13.1 RSS: Quick History

Both a varied history and some controversy surround RSS. Rather than spend time on this in the book, I'll point you to a Yahoo group's RSS Development Group message that details much of it (at http://groups.yahoo.com/group/rss-dev/message/1136). The concept of a providing data in a defined format in order to support news feeds and data channels didn't originate with RSS. For instance, the use of RSS—in any format—was predated by Microsoft's CDF, which I used at one time to provide data channels of my web sites.

RSS originally stood for RDF Site Summary and was a vocabulary of RDF/XML developed at Netscape several years ago for the company's implementation of channels. The use of RDF in RSS, as well as in earliest implementations of Mozilla, was due to R.V. Guha, an original RDF pioneer. When Netscape's attention was diverted to other matters — or should we say when AOL decided the support for RSS wasn't an effective profit center — RSS was left orphaned. This actually had a physical impact, because one result of this abandonment was Netscape pulling the DTD for RSS 0.9/0.91 that people used to use to validate their RSS XML.

Other people and companies took up the interest in the specification, including Userland's Dave Winer, a principal contributor to the earliest RSS specification. It was Winer's and Userland's use of RSS that sparked a growth of RSS, first within Userland, then eventually outside of the company.

These versions of RSS did not depend on RDF, but instead were based on an RSS-specific XML vocabulary. The specification released in December 2000, RSS 1.0, is based on RDF. One reason for switching back to an RDF base was to inherit the rich extensibility built into RDF, including the use of namespaces to handle element collision. The RSS Group saw in namespaces the answer to the problem of how to extend RSS without having to continually release new versions of the specification. Another reason is that the information included in RSS feeds—such as article title, author, excerpt and so on—is a rich source of information. By implementing RSS within an RDF framework, there's hope that the information can be merged with other RDF vocabularies and uses.

Today, RSS 1.0 is solidly RDF based. However, non-RDF RSS documents are currently in use throughout the Internet, primarily based on Userland's current RSS specification. The most recent such non-RDF specification was RSS 2.0, released late in 2002.

For a more detailed and comprehensive look at RSS and its history as well as non-RDF implementations, please see the O'Reilly book Content Syndication with XML and RSS, by Ben Hammersley. Ben covers both branches of RSS, the RDF and the non-RDF versions.