A large-scale platform for shared virtual worlds circa 1991
Retrospective: 15 October 2001
In 1990, few knew what the term virtual reality meant let alone what the letters VR represented.
Even in the twenty-first century, most people mistakenly believe this to be just games.
The aim was to progress technology, ultimately making computers easier to use.
These designs included all the trimmings that by the end of the 1990’s became obvious to any large scale project: distributed system, shared memory, platform independent programming languages for customization, extensible protocols, network transparency for applications, fault-tolerance, etc.
Some unique characteristics included things still uncommon in the early twenty-first century: user-centric view of data, arbitrary information engine (handles data beyond just 3D or even n-Dimensional) and a programmable protocol for on-the-fly optimizations, among many other things.
The goal was to manipulate data in a multi-user environment distributed over a potentially wide area and permit the users to traverse, share, negotiate, and intuitively understand the information being dealt with. Such an environment would have allowed immersion such that the user would have actually participated in those manipulations rather than just wait for a result after status information has been generated.
The work being done during the early 1990’s was to handle all of the issues related to this problem. Our focus in the collection of papers below was the design, specification, and implementation of a very high level operating system. The OS would serve as a platform for the tools and applications that could make virtual environments a reality.
Our solution may been viewed as using artificial intelligence tools, operating system features, distributed processing techniques and elements of the much hyped category of virtual reality.
These are not papers about virtual reality but a detailed system design. The design allowed users to manage information and provide researchers with a tool to understand and develop true virtual reality systems.
The MOSES Project goal was to implement the foundations of such a system whereby the infrastructure would last. It was intended to withstand multiple generations of research and development. The elements providing an intuitive understanding of information would come from outside the foundation system as an application to this system. Our project focused on the techniques used for developing that foundation and how applications might have use it. The tools and techniques for user manipulation of information were to be topics of additional papers.
MOSES stands for the Meta Operating System and Entity Shell and was to be the high level system platform. The intermediate and end results of the MOSES Project were to become free software. Ideally, the hope was to influence future system designs and establish a working communications protocol with an upgrade path designed as part of the protocol specifications.
* * *
Below are the original MOSES documents drafted primarily between January and October of 1991. Many topics emphasized are now taken for granted.
From a contemporary perspective, MOSES Digested: The Meta-Operating System And Entity Shell, Ten Years Later (October 2001) should be all you need to read.
These are the original documents converted to PDF (from LaTeX), including some not identified in the HITLab document archives: