August 2012 Archives

API Madness

This week, the inter webs went all aflutter when Michael Sippey of Twitter announced the Changes coming in Version 1.1 of the Twitter API.

In general, Twitter is seeking to more tightly control the user experience and discourage active development of third-party client applications. Yet it seems like so much of the success of Twitter comes from the origin in lack of control. It was simple and the users built most of the conventions that Twitter relies on.

For a service like Twitter that is so simple and basic, will attempting to make it into something different end up killing it off? Will App.net or something else be the Facebook to Twitter's Friendster or Myspace?

Even though much of the use of Twitter is on its own website, it seems like the most active users, and the reason that the service became successful comes form client software, all of which came from third-parties. Twitter's official clients were originally written independently by Loren Brichter as Tweetie and then acquired (and then apparently left for dead.)

As Twitter is trying to build itself into a business, it's also changing to dictate how the service is used, rather than building on the conventions that have evolved.

Web communities tend to take on their own unique and individual character and personality. Some, like Metafilter or Reddit are largely supportive and collaborative. Others, like 4chan or Funnyjunk take on personalities that are more anarchic or antagonistic. The communities that tend to have stronger community values are the ones who tend to have stronger moderation enforcing community norms, whether that is individual moderators like at Metafilter or the community norms that Twitter's users established. In particular, the @username convention and the #hashtag convention both came from use, not from Twitter.

Image uploads were supported by third-party clients long before Twitter launched it's own image hosting service.

And while if hoping to extend the Twitter service and sell it to advertisers, it makes more sense for it to be a website rather than a service that works across different software. But it seems more likely to alienate the user and developer ecosystem that Twitter enables. And because Twitter as it is today provides tremendous value to the users and developers, trying to recapture some of the value from the users and developers, rather than sell those users' attention to advertisers seems like the better way to capture value, because it will encourage the users to use the site more.

By carefully and narrowly designating what the Twitter service is, rather than listening to what the most active users want, is Twitter going to be driving its most active users and third-party developers away from its service?

The most-active Twitter users seems to interact with the service mainly through Tweetdeck*, Tweetbot, or the rapidly stagnating Twitter apps rather than the website.

*Yes, Twitter own Tweetdeck, but it seems to be a vastly different experience than the Twitter website.

Developer Rules of the Road,

Terms of Service and Display Guidelines, which will become display rules.


Marco.org, Interpreting some of Twitter’s API changes: "I sure as hell wouldn’t build a business on Twitter, and I don’t think I’ll even build any nontrivial features on it anymore."