Setting the record straight on Aaron Swartz's contributions

I don't have more to say about Aaron Swartz's death; I knew him a little, but felt his loss keenly. As coverage appeared, however, I found myself concerned about his legacy. Aaron did so much in such a short period of time, but several of his accomplishments have been glossed over in a way that distorts his contributions.

RSS co-inventor: no

Aaron co-authored the RSS 1.0 specification. Impressive enough for anyone at any age, Aaron was perhaps 15 when he accomplished this. But he didn't co-develop RSS itself, as most mainstream press accounts have stated. RSS originated at Netscape (as version 0.9) to handle channels on its portal site, based in part on previous work by Microsoft and others. Netscape revised the spec to incorporate ideas from Dave Winer; Winer revised RSS further, incorporated it into his firm's products, and evangelized its adoption. He also updated the spec to version 0.92 to add audio enclosures, which led to the spectacular growth in what rapidly became known as podcasting. (That was partly due to work by Winer and Adam Curry to demonstrate automatic delivery of podcasts to iPods and other MP3 players.)

RSS 1.0 was an entirely separate effort spearheaded by O'Reilly Media with the participation of many other interesting Internet media and syndication firms, and which used RDF (Resource Description Framework) elements, which form the basis of the Semantic Web. RDF comprises both ends of a link (not just a hyperlink pointer) and the relationship between them. RSS 1.0 was a dead end in itself, though still in wide use because many blogging system generate RSS 0.92, RSS 1.0, RSS 2.0, and Atom syndication files. But Aaron and others went on from RSS 1.0 to work at great length on RDF at the World Wide Web Consortium (W3C). The Semantic Web remains an important concept, still not fully realized, and Aaron was a critical part of that evolution.

Winer's RSS went from 0.92 effectively to 2.0, the rights to which he transferred from his firm to the Berkman Center at Harvard, which then released it under Creative Commons license in 2003. The Atom spec was meant to replace RSS with something fresh and designed around how blogs and commenting were being used, but has ultimately only complemented RSS. The RSS Advisory Boardhas become a kind of keeper of the flame making minor tweaks and having been formally assigned by Netscape the 0.90 and 0.91 spec and by Yahoo the Media RSS spec.

Reddit co-founder: yes

It's a minor quibble, but has provoked many arguments over the years. Reddit was founded in June 2005 by Steve Huffman and Alexis Ohanian. There are varying accounts by the two of them and by Aaron over the years as to when Aaron became involved, but it was after the start. Aaron moved Reddit from lisp to python, and later released the web.py script into the public domain that was the framework for running Reddit. (Reddit later moved to other code.)

Aaron's existing firm, Infogami, merged with Reddit in late 2005 at the suggestion of Y Combinator, which had funded both. Ohanian often calls this an acquisition by Reddit; Paul Graham of Y Combinator has repeatedly called it a merger. Condé Nast acquired Reddit at the end of October 2006. Aaron moved to San Francisco and worked out of Wired's office (also a Condé Nast property) until he was fired or asked to resign after an extended vacation and (Aaron said) convalescence.

Aaron didn't technically co-found Reddit, although he claimed in May 2007 that he had worked on it earlier than Ohanian and Huffman agree. Ohanian posted a Google+ item in 2011 trying to clarify with a timeline when Aaron joined after his arrest for the JSTOR incident pointing to a 2005 Wired story with quotes from both Aaron and him in it. But it seems clear that Y Combinator's head, Paul Graham, anointed Aaron a co-founder after the merger, since the three were founders of their respective firms. The press release from Condé Nast announcing the acquisition (PDF) also called him a founder; that would have been approved by Reddit before being released.

web.py is one of many pieces of code that Aaron wrote and released either into the public domain or via broad Creative Commons licenses. Between that code, his early involvement in Creative Commons, RSS 1.0, and RDF its likely that a good hunk of the Web pages in existence have some connection to his work.

Aaron's reach was huge, and his loss incalcuable. But his work should be accurately remembered.