OPML 2.0 Draft Spec revealed
Dave Winer has revealed the draft of the OPML 2.0 specification. Woo!
I’ve had a quick read and most of it makes sense. But I have a comment to make on something that bugs me, as a developer of OPML-powered products. I’ll eat my dinner, then drink a coffee and post it on the Yahoo! Group which has been set up to manage the feedback to all this.
First impression though, is that I’m glad to see the ‘include’ type attribute.
I have a comment about the usage of the ‘link’ attribute, which may or may not be linking to an OPML file for inclusion or it may be a link to a website. Checking for .opml is a bit of a pain, when many people use php or whatever to generate their OPML. But I suppose it should be backward compatible with 1.0 for all those opml files out there which use ‘link’ to point to both web links and opml inclusion links. This new spec should get some good traction. I’ll support it. But I personally think ‘include’ should link to opml formatted files only and link should always be a link to a site. Maybe through extending OPML we can provide type flags for links to mp3 files, or video files or software updates, then build clients around this to consume that ‘opml enclosure’ of sorts.
More thoughts later…. meanwhile, have some icons 
i thought the generation of code was done in OPML and not in PHP. I am the first to admit that I am a newbie with OPML and am thrilled that I got it working with the wordpress root to give you an idea of where I am at.
Since I live about 1/2 an hour from Dave, I need to go some meetings that Dave sets up in Berkeley. I would be interested in more posts from you
I meant that people use scripts to generate xml formats. EG: php is mostly used to output HTML. It has .php at the end of the url, but it’s still ‘HTML’
That’s why some opml reader developers have to ‘klooge’ a script generated opml file by appending ‘?.opml’ on the end (or ‘&.opml’, if you already have a query added to the url)
more thoughts soon 😉
I’m very excited about this release and that we catch issues right now.
treedia.com has become podcast.com?
treedia is the codename and testbed for the system that will power podcast.com – there I said it 😉
things will be going up a gear from monday
blogs, podcasts, feedback and conversations aplenty!
😀
oh and trees! lots of trees!
opml = trees
rss = vines
items/enclosures = fruit
tags+ratings = roots
a semantic tree of multimedia
I noticed the trees at Treedia 🙂