COBEA shares the major goals with the existing architectural frameworks for event handling in large distributed systems [2, 14, 18, 23, 24, 25]. The CORBA Event Service and the Cambridge Event Paradigm were reviewed above. Despite its weaknesses, the CORBA Event Service is nonetheless influential. Some recent work and products [7, 18, 10] have extended it; Expersoft, Iona, Sun Systems and Visigenic Software have developed commercial CORBA-compliant event services. The OMG TELECOM SIG has issued a Request For Proposal (RFP) on a notification service which has received several responses [18]. The proposed Notification Service must address issues such as filtering, assured notification delivery, security, QoS, and notification server federation. The Notification Service, however, is based on the indirect event communication model, and does not address implementation issues.
Work in real-time event notification [7, 19] has produced useful designs and implementations which use real-time threads for event publication in order to prevent priority inversion. Performance has been a major emphasis. [7] in particular, also address issues such as event filtering and correlation. However, its filtering and correlation mechanisms are not as powerful as the Cambridge Event Paradigm [2]. Moreover, filters can only be placed at the event channel.
Work in active databases is of direct relevance to the research on active systems [4, 5, 22]. In an active database, events include time events, start- and end-of-transaction events, operation invocation events, abstract events (events signalled from outside the database system). These are monitored and conditions are checked before actions are triggered upon event occurrences. The concept of composite events is introduced in active database for events correlation. Most techniques used for implementation are designed for a centralised database, thus do not address directly issues required for a distributed implementation.