>"The most impressive identity solution I saw ...... The system was able to move organisational structures with drag and drop. ....."
This is the key, User Interaction Design. The whole long-term-success lies there.
A campus enterprise architecture is the right approach for OpenACS/.LRN, since we are actually the only real web development framework out there that has a vertical application for the educative market, which is .LRN. I've been pretty much thinking and researching about this for a while, and yet, nowadays we are quite in real advantage to other solutions, but first of all, no body knows that, and worst than that is that no body does anything significant to make the things happen. Probably the worst mistake that .LRN and any other open source project might do is to invest in development, but leave no resources for project management and marketing.
If we ever deliver a product that is a pleasure to use by the administrators, deans, students, professors, teachers (and not to the *users* without a deterministic profile), then we have excellent chances to grow the adoption.
The product is already there, the organization is already there, the excellent developer community is here, what we need is to expose that, the obvious step is to do real marketing pretty fast. Improvements to the actual core can go in there while we market.
I agree with the last-chance comment, but probably marketing, project-management and user interaction design are the real priorities, while ensuring to create a real process for supporting innovation in a consistent manner.
