N.Z. Bear]2 says:
We’ve had a great deal of useful and productive discussion in the forum, but it seems that some practical experimentation would be of use to us as well at this point. We seem to have reached a rough consensus on what data we want to track — and are getting bogged down in the many, many possible approaches of how to track it.
At Dean’s suggestion, I’ve gone ahead and taken a rather quick-and-dirty approach to our encoding problem. I’ve developed a specification which shows how to encode our general schema’s metadata using only HTML [meta] tags. Along the way, I’ve also “Dublin Core-ized” our data schema, and tried to use DC tags wherever possible and appropriate.
What we’d like to do is get several (as many as possible) volunteers to apply this specification to their own weblogs, thereby beginning to actually ‘publish’ real metadata. At the same time, we call on everyone who is codingly-inclined to begin examining approaches for grabbing, parsing, slicing, dicing and presenting back this very same metadata.
In other words, enough talk. Let’s do some hacking.
Please note that this does not imply an end to our discussions on appropriate encoding formats for the final specification. Depending on what we learn through this experiment, we may decide tags are one of many encodings we support, or perhaps decide they are unworkable entirely. But I think we’ll learn a great deal by the attempt. (And to be frank, I personally believe we’ll be best served by supporting many encoding formats, and only strictly dictating the actual data to be tracked itself. But that’s a discussion for another day).
[ via David Gammel]