Object communication usually uses message passing to transfer the request and the reply. However, in business applications many of the object requests are in fact events which trigger other business objects, and they usually form sequences. Therefore, in business applications, many types of object communication will be easier to be presented by the event model than message expression. Following Steven Cook’s argument [1], message sending is an over-specification for the purpose of specifying stimulus-response behaviors. Therefore, communication with events will be a better choice when the broadcast happens very frequently.[2] On the other hand, in dynamic invocation, the major source of the target object interface requesting is coming from an interface repository that contains the object interface syntax information only. [3] Retrieving the information needs to go through a naming or trading service to find the object reference. Actually, such information is not enough for the dynamic call when the object does not have knowledge about the target objects since the major location for the logical search is the trader service. In CORBA’s Trader, there is no standard location or format for these data. [3] Therefore, more invocation information will be needed when the object is using the dynamic call.