The Grievances Were Real, but They Were Not the Point

By mid-2003, RSS had accumulated several years of genuine technical debt. Dave Winer's RSS 2.0, released in September 2002, was a capable format — simple, widely deployed, good enough. It was also frozen. Winer had transferred the specification to Harvard's Berkman Center and declared it closed to further modification, a decision that protected stability for publishers and aggregators but infuriated developers who saw fixable flaws and no mechanism to fix them.

The complaints were specific. RSS 2.0 had no namespace-based extensibility built into the core. The format of dates was loosely specified and implementations varied. Content encoding was ambiguous — the relationship between <description> and arbitrary HTML inside it was unclear enough to cause real interoperability failures. Internationalization was an afterthought. None of these problems was fatal, but collectively they made RSS frustrating to implement correctly, and there was no standards body, no working group, no revision process to address any of them.

Terminal window on a dark monitor showing a raw Atom feed in XML, angle brackets and namespaces visible, person's hands on keyboard
A feed is a text file sitting on a server. Between the publisher and the reader there is a fetch, and nothing else.Photo: Tima Miroshnichenko / Pexels

What crystallised opposition was not the bugs themselves but the governance arrangement that made them permanent. A single individual, working outside any open standards process, held veto power over a format that was rapidly becoming critical infrastructure for the web's information distribution layer. Sam Ruby — a developer at IBM with deep involvement in the XML community — opened a wiki in June 2003 and invited contributors to sketch what a new, cleanly specified format might look like. Tantek Çelik and Mark Nottingham, among others, joined early. The project was first called Echo, then Pie, and eventually Atom.

From Wiki to RFC

The Atom community made a deliberate choice that distinguished it from RSS at every level: it would seek formal standardisation through the Internet Engineering Task Force, the body responsible for the protocols underlying the web itself. This was not merely a procedural preference. It was a statement about what kind of thing a syndication format ought to be — a public, collectively maintained standard, not a specification under private stewardship.

From the record

Chronology

  1. June 2003Sam Ruby opens a wiki to sketch a new syndication format; Tantek Çelik and Mark Nottingham contribute early
  2. September 2002Dave Winer releases RSS 2.0; in July 2003 the specification is transferred to Harvard's Berkman Center and declared closed
  3. December 2005Atom Syndication Format published as RFC 4287 by the IETF (authors: Mark Nottingham, Robert Sayre)
  4. 2007Atom Publishing Protocol published as RFC 5023

Key governance contrast

  • RSS 2.0closed specification, single steward, private veto; later transferred to RSS Advisory Board
  • Atomopen IETF working group, public archive, versioned RFC, traceable issue process

The IETF process took several years. The Atom Syndication Format was published as RFC 4287 in December 2005, with Mark Nottingham and Robert Sayre as the primary authors. The specification was precise where RSS 2.0 was vague: dates required RFC 3339 format, content was typed explicitly, and the relationship between summary and full content was unambiguous. A companion protocol, the Atom Publishing Protocol, followed as RFC 5023 in 2007, describing how clients could create and edit entries over HTTP — something RSS had never attempted.

The technical improvements were real and substantive. But the more significant outcome was structural. Atom had a published, versioned specification in a public archive. It had an IETF working group process on record. Anyone who implemented it could read the same authoritative document, file issues through a traceable process, and cite chapter and verse. RSS 2.0, by contrast, remained exactly what it had been: a useful, stable, privately stewarded document.

Printed copy of RFC 4287 on a wooden desk, a hand resting on the first page, coffee mug beside it
Specifications are argued over in rooms like this one. Atom's took a working group two years; RSS 2.0's had a single author and a closing date.Photo: Kindel Media / Pexels

The Détente That Followed

The format war that many predicted — one format displacing the other — did not arrive. Feed readers simply supported both, and most publishing platforms emitted both. WordPress, which powers a substantial fraction of the web, still offers RSS 2.0 and Atom endpoints simultaneously. The RSS 2.0 specification was eventually moved from Harvard to the RSS Advisory Board, a community body, which reduced but did not eliminate concerns about private control.

What the Atom split produced, ultimately, was not a winner but a clarification. The syndication ecosystem now had two formats with genuinely different governance philosophies coexisting without either destroying the other. Atom demonstrated that a web syndication format could survive an open standards process and emerge more rigorous for it. RSS demonstrated that a frozen, simple format could remain useful for decades without institutional backing.

The argument about who controls a specification — and by what process it may change — did not end in 2005. It surfaces in every subsequent syndication debate, from the JSON Feed specification Brent Simmons and Manton Reece published in 2017 to the governance structures of ActivityPub and the AT Protocol. The Atom split was the first time that argument became explicit. It has not stopped being relevant.