The aggregator I use at the moment (RssReader) is pretty strict. That’s as it should be: it’s definitely the Way of XML and to a certain extent the Web in general is taking up that approach with the move towards web standards. Recently I’ve hit several blogs whose feeds have been rejected by RssReader for various reasons: even, in one case, just flat invalid XML caused by a malformed CDATA section.
When RssReader sees an invalid feed, it gives you a link to a feed validator put together by Mark Pilgrim and Sam Ruby. There you can see exactly what went wrong, in context, and get some help in fixing it. Everyone should run this on their own feeds once in a while, in the same way that they should be running the W3C validators on their web pages and CSS.
What’s sauce for the goose… you can validate the feed for this blog.
(Of course, the first time I tried it myself I got an “Unexpected Content Type” error message and had to delay this post until I had fixed it. D’oh!)
[2019-05-02: removed the validation link]