I have decided to consolidate Twitter, LiveJournal, and Facebook.
- Twitter is primarily a dumping ground for quick one line thoughts. It doesn't provide a good forum for comments or discussion.
- Facebook has a large social network with a very active user base. It lacks threaded comments for discussion, discourages long posts, and is zealous in response notifications.
- LiveJournal provides an excellent ground for longer posts, comments, and discussion. Notifications are context aware and opt-in. It also supports OpenID for leaving comments.
To combine these three distinct systems I'll be creating a feed aggregation and republication system.
- The first step will be to consume the feeds. LiveJournal provides a feed of posted entries, as does Twitter.
- The second step will be to provide interfaces to publish the aggregated data. Both Facebook and LiveJournal provide APIs to post entries.
- The third step will be to specify what, how, and when data will transferred from consumer to publisher. This will be by:
- Immediately posting new content from one or more sources to one or more publishers, or
- Bundling content from one or more sources into one or more posts after meeting one or more criteria:
- After X unit of time since last publishing,
- After X unit of time since new data to publish, and/or
- After X new posts.
The third step will be the biggest with the most complicated rules, but even so it's relatively simply. I'll try to make use of OpenAuth and OpenID where possible, as well as any other open standards and packages already available.
There will still be a problem with splintering comments, but that's a much bigger problem. I have a few ideas on how that sort of thing could be solved, but it would take work from each location to enable sharing of comments across applications. There would have to be a way for an API to insert comments into LiveJournal while automatically attributing it to the Facebook user that made it there, and vice versa. That's a whole other can of worms.
In time I will be able to expand upon this system by including items republished from my own feed reader, as well as a feed from an image gallery (yet to be created as well). I might even be able to expose the underlying central feed, give access to it via OpenID or another authorization token system, then add permissions to content directly on the source. It's an idea, but first step first.