OPML seems to be at an interesting point at the moment. Amyloo wrote today that she felt almost nostalgic for the heady excesses of last year, and that OPML seems to have just reverted back to being a 'feed thing'.
I have to admit to being a contributor to this, since the energy I was spending on OPML has been transferred to the Semantic Web.
I still spend a lot of time at an outliner, mostly the OPML Editor outliner. A while back, I came up with an OPML-to-RDF standard that can serialize any amount of OPML as RDF. This means that we can use some cool Semantic Web tools to query OPML data.
To those of you somewhat mystified by the Semantic Web but who grok outliners, let me say this. Currently we have includes, links and feeds as built-in types in OPML documents. Imagine if you could have a thousand or a million or a billion different 'types' of thing all link together. Imagine if you could put in to a glorified outliner people, places, things, ideas, pictures, symbolism, trust, relationships and every other facet of human existence, and that outliner could do clever things with them. That's the dream. We are making baby steps towards that dream, and even if it doesn't happen, we are having fun in the process.
The syndication sphere deals with the core essence of something - the post, it's title, the feed it comes from. All of these things are the bare necessities of what it takes for an aggregator to work. An aggregator deals in blog posts, and sometimes enclosures.
The Semantic Web is built to handle everything else.
Outliners are intereting, but not as an end in themselves. Outliners are just an efficient way of getting data out of your head. Once that data is out of your head, it becomes more interesting what you do with it. Here, we leap away from OPML and outliners and in to other realms - although OPML is still a useful format for output. I still wish Google Reader supported Reading Lists properly, for instance.
