I just removed my RSS .91 feed, which should discommode almost nobody — I got about 15 hits on that feed over the last week, most of which were from Web crawlers. Conversely, I got seven hundred or so hits on my RSS 2.0 feed. Thus, I’m not too worried about discommoding people, and I’ll point index.xml at index.rdf just in case. Administrivia done; read on if you care about why I’m making the change. (Hey, he rants about things other than politics.)
The last few days have seen some occasionally heated, occasionally peaceful discussion about Movable Type’s use of RSS 2.0. It was triggered when Dave Winer made the off-hand claim that Movable Type does “funky” RSS. The concise version of the question is simply whether or not Movable Type should wedge some RDF into their RSS 2.0 feed. There is no question but that the spec supports what they’re doing. Dave wrote the spec. Nonetheless, he says that their application of the spec is “totally wrong.”
I would say that it could be improved. For example, the RSS 2.0 spec has <pubDate> as an optional element, and Movable Type uses <dc:date> instead. As a scripter, I prefer working with the latter, since <pubDate> uses RFC 822 date format and is harder to parse than the ISO 8601 date spec used by <dc:date>. Leaving out <pubDate> and putting in <dc:date> is valid RSS 2.0. On the other hand, it wouldn’t hurt to put <pubDate> in as well and it might make some aggregators happier.
However, “it could be improved” is a far cry from “it’s wrong.” And bitching about something in public, then refusing to explain complaints is a terrible way to get things done. From all accounts, Ben and Mena Trott (the people behind Movable Type) are nice friendly folks. If I’d been in Dave’s shoes, I’d have emailed them and suggested adding a couple of fields to the default templates. I bet they’d have done it — why not?
Instead, we get paranoid fantasies in which the Trotts stuck in the RDF in order to gain a competitive advantage. That’s nuts. The only way that’s a competitive advantage is if there are aggregators which will do extra things with the extra information… and if that’s true, then I can’t see how the Trotts can be faulted for taking advantage of that. Nothing’s stopping Dave from doing the same thing.
Long story short, I don’t care to support Dave’s formats. I don’t like the way he writes a spec which permits namespaces, then implies that any use of namespaces is bad. So… no more WinerRSS for me; I’m content with RSS 1.0. And, from the evidence of my Web traffic logs, so are the vast majority of my readers.