Aaron Swartz: The road to RSS 3.0. While I found this humorous, it demonstrates some unstated assumptions about the core of site syndication.
First, note the absence of 'backwards compatibility' in the RSS 1.0/2.0 comparison. The RSS 0.9x/1.0 split doesn't bear a lot of resemblance to, say, HTML 4.0/XHTML 1.x. Everything is smaller: the number of elements and attributes, the amount of divergence between 0.9x implementations, the number of clients and servers and their complexity, the number of unique documents and generators (people or scripts). See the comments in this post on Sam Ruby's weblog for some good discussion.
Mark Pilgrim writes there: “It appears that RSS 2.0 will be much like HTML 4.0 Transitional. Lots of cruft built up that no one is brave enough to deprecate, raising the barrier of entry for developing news readers, and causing confusion for newbies trying to learn by example.” At first glance this seems obviousthere's plenty of cruft and breakage caused by the 0.91 to 0.92 transition which will stick around in a hypothetical 2.0 transitionbut the interoperability problems have already been solved. RSS 2.0's core is small. Breaking stuff is bad; needless complexity (RDF) is bad. The audience of tool vendors is much smaller than the audience of people who simply use RSS as opposed to writing it.
Mark has the right idea in his next comment: “I guess we're going to need a widely publicized 'RSS 2.0 Best Practices' document to accompany the spec. Maybe even a weblint-type validator that not only validates but recommends usage patterns.”
Second, more than its unstated goal of stripping RSS to unusability, the sample RSS 3.0 feeds do not allow for including unstructured, hyperlinked text within a feed. Like it or not, I don't want to follow individual article links. The very benefit of writing a syndicated weblog for me is that per-post, they're unstructured, but in summary form (posts, calendars, feeds), they're structured and aggregatable, sortable, and easily navigable.
Nobody asks a newspaper reporter to limit him/herself to one quotation per article, or to make a reader change contexts to move from headline to article. I'd rather the entirety of the content I'm subscribed to (within reason, up to a few paragraphs per entry) were syndicated. We're perfectly capable of visually skipping unimportant information.
As I'm exercising each morning I sit or stand looking at my PowerBook, hitting the space bar or saying “next page” to it as I read through the day's accumulated news. If the entire feed consisted of snipped entries with ellipses at the end, I'd have to stop and click on links all the time.
RSS 2.0 appears to be heading in the right direction. Sjoerd Visscher has a succinct summary of the important backwards-compatibility features of an ideal RSS 2.0, which don't serve to limit its expressivity.