How Not To Publish RSS

More and more web sites are publishing RSS feeds of their useful data. But if you’re going to publish RSS feeds then it’s important to get it right. The Transport for London site is a good example of how not to do it.

Whilst trying to get the latest news on the current Northern Line unpleasantness, I ended up on this page on the TfL site. “Oh look,” I thought, “an RSS feed. I’ll add that to my Bloglines subscriptions.” But it wasn’t as easy as it should have been to track down the actual address of the feed.

My first guess was that the link labelled “RSS news feed” would link to the feed. But that turned out to be a link to a page that explained what RSS is. My next guess was that the orange “XML” icon was a link to the feed (that is, after all, the way they usually work). But no, that was just a graphic with no link at all. That, in itself, is a major break with standard practice. Then I looked for the “autodiscovery” headers that can be added to a web page to point to related RSS feeds. They were missing too.

At that point I was stumped. I’d exhausted all possibilities. Realising I had probably missed something, I retraced my steps. Remember the first link that I looked at? The one that contained an explaination of what RSS is? Well on further investigation it also contained links to not one but two actual RSS feeds. Finally I was able to add the feed I wanted to my feed reader.

Now I consider myself pretty clued-up on how this stuff works. But it took me a good couple of minutes to track the feeds down. The majority of users would probably not have my tenacity. TfL might be wondering why they aren’t getting many subscriptions to their RSS feeds. I think it’s obvious why.

I don’t think it’s hard to follow current best practices when implementing features like this. I wonder why so many companies get it so wrong. As long as sites have non-standard ways to access these features, users will just get confused and not use the features.