$1,400 stimulus check timeline Apple to discontinue iMac Pro Intel's Thunderbolt International Women's Day Meghan Markle, Prince Harry and Oprah Tesla's Full Self-Driving Beta

Facebook wastes no time putting FriendFeed to work

The newly acquired start-up has become the face (pun intended) of not just Facebook's further development of real-time streaming technology but also its forays into the open-source community.

PALO ALTO, Calif.--Facebook has unleashed a Tornado, and it's hoping that some eager engineers will go catch it.

Earlier this month, Facebook released the open-source Web server framework called Tornado, which powers the real-time streaming behind its latest toy, social feed aggregator FriendFeed. And on Wednesday evening at the office that most recently housed the FBFund incubator program, senior open programs manager David Recordon and director of products Bret Taylor held a "tech talk" to pitch Tornado to a crowd of several dozen interested members of the Web development community.

"We had actually been planning on open-sourcing (Tornado)" prior to Facebook's acquisition of FriendFeed, said Taylor, who had served as CEO of the start-up. "When we got to Facebook we thought it was a really good opportunity to do it."

The slant of Wednesday evening's talk (which was quite technical, so I won't be going into significant detail): if you're dealing with real-time, streaming content, Facebook thinks Tornado is for you. And if you've been listening to anything that Facebook has been saying recently, it believes the real-time Web is the future for everyone--not just its own company.

"FriendFeed's a real-time system," Taylor said as he described how the Python-based Tornado framework's non-blocking nature was ideal for real-time Web services. "Essentially, every active user of FriendFeed maintains an open connection to the FriendFeed servers."

Both Recordon and Taylor are recent arrivals at Facebook: Recordon joined Facebook last month as its resident open-source guru, and the company had acquired FriendFeed a few weeks earlier in a deal that brought on board both a top-notch engineering team (its founders, including Taylor, were Google veterans) and cutting-edge technology for amassing and indexing real-time Web conversations--so cutting-edge, in fact, that it was unclear as to how the mainstream would ever actually accept it.

At the time, there were questions about what, exactly, Facebook would actually do with FriendFeed. In the meantime it's become clear that acquiring the would-be Twitter rival allowed Facebook to leap ahead with some of its development of new, real-time-focused features as well as to enhance existing ones with FriendFeed's technology and brainpower.

Open-sourcing the technology doesn't have an obvious financial end for Facebook. But it will ideally mean that some of the developer community will be marching to Facebook's beat, at a time when the company continues to compete with the far smaller Twitter for a majority share of what's come to be known as the real-time Web.

As for its Python foundations, Taylor said that FriendFeed had been looking to build Tornado in a manner "sophisticated enough that we could do all the things we wanted but well known enough so that a new engineer could theoretically understand our code base right away...Python has a lot of its flaws, I wish it had real inline functions like Javascript, but for all of its flaws it's actually pretty nice to use in practice."

Taylor told me afterward that no concrete plans have been put into action as to which Facebook features may be getting a FriendFeed makeover (so as to speak) but hinted that one getting talked about for some enhancement from the former FriendFeed team is Facebook Chat, the site's instant messaging client, because of its obviously real-time nature.

Tornado isn't the first technology that Facebook, still criticized by some of the open-source community for its heavy reliance on proprietary technology and a login wall, has released as open-source code: well over a year ago, the company released the code for a significant portion of its developer platform.