Locales: supporting large multiuser virtual environments

Creators of multiuser virtual environments naturally desire to make them large in spatial extent, large in numbers of objects, and large in numbers of users interacting with the environment. However, doing so creates several problems: efficiently managing the flow of large amounts of data between large numbers of users, representing precise position and velocity information about objects arrayed across a large volume of space, and allowing designers to create parts of a virtual environment separately and combine them together later. The concept of locales is based on the idea that even in a very large virtual world, most of what a single user can observe at a given moment is nevertheless local in nature. You would expect a large virtual world to be large primarily because, like a city, it combines a large number of relatively small, localized activities not because it contains very large individual activities. Locales divide a virtual world into chunks that can be processed separately. This division is purely an implementation issue-it is not apparent to the user. A user sees several locales at once-generally the locale containing the user´s point of view and those neighboring it. The user does not see any seams between the locales nor any abrupt changes as the point of view moves from one locale to another, thereby changing the neighborhood set