About Twitter…

You’ve probably noticed that Twitter recently announced a few changes to how they’re going to handle access to their API (part 1, part 2). This wasn’t particularly well received in tech circles. It’s certainly no coincidence that the more drastic changes and restrictions were announced mid-August when business is slow and people are away on holiday, but the public backlash is palpable nonetheless. For some good commentary on the situation and speculation where Twitter might be headed with this new direction, you could do worse than reading what Marco Arment, Ben Brooks, Rafe Colburn, Shawn Blanc, Watts Martin, Kyle Baxter and MG Siegler had to say. One common strand of argument is that by restricting access to its API and by extension limiting third-party developer’s ability to innovate, Twitter is biting the very hand that contributed significant features (such as @replies, #hashtags and even the logo bird). This is all the more troublesome because innovation from inside Twitter appears largely stagnant. They bought a couple of amazing third party Twitter clients and made them their own, only to subsequently abandon them (i don’t hold out much hope that we’ll ever see an update to Twitter’s official Mac client). Instead of innovating and improving their product, they seem determined to close off their platform in a desperate attempt to defend their existing turf.

It’s a bit early to foretell the eventual fallout of these changes, but a few early signs emerge. While early commentators began to worry that Twitter is effectively trying to shut down third-party Twitter clients, actual Twitter client developers don’t seem particularly concerned, or at least put on a happy face for fear of potential repercussions (exhibit a, exhibit b). Nevertheless, early ramifications of the policy changes are already manifesting: Twitter cut off Instagram‘s and Tumblr‘s access to friend-finding functionality.

Amidst growing discontent among techies, a number of interesting Twitter alternatives are rearing their head, most prominently Dalton Caldwell’s App.net, which positions itself more as a Twitter-infrastructure-replacement rather than a social-network-replacement and distinguishes itself through its for-pay funding model – joining the service costs $50 and has subsequently drawn both (brilliant) satire and criticism for being exclusionary.

What strikes me as most interesting about App.net is how they’re going about building their platform. Normally when tech mavens grow discontent with their tools, services and infrastructure and decide to do something about it, there’s a strong impetus to drastically re-invent and re-conceptualize the thing they’re trying to improve upon – re-inventing the wheel, so to speak. App.net seems to be taking a much more humble and pragmatic approach: rather than reinventing Twitter they seem determined to rebuild it. They acknowledge that Twitter as a platform worked pretty well a few years ago and they are building from that foundation. At this point App.net looks a lot like an alternate-reality version of Twitter, the kind of platform Twitter might have turned into had they decided to pursue the Twitter-as-a-platform model rather than the Twitter-as-a-media-company model back around 2010 (for an interesting take on that decisive fork in the road i highly recommend reading what Alex Payne had to say about the future direction of Twitter back in 2010). However, in building a Twitter competitor on similar principles a number of concerns remain: App.net being a closed, centralized platform, ultimate control over its future direction still rests in the hands of the platform provider. It is only a matter of time until the visions of the platform provider and its user base diverge (just as is happening right now in Twitterland). Considering that early App.net adopters are piling up hopes and expectations in a platform that’s still in alpha and far from finished, disappointment might settle in much sooner than with organically grown platforms.

Tent.io is another potential Twitter alternative that’s been gathering a lot of attention lately. Where App.net tries to establish a Twitter alternative based on a similar centralized architecture but with a different funding model, Tent.io proposes an open, decentralized protocol for social networking (they are taking the reinventing-the-wheel approach i was alluding to before), but as Dave Winer rightfully points out, “formats and protocols by themselves are meaningless”. As far as i can tell there isn’t a single implementation of the Tent.io protocol, so it’s little more than a pipe dream at this point.

I personally think it’s a good idea to replace an increasingly locked-down centralized service with a decentralized alternative, but i’m not quite convinced that Tent.io is the solution in this case. I agree with Brent Simmons and Dave Winer that RSS (no, it’s not dead) would make a decent starting point for building an open alternative to Twitter. Unfortunately there are two big user experience problems with RSS that no one has successfully solved yet – discovery and subscription. Twitter solved both of these problems magnificently: they made it very easy to tell someone how to find you by popularizing the now ubiquitous @handles and they made it very easy to subscribe to your Twitter feed by simply clicking one button. These might not seem like intractable problems to solve, but let’s not forget that RSS failed spectacularly in both areas in the past, and not for lack of smart and talented people trying. The question remains whether it’s even possible to tackle these problems in an open and decentralized fashion.