Where to Start
In order to publish a web feed, there are two things you need to do:
1. publish an XML file on your web server that adheres to one of the many available web syndication formats (RSS, RSS 2, Atom, etc)
2. place a related meta tag at the top of the relevant page.
For example, if you were going to publish a coupons Atom feed, you would place the xml file at http://www.example.com/coupons/feed.xml, and on the pages in the /coupons/ directory, you would add a meta tag in the HTML header like this: <link rel=”alternate” type=”application/atom+xml” title=”Example Coupons Atom Feed” href=”http://www.example.com/coupons/feed.xml” />.
The Meta tag tells web browsers (Internet Explorer, Firefox, Google Chrome, etc) as well as bots that there is a related feed (for our example, in Atom format) that corresponds to the content on the page, and provides a link for them to follow. Each web syndication format has a corresponding meta tag specifying it’s type.
Useful Resources
The XML file provides the actual feed data, and must follow a specific format and syntax, which can be researched quite easily, with examples, on Wikipedia or a related resource; see the Atom entry at Wikipedia: http://en.wikipedia.org/wiki/ATOM.
Each protocol should also have a more detailed documentation page available on the web, for Atom feeds, visit here: http://tools.ietf.org/html/rfc4287.
Alternately, it is often useful to view the source of an actual XML feed file, as a live example of how one should be constructed, check out the Google webmaster central blog’s feed: http://googlewebmastercentral.blogspot.com/atom.xml.
Before Publishing
Although many people refer to all web feeds as RSS feeds, RSS is only one of many web syndication formats, and many people these days (including some departments at Google, like all Blogger blogs) have been using Atom feeds in place of RSS. All syndication standards are similar to each other, and really the choice of which to use comes down to personal preference – on the user end it doesn’t really matter which feed type is found on the server, because most user applications (feed readers, browsers, etc.) interpret them the same way.
The contents of a XML feed file can be automatically generated, if the data is readily available like in a database, and works best for pages that have fresh content regularly. Alternately, the XML file can be manually built and updated with each new version of a page, which often works well for semi-static pages.
For example, the WordPress blogging software automatically generates the XML feed file when it is requested, injecting the most recent blog posts available from the database into the file.
Whether the pages you wish to publish are built manually or generated on the fly, Atom or RSS, the syndication syntax is easy to grasp, and should be fairly easy to put together for most developers.
