By now you may have heard that Tweeterboard has shut down their service. The reasoning from Gene Smith was:

I realized that I no longer have the time, energy, interest or technical abilities to maintain it. So I’m shutting it down.

As the developer of YackTrack, I understand the difficulties of developing on top of Web APIs. Part of the difficulty is that many sites do not intend developers to create applications that are truly novel. For instance, Twitter allows you to get the last 20 tweets from the public activity feed. You can also get the last 20 tweets for a specific user or a user’s replies and direct messages. For some reason, they have greatly restricted the number of tweets available from the API. The only other way to get a significant amount of information is using the XMPP interface, which has been shut down to a majority of third party applications lately. Given  the reasoning behind the Twitter issues of late, I am OK with limiting the API. It is the limit of the functionality that is a problem. Steven Hodson of WinExtra has a good post on this problem as well. There is one paragraph that explains part of the problem of working with someone else’s API:

As bad as this could be for Twitter it shows how very shaky the ground can be for developing against web based API’s. Developers that get excited about this new territory and being able to transform someone else’s ideas into something new by using these APIs jump in with both feet but totally at the mercy of the service providing that API. It is an API that can change overnight or it can literally disappear if the service decides that like Twitter they need to make radical changes.

The point here is that with the limitations of the API, people tend to work around issues. Because Twitter has been around for “so long”, many applications have been written to enhance the experience. Tweeterboard was like the Technorati of Twitter. It used an idea close to the concepts of Google’s page rank to determine what Twitter users were worth following.

There are several third-party clients that use the API or XMPP interface of Twitter as well. Twhirl is one of the more popular clients that is used, and there has been no end to the complaining on how Twhirl was broken recently with the API limits. The real problem was that Twitter changed the API for determining the rate limit. They were also having problems with changes to the update API. Obviously, this is bad, and Twhirl was feverishly answering tweets about the issues as well as working towards fixing the problem.

The real problem here is that if the API and XMPP interface do not get stabilized and usable soon, we will see more services like Tweeterboard close down. Most likely this will not impact something like Twitter search engine Summize, but it will definitely impact some smaller clients or one-developer applications. If there are enough problems, people will not feel that creating the new application is worth it.

Lastly, if the API does not stabilize soon, other players can capitalize on the situation. We already saw the release of earlier this week. They do not have as many features and do not have an API. However it is open source, so they may get some development help. If people are frustrated enough, they may start making applications for the platform. If that starts happening, then people will not care that other Twitter applications are shutting down. They may not even be there to notice.