How Simple Sharing Extensions Will Change the Web
posted Saturday, 26 November 2005
I've been studying Microsoft's proposed new RSS extension, Simple Sharing Extensions (SSE), for a few days now. Authored by Groove's Jack Ozzie and George Moromisato (pictured in this article towards the bottom), Simple Sharing Extensions has two big things going for it.
-
One, it's being personally pitched to the world by Microsoft's CTO Ray Ozzie with an openness and participatory approach that almost seems startling coming from Microsoft. -
Two, the draft spec is really good.
SSE is elegant, is itself simple, and provides an essential solution to an important problem with the increasingly two-way Web: Though RSS is a resilient, powerful, and ubiquitous protocol for Web content distribution, RSS has a one-way generic view of content flow. Without a standard way to process mutually published and edited content, there can be no shared perception of the unique pieces of information that pass between users and applications. These chunks of information are what RSS and SSE call items, and can represent calendar entries, blog posts, eBay auctions, podcasts, or whatever. With SSE added, RSS is still one way but we now have an easy way to collaborate separately on the same content (a wiki article, inventory list, or stock quote.)
Having no way to atomically perceive whether content is new or changed makes robust behavior, or truly interesting capability, impossible without resorting to hacks or non-standard techniques. SSE understands that more than one entity can be the "publisher" of an item and makes this possible to handle in RSS, which without SSE has a view of only one publisher, many content subscribers.
By itself RSS provides fall-down-the-stairs easy content sharing and it scales well and encourages loose coupling. But this radical simplicity actually forms a barrier to entry for sophisticated and higher-order content distribution scenarios like mash-ups, supply chains/rings, web-scale SOAs, etc. But RSS did enable a truly service-oriented web and gave everyone workable planetary scale content syndication that virtually every blog, wiki, and web application provides today. And with SSE added, RSS can now form the fundamental basis for the rapidly growing Web 2.0 information ecosystem.
The masterstroke is that SSE rides inside RSS. The massive collection of feed readers, RSS aggregators, tools, scripts, utilities, and infrastructure doesn't have to change a whit to deal with SSE markup that appears inside an RSS feed. This gives SSE feeds backwards compatibility and a piggyback ride on top of the almost omnipresent world-wide RSS infrastructure. Straightforward upgrades of existing RSS handlers can provide a smooth and simple adoption path for most, since they can test if a feed contains SSE and do additional processing that makes two-way conversations smarter, instead of the unintelligent content feedback loops that occured with plain RSS. Or they can ignore SSE markup in an RSS feed if they don't care about it and just use the content they find the old way.
SSE is still so new (and forming, it's not even at 1.0 yet) that virtually no software exists yet except for what prototyping Ray Ozzie and company did to validate their ideas. But expect to see SSE become very popular, and quickly, since it solves such an important problem. I anticipate that RSS+SSE will become the most popular way to syndicate Web 2.0 content and to glue applications together. I've been taken to task occasionally for my prediction that RSS will be the fundamental Web 2.0 protocol, and with SSE my prediction is reinforced. It's so good, so simple, so right that I think you'll see the Web services of yore become almost completely eclipsed by it. RSS has reached the tipping point in terms of hundreds of millions of available feeds, tool support, mindshare, and the all-important success factor: critical mass. SSE completes RSS in so many ways. I'll be talking about this in more detail in the near future but I wanted to encourage you to study SSE, promote it, and implement it. Now is the time, it's early yet...
Other excellent coverage of SSE: -
Charles Cook does an impressive job analyzing how SSE does replication and conflict resolution between items.-
Lance Knobel covers the exemplar of calendar sharing that Ray Ozzie used to introduce SSE.-
The Inside Microsoft Blog discusses the SSE announcement and how Dave Winer has asked Google to use its position to make SSE big, quickly.-
Read how Niall Kennedy is already building SSE exporting into NetNewsWire.-
Don Dodge did a good number on SSE and discusses Microsoft's encouraging use of a Creative Commons license for the specification.-
Update: Infoworld's Jon Udell recently did a bang-up job of analyzing SSE.
Do you think RSS+SSE will become the dominant Web service standard?
Technorati: web2.0, rss, sse
links: del.icio.us
SSE is interesting. I think defining it as an RDF vocabulary would make it
a stronger, better defined and more extensible standard. This need not
involve any more complexity.
