Key Project Draft Specification
October 1996
by Paul Mclachlan
aka subtle
Introduction:
Key is intended to be used to implement a Multi-participant Virtual
Reality, in a similar vein to LambdaMOO, except, well... better.
System Goals:
- The system must support multiple participants interactively.
- The system must be practical. It must not require large amounts
of bandwidth nor be dependant upon the client having a particular set of
capabilities.
- The system must be capable of implementing a Virtual Reality.
- The objects in the system must be completely user-definable. Not only
their appearance, but also their behaviour. The system must, therefore,
be programmable.
- The system must be able to support graphics and sound for those clients
with the ability to display them.
- Ideally, the system could be 'distributed' to reduce bandwidth requirements.
- The system does not necessarily need to be small. It is more important
that it is flexible and powerful than efficient in memory or disk space
usage. CPU efficiency is important, however, system 'lag' must be
kept to a minimum.
System Layout and Interface:
"remove not the ancient landmark which thy
forefathers have set"
It is important to realise that the concept of a VR is not entirely
new. There are many such systems, active and successful today which
have established a certain precedent with their user-interfaces.
While most of them have been haphazardly designed, there are some fundamental
concepts that all (or almost all) of them retain. For instance, the
concept of the say and emote. The idea that the virtual reality is broken
into distinct 'rooms' - a way in which normal space is not.
Without good reason, I do not intend to depart from these established
standards.