Saving the FailWhale

I so want to befriend FriendFeed. It would be the right thing to do: Permanently screw with the commenters who correctly believe I’m an arrogant elitist who wants to maintain my flimsy status at the bottom of the C-List. Truth be told, I despise social media for the wasted time I’ve spent ignoring the next phase only to pathetically cave as soon as I see the tide of history passing me by.

34 days ago I carved FriendFeed a well-deserved new one as Twitter showed signs of leveling off at barely useful. Back then, IM and Track were protected from withering eyes by the random failures of the whole system; today the complete absence of real time network effects shifts the blame squarely on the the social media crowd and their inability to consolidate value. Summize has emerged as an adequate bandaid, but in so doing also suggests using FriendFeed’s Search function to similar effect.

Then the dominoes began to fall: Dave Winer injected some real emotion into the new cloud, recognizing as he frequently has done over the years that the power in disruptive transports depends on real human concerns to trigger the bootstrap phase. I think he finally realized that the decentralization evangelism didn’t fundamentally resonate with those who watch Dave for clues to the next direction. When he started engaging around his issues of A-List, pro blogging, and so on, people were forced to follow him into the silo.

In engaging with FriendFeed without subterfuge or TinyURL masking, Winer lent much needed credibility and validation to the cause. This wasn’t about Twitter’s failings. Honestly, I don’t know what it was about but it felt real, a nod to the new service that recalled his seminal validation of Twitter early in its run up. To be sure, Dave also pushed Pownce at a midway point to no particular effect. But this one felt different.

So I jumped in last night and started reading through the back archives. Where I thought I’d find a cacophony of trolls, I found something not all that different from the comment stream on my Saturday TechCrunch posts, except a bit more focused and less redundant. Perhaps the time spent out of the loop also spared me the learning curve of a threaded conversation, but again, not sure but better than expected. Maybe the new old guard of FreindFeed had begun to establish terms of engagement, subtle rules for the road rage that social networks incite.

Of course, I quickly retreated to Twitter, like a failed spouse trying to reclaim the magic. Having taken all this far more seriously than many and endured the burden of the messenger, I found Twitter largely intact but strangely devoid of something that I couldn’t quantify. Is it like the fall-off of traffic after the Bubble burst, or the return of the traffic on 101 in the past two years but not the frenzy? Do I think Twitter’s time has past. Only that am I sure of, that that isn’t true. Twitter will get fixed.

I still feel emphatically that FriendFeed is an iterative child of Twitter, an evolution of the Facebook activity stream without the confusion of the hoarding vibe of that service. The conversations may be siloed, but not because of business model concerns but rather a genuine question of how to reinsert the flow back into Twitter or some other originating source. My intuition suggests TinyURL could play a role in bundling up personalized versions of comment threads based on some follow/hide/search algorithm, where pruning of rude and time-wasting behavior weights the way things are aggregated.

But the current FriendFeed architecture is not granular enough to support the signals necessary to create sufficient value to justify the investment. Comments are not time stamped, requiring careful reading in context to identify a recent comment from a day-old one. This is an artifact of the lack of real time XMPP flow, which seems likely to appear soon. The explicit voting of the current system would be augmented, and perhaps replaced, by a more implicit filtering based on previous behavior and the user’s perception of the commenter’s credibility, insight, or entertainment value.

The endgame is not yet fully engaged, but the tiering of the FailWhale has produced a competition on a far more level playing field than I anticipated 34 days ago. The drip drip of Twitter API pauses and repetitive displays of old FriendFeed messages in Twhirl’s omnibus displays reminds me of the immaturity of both platforms, but the realization that the FailWhale serves all players in this rapid buildout augurs well for the near future.