Intriguing stuff that Wright is doing. I’m not entirely sure I buy the data protection argument although I realize it’s a bit of a tricky sell – as they point out, it’s better to start from scratch with these kinds of specs than find work-arounds. However, it’s also nice to start with a gazillion object asset base and existing code so you can get a learning module up and running fast, with reasonable fidelity, deep options for avatar expression, etc.

Data security can be handled in many open virtual worlds by holding the data “off grid”. I don’t deny the point, but it’s not to say that this is needed because others can’t manage the same things.

I’m curious about the ‘viewable/not viewable’ function however, because it makes a lot of sense if you had courseware that people aren’t taking concurrently. Courseware objects should only become available as you reach that part of the lesson – I’m going to put some thought into that, because it’s a good point and a really useful feature.

And having said all that – Wonderland, being built from the ground up, can be “built to spec”. But then, so was Croquet no? I can never quite tell where that’s at. In any case, a useful post in particular because of how it points out what the needs are for educators.

(And yeah, the system we’re working on will handle some of this functionality plus integration with Web-based learning objects but that’s another story).