Here's what I think Microsoft should fix in the Windows RSS Feeds API:
- Support feed and item ids. The item.id property does not reflect the RSS <guid> or Atom <id> of the item/entry in the original feed. Plus, it's an integer -- not the right type for a URN. If you want to get the actual feed or item/entry id you have to parse the XML. Make it easy to get the real id if there is one
- Item objects should have description and content fields. Currently, the object model can't completely repesent Atom, Wordpress style RSS or Typepad style RSS -- they have both description and content elements. So, add description and content and get the mappings (and content model) right for Atom, plus regular, funky and 1.0 style RSS.
- Allow parsing of feeds from file system. Apparently, you can only parse a feed named by URL and you can't parse feeds on your hard disk. That can't be right. Make it possible to subscribe to and parse feeds from the file-system.
- Cough up a spec for the Common Feed Format extensions. The normalized format used to store all feeds includes some new elements in the same namespace as the Simple List Extensions (SLX), but they're not mentioned in the SLX spec. Fix that.