Here is what I had in mind. SloanSpace (our instantiation of dotLRN) now serves as our common platform for courses (on-campus and distance) and online communities (students, research, alumni, industry partners). There is a lot of flexibility in the platform. It can be used by colleges and universities for basic course management functions but it can also be used as portal system for online collaboration.
dotLRN should have appeal in the developing nations because commercial products are simply too expensive. But free software is not enough. We have to take into account other barriers to adoption. Most institutions in the developing nations cannot afford the hardware, infrastructure, and support costs associated with implementing a platform like dotLRN.
What I envision is that the grant would fund all the costs associated with installing and running dotLRN for one or two years. This would include hardware, some base infrastructure (e.g. internet line charges), compensation to some staff to learn the system and support faculty, compensation for faculty to learn how to use dotLRN, compensation to the project director, etc.
I am thinking of approaching a couple of Foundations initially and also some vendors (e.g. Sun, IBM, Dell) for things like hardware.