This paper presents an algorithm for concurrency control in real-time groupware systems. Groupware systems support multiple users working on a shared task and require fast response times and fine-granularity data sharing. The algorithm allows users to directly operate on data without locks, ensuring responsiveness and simplicity. It is based on operation transformation, where conflicting operations are transformed to maintain consistency across all participants. The algorithm is fully distributed, resilient to site failures, and does not require locks, allowing data to remain accessible to all group members. The algorithm is applied to a group text editing system called GROVE, where participants can modify a shared outline. The algorithm ensures that operations are executed in a way that maintains consistency across all sites, even when operations overlap or conflict. The algorithm uses a transformation matrix to resolve conflicts by transforming operations so that their combined effect is the same regardless of execution order. The algorithm is proven correct by showing that it maintains the precedence and convergence properties required for correct groupware systems. The algorithm is suitable for environments with replicated data and distributed control, and it does not rely on total ordering of operations, which can be inefficient. The algorithm is also robust to user actions and can adapt to changes in the system. The paper discusses the challenges of implementing groupware systems, including the need for real-time notification, wide-area distribution, replication, and robustness to user actions. The algorithm is shown to be effective in resolving conflicts in overlapping operations, such as insert and delete operations on a shared text outline. The algorithm is also applicable to other groupware systems, including hypertext systems, where users frequently apply associative access techniques rather than accessing objects by name. The paper concludes that the new emphasis on groupware suggests interesting and challenging frontiers for future research.This paper presents an algorithm for concurrency control in real-time groupware systems. Groupware systems support multiple users working on a shared task and require fast response times and fine-granularity data sharing. The algorithm allows users to directly operate on data without locks, ensuring responsiveness and simplicity. It is based on operation transformation, where conflicting operations are transformed to maintain consistency across all participants. The algorithm is fully distributed, resilient to site failures, and does not require locks, allowing data to remain accessible to all group members. The algorithm is applied to a group text editing system called GROVE, where participants can modify a shared outline. The algorithm ensures that operations are executed in a way that maintains consistency across all sites, even when operations overlap or conflict. The algorithm uses a transformation matrix to resolve conflicts by transforming operations so that their combined effect is the same regardless of execution order. The algorithm is proven correct by showing that it maintains the precedence and convergence properties required for correct groupware systems. The algorithm is suitable for environments with replicated data and distributed control, and it does not rely on total ordering of operations, which can be inefficient. The algorithm is also robust to user actions and can adapt to changes in the system. The paper discusses the challenges of implementing groupware systems, including the need for real-time notification, wide-area distribution, replication, and robustness to user actions. The algorithm is shown to be effective in resolving conflicts in overlapping operations, such as insert and delete operations on a shared text outline. The algorithm is also applicable to other groupware systems, including hypertext systems, where users frequently apply associative access techniques rather than accessing objects by name. The paper concludes that the new emphasis on groupware suggests interesting and challenging frontiers for future research.