The fun part about doing an RSS feed is that its such an (unfortunately) politicized process. There are a lot of people involved with the process, and instead of coming to agreements, they just went ahead and implemented stuff, and then left cleaning up the mess until later.
There are two sides to a market like this: creation/serving and readers/clients. From my perspective, I wanted to make sure that our feed was "valid [feedvalidator.org]", and would be read by most readers/aggregators, but I also wanted it to be as descriptive as possible, because I didn't want to miss data that readers I didn't know about could use.
In a drastically oversimplified sense, the basic argument is based on RSS never being well defined in its early days. Dave Winer [scripting.com] and Netscape defined a early version of the spec, at first the 0.9x line, and then transformed by Netscape into the incompatible 1.0 line, when they decided to go all out XML crazy. That got more complicated when Netscape dropped RSS as something they cared about as they stopped doing portals. The main difference was that the 1.0/XML version allowed namespaces for extensions, and the 0.9x line didn't. 2.0 theoretically resolved that by allowing namespaces.
Current politics, and what makes 2.0 slightly difficult to figure out, is about whether and when to use those namespaces. This goes back to a never stated disagreement about what the purpose of RSS is. Originally, as envisioned by Dave, it's supposed to simplify notifications about updates to web sites. For that purpose its a very lightweight spec that is mostly about URLs, headlines/title, and maybe a short description. As aggregators started to appear, however, other people started using them as their principal method of following a site. I enjoy Atrios [blogspot.com] and Calpundit [calpundit.com] more because I don't have to jump back and forth between NetNewsWire and countless open windows in Safari. I'm less likely to read the full item on sites where I need to do that jumping. RSS doesn't have a native method to include a full post, it depends on the "content" namespace extension to do so. The current real, and stupid, debate is over whether it is "more" correct to use a namespace extension for tags that exist in RSS 2.0, but are optional. The debate is stupid, because the only people who care are those writing readers and aggregators. But they already have to support both if they're going to support version 1 and 2 RSS, so the extra complication is already built in. Deal.
And finally....Jon and I have disliked, and continue to dislike, third party aggregators that just grab the strip and put it into somebody else's environment. Since we support the site primarily off of merchandise these days, it's important to us that readers find out about what we're doing in that area. To that end, I'm happy to have our own feed that includes comics and news items, so that those who can't stand the site's design, but really want to buy Sporkle [goats.com] shirts can find out that we have them available.
Work is the curse of the drinking classes.
-Oscar Wilde
