I think we should probably stop calling it syndication APR 13 2004
"Syndication" (via RSS and Atom) is about to hit the big time. It's getting a lot of coverage in on/offline technical publications and will soon be covered in more mainstream glossy magazines and newspapers. Millions of people are now using RSS and Atom to syndicate their Web sites. Large media organizations like the BBC are syndicating their content via RSS. Amazon is syndicating lists of their bestselling items via RSS. Syndication is booming. Syndication is why RSS and Atom use is skyrocketing. Say it with me: syndication!
But is syndication really what everyone's all excited about? I don't believe so.
When the BBC (for example) provides content (headlines with story summaries, dates, and links back to full stories) for publication on other sites, that's syndication. This is what's happening with RSS on My Yahoo! and the purpose for which RSS was first developed at Netscape. Weather.com syndicates their five-day forecast (for a fee). Offline (where the syndication idea originated), United Media syndicates comics like Dilbert to hundreds of newspapers. Republishing is a distinguishing feature of syndication. When content is syndicated, the reader is getting the content from someone other than the producer. The BBC provides content to an online regional UK newspaper which is then read by that newspaper's readers.
BBC content --> regional UK newspaper --> readers
But things have changed since Netscape introduced RSS. RSS and Atom feeds are now largely read directly by people with newsreaders. The BBC provides their content in RSS format, a reader accesses the file from the BBC's server, and reads it.
BBC content --> readers
Hmm. So, people access documents written in a markup language that have been published on a Web server with a software application. If this seems familiar to you, it should. It's called Web browsing and has nothing to do with syndication. RSS readers and newsreaders are just specialized Web browsers, nascent microcontent browsers if you will**.
If not syndication, then what makes RSS and Atom so compelling in comparison to plain old HTML pages? The data contained in an RSS/Atom file is more specialized and structured than in an HTML file***. An HTML page has a title, maybe a header, some paragraphs, and perhaps a couple of lists. That's all a page can tell the browser about its information. When a newsreader loads an RSS file, it knows quite a bit more about the content contained therein. It knows the file contains 15 items (an item is typically a news story or weblog post) and each of those items has a title, a description, a link, maybe some categories, when the item was published, etc.
Using this information, the newsreader can then display the content in these files in various ways that are helpful to the reader. It can tell you at a glance that you have 68 unread news items; it can aggregate items from several RSS files into a new "feed" (perhaps a feed on biotech); it allows you to skim literally thousands of different items from hundreds of different sources sliced and diced in a myriad of ways. RSS and Atom treat the items contained within a file as first-class citizens.
So, that's the big deal about RSS and Atom, not syndication (although RSS/Atom can be used for syndication). I figure that if we technologists, publishers, and journalists are going to get all excited about this stuff and evangelize it to others, we should make sure we know exactly what we're so excited about.
**As an aside, what are now called RSS readers and newsreaders will eventually evolve into microcontent browsers (bad name for a good idea). I talked about such applications last year in relation to Safari and Sherlock:
A web browser is a tool for people to get information from the web. Much recent effort has gone into developing other interfaces through which to do just that. With Watson, Sherlock, and NetNewsWire, you "browse" the web for specific kinds of information with interfaces custom built for each task. Why the distinction between regular web browsing and web browsing using specialized interfaces for structured data?
With Amazon offering product information and the availability of other non-news information via RSS and Atom, the term newsreader is already a misnomer. When more people start publishing content that doesn't fit the title/description/url format (recipes, movie reviews, photos, music playlists, etc.), "standard" formats will start to spring up (some have already) and the browsers will need to support them in some fashion. (This requires that the publishing tools support these new formats as well, which they eventually will. The whole ecosystem -- readers, publishing software, publishers, browsers -- will move along in fits and starts, just like it did with RSS.)
***This isn't strictly true. Valid XHTML files are XML and there's no reason why you can't make an XHTML file that contains headlines, dates, and summaries and use them like people are using RSS/Atom files. Tantek and co. are discussing something similar with XOXO, using XHTML to "semantically [express] Outlines and Blogroll-like subscriptions in an XML format that is both renderable by browsers and parsable by strict xml engines". But for now, let's assume that RSS/Atom files are more specialized and structured than HTML/XHTML files, if only because of current convention.