OPML reading lists scenarios
I agree with Alex Barnett :2006 will be a big year for OPML. The format is getting more and more traction, ideas start to flow and applications are just around the corner. On my side, I’m getting really excited about it.
Recently Dave Winer proposed of OPML reading lists, lists of RSS subscriptions you subscribe to, making channels enter and exit your aggregator as the lists change.
This is totally gonna rock and will work from more than one point of view:
- As a personal centralized subscription system for the user who maintains his subscription list (public or private) on a public server. Think del.icio.us for RSS. This way you’ll be able to seamlessly switch environments and aggregators: for example you could use a desktop aggregator that reads both a local list of private/authenticated feeds and an online OPML subscription list of public feeds. When you don’t have your computer at hand and need to use an online aggregator you automatically find all your public channels there because you set your account to use the same OPML subscription list. Ok, you won’t have your private feeds but won’t need to manually sync the two either. What’s cool is that if you subscribe or unsubscribe some channels, the online aggregator knows how to update your OPML reading list on the other service, via your account and a public API. When you get back to the desktop aggregator the list of public feeds is automatically updated to the newer version. Cool, isn’t it?
- As a social subscription system: e.g. Richard MacManus maintains a list of feeds he suggests everybody interested in Web 2.0 to read. You subscribe his OPML reading list and then, as he changes the list after discovering new sources, you automatically get interesting content to your aggregator. When you decide that one of those feeds is a must-read you simply copy it to your list so if Richard drops it (maybe because it became off-topic over time) you don’t lose it.
We can go further by adding a ranking system to the items of a subscription list. If I starting to get interested in Web 2.0 but I don’t want to be immediately overloaded with information, I could tell my aggregator to download from Richard’s list only those channels he rated over a certain score.
Even further: I subscribe to multiple OPML reading lists and I can also give a reputational ranking to the maintainers. This way, the score of a channel can be the sum (or average, or maximum) of the scores of every occurrence of it in multiple lists, where the single score is my personal ranking of the maintainer multiplied by the score he attributed to that feed. And those ranking systems, if smartly implemented, could even lead to presentational changes on the aggregator side: how big and bold and towards-the-top they will be will depend on the ranking. Think about it as your personal memeorandum.
Differently from all-personal ranking system this will work because it won’t require all that work on your side but will rely on the social attribution of the scores, corrected by your personal trust in the few, reputable maintainers. And the weights you give to each maintaner will avoid the flatness of the social tagging system.
As an example, the same ranking system could be used to filter only most relevant content when bandwidth and screen real estate is an issue, like with smartphones. Or as a filter to get your custom ten most important news notifications of the day via SMS.
What the web needs to get to this is a centralized OPML reading list web service that is: free, fast, easy, cross-everything and has a complete API for reading and writing lists: people will start developing bookmarklets, Firefox extensions and browser toolbars. And, or course, we need one major aggregator supporting it. Then all the others will rush to support it, and news reading will experience a quantum leap.
