Hacker News new | ask | show | jobs
by jilebedev 5083 days ago
What pain point exactly is this trying to solve? That twitter is ad-spammed? I'll admit I rarely use twitter, but I can't recall ever seeing ads on it. (I can, as a side note, recall how /slow/ twitter pages load.) That twitter changes its API frequently and doesn't give adequate notice to the people who develop for it? How exactly does a user interact with app.net? I signed up for a twitter account because I wanted to hear a lot about a beta for a video game. How does this work with app.net? Are only broadcasters paying members?

Kudos for taking on the big dogs, though.

1 comments

To me, Twitter's pain-point is that it's a platform that can kick off third-party applications, doesn't have a way to extend its client, and doesn't allow integration with similar services (last I checked its ToS). As a result, it's software I can't fully control.

And I think that's the same pain-point with most of the Web right now. We can't really evolve our web applications because we don't control them fully; we have to accept the terms of the platform, which tends to be advertising driven, which means they want to constrain the UI.

If you're tired of Twitter, you can't direct your Twitter interface to a new service. If you want to add a new feature, you have to build an entirely new interface and hope they never revoke your app's access. If you want to send a private message, you have to share it with their logs.

This idea that we're building platforms is insane. Listen to Eben Moglen (founder of Freedom Box) talk about this; you'll enjoy it. As he puts it, a platform is a place that you can't leave. That's our pain point right now. We need services that you choose and web apps that are compatible, not walled gardens. Not for our applications, anyway.

So losing the advertising model has a pretty distinct advantage: it gets us away from the platform model, where they have to control the way you access the service so they can deliver the ads. I can't say how the business model fits into that, but the engineering side definitely wins.