Blog Post

What Would a More Open Twitter Look Like?

As Twitter struggles with repeated outages on the network (which, to its credit, the company has been very transparent about), some are asking whether it isn’t unwise to rely on a single company for a communications platform that has become such an integral part of so many people’s lives. So what would a more open version of Twitter — or a world that relied less on it as a single point of contact — look like?

It might look something like the idea proposed by Dave Winer, inventor of the RSS standard. He’s written often about the idea of “a decentralized Twitter” and how he’d like to see someone develop a Twitter app based on open-source standards, which he compared to the Apache open-source server software that lies behind a majority of the websites on the Internet. More recently, he’s talked about how developers of Twitter clients — whether they be apps, or services such as WordPress (see disclosure below) or Tumblr — could build an “emergency broadcast system” that would function even when Twitter is down.

Distributed Twitter Model Proposed by Dave Winer. Image courtesy Dave Winer

Here’s how such a system could work: Clients such as Tweetdeck, Seesmic and Brizzly would communicate through the Twitter API (application programming interface) just the same way they do now, but when the network goes down, they would store tweets — either on their own servers or at some neutral location — and each client would be able to get updates from the others. When Twitter comes back up, the tweets get sent to the network as usual, along with a record of what was tweeted while Twitter was offline, so they can push out an updated activity stream. All this takes is co-ordination among the various developers of Twitter clients and a standard to work around.

Part of that standard could involve Montreal-based developer Adrian Evan Promodrou’s Prodromou’s, which is similar to the blogging software company WordPress. It provides Twitter-style microblogging software that can be run on a server by any user or company, but also has a hosted version with additional features. Promodrou has also developed an open standard designed to enable communication between microblogging clients called OStatus, which the company says is based on other publishing standards such as PubSubHubbub, WebFinger and Salmon.

Howard Lindzon, CEO of the investing-related service StockTwits (see disclosure below), says he and his co-founder Soren Macbeth see a future in which there are multiple versions of Twitter. Stocktwits pulls information from Twitter and also pushes information into it, Lindzon said, adding that if Twitter were to go away tomorrow, “I would miss the musings of a few hundred people that I follow, but so many companies would step up fast and a new leader would quickly emerge.”

Whether Twitter likes it or not, the service’s continuing network issues — as well as its growing dominance in the marketplace for real-time communication — makes it likely that more StockTwits and Status.nets will emerge. And that is as it should be. If the company is smart, it will do what it can to become the engine that helps to power these different ventures, or at least plays nicely with them, rather than seeing them as competition and trying to shut them down or block them off.

Post and thumbnail photos courtesy of Flickr user watchsmart

Related content from GigaOM Pro (sub req’d): Lessons From Twitter: How to Play Nice With Ecosystem Partners

Disclosure: Automattic (maker of and StockTwits are both backed by True Ventures, a venture capital firm that is an investor in the parent company of this blog, Giga Omni Media. Om Malik, founder of Giga Omni Media, is also a venture partner at True.

18 Responses to “What Would a More Open Twitter Look Like?”

  1. A decentralized Twitter is possible although it is hard to build. I did this as my final degree project in CS:

    It works over a Pastry overlay and uses its KBR and DHT features. It has a partial Twitter API support (tested with Spaz and Buzzbird) and it is an state-of-art implementation, so it has lots of room for improvements :)

  2. As covered in previous comments, technical solutions to this problem already exist (e.g. OStatus). The central question is motivation. Why isn’t Twitter, a single, commercial service provider good enough? An analogy helps here. What if email was provided in a similar way through a centralized, commercial platform. What would you lose?

    It turns out you would potentially lose a lot. I don’t pay directly for email, but it is a service provided through my ISP. In return that ISP provides a very straightforward service, namely data routing. They don’t read my email and index the content into some “user profile”. They don’t “own” the contents of that data in any sense. They don’t “inject” marketing into my “email stream”, etc.

    Does Twitter do this with the content I publish through them? I don’t know. They certainly have the legal right to. After all, they have a comprehensive license to use the data and do whatever they want with it.

    A second question is innovation. In digital communication we have learned that innovation goes hand-in-hand with open. Is there a reason to think that this will not be true here?

  3. An open Twitter? Ugh. I want a “closed” Twitter. That is, something like a constantly re-constructed virtual private network I can suddenly have private tweets, private newsfeeds, then switch back to the public feed.

    If Twitter actually makes real money, there’ll be no need to wonder if we should have a second one. Money guys will build it at that point.

  4. Ernest Nova

    All your tweets are belong to us. All others use SMS!

    Dave Winer’s idea is “hokey” retrofit. Dave Cridland and Brad have the right proposal – a federated messaging system with store and forward capability where each party/company can run their own server in their own domain. XMPP is the right vehicle but without the presence update traffic of the core platform- this is what pubsub does. It also make the identity name space distributed like email addresses.

  5. Great post. When you look at what Twitter is trying to do (1) completely host a new communications platform. 2) add value on top of that platform), I can’t think of anyone else doing something so ambitious. I don’t see why they don’t hook up with someone like Amazon for the first part.

  6. Twitter be replaced, by some sort of openly federated network? A laughable notion.

    After all, did Compuserve get replaced by email? Did AOL get supplanted by the web? Did the military and governments of every major power suddenly pick XMPP for IM?

    Next you’ll be suggesting that a company like Google would welcome openly federated IM, or contribute heavily to the underlying technology of an openly federated Twitter alternative itself.

    Sheer foolishness! This sort of knee jerk journalism is terrible.

  7. Hi,

    You mean like this thing that was invented in the 80’s and became practical n the 90’s, that contains multiple protocols to allow a pretty flexible feature-set and even had business-intended desktop management apps, um, called, um, Short Messaging Service!

    The great problem is the almost ideological stand that the largest of traditional/mainstream media have taken, e.g. esp. BBC, Guardian, CNN to exclusively hype [to non-early-adopters] the one single proprietary service at the expense of open standards and plural alternatives.

    Yours kindly,

    Shakir Razak

  8. Hey, Matthew, Evan Prodromou’s name isn’t Adrian Promodrou. Are you writing a spoof article? In that case you should rename too! (Seriously, the rest of the article is great, and it’s way past time this stuff hit the mainstream.)

      • It’s still wrong. “Prodromou”.

        There are more than 20,000 sites on the public web using StatusNet already.

        Also, because OStatus is a suite of existing standards, there are a number of sites that are already partially compatible: Posterous, Tumblr, Google Buzz, LiveJournal, WordPress, Cliqset. I follow dozens of people from different platforms in my own StatusNet instance.

        We’d love to see more participation in this open and federated social web. Probably the most important thing that people can be doing today is establishing their presence in the Open Web, and encouraging subscription through open standards instead of closed systems.