I had dinner last night with Jim Moore, one of four people who are part of the new venture capital fund that's committed to RSS. They don't have a blog yet, we talked about that, among many other things.

I've known about their plans for quite some time, to me it's like Jean-Louis Gassee's announcement of Be, back in 1995, a project I also knew about before it's announcement, and one which surprised me because it was so enthusiastically received by the industry. I wrote this DaveNet piece, welcoming him back.

The same thing happened with the RSS fund, I knew about it, didn't think it would stimulate much interest, and was surprised when it did. I wanted to know, from Jim, why.

He explained. RSS is more than a format, it's an approach to creating systems. He likened our web content management software to corporate publishing platforms, like Atex, which were cumbersome, hard to use, and even harder to manage. Connecting them to other manufacturer's systems was so hard that people practically never did it. The whole point of RSS, Jim argues (imho correctly) is to make connecting systems together so easy that users can do it themselves, without any help from system managers or vendors. This is a brilliant observation, in all my years thinking about RSS, I had never approached it from this direction.

Then we talked about a question that Jason Calacanis asked at Tuesday's OPML meetup in NY, one that I answered in a less than opportune way. Jason wanted to know if it was possible to use the OPML Editor to browse structures produced by databases. I explained that the environment that the OPML Editor runs in is ancient and rich, and among many other connectors it has an ODBC interface, so yes, it could easily be done, and there are developers who would happliy do it (for a contract fee of course).

This is true, but it missed an opportunity to exhibit the way of thinking of RSS and OPML that Jim had talked about. The correct answer is, yes, of course, just put an OPML cap on the database service, something that can be done in an afternoon of scripting (including a couple of coffee breaks, and one or two false starts) and then you can plug your database into any tool that understands OPML. Or if you prefer, create an RSS cap, and plug it in just as easily, but without the hierarchy (assuming there's only one level to your data).

But forgive me for missing this opportunity, because Odeo and Apple missed an even bigger, related, one (but of course there's still time for them to correct that). They're not expecting to see it, so they don't. I know that Adam Curry has urged Apple to see it, as far as I know, no one has pointed the opportunity out to Odeo. Last night I got Jim to look at it, and it wasn't easy, because (my opinion) he was't expecting it, so maybe you weren't either.

The podcasting community did more than bootstrap a new medium in record time. As if that wasn't enough, it also bootstrapped a directory of podcasting resources and podcasts themselves, and a way of doing directories in communities that goes way beyond podcasting in importance. It's a structure of interlinked OPML directories, maintained by the community, linked together by the architecture of OPML. It's the small pieces loosely joined philosophy applied to directories, and it works. There is no official top level of the directory, but many people think of Adam Curry's ipodder.org directory as the top level. For many months Adam toiled over this directory, using an early version of the OPML Editor that I created for him. Many of the people who create directories that Adam links to actually code the OPML by hand, that's how dedicated and visionary these people are.

It's a wonderful thing! People in the community want the newcomers to podcasitng to use their work, and help them do more. But so far they've taken the Atex approach, one-corporation doing everything-for-everyone. It can't possibly compete with the structure the community has already built. I'm committed to making it work, and I think we have the strength to do it, because so many people are pulling in the same direction. Last night I got Jim to think about his OPML search engine (in development) as a way of searching the podcast directories. And of course any other public directory structure that uses OPML. This way his work, will immediately be relevant to a small community of enthusiastic people, which is exactly how these things get started.

Okay, this is clearly something I need to podcast. And we'll certainly talk about it at tonight's meetup at Berkman Center in Cambridge, 7PM.

# Posted by Dave Winer on 7/14/05; 5:00:00 AM - --