Livejournal: A Lost Cause

This is how it's going to play out. SUP is going to squeeze their users as hard as possible to increase revenue. They're going to burn through all their goodwill without noticing or caring, because there's a long lag between pissing off your customers and actually losing revenue. As dissatisfaction mounts, ex-users will finally be motivated to implement the necessary infrastructure to make it easy to migrate away from LJ. At some point the software and service offerings - along with the number of high-social-status charismatic people who have bailed - will pass the tipping point, and there will be a mass emigration. SUP will finally notice the loss of revenue, do too little too late, and eventually unload Livejournal - which will by then be a part of the Internet's cultural history instead of a going concern - to somebody else in a fire sale.

Livejournal is a lost cause. You cannot save it; do not try. The only choice you have is when to leave. If you leave early, the support for aggregation elsewhere will not be as good; if you leave late, the odds increase that all your supposedly private posts will have been sold to the highest bidder.

It's trivial to interoperably set up shop almost anywhere today with public posts; the problem lies with authentication. Right now, I do not believe there is any off the shelf software capable of aggregating RSS or Atom feeds with OpenID authentication, which is unfortunate. It looks like it's going to be at least a mild pain to implement, because using OpenID authentication requires creating a session with cookies, and scripts which fetch web pages are generally not designed with a session paradigm - they expect to make a single request and conduct any required authentication as part of that on a one-time basis. I'm going to suck it up soon and implement this.

As far as publication, it looks like wordpress and movable type have an openid plugin which works for comments. It's not clear whether it supports restrictions on viewing posts, but it probably wouldn't be too hard to add once the openid consumer support is there.

Am I missing something? Do you think I'm wrong about the way it's going to go? You'll have to, uh, email me, because as you can see below, I still haven't written the comment script yet....