« The Human Flower Project | Main | Letters from the UK »

Social software loosely coupled

Adina Levin notes, in response to a post by Brian Dennis, that social software is loosely coupled, not within any individual tool, but in the combination of tools (and I would add that this includes tools that link or aggregate output from other tools). In a comment on Adina's post, I note that Nancy White made this explicit for me in a phone conversation a couple of years ago, in feedback on a presentation I was pulling together for an online conference. The emphasis is on the social, it's really all about communication. A good example of all this is Joi Ito's toolkit, which includes a weblog, a wiki, a chat room, and accounts on various systems that give visibility into social networks, including flickr, a site where you can share photos with members of your social network. People like Joi, Marc Canter, and yours truly (among many others) keep working and playing with a stew of social software products and innovations, picking combinations that work for us. When my company consults about communication or builds a site, we suggest a combination that we might build, or that might involve integration of tools that already exist.

This brings me to an important point about the evolution of software in general, especially in the Open Source world. We're getting away from the idea that you build one product to suit everybody, which is the Microsoft/shrinkwrap model, and it made sense when we didn't have the kinds of information networks we have now. You hear complaints about the bloat in Microsoft products, but that bloat is inherent in an approach where one product has to serve many different users with many different applications for that product. The requirements are crazy complex in that case.

However in a network environment you can have a bunch of coders building innovative products for very focused applications. Sometimes the products are built as a core technology, which software consultant-developers customize and integrate into fairly seamless systems to meet specific customer requirements. You can do that with Open Source (or applications where you can acquire access the source code, even if the license is not exactly according to Open Source principles).

My own company sees this as the wave of the software future, so in our consulting and developing, our approach is to do collaborate with our customers on very clear requirements and architecture, then determine whether to build new solutions, integrate exisiting solutions, or some combination of both. We think this is the wave of the software future. I've been working with like-minded consultants and developers in Austin to build an Open Source Business Alliance, and part of our intention is to educate potential customers about this approach. We can build custom systems with a high degree of efficiency and overall lower costs.

There's a community aspect, too. Nonprofits can acquire robust but still affordable systems, and find knowledgeable volunteers to support them. Open Source tools have many adherents throughout the world, in large and small communities, so you're more likely to find a knowledgeable local resource to install, customize or develop an Open Source package. And they can get at the source code and make direct changes, if necessary – something you can't do with proprietary, closed-source software.

Getting back to the question of social software as centralized vs. loosely coupled, the essential point is that it's both, and that's true of other kinds of software, and it's true of the social and political future, too, I think: we can be more decentralized in many ways, thanks to network approaches to social and political engagement, but we still need centralized approaches for some things. To build software that's effective and bug-free, you have to have some degree of centralization and management, but once developed the software can be loosely coupled with other pieces, and the ongoing enhancement, integration and support can be somewhat decentralized. And in politics, we can have decentralized activism, discussion, and debate, but when it comes to making decisions or managing government organizations, we need approaches that are more centralized.

jon posted this at 8:09 AM
Share on Facebook| email to a friend

 Subscribe to RSS feed for Weblogsky