Federated Social Web: Top 10 of 2010
For those of us who have been working on federating the social web, 2010 has been a busy and energizing year. The idea has jumped from the speculative blog posts of tech influencers into code, meetups, and attention from everyday users.
(If you're not familiar with the term “federated social web”, you can review my blog posts, “What is the Federated Social Web” and “Features of a Federated Social Web.” The short version: it's people connecting across social networks.)
Here are a few of the FSW developments that I think have been important in 2010. The list is in order from least important to most important, and all opinions come from yours truly only. My criteria for inclusion were influence on future uptake of federation technologies – positive and negative. I didn't exclude events or developments that my company or I personally was involved in; it would be a pretty short list in that case.
Demise of Google Wave. In August, Google announced that it was halting development of its innovative collaboration tool, Google Wave. Announced at Google I/0 in 2009 and welcomed with over-the-top praise from the technology community, the project failed to get significant user or developer traction. Ostensibly Open Source and federated, the code was unavailable for end-user adoption or experimentation and the XMPP-based protocol was so low-level that it was unintelligible to third-party developers.
What does the end of Wave mean for the FSW? On the positive side, it clears out a confusing, distracting element from the conceptual landscape. But it's also soured some thought leaders on the federation pattern. Most important, I think, are the lessons learned: the value of having multiple implementations; the need to make protocols easy and modular; the importance of engaging existing players in the space. Finally, it shows what a tiny window new technologies have for success: Wave went from hero to zero in only 16 months.
Cliqset and StatusNet Salmon interop. In late March, the first two implementers of the Salmon protocol, Cliqset and StatusNet, demonstrated interoperability between their services. Throughout 2010, StatusNet and Cliqset users sent messages using the protocol across the implementation and social network boundaries. I'm particularly proud of the fact that the protocol's main developer, John Panzer, demoed Cliqset-StatusNet interop at Google I/O 2010.
Time will tell if this was a first step or a dead-end for Salmon. It's a crucial part of the OStatus stack, but its encryption requirements, and concerns about spam, have turned off developers and Open Web advocates. Unlike Activity Streams, Webfinger, and PubSubHubbub, Salmon hasn't had the kind of quiet, invisible adoption that helps push protocols along.
Mark Zuckerberg keynote at F8. In April, at Facebook's annual developer conference, F8, CEO Mark Zuckerberg announced a number of new initiatives for the Facebook platform. Two, however, took on larger significance. First, he announced a goal that Facebook would provide a social layer for the Web. Second, he announced a loosening of privacy controls on third-party developers using Facebook data.
The two announcements set off alarm bells through the Web – not necessarily well-thought-out, but influential. The combination of a winner-take-all vision of the social web with arbitrary changes to default privacy settings, however innocuous, caused serious second thoughts. The ensuing concerns directly inspired a number of new federation projects, including the high-profile Diaspora (see below).
Google Buzz. In February, Google announced its new social aggregation and microblogging site, Buzz. Integrated with GMail, and leveraging a user's existing contact list to kick-start the social graph, the service leaned heavily on Google's huge userbase. Most importantly for the FSW, Buzz's APIs include most of the parts of the OStatus stack, like WebFinger, PuSH and Activity Streams, so that's it's possible to follow Buzz users from OStatus-enabled applications. A side benefit is that Google employs most of the supporters and developers of these protocols, including Bret Slatkin, Brad Fitzpatrick, Chris Messina, Joseph Smarr, and John Panzer.
I can't overstate how important the launch of Buzz was for us at StatusNet as we were designing OStatus. Knowing that there was at least one other compatible implementation in the wild, with millions of users, that shared data openly the way we were planning to, helped us immensely. I think it's also been helpful for other FSW developers as a testbed and proof that someone gets it.
Google Buzz may also, however, have a big negative effect. Some industry analysts have written it off as a failure. Its integration with existing social graphs inside Google apps caused privacy concerns that resulted in a class action lawsuit. If Google does another social network effort (the rumoured “Google Me” or “Google +1”), it consequently may not look or work like Buzz. This may mean that the FSW effort won't have its most powerful industry ally, which would be a big loss.
Delicious “Sunset”. In December, a leaked screenshot from an internal presentation at Yahoo! showed, among other services, popular social bookmarking site Delicious.com scheduled for “Sunset” (whatever that means, it's probably not something good). This venerable service, with dozens of imitators, helped form our understanding of how social software could work, and introduced tagging as a content-organization tool. With all its staff laid off, the millions of bookmarks and social connections stored in Delicious's systems are facing oblivion.
This is an opportunity and a potential distraction for the FSW. It's an opportunity because a distributed, multi-platform social bookmarking network isn't susceptible to the business requirements of any particular corporation. Delicious's “sunset” is thus a great cautionary tale for those who pooh-pooh the need for social network federation.
On the other hand, it's a distraction because there's a mistaken expectation that federation can “save” Delicious. Even if a distributed social bookmark network launches (stay tuned), and many people move their Delicious accounts to it, there will still be millions of Delicious accounts from inactive or infrequent users that will never be converted. Federation is robust, but it can't fix past mistakes.
SWAT0. One of the concrete outcomes of July's Federated Social Web Summit (FSWS -- see below) was the Social Web Acid Test, Level 0 (“SWAT0” for short). SWAT0 is a test for any federated social web technology or protocols. It describes a simple usage scenario – friends sharing, and commenting on, photos across the Web – with the additional requirement that participants be using different networks and, preferably, different server software.
Most importantly, SWAT0 required that we meet its goals by September 30, 2010 – about 8 weeks after the test was announced. Surprisingly, we had four demonstrations of this interoperability by the time the SWAT0 deadline was closed, and potentially a few more since.
SWAT0 was important for a number of reasons. First, it gave us a collective challenge as federated social web developers and enthusiasts, with an emphasis on collaboration and interoperability. Second, it made us focus on real solutions, real fast – not cryptographically provable but unimplementable solutions posited in some imagined future network. Finally, it emphasized the importance of getting technologies into people's hands today. We intend to continue developing the SWATx series of tests, encouraging new entrants to show SWAT0 successes with their software and protocols, and making right-now the time to build a federated social web.
W3C Federated Social Web group. The momentum of the FSWS and SWAT0 test attracted the attention of the W3C. The organization had previously had a working group focusing on surveying social network technologies, and its final recommendations included the launch of a Federated Social Web group. That group got its charter in November and was announced in December. (I'm fortunate enough to serve as co-chair of the new group.)
The decision by the W3C that social web federation is an important enough issue to put its resources and reputation behind is a momentous one. It sets an important focus on the issue, and will introduce it to industry players that might not have otherwise been involved.
One danger is premature standardization before the demand for federation is established. Making “just enough” standards to meet the interoperability needs of developers and users, while not stifling innovation, is a delicate balance to strike. Hopefully, if the participants in the new group can stay aware of the need for balance, there's a good chance this can be a successful effort.
Federated Social Web Summit. In July, almost 50 developers working on federated social web projects from around the world came together in Portland for a single-day summit. Participants included representatives from open source projects StatusNet, Diaspora, Appleseed, Project Nori and OneSocialWeb as well as open web advocates and developers from VZ, Facebook, Microsoft and Google.
There were some strong objections to the invite-only format, but I think this event was a turning point for the FSW. First, it brought together the main people working on this problem from around the world. Second, it put an emphasis on the end goal (federation) rather than the underlying technologies or software. Lastly, I think it ended the era of the lone mad genius who was going to save the social web with science. Interoperability and real-life usage were the main focus, and I think they'll remain so for the future.
OStatus. In March, my colleagues and I at StatusNet announced OStatus, the next generation federation protocol suite that we would be using for our software. Unlike our previous home-brew system, OpenMicroblogging, we'd worked hard to integrate the existing protocols PubSubHubbub, ActivityStreams, Salmon and Webfinger into a pleasing whole. We'd taken time to socialize the idea with colleagues who were working in the same area, and most of all, we'd implemented it. We could show that social networks could be integrated with an open protocol stack.
OStatus has become the protocol suite to use as a base for new distributed social networks. Many of the SWAT0 tests used OStatus, and projects like Friendika and Diaspora are implementing OStatus for interop. The protocol suite leaves some important use-cases open – notably, distribution of private information – and its name ties it probably too closely to our software. But its use of well-known developing protocols has meant at least partial, and mostly unintentional, support by sites like Posterous, Tumblr and WordPress.
Diaspora. In the wake of the F8 keynote, a group of four students at NYU announced a kickstarter project to create a distributed social network. Unlike other mad-genius announcements, they managed to raise US$200,000 to fund the project, with an unprecedented level attention from technology and mainstream media. Announced in June, the team released code by September and started a demonstration service in November.
Diaspora's success has been a huge inspiration for people working in the FSW. I'm not sure what's made it more prominent than other FSW efforts – the great media coverage, initial Kickstart funding, momentum from Facebook privacy issues, or just having four smart and enthusiastic people with their heads screwed on right. Whatever it is, Diaspora has become synonymous in the public's mind with the federated social web.
The stakes are high for Diaspora. A high-profile failure could be a huge setback for social web federation – essentially dooming its prospects for the consumer web. A high-profile success can potentially be the engine for a virtuous cycle of growth. Either way, I think Diaspora's a giant factor in the Federated Social Web for 2011.
For those of you who read this far, here are some of the also-ran events that I considered for this list.
Twitter's announcement of its own client applications
GNU Social combines with StatusNet
Launch of Friendika
Launch of One Social Web
Rollout of WebFinger at Google
Continued development of RSSCloud
Massive rollout of PubSubHubbub by Superfeedr
Did I miss anything? Let me know in the comments below. I'll have some predictions for 2011 in the new year. Watch this space!