Why aren’t UserLand doing RSS Auto-Discovery right?

UserLand has announced support for RSS Auto-Discovery for Radio. This is great, but what is not so great is that they have implemented an extremely simple parser for the RSS link tag.

Here’s an example of a link element that Radio’s aggregator understands:

<link rel=”alternate” type=”text/xml” title=”XML” href=”http://www.scripting.com/rss.xml”>

In order for it to work it must be formatted in exactly this way. It’s totally brutal. If the attributes aren’t all present, it fails. If they aren’t in the correct order, it fails. If the whitespace isn’t exactly as above, it fails. If the attributes aren’t quoted, it fails.

Radio is too big of a tool in the weblog community for this to be good enough. Dave and the rest of the guys should really make an effort at doing this RIGHT. Doing it this rigid is going to lessen the effect of the tag.

[Update: Exchanged emails with Dave. Although it seemed to me from what Dave wrote about the RSS Auto-discovery that the way Radio and Manilla handles the RSS auto-discovery tag wouldn’t allow for proper XHTML formatting of the link tag, it does indeed work even when adding the mandatory ” /” on the end of the link tag. (see the previous posting about RSS auto-discover to see what I mean)]

[Update 2: Nice to see that Dave has updated the description on his support for RSS Auto-Discovery for Radio page. Now he explicitly states that it will work to add the necessary characters to make the link tag validateable XHTML.]