kyleandkelly.com

It seems RSS is badly misunderstood, which is not surprising, given the confusing version numbers out there (0.9X, 1.0, 2.0). In short, many (all?) of the SitePoint editor’s points are not applicable to modern RSS feeds. Even the responses neglect to mention the points below.

To begin with, I’m going to assume that RSS versions 0.9X are legacy - water under the bridge. That leaves RSS 1.0 and 2.0, which is important because both of these versions are extensible.

My experience is with RSS 1.0 (others may be able to speak to RSS 2.0), so let’s clear up some of the confusion:

1) RSS 1.0 is extensible, so you can use it to create a feed of whatever you want. Extensions have been written for streaming media, MP3s, etc. And if you want to create a feed for something new, just create your own extension.

What does this mean? All of these arguments about RSS being only for headlines with descriptions, etc. just doesn’t hold water for modern RSS.

2) One of the side effects of the extensibility is that someone’s already written an extension for encoded (encoded means HTML, XHTML) content. Which means that, yes, you can include advertising, or position your images, or whatever you can do in HTML.

Anyone thinking otherwise is just plain mis-informed, and I’d recommend a trip to Ben Hammersley’s RSS feed to get back in touch with reality. Ben wrote O’Reilly’s Content Syndication with RSS, so he knows his stuff. His feed contains cartoons, advertising - the very things the SitePoint editor claims aren’t possible with RSS.

The caveat to this is that some of the more primitive RSS aggregators (most of the Windows ones, including FeedDemon) only read the description tag, which by spec is text-only, so they miss out on the encoded content. But that’s bound to change as they mature. NetNewsWire is pretty much the gold standard among aggregators - if you want to see what a mature aggregator would look like, that’s the place to look.

To wrap up, with RSS 1.0, points #1, 4, 5 aren’t applicable. My guess is that you could also address point #3, potentially via some clever extension of RSS. Others have adequately addressed the folly of point #2, though I may point out that the bandwidth difference between an HTML newsletter and HTML embedded in RSS is moot. RSS may even be more efficient, as a good RSS feed only sends back what’s changed since the user last checked, thereby distributing the load.

Maybe (hopefully) this conversation will serve to better educate people on the possibilities of RSS.

written by Kyle