A sample chapter from the new O'Reilly book on blogging. [Scripting News]

Including important instructions on Backing up your Radio. If I'm not mistaken, a lot of content on Radio Userland by the same author is available here.

Weblogs have a potential for group-forming like no other medium. However I'm convinced that much of it to this day remains untapped. I'd like to explain an idea that I have been bouncing around for a while. It might well be a reformulation of what others have said previously. I believe that implementing this properly would give a nice boost to the blogosphere's social aggregation capability.

Basically the goal is to push the threshold for group creation to an unprecedented low. I think Reed's Law should be refined to state:

The value of a group-forming network increases exponentially with the number of people in the network, and in inverse proportion to the effort required to start a group.

Here's a sample motivating scenario. Not long ago I wrote an item on professions in the blogosphere. The post caught the interest of other bloggers. A few replies came here and there. If you search diligently enough you'll find them, but it's not easy. Presumably, those who have taken part in the discussion would like to hear about it if the topic comes up again, but currently this will only happen by chance. This kind of situation is very common.

The topic is pretty narrow. It wouldn't make much sense to start a Yahoo! group on this. Still, it would be nice to somehow be able to make it into some kind of "focal point" for interested people. If this were very easy, this would allow for quite fine-grained knowledge classification, which would be a boon to those who care for and closely follow particular topics.

Now, the idea is this. When I come across a post on an interesting theme that seems like it might have lasting value, I want to be able to

1. Create a topic, with a title of its own and a definition or description in plain English (which may contain arbitrary hyperlinks). Just "where" the topic is stored is unimportant. The important thing is that it is a public entity.
2. Subscribe to that topic. Subscribing has two effects: it adds the topic to a personal topic list of mine, and it means I'll get posts by other people on that topic in my RSS aggregator because each topic is associated to a shared RSS feed.
3. Post to that topic whenever I talk about it in my weblog. This has to be *easy*, like checking a box or selecting from a drop-down menu displayed under the box where I write my posts.
4. Access an archive of posts on that topic somewhere on the Web.
5. Let anyone edit the description of the topic when important things are added to the "state of the art" on the topic, or when other related topics spring out of the discussion, to let people know where the conversation has branched off.

Basically, from where I stand, this sounds a little like a witch's brew of liveTopics, standalone TrackBack, and this peculiar brand of editable web sites known as wikis.

I haven't worked it out in detail, but wouldn't it be possible to hack a beta of this together as follows?

1. Create a public topic database server.
2. Let each topic "know":
   1. Its name.
   2. The address of the web page defining the topic - this is an easy way for people to associate the topic with a Wiki page.
   3. The address of a particular RSS feed associated to the topic.
   4. The address of the web page that archives posts to that feed. That page could look just like the KMPings page.
3. Make it easy to create a new topic and register it with the server, by asking minimally for element (a), keeping element (b) optional. Upon registration, the server generates (c) and (d).
4. Modify weblog software to make it easy to post to a public topic. This is the hardest part.

Here is a snippet of a conversation I had recently about why I use Radio over Manila: