I’ve been meaning to write something about the Twitter API changes for a while. For one thing, they directly impacted me. I had started off writing a post while I was on planes that ended up coming out whiny and childish so I abandoned the piece1.
I was in the process of developing a Twitter iOS application that, while not directly a third party client, was still close enough to one that I felt it would be a more prudent use of my time to start focusing on other projects. This had consumed a great portion of my time, to the point that I was forgoing client work in order to develop this side project. While it wasn’t a ton of lost money, and while it’s also not Twitter’s problem that I decided to take a risk and work on side projects, I did have to live off of my intern’s wages alone for 4 months and it sucked. That said, the sunk cost was not enough for me to continue working on the project in light of the API announcements.
Twitter’s behavior today has made me feel very, very, very good about my decision.
I quipped about this earlier (humorously enough, on Twitter), but I want to talk about it a little more here. This is an interesting piece of news for a few reasons.
- This is the first public precedent for how Twitter is going to respond to developers looking to work against their user token caps. And while we will most likely never actually know what Twitter asked of Tapbots, nor what Tapbots asked of Twitter, what we do know is that neither party liked what the other had to say. At that, it’s pretty safe to assume that Twitter’s “best interests” are just as divergent from 3rd party dev’s “best interests” as we had feared.
- On the surface, this registers simply as Twitter enforcing their new rules. But there is a subtext here that I don’t find to be all that subtle: Twitter is not content with passively making client development difficult through these new rules. Twitter is actively making development of third party clients difficult. Not with pronouncements and guidelines, but by taking physical action against developers.
That second point bears some discussion. Tweetbot for Mac is not a finished product. It’s alpha software2, still in the process of being tested, vetted, beaten up on, and run ragged. While people were testing and writing software before the concepts of “alpha” and “beta” even existed, and the ability to perform those types of tests are by no means a right, the actions that Twitter took today are actions that directly impede not just the success of a product, but also the creation of the product thereof.
Nobody can make the argument that Tapbots is not a group of intelligent, talented, and successful developers. It should be safe to assume, then, that they know what is best for their development workflow and what steps they should be taking to best succeed in the development of new products. They felt that an Alpha and Beta testing phase would be important, and when they recognized that the new API guidelines would impede their ability to do so effectively or could potentially harm the future of their finished product3, they requested some sort of stay of execution, which Twitter denied. In other words, Twitter consciously and deliberately took direct action whose end result was to interfere with and impede the value gained from a particular aspect of the Tweetbots workflow.
Or in plain English, they are intentionally making it harder for Tapbots to make their new Twitter client.
Nobody can begrudge Twitter for wanting to run a successful business, and if you’ve listened to any of the myriad 5by5 shows4 talking about this lately then you’ve no doubt heard all the arguments that could be made both for and against Twitter for the path that they’re going down.
But at this point, it’s becoming pretty hard to argue against the idea of Twitter’s One True Message being that the era of the 3rd party Twitter client is coming to a close. Very, very quickly.
I’m still glad I started it, though, because it gave me a moment to introspect on the path I was on as a software developer and I gained a lot of insight.↩
Though a hell of a lot more stable than a lot of final software I’ve used↩
This is most likely the issue; I would guess that Tapbots didn’t want all of the people who used the beta and then decided they didn’t wanna fork over money for the finished product to count against their cap. Either that, or there’s a whack of people testing this app.↩
Listen to any of the last 2 or 3 episodes from Marco or John Siracusa and you’ll get an idea.↩