In March 1999, Dan Libby of Netscape introduced RSS as alternative way of bringing content to users; it stood RDF Site Summary or subsequently, Rich Site Summary. Just how “rich” it would be– how much structure through the RDF Resource Data Framework to employ– was open to discussion. The basic premise of syndication was to deconstruct web content content into its logical components: headline, title, publication time, author, content, in order to be re-assembled by the user. It was a bit of a fluke for its time: websites were trying to be ever more flashy and interactive, and RSS undermined that experience somewhat.

RSS would be better promoted on a content channel of short, simple, and frequent items– the weblog. And there was one person who was not only already broadcasting his thoughts multiple times a day on his weblog, but had the the means and the opportunity to tinker with the protocol– software entrepreneur Dave Winer. His ScriptingNews weblog, which intermixed industry news with his personal views, had built up a impressive readership. Many of his early readers would ultimately start weblogs themselves and hail Winer’s influence.

The big value from syndication would come in aggregating multiple channels, and that would only be possible with universal standards. Software developer Mark Nottingham set up a mailing list called syndication, still accessible via YahooGroups. Rael Dornfest, a senior developer for O’Reilly Media (today the CTO) created the Meerkat aggregator in March 2000, and shared it with the list when he joined later that year. Other names I recognize from the archives are folks still quite active in the RSS world today: Stephen Downes, Aaron Swartz, Phil Wolff, and Kevin Burton, who now runs the TailRank aggregation/ranking service.

Missing from the list, for the most part, was the news industry. They had already published their own standards, ICE (Information Content Exchange) for syndicating content, as well as the NewsML markup for describing the stories. ICE was used by news companies (such as Reuters and Ziff-Davis) to distribute news to client organizations, and had been developed in 1997 by a consortium of 80 content producers and software vendors. Winer had his competitive disagreements with the ICE standards process, but they were calmed when Laird Popkin joined the World Wide Web Consortium’s xml-dist-api list and responded to his concerns. Popkin was, and still is, the chair of the ICE Authoring Group. He subsequently joined the syndication and offered his help in both places, posting a document RSS in ICE.

One of the key differences was that ICE was fundamentally about syndicating content, while the nascent RSS team wanted to syndicate links. The original publisher, after all, was expected to have a host of services on their own website, and that’s where the user ought to go. As Winer had written in 1999: “Stories can live and grow while new information is obtained. Comments from readers can add new facts and ideas and link to other related stories.” (see Constructive Media).

It was this type creative vision which attracted developers at some of the emerging media outfits, like O’Reilly Media and the Motley Fool to the list. A number of discussions centered on the legal and economic ramifications, but the focus overall was on the technology. Here’s some of the ideas that contributors brought to the list the very first day it was launched, July 2, 1999:

Mark Hamer: “I dream of having an application that consolidates, filters, archives and prioritizes all the information available from various ‘weblogs’ and sites like Scripting News, and presents it to me with an easy to use interface.”
James Carlyle: “My main thrust at the moment is categorisation of content, in the expectation that in the near future we will see thousands of channels.”
Carmen: “Right now between my built-in list, Ian’s list, and Dave’s list I have about 250. This is way too many for a normal user to sort through in order to decide what kinds of news they want. We need standard (yet extensible) categories.”

ICE had been mentioned on the list, but no one considered it a solution they were looking for. The above ideas all pointed in the direction of building a protocol that would extensible for future uses. Rael Dornfest wrote a post picking up on this idea, “We’ve already seen the requests for RSS support for categorization, aggregation, threading, discussion, job listings, and so on.” He continued that he felt that encapsulation would be the best way forward to evolve the standard. This would be done using XML namespaces, that aspect of XML which, in short, allows for the same element names to be re-used. The encapsulation of standard protocols is fundamentally the guiding philosophy of the Internet; it is how the Internet works. The HTTP protocol transporting this document works over the TCP/IP designed in the 1970’s. The development of TCP/IP took half a decade; Dornfest’s subesquent call to develop RSS 1.0 took less than half a year for the specification to be completed (see the rss-dev mailing list). In RSS 1.0, if a publisher wanted to enclose metadata that would only mean something to an RSS reader which understood DublinCore or NewsML tags, they well could.

Dornfest’s post opened a fissure that many saw cracking. Winer wanted to keep the standard simple and continued to advocate such; he was miffed that the name “RSS” would mean something else. The story of the “RSS fork” over the 2^nd half of 2000 has been chronicled at length, and the larger timeline is summarized here. Winer re-asserted his standard, updating it to be version 0.92. In September 2002, Winer came around and added support for namespaces with RSS 2.0. He also declared it the last version for Really Simple Syndication. A new project, first called “Pie” and briefly “Echo” and finally called Atom by vote, was initiated as a way to take RSS forward. And here are comparisons by Tim Bray,author of the XML specification; by Mark Nottingham, by Eric Lunt, co-founder of FeedBurner; by Danny Ayers, author of a book on RSS.

Needless to say, the different branches inspired a plentitude of comparisons. If only there were as much written work comparing what-is and what-might-have-been. That’ll be covered in the next part.