Why do we need Twitter? I don’t mean in the sense of my previous question from 2008 (i.e. what’s the point?), I mean why is it necessary to have a single company controlling “microblogging”? It’s a question that hadn’t occurred to me until I saw Matt Gemmell’s tweet this morning linking to Brent Simmons’ suggestion for a decentralised Twitter-like service:
Since Twitter insists on (1) not letting us pay, and (2) selling us to advertisers, we should probably replace it. http://t.co/ryfM0xwW— Matt Gemmell (@mattgemmell) June 30, 2012
The discussion is prompted by Twitter tightening its control of the service by restricting what third-party clients and developers can do.
It took me some time to accept Twitter as even vaguely worthwhile. One unexpected way that I’ve found myself using Twitter is as an RSS reader replacement (these days most sites post links to their articles on Twitter), particularly since Google crippled Google Reader’s sharing in order to try to drive adoption of Google+. The fact that Twitter is effectively just doing what RSS feeds and aggregators have been doing for years (microblogging is just blogging with a higher volume of shorter messages) raises the question of why it can’t use a similar decentralised architecture?
Like RSS feeds, Twitter-like feeds could be published anywhere on the open web as long as they conform to some standard format (e.g. RSS with a 140-character limit). People who don’t have a web server to host their feed could use one of a number of competing services just as they might choose Tumblr, Blogger or WordPress to host their blog. In theory clients could fetch the feeds directly from the source although in practice, as with RSS feeds, there would be some efficiencies and other benefits from using intermediate third-party aggregator services. Searching could be handled by Google or another search engine while notifications of mentions could be handled with blog-like pingbacks.
The benefits to a decentralised system based on open standards are that it would allow different services and clients to compete on implementation and would be resilient to censorship. Service providers could experiment with business models such as different advertising strategies or paid access for a premium service. Client developers could innovate in how they present information without the risk of being blocked from accessing the network. Network admins could set-up their own sub-networks based on the same standards, such as a Yammer-like network restricted to a company’s intranet.