Distributed object technology has now reached the point where it is possible and highly desirable to take advantage of the coherent framework which it offers. We are currently studying how best to make the transition to a distributed object model for the next version of the ADL testbed system. We are approaching the problem from three simultaneous directions, which include:
We are following industry standards as far as possible. There are currently two major industry contenders for distributed object technology: CORBA from the Object Management Group (OMG), a consortium of over 500 companies, and DCOM/Ole from Microsoft and its partners. We have opted initially for CORBA, but recognize that both standards are likely to remain in parallel for the foreseeable future. Conversion tools between the two are already an integral part of commercial CORBA offerings, so we do not anticipate great difficulty in supporting both standards. We are also fortunate in having been able to obtain DEC's Objectbroker CORBA implementation as part of DEC's free software distribution to universities.
In addition, we are looking at how ADL's CORBA environment might connect to Stanford's Infobus, and how ADL's user interface relates to Stanford's Digital Library Integrated Task Environment (DLITE), which is firmly based on the distributed object paradigm.