March 13, 2010

Activity Streams Aim to Be DNA of the Future Web

When the well-respected open source advocate Chris Messina announced he was joining Google in January, many folks were concerned that his being absorbed in to the big company Borg would mean a cessation or redirection of some of his projects targeting the next generation Web, possibly in exchange of proprietary efforts to promote the company's products. Today, at the South by Southwest Interactive event in Austin, Texas, he spoke on how he and others in the community, both at Google and outside of it, are working to bring more meaning to our social networks, activity, and feeds, through extending today's data portability standards to include more information and more relevance. Messina walked through a history of the Web's publishing, from static portals of a decade ago, to today's RSS and Atom-powered sites, and suggested a future with even more information, based on streams, that tells a story.

Messina, after expressing his excitement about working with a team he believed was leading the industry in things he cares about, including the Data Liberation Front, letting data move from one site to another, said he was focused on what he called "generative structures" that were the underpinnings and DNA of how information is shared, updated and transmitted.

As he recounted, in 1999, portals ruled the Web, and people, myself included, would put their data in sites like My Netscape and My Excite, customizing these sites with headlines from third party services, primarily tapping RSS, which offered the headline of a story, a link and its description. In an era when publishers wanted to not give away their data, it was "the best we could do", he said. By 2005, a new extension of RSS was promoted, called Atom, which was still the essential concept of syndicating data from one Web site to another, but also adding an author and an identifier for the atomic bit of information.

Now in 2010, little has changed. Most news feeds of today, be they on Facebook, on customized portals, or the headline and link model that dominates Twitter, are fairly simple, and they don't indicate intent. As Messina said, "It's not all that different from the last 10 years, and that gets kind of depressing".

But what has changed is the increase of sharing rich media in these places, on platforms designed for "dead tree media". He said we should be able to show what we did, who with and why we were doing it, and that needs to happen through new richer formats for the social Web.

Activity Streams, an extension to the Atom Feed format, is looking to accomplish this by extending Atom and RSS with new aspects, including a verb and an object type. The world of FriendFeed, which supported a unified feed of 58 different services, where people could have one single stream that represented their identity online helped guide much of ActivityStreams' framework. As entries flowed to FriendFeed, they largely represented actions of posts, shares, bookmarks, reviews, from different sites. But ActivityStreams is aiming to do more than just syndicate data from one home page to another, as RSS and Atom have done for a decade, but also display intent and meaning.

"If your goal is to help people produce meaning, knowledge and culture, you have the basics for a pretty compelling social application and can motivate people to act," Messina said.

But with more streaming of information from many different sites, it can exacerbate the assumed problem of information overload - and tools need to be further developed to help us consume the data.

"We snack on information. It may feel like overload, but the tools haven't caught up," Messina said. "The solution to data overload is more metadata and we are at that point where can start generating that. We take the basic construct from 1999 and weave in some additional information - data about data."

As I outlined in my summary of DeWitt Clinton's talk on Google Buzz at the beginning of the month, Activity Streams are playing a big role in this new network, and these streams are intended to be open, not just for a company like Google, but other social networks that are tracking individual's activity and intent. The goal is to make discovery of intent data ubiquitous and transitive between sites, in the same way that RSS and Atom focused on publishing of data from one site to another.

Messina called the work on Activity Streams as iterative "baby steps", but ones that focus on getting today's rich media activity a home with a rich experience, and to make this process easy for service providers.

"If you have a Web site that has people doing things on it, and they have a feed they are taking with them, it is fairly trivial to add ActivityStreams information," he said. "Essentially we have the verbs and object types represented."

You can find out more on the continued development of ActivityStreams at