, with (probably) some learning algorithm, or Neural Net or something driving it. Correlating data could also help define what I like (or what I DONT like) in a better fashion.

    Scanning 20 feeds as opposed to sites does save us time, but for me with ever so many interesting feeds, the time saved seems small when compared to how much it could be saved, and how much information I could read if there was a more intelligent algorithm at work.

  3. IO ERROR says:

    I think you’re going to find Google Desktop’s use of feeds quite interesting. Install it, browse for a couple of days, and then check the Web Clips and News sections.

  4. Jalenack says:

    I think you’re looking for FeedLounge. It’s already the next killer internet app. Too bad it’s still in Alpha testing and on invitation-from-founders only. Brought to us by WP’s own Alex King!

  5. Greg Linden says:

    Hi, Carthik. That’s definitely what we’re trying to build at Findory.

    Sorry to hear that you don’t find our search and favorites UI easy to use. Could you elaborate? What is it about our search that makes it hard to find blogs you want to read? What is it about subscribing to blogs on Findory that is not intuitive? What could we do to make it easier?

    We do want Findory to be the intelligent, helpful reader that you desire. I’d appreciate hearing your thoughts on how we might improve it. Thanks, Carthik.

  6. Sameer says:

    Have you tried Gregarius, it now supports amphetarate. It doesnt solve your problem, but it is a really cool web based feed reader. (sort of like wordpress for feeds)