The real-time collaboration in Croquet is cool. It provides a very different way of structuring applications that will allow things that nothing else can. The croquet team is working hard on this aspect. But we’re just begining to consider the implications of shared persistence. I think this is just as radical in itself, and will inspire truly extroadinary software when combined with Croquet’s other aspects.
The realtime collaboration mechanism in Croquet combines with the late-binding development/deployment platform and 3d UI to allow people of varying skills to directly create and modify their information environment. This makes it easy to create good work, but says nothing about publishing it. Publishing isn’t nearly easy enough on today’s Web. Even if you are comfortable creating content, most people don’t have free access (both libre and gratis) to a fixed IP address and a server. Shared persistence will truly allow folks to build on each others work. I’m talking about both simple textual content, but also data, dynamic applications, simulations, and live models. Make it easy for people to make all sorts of content that is directly used — not just quoted or linked to — by others. It’s all public, with varying definitions of “public.”
As with any disruptive technology, one challenge will be to build critical mass. Content is one key. I’ve long felt that an issue with developing next-generation information-management systems has been the difficulty in developing huge ontologies that are meaningful to large groups of people (and not just isolated specialists in a lab). But I’m now starting to think that public data will soon be necessary for current-generation technologies, too. For example, security models involving trust networks and delegation of authority are fine for static trees of organizational membership. But in the real world, we’re each part of a mesh of many constantly changing virtual organizations. I wonder if such networks won’t collapse under their own weight unless we find a way to distribute the maintenance over the membership. That means some sort of data creation and reuse that crosses organizational boundaries. This is public content reuse.
I think the only way to move forward is to “Publish ’em all and let Google sort ’em out.”