Another Croquet-based project has been launched this week. Cobalt is the first such coding project to be done by Duke University, which is also the home of the Croquet Consortium.
Consortium director Julian Lombardi has been tireless for over five years in establishing institutional housing for Croquet. Running any public and open project is difficult. At the Collaborative for Croquet we only had to coordinate three people and we made our own decisions, and it still took months of work to get incorporated and put out a release on a hosted Web server. Doing this on an institutional scale is orders of magnitude harder. Having a non-commercial home is crucial. Julian single-handedly got the pre-consortium CroquetProject.org site set up at the University of Wisconsin, and housed the Jasmine pre-release there, even as most of the key developers were working through Hewlett-Packard. This was critical in keeping Croquet going when HP was sorting through its commercial proxy fight and executive changes.
Then Julian got the Consortium stably housed at Duke and the various OpenCroquet.org sites and mailing lists there, bringing five other educational institutions into the fold, losing only U.Wisconsin, and providing a place for Intel to join and HP to return to. Institutional housing has also been critical in arranging grant and award funding. The Mellon award has been one success for Duke, even as Julian vigorously pursues others for the Consortium and it’s partners.
Julian is a technical guy. And an artist. The challenges of getting technical work done with all that institutional stuff going on makes limiting the scope of a technical project a good strategy. Cobalt is all about how we interact in virtual worlds with things, people, and even worlds themselves. Julian and Mark McCahill had started exploring this with their C5 papers(*) even before there was a real Croquet release, and have been busy doing all the institutional stuff since then. Both have done this kind of technical work before with other technologies, and they certainly could have continued the work with any of the non-Croquet virtual world tools in use today. This User Interface work isn’t really directly about the Tea-time scalable distributed system infrastructure. And yet, the dynamic language development environment of Croquet is so productive, they did choose to do this side project using the SDK.
And of course, going with Croquet allows the work to potentially be applied to Croquet worlds and to take advantage of new Croquet developments, if desired. No doubt there will be advances made in the Cobalt project that will be incorporated into the tea-party series of Croquet releases (Jasmine, Dormouse, Hedgehog, Jabberwocky…).
I’m looking forward to seeing this happen. The 2004 People, Places and Things UI papers has been one of the big influences on the development of the UI for Qwaq Forums over the years, and is even now being studied by the Intel Miramar guys for the future Intel/Qwaq product.
Scanning the recent activity on the croquet mailing lists, there seems to be a huge demand on the part of non-developers for packaged Croquet applications and for infrastructure for connected worlds, 3D wikis and meeting spaces. But this isn’t the space Cobalt is acting in. The architects of this project have begun with a UI design exploration, another area in need of attention. And I think this scope is shrewd use of the architects resources. I think the whole global-community, visiting-worlds thing is yet another step to happen, either institutionally or otherwise. But the Cobalt project is an exciting alternative on the way there.
(*) “User Interfaces for Self and Others in Croquet Learning Spaces” and “User Interfaces for Places and Things in Croquet Learning Spaces” in C5 proceedings.