Replicated architectures for shared window systems: a critique

Abstract

Replicating applications in a shared window environment can significantly improve the performance of the resulting system. Compared to a completely centralized approach, a replicated architecture offers superior response time and reduces network load. To date, however, these advantages have been overshadowed by the equally significant synchronization problems associated with replication. In this paper we document these problems and show that the most frequent synchronization problems can be solved without changing existing software. We further indicate how some of the limitations of the resulting system can lx removed by making applications or system servers collaboration-aware. Finally, we point out where general system support is needed to address the remaining deficiencies.

Issues and experiences designing and implementing two group drawing tools

“…Two architectural alternatives for constructing distributed groupware are the centralized and replicated approach (Ahuja, Ensor & Lucco 1990;Lauwers, Joseph, Lantz & Romanow 1990; WScrawl, the participant process is simply the X window server. The advantage of a centralized scheme is that synchronization is easy, as state information is consistent since it is all located in one place.…”

“…We believe it has been used as a reason not to implement multiple cursors, and for choosing some architectural styles over others (eg Ahuja, Ensor & Lucco 1990;Lauwers, Joseph, Lantz et al 1990;). …”

Sharing views and interactions with single-user applications

“…The software must be available at every workstation; there may be contention for physical devices; states local to a specific machine may be referenced; it may be impossible to guarantee consistent views across workstations using this architecture. A good critique of the issues in replicated architectures is provided by Lauwers, Lantz and Romanow (1990).…”

Multimedia Conferencing

“…Because the window management facilities are not available within the Viewshell window, graphically-oriented programs cannot be used. For this reason, Lauwers et al lobby for centralized computer conferencing architectures, but they conclude that modern window systems make this task very difficult [25,23].…”

“…Noteworthy conferencing implementations include work at SRI [1], the Rapport system at Bell Labs [2], several MCC efforts [10], the Olivetti-sponsored research of Lantz and Lauwers [25] as well as the Pandora project [19], and Sakata's work at NEC [32].…”