Social network App.net to shut down, open-source its platform

The developer-centric alternative to Twitter and Facebook might give rise to something new—but it probably won't be a social network

Social network App.net to shut down, open-source its platform
Bryan Mills (CC BY 2.0)

App.net, the microblogging service launched as a paid-subscriber alternative to ad-supported systems like Facebook and Twitter, has decided to close its doors and release its software as open source.

In a blog post, App.net cited diminishing revenue—a lack of subscribers—as the reason for the shutdown. Users have until March 14 to export their data, and at some point (it hasn’t been specified when) all of the code underlying App.net will be released as open source.

The project was widely regarded as a brave idea, but it was at odds with the accepted economics of social networking projects. It chose to only partially open-source its code base and didn’t generate enough critical mass to make it self-sustaining.

If you build it, they will come—or will they?

The original idea behind App.net, as Simon Phipps explained in a skeptical 2012 InfoWorld blog, was to crowdfund an effort to create a messaging platform that could host many kinds of apps, with Twitter-like microblogging as one of the most prominent. Since users paid for the privilege of accessing the service, it theoretically would be immune to the ethical quandaries of an ad-supported service. It was also meant to be more appealing to developers frustrated with Twitter’s tooling.

After an initial burst of interest, App.net in 2014 had enough customers to remain online but not hire full-time staff. The company had chosen to open-source only part of the code base and was perceived as unwilling to commit completely to an open source model and thus stimulate further adoption.

App.net’s approach stood in contrast to Diaspora, another open source social networking project. App.net had one central piece of closed, hosted infrastructure to run the service, with a number of open projects running on top of it. Diaspora provided all the code as open source, but left the burden of running it to users (some of whom have provided hosting for Diaspora nodes as a service).

Neither App.net nor Diaspora attracted a sizable audience—including the developers who were meant to be the primary users and evangelists for those systems.

Never underestimate the power of inertia

Despite its commercial nature, Twitter remains a chief venue for devs to connect with each other and obtain quick answers to shouted-out questions. For most people targeted by App.net, the immediate utility of Twitter—and the fact that everyone was already using it—outweighed any concerns about the commercial nature of the platform.

App.net’s shutdown notice hinted at the company realizing it had banked too heavily on developers as drivers of the business, rather than lay users. “Ultimately, we failed to overcome the chicken-and-egg issue between application developers and user adoption of those applications,” wrote App.net founder Dalton Cadwell. “We envisioned a pool of differentiated, fast-growing third-party applications would sustain the numbers needed to make the business work. ... [B]ut that initial excitement didn’t ultimately translate into a big enough pool of customers for those developers.”

One possible model for what App.net had in mind is Box. That enterprise storage company has focused on providing APIs for developers, allowing businesses to build their own storage and content-management functionality, with regulatory compliance already built into the platform. Box works because it addressed a genuine need and provided tangible conveniences; for most people, App.net’s value was more nebulous.

The next (and last) step for App.net is to offer all its infrastructure as open source. Previously, the company open-sourced key projects that ran on top of the service, such as the alpha microblogging client, but not its full underlying platform. One possibility is for App.net to go in the same direction as Diaspora—with the ability to be self-hosted, in much the same manner as a WordPress installation.

Will people ditch Twitter for an indie, bootstrapped alternative? Probably not, when Twitter remains ubiquitous, easy, and already populated by the people they want to reach. The more likely scenario is that others will repurpose App.net’s code into a useful option—a DIY service platform, for instance—and save the pieces for other projects. The big lesson is that it takes more than providing an alternative to get people to switch to it.