Loath To Improve Itself, Could Twitter Be Disintermediated By A Protocol?
John Herren, known by some for inventing TagCloud.com, is a talented developer who has spotted opportunity at the intersection of Twitter and what he calls "Web hooks." In a blog he posted earlier today, he offers some real-world examples of how easy it is to trap Twitter's e-mail notifications for events that can trigger any business process. The same could go for an
John Herren, known by some for inventing TagCloud.com, is a talented developer who has spotted opportunity at the intersection of Twitter and what he calls "Web hooks." In a blog he posted earlier today, he offers some real-world examples of how easy it is to trap Twitter's e-mail notifications for events that can trigger any business process. The same could go for any site (eBay, Facebook, etc.). If there is such a thing as Web 3.0, he may have scratched its surface. Bigger picture: Could social networks like Twitter end-up disintermediated by open protocols?After reading Herren's blog entry, you can't help but ask yourself if the vaunted API that so many social networks offer in order for developers to programmatically unlock some of the networks' inner workings might also lead to their doom.
Think about it. Once a service like Twitter reaches the point at which the majority of interactions with it are taking place through third-party clients (via the Twitter API) like Twhirl, TweetDeck (my personal favorite), and Twitterific, what's to prevent those clients from effectively talking directly to each other instead? Take out the middleman. That next step (really, just a baby step in the big picture given all the open source developers looking to scratch some itch) would only require (a) that Twitter be turned into a protocol instead of a service and (b) that each of us have our own servers for communicating (Twitter-style) via that protocol.
Such a server could run on a physical server, a virtual machine, our desktops or notebooks, or even our smartphones. Multitenant cloud-based services could spring up all over the place that offer you your own turnkey Twitter server that talks to all the other Twitter servers. Before you laugh at the idea, consider this: this is pretty much exactly how Internet e-mail works over an open protocol called SMTP.
Architecturally, distributing Twitter in this way has several advantages. For starters, the famous FailWhale (a sign that Twitter is, for whatever reasons, unresponsive), would pretty much go away. Your Twitter server or mine might occasionally be out of service but, at worst, it would only slow the delivery of certain messages to certain people.
Another advantage would be performance. The distribution of Twitter across thousands if not millions of servers would mean that a single server like Twitter.com would probably never be overwhelmed to the point that we have to wait for the Twitter service to be upgraded.
No more API limits: today, each account has a limit to the number of allowable API executions. The folks at Twitter.com get to decide what this limit is. When was the last time someone got to tell you how many e-mails your server is allowed to send out?
Then, there are all those obvious improvements that could make Twitter.com more useful. For example, an easier way to track threads and conversations. If every Twitterati had a dime for every great idea s/he came up with for bettering Twitter, there would many more millionaires on the planet. That's how starved we are for improvements to Twitter.
But the folks at Twitter are wholly disincented from making investments beyond what's required to keep the service working. Why? Because they'll never get a return on that investment. Whatever Twitter is worth to whoever buys it, that worth won't change one single iota as a result of any new enhancements. That's because Twitter.com has no competition. Unless, of course, an open Twitter protocol shows up.
Server Market SplitsvilleJust because the server market's in the doldrums doesn't mean innovation has ceased. Far from it -- server technology is enjoying the biggest renaissance since the dawn of x86 systems. But the primary driver is now service providers, not enterprises.