Distributed Tomboy ahead?
Tomboy (author’s blog) is a very nice little app.
When in Gnome, I use it to manage random snippets of information such as to-do lists. I also use it as a resource aggregator, in the sense that I collect info about certain subjects. A kind of place where seeds of larger, more structured content are allowed to grow.
It even has a plugin system with a small number of available extensions. So, very slowly, people start using it as an information management tool. Which is nice. As Wikipedia founder Jimmy Wales explained at FOSDEM 2005, you don’t need a lot of features to build a nice structure. You just need people willing to craft it. [Which was a nice point: don't try to build everything into the software. Let the people, the community, be the features. They have common sense.]
A good example of this is a category system. Instead of building extra UI functionality to accomplish some kind of name space system (or a hierarchy, or tagging / folksonomy), you could use the existing UI to do so. In the same way that in Everything2, everything is a node, in Tomboy everything could be a note. The UI for making a ‘name space’ would then be to type in a note: (is a name space). Or something more catchy. All notes referenced from the ‘name space parent’ are then part of that name space. Tagging could be done like so: (tags: private, interesting, new).
On the mailing list, people are now talking about extending the rather limited plugin API. Folks are planning to implement content sharing, like including notes from other people or pulling in notes over HTTP.
This is where things get both interesting and very fragile; name space collisions and double note names are just around the corner. If one could conceive a fitting ‘information architecture’, then Tomboy would suddenly become a distributed wiki editing app.
Of course, the main goal of any solution to this problem should be simplicity, because that’s what Tomboy really is about.