No later than yesterday morning I was discussing on the phone with Simone about how we could create xml-based backups for ideaTools sites. The idea was creating directories on the server for each data structure and store xml files there, then create one of more indexes that would allow the whole site to be rebuilt or imported into a different CMS.
A few hours later I was reading what UserLand is doing with their "blog browsers format": just perfect!
Our challange is that we are not managing only weblogs with IdeaTools, but also sites containing discussion groups, events, products in e-commerce catalogs, stories, shopping baskets, users' profiles and other kind of data.
The good news is that thanks to RSS 2.0, we can quite easily create our own namespace and manage all kind of data we store in our sites.
To create site indexes (meaning files describing the whole structure of a site) we are already using opml outlines, which are just perfect to describe where each story is published in a complex web site (for example, this is the opml file of this site).
From here, I wonder how much will it take to create specialized browsers for sites that are different from weblogs....