The future of online community

I used to think I knew what online community was all about. I thought it had something to do with discussion forums, like the one here at InfoWorld I've recently tried to colonize. Having spent too many years, keystrokes, and brain cells debating the pros and cons of various discussion technologies, I'll just cut to the chase. This WebX thing is not working for me. It's not simply that the software mangles URLs, doesn't preview messages, and handles topics and threads in a way I find awkward. What's broken, for me, is the idea that an online community is a place where people gather, and a centralized repository of the discussions held in that place. In that model, I've concluded, the costs are just too high. It's expensive to join. It's expensive to participate, because interactive discussion demands a lot of attention. And it's expensive to leave, because the repository has your data, and may or may not (probably won't) preserve its linkable namespace or hand the data back to you in a reasonable form.

The weblog model reduces all these costs. It's single sign-on: just log into your own blog software. There's less pressure to participate: you can acknowledge other blogs that comment on your stuff, or not. You control the data and can, if you choose, ensure that your namespace persists.

There are tradeoffs, of course. People do miss the feeling of direct interaction. Comment trails attached to blog items are one attempt to recreate the feeling of a discussion. Trackbacks/pingbacks are another. For me, neither quite manages to restore that sense of place and belonging that is lost when you switch to blogging's more loosely-coupled mode of interaction. But I think we'll get there. And when we do, virtual community is going to be even more virtual than we think of it today. For a couple of years, Steve Yost has been pushing the idea of ThreadML -- that is, a way of representing discussions as portable XML objects. When I went back and looked at the column where I first mentioned Steve's idea, I found it to be a quilt woven from many threads. It began with a wonderful essay posted by John Faughnan to my newsgroup -- which I'm glad I quoted in the column, because the newsgroup is now defunct. The column went on to weave in discussion at Steve's QuickTopic site, on the Yahoo Groups syndication list, on Rael Dornfest's weblog, and elsewhere.

Another longtime contributor to my former newsgroups, Bruce Elrick, popped up yesterday on WebX:

Also interesting is the concept of "home". For a long time that home for Jon was Byte, including the newsgroups where he started developing a loose-knit community. Then he struck out on his own with his web log (I can't bring myself to write the contraction) at Radio Userland. Yet now that he has steady work with Infoworld, he's chosen to align his web log with that.

In the long term is that what the community will focus on (I'm sure some of Jon's following came from the Byte community itself)? I think not. At this point, now that the community has fashioned, it will be around Jon, not his "home" at InfoWorld.

I think the work dynamics of the IT world are a window on the future of work dynamics for the larger world.

If there's a community "around" me, it's only in the same sense that there is (or can be) a community around everybody. Take a look at John McDowall's latest blog map. Every node sees a community around itself. Now granted, my vocation enables me to spend a lot of time writing and linking, so I wind up being a more-connected node than most. But we all want to be, and need to be, engaged with multiple and overlapping communities. And when the costs of joining, participating, and leaving are lower, we can be. Of course there are, and will continue to be, vibrant and successful newsgroups and discussion forums. But I'm convinced that destination sites and centralized message stores are not the future of online community. Blogs are. They solve a bunch of problems. They also create a few new ones, but these feel like really good problems to tackle. As Bruce implies with his comment about work dynamics, ad-hoc assembly and loose coupling will increasingly characterize both social and technical architectures. I'm done debating how to display discussion threads. I want to figure out where this new stuff is going.

Former URL: