RSS Email Addresses

A few months ago, there was some discussion on the RSS Advisory Board mailing list regarding the format an email address should take in author, webMaster and managingEditor elements. I thought it might be a good idea to investigate how people were actually using these elements in the wild, but it wasn’t until recently that I had a chance to put together some tests.

I started with a little over 20,000 feeds gathered from a variety of sources.⁠1 I figured that would constitute a reasonable representation of feeds on the Internet. From each one I extracted up to three email samples, ignoring any repeats of an element (on the assumption they would likely all be using the same format). This process resulted in around 10,000 actual addresses.

I knew from experience that there were a fair number of address formats in common use, but the variety I encountered in these tests was positively staggering. What follows is just a sampling of the forms returned.

As diverse as this list is, it could be argued that most of these examples are perfectly valid as far as RSS 2.0 is concerned. The spec just says that these fields should contain an email address – it doesn’t say what form that address should take.

That said, in RSS 0.91 the suggested format was specifically listed as userid@host (Full Name), and the examples in the 2.0 spec all still follow that suggestion.⁠4

As far as I’m concerned, that is clearly the format that feed producers should be using – and most do. However, the reality of the situation is that there are still a good deal of feeds inventing formats of their own (the author name by itself being the most common), and feed readers need to be prepared to deal with those too.

The good news is that most of these weird formats are fairly rare. The bad news is that if you’re really intent on parsing everything, there’s no end to the variations you can expect to encounter.


  1. The Syndic8 feed list, the CNET Blog 100, the Share Your OPML Top 100, and a selection of feeds from my personal subscription list.
  2. It’s worth noting that, despite appearances, neither of these two samples is a valid RFC2822 address.
  3. This is how the sample appears after XML processing. These are not XML character references.
  4. Assuming you can accept that the parenthesised name is optional.