One of the projects demonstrated by Andy Turner at BarCamp Grand Rapids was Mapufacture. Developed by Andy, with Guilhem Vellut and Mikel Maron, the site functions as an RSS aggregator with a difference. As well as pulling in the latest content from feeds, it also extracts GeoRSS data from those feeds and plots the results on a google map. You can read an announcement here.
Mapufacture is one of a number of apps that are demonstrating one of the key ways that I hope the web is going to move on from the ‘Web 2.0’ phenomenon, showing how open data can begin to be leveraged using tools designed for average users. Open APIs are all well and good, but it’s only as their potential is opened up to non-traditional users that I think we’ll really begin to see the potential.
The first generation of mashups, created by coders, usually for coders, are an important testing ground and serve to show people the power of open data. But once it becomes easy for people to realise their own ideas of how to combine data we suddenly have a much larger population of innovators.
Aside from Ning—which, with its emphasis on “apps,” may appear a bit heavier than most people think they need—there aren’t many examples in the wild that do more than aggregate RSS in various ways and/or plot data on maps. But the more advanced tools can’t be far off, can they?
Tags: barcampgrandrapids, georss, maps, mapufacture
What about tools like MeeboMe + WordPress? Embedding an IM widget in your blog sidebar by drag-n-drop (if you’re a user of the WordPress.com service) seems accessible to non-techies, and it represents an interesting mashup of sorts. Since MeeboMe is based on Jabber/XMPP, it seems meets the “open data/API” requirements. Or am I misunderstanding what types of tools you’re looking for post-Web 2.0?
Yeah, meebo looks like a pretty good service and a good indication of what’s possible, but I’m more looking for much more generalised ways to build things. Ways to take two or more almost arbitrary pieces of content and produce a new thing of your choice (restricted by the nature of the data, naturally) from them.