mnot’s blog

Design depends largely on constraints.” — Charles Eames

Tuesday, 25 March 2003


RSS standardization (again)

Jorgen hits a subject that’s of great interest to me; RSS standardization. I originally started the Syndication list to get RSS moving towards some sort of
recognized standard; more recently, my effort to register an RSS media type was stalled by the lack of a stable spec published by a recognized group.

More recently, there was an informal meeting of the minds in San Francisco last (northern) autumn regarding this topic. There was good participation from the blogging, RSS weenie and standards weenie crowds, and we had pretty broad consensus that A Standard Would Be A Good Thing.

Here’s the rub. Normally, standards don’t ultimately matter; the market will decide, etc., etc. Sometimes, however, they can make a difference; when there are a number of solutions to the same old problem rubbing up against each other, the friction can hold all of them back. People don’t want to walk into a family feud when they adopt a format; the respective advocacy and unreasonable positions of all sides tends to make normal folks back away slowly and carefully. Standards can help this by “helping” the market to settle on one solution.

I think this is the case with RSS. I have been SORELY tempted over the last year and a half (or so) to grab one or another RSS candidate and push it to a standard.

Unfortunately, they all suck. RSS 1.0 is fixated on RDF, to the exclusion of all else. Don’t get me wrong, I LOVE the RDF model, but the syntax… well, let’s just say I’d rather change a diaper any day. It also doesn’t help much for software simplicity when a format is capable of expressing any semantics known to god or man.

On the RSS 2.0 side, things are a little simpler, but there’s lots of Dave-specific cruft like “cloud” in there, and more importantly, it really mangles the use of namespaces. Not good.

So, where does this leave us? I think there are two potentially good paths that we could walk down. Rather than show my hand now, I’d like to do a little work on it first, to see which pans out. All that I ask you, dear blog reader, is to wait, and bother me in about a week if I haven’t come up with anything.