I’ve never seen this properly studied, and only rarely taken into account by software developers: the opportunity cost of committing to one service or program over another. In a word: Why is it software that’s in charge, not the data itself?
An obvious one is Twitter vs Jaiku. Which one to embrace? Jaiku actually has more features in a way than Twitter, but more people are on Twitter. And, perversely, because one of Jaiku’s features is being able to easily include your Twitter stream into Jaiku, it makes more sense to stick with Twitter as your main presence/communication service, since those updates will automatically feed into Jaiku. Jaiku loses out because it’s better.
But usually it’s a starker choice: choose one program or service over another, and you’ll find it harder and harder to reverse engines and try another. I’ve had two versions of this blog going, one on TypePad and one on WordPress, because I can’t decide which is the better service. It’s a lousy solution and often ends up confusing people and diluting the conversation. I haven’t committed to either yet, a makeshift solution made easier by tools like BlogJet, which allow me to post to both blogs, and the import/export tools that both blogging services provide. But it’s still a dumb compromise.
Worse is the commitment one makes to software. I love PersonalBrain, but I also love mindmapping tools like Freemind. And outliners like MyInfo. I also want to explore stuff like Topicscape. ConnectedText has potential too. But because I want to use them in the real world, with a real project, I don’t want to find that by committing to one I’m foregoing using the others. But that’s inevitable. There are import and export tools available to make it easier for these kinds of about turns (or occasionally starting out in one simple program and then moving the data to something heftier when the data gets too big).
But surely there’s a better way of doing this — by making data so open that we can easily move it between programs without these hurdles? Instead of the programs being the dominant tool, they become servant to the data? A case in point: I want to look through all the blog postings I’ve written in the past five years. I want them somewhere I can see them, but also some way I can index them, and view them in different ways. I want to be able throw them at a Bayesian filter to look at the language I use, the topics I choose, the arguments I present. I want to be able to view all the data as a big mind map, or a treemap, with the categories and tags as branches. I want to be able throw them at a Wiki builder so it becomes one big Wiki without me having to do anything fiddly. I want to throw all the posts into a PersonalBrain, where the links between articles turn into links between thoughts. Then I want to throw all my emails into the mix and see what pattern they make. I want to move between all these ways of looking and manipulating my stuff without me having to worry I can’t ever go back.
In short, I don’t want to commit to one program. I want my data to be in charge, and the programs themselves conform to the data, not the other way around. Perhaps this is impossible. But why should it be?