I had a quick IM chat with David Gratton last week, when he was asking me what I thought of content package specifications. My initial from-the-hip reaction was along the lines of "gah! metadata for metadata's sake" and that just getting content Out There was the goal, not encapsulating it in layer after layer of helpful metadata.
Then we spent a couple of minutes hashing it over. If there's a requirement that a set of content needs to be ingestable in a system, a package begins to make sense. A system then only needs to know how to ingest stuff that meets a given specification, and all kinds of workflow opportunities open up. I'm skeptical about the benefit to the end user (students, teachers, etc...) but the value to the Institution (or higher) is undeniable.
Then, David writes a blog post this morning, where it all becomes clear. Content Packages are really a way for content producers to bundle up various bits that make up the experience of interacting with their content. The indivdual bits of content, the metadata that describes each one, the metadata that describes various paths through it, interfaces to present the content to the user, potentially code that interacts via an API to communicate with other systems and users, etc...
David is approaching from the angle of the music industry, specifically through the awesome Project Opus. Content Packages as replacement for the dying CD industry (bits are cheaper than atoms). The XIPF project (Extensible Interactive Packaging Format) will be building on MPEG 21 to define ways to share content experiences (albums, etc...) and they're planning on working with the education community so it's not just about building the next 8 Track specification.
If this works out, when you buy a digital album, instead of simply getting a set of tracks and maybe embedded cover art, and maybe a PDF of the liner notes, you'd get an XIPF package containing the full experience (tracks, cover art, liner notes, lyrics, embedded interfaces to community features, etc...) all in one shot. It'd be cool to see Apple get on board so when I buy albums on iTMS it comes in a standard format, as they will from Opus, et al.
It's interesting that the XIPF wiki doesn't mention either IMS CP nor SCORM as existing models, but a fresh start with an extensible model from the ground up will be nice anyway. Hopefully there will be some form of interoperability between the camps.
So, if I look at content packaging as more of an experience than simply as a "content cartridge" then it makes more sense.