Flexible View Notifications from Replicated Model Objects in Distributed Collaborative Applications

        Many distributed collaborative applications need support for consistent distributed shared state. This requirement can be well supported with a multi-user extension of the Model-View-Controller (MVC) application architecture supporting replicated model objects and the ability for view objects to take consistent snapshots of attached model objects. However, most collaborative applications also require highly responsive views of the distributed shared state, since end-users need immediate visual feedback for input gestures that change the shared state. It is difficult to support both the requirements of consistency and responsiveness with a single view notification semantics, particularly in high or variable latency environments such as the Internet.

        In this paper, we present MVC-based application framework that supports replicated model objects and four novel kinds of view objects to support a range of consistency and responsiveness requirements of collaborative applications. Applications can use pessimistic view objects to obtain strict consistency as well as a guarantee to losslessly receive all updates in order. R-pessimistic views weaken the guarantee of losslessness in order to be more responsive than pessimistic views. Optimistic view objects further weaken the guarantee of consistency to support highly responsive views of interactive shared state. Finally, adaptive views dynamically switch between pessimistic and optimistic behaviors based on the prevailing concurrency and network latency conditions. We present the semantics of these view objects as well as guidelines and examples for effectively using them. A single applications can mix and match these view objects to obtain the desired behavior. Experience with building real-world collaborative applications using our framework demonstrates the benefit of supporting these various types of view objects for deploying applications on high-latency networks such as the Internet.

By: Guruduth Banavar, Kevan Miller, Atul Prakash, Robert Strom, Michael Ward

Published in: RC21401 in 1999

This Research Report is not available electronically. Please request a copy from the contact listed below. IBM employees should contact ITIRC for a copy.

Questions about this service can be mailed to reports@us.ibm.com .