A few weeks ago I wanted to release an update to my own blogging engine and, because there was a minor data change I thought that it would be nice to export the data before running the update, change the files and then re-import it afterwards. As I looked into this more I decided that I really should do it in a standards-compliant manner (if possible) and so then set out to: A) talk to people about it, B) rtfm some of the existing formats and C) write down my requirements.
Very seldom does someone come along who has simple ideas that can work coupled with what it takes to make those ideas reality. Darren is one of those people.
BlogML may not be the best thing since sliced bread, but it's an idea that is brilliant in its simplicity. Why should a task - importing and exporting blog content - be shoehorned into a spec that was never meant to do this in the first place? Isn't that the reason that atom was designed itself? To rehash the failures and stick with the successes of past specifications dealing with blogging? So is BlogML IMHO to import/export of blog content.
Keep up the good work, Darren. Although we live on opposite sides of the globe, you are always welcome to ping me with these crazy ideas of yours, if for nothing else than to talk them out.