Smarter Feedreading
I’ve gone through this myself. At one point last year, I realized I was spending maybe two hours a day trying to keep up with all my feeds – to no avail. I was well informed on many issues, but there was a certain amount of angst as I saw the unread count spiral out of control despite my best efforts. So I set up a ‘quarantine’ folder, put a whole ton of feeds in there, and told NNW to not update them. Phew! It felt good, but it still requires vigilance, considering I want to be able to add new feeds from time to time. So I must still prune the bad feeds, and think hard about how much value certain feeds have to me.
I definitely prefer less-frequently-updated sites now. I love reading Gizmodo, but sites like it will kill my news inbox, so they’re something I visit on the web when I have time, maybe once a week. I had high hopes for BuzzFeed, seeing as it was meant to be a filter of sorts, but it could be the next quarantine because I can’t deal with 17 celebrity “buzz” items a day.
So I wonder if the next stage of news reader development might attempt to make it easier on those of us who want to spend less time reading feeds. There are two ways I can imagine this happening. First is some rethinking of the ‘unread items’ count because not all feed items are of equal value. At the bare minimum, one could choose some feeds whose counts should not be, er, counted in the total. But there may be more graceful ways to do that, based perhaps on choosing an image rather than a number to communicate the urgency of the feed pile: a big stack of paper or a clean desk, depending on the number and importance of incoming feed items.
How to determine importance is a good question. This could be based on one’s own activity with the feeds, perhaps. If one clicks a lot of links, or posts items to delicious or to one’s weblog from that source, it should be ranked higher. Plus a manual weighting mechanism could help. Facebook’s mini-feed system would be something to look at: you set some weighting in the preferences, and then it will only show stuff you don’t want to see as much if there’s nothing of the good stuff to show. Another way to go would be to use social data. NNW would have a good head start here because of NewsGator, who already measure how many links are incoming to individual items.
That brings me to the other thing: duplicates. As an apple nerd, I certainly hear about it when Apple announces a new product. I hear about it many, many times over. If my reader could tell via text or link matching when a bunch of items are all fundamentally the same thing, that would be great. And then, grouping by topic could be done on the fly, showing the x number of articles that link to the news in question. That would make it easy to mark them all as read right away. The data on which feeds were often posting duplicates would be handy to have during the pruning phase of feed readership.
Worth mentioning: I am so far from being a programmer I have no idea how impossible or not these ideas are.