Namespace impasse

I mentioned in an update to my previous posting that there are grave problems with declaring namespaces in RSS 2.0. The specification has a logical inconsistency at its heart. It allows you to add extra elements only if they are defined in a namespace. But its core elements can't be defined in a namespace, because if they were, it would break compatibility with previous versions.

As you can imagine, this has been quite an upsetting discovery. A lot of impassioned technologists have spent the past week or so banging their heads up against this brick wall and frankly it doesn't look as though they're going to be able to fix it. But it doesn't mean that RSS 2.0 can't work. It just means it's a bit of bodge. And, as Joshua Allen noted, none of the alternatives have squeaky clean track records, either.

So in the absence of a sufficiently compelling replacement, will 2.0 win the day? This is the only question that really matters to the majority of us users, but unfortunately I'm not picking up a strong sense of a consensus forming among the experts. This may be because many of the key players are traveling to conferences, which has kept them away from their weblogs (but perhaps in a position to make progress in face-to-face discussions). The upshot is that I'm going to hold off for a while longer until I see some signs of firmer direction. A good resource for keeping track of developments is the RSS2-Support discussion group. In the meantime, RSS 0.92 still works fine for most current purposes, so life as we know it goes on. It's only development of new applications that is hanging around in suspense.