Ever since FriendFeed was sold to Facebook, we’ve been told over and over again that the company and its community were toast. And as if to underline the fact, FriendFeed’s access to the Twitter firehose was terminated and vaguely replaced with a slow version that is currently delivering Twitter posts between 20 minutes and two hours after their appearance on Twitter. At the Realtime CrunchUp, Bret Taylor confirmed this was not a technical but rather a legal issue. Put simply, Twitter is choking FriendFeed to death.
What’s odd about this is that most observers consider FriendFeed a failure, too complicated and user-unfriendly to compete with Twitter or Facebook. If Twitter believed that to be the case, why would they endeavor to kill it? And if it were not a failure? Then Twitter is trying to kill it for a good reason. That reason: FriendFeed exposes the impossible task of owning all access to its users’ data. Does Microsoft or Google or IBM own your email? Does Gmail apply rate limiting to POP3 and IMAP?
So the reason Twitter is killing FriendFeed is because they think they can get away with it. And they will, as far as it goes, as long as the third party vendors orbiting Twitter validate the idea that Twitter owns the data. That, of course, means Facebook has to go along with it. Playing ball with Twitter command and control doesn’t make sense unless Facebook likes the idea of doing the same thing with “their” own stream. Well, maybe so. That leaves two obvious alternatives.
The first is Google Wave, which offers much of the realtime conversational technology FriendFeed rebooted around, minus a way of deploying this stream publicly. The Wave team seems to be somewhat adrift in the conversion of private Waves to public streams, running into scaling issues with Wave bots that don’t seem to effectively handle a publishing process (if I understood the recent briefing correctly.) But if Waves can gain traction around events and become integrated with Gmail as Paul Buchheit recently predicted, then an enterprising Wave developer might write a bot that captures Tweets as they are entered or received by Twitter and siphons them into the Wave repository in near realtime.
Note that this doesn’t presume access to Google’s version of the firehose. As with Microsoft, no one has publicly described the terms of service, including whether either licensee can allow third (or fourth) parties to derive services such as Track from their copy of the stream, or indeed how long they can archive tweets if at all. But that doesn’t preclude a Wave developer from leveraging each user’s polling rights just like any third party Twitter app or, even simpler, warehousing the stream before it enters the Twitter stream. Placing this initiative at the Wave bot developer level also engages an App Store-like critical mass that may prove significant in the early buildout of Wave momentum. Destroying the false premise that Twitter owns the data to begin with is an added bonus. Or it will smoke out some details of the Google/Twitter licensing deal.
The same could be true of Microsoft’s deal for the firehose, but here, as with Google, Twitter may not want to risk flaunting ownership of a stream that can so easily be cloned for its enterprise value. And as easily as you can say RSS is dead, Salesforce Chatter enters the picture. Here’s one player Twitter can’t just laugh off. First of all, it’s not Twitter but Facebook Benioff is cloning, and a future Facebook at that, one where the Everyone status will be built out as a (pardon the expression) public option. This free cross-Web Chatter stream will challenge Facebook’s transitional issues from private to public, given that Salesforce’s cloud can immediately scale up to the allegedly onerous task of providing personalized Track on demand.
It’s likely this pressure can be turned to good use by Facebook, unencumbered as they are by any licensing deal with Twitter. Instead, a Chatter alliance with the Facebook Everyone cloud puts Salesforce in the interesting position of managing a public stream with Google Apps support, which eventually could mean Wave integration. Where this might break first is in media publishing, as Benioff noted at the CrunchUp. Twitter’s leverage over its third party developers could be diluted significantly once Salesforce offers monetization paths for its Force.com developers. So much so that this may call Twitter’s bluff with FriendFeed.
But FriendFeed has always been more of a tactical takedown of Twitter than an actual competitor, a stalking horse for just the kind of attack Twitter seems most afraid of. No wonder the speed with which Twitter is introducing metadata traps to lock down the IP before a significant cloud emerges to challenge its inevitability. Lists, retweets, location — they’re all based on raising the rate limiting hammer to discourage heading for the exits. It’s not that retweets reduce the functionality of the trail of overlapping social circles, it’s that they lock them behind the Wall.
By aligning first with Facebook and adopting Twitter’s look and feel but not its metadata wall, Chatter developers are less threatened by a Twitter embargo in realtime. Benioff doesn’t need Twitter’s realtime conversation if he can build out Facebook’s, by retaining Facebook Connect control of the private stream while accelerating the Everyone status cloud and co-mingling it with Chatter + public notification and Wave bot integration. Add Silverlight to iPhone streaming to Moonlight recompiling, and developers have a powerful porting path that will quickly populate the Chatter Store.
Strange bedfellows indeed, but that’s the risk Twitter is taking in cutting off FriendFeed’s oxygen. Crushing a weak opponent’s windpipe does little to consolidate your advantages, but it leaves an indelible impression in those potential partners who are calculating a complex algorithm of time to market, stream access, platform elasticity, and user trust. But when you add a platform (Salesforce) which has continued to stay ahead of expectations while mining the social wave, the balance of power shifts dramatically. When you see Seesmic CEO Loic LeMeur walk out on stage at the Microsoft PDC with a Silverlight demo he bootstrapped in 2 weeks, ditto.
I don’t expect anyone from Twitter to answer the simple question of when will Twitter give FriendFeed the same access they provide other third party client vendors. For now, it’s frustrating to not see the flow of Twitter messages in realtime, but over time we’ll build tools on top of FriendFeed to take such embargoed messages private. Once inside FriendFeed, the realtime conversations that result are just the kind of high value threads Chatter will support, Wave will accelerate, and Silverlight will transport. Keep up the good work, Twitter.