... makes for a quiet blog.
I'm currently helping a client migrate several terabytes worth of data into their SharePoint server. It's a fun problem as there's a Siebel CRM application that most of the business folks use, which of course needs to integrate with SharePoint. Naturally you'd think it'd be easy (heh heh). The issue was every document was being stored in Siebel as an attachment to entities and activities, which makes it really hard to do document management things such as versioning. They developed a system that sort of worked, but when a big government agency asks you about your document management strategy... well, then attachments don't look so good.
The document management and versioning answer turned out to be SharePoint, but the trick was how to match Siebel's relational view of the world with SharePoint's hiearchical (as in web addressing) view. It turned out the business organization had a natural encapsulation within Siebel that lended itself very well to a URLs hiearchical requirements. So we developed an addressing scheme the Siebel developers could generate from Siebel data, transform it into a URL using predefined rules and thereby point users to the right spot in SharePoint.
And did I mention a few hundred pages to document this all? Just adding to the terabytes!