The author presents a personal perspective on where research efforts are most urgently needed to advance the fundamental foundation for real-time distributed computing. At the top level the biggest fundamental research challenge is to enable high-level high-precision real-time distributed software design and programming accompanied by cost-effective guaranteeing of acceptable response times. This top-level goal cannot be realized without successful development of some key component technologies such as programming model and API and middleware possessing effective resource management capabilities. The state of the art and the promising research directions established by the community in those component technology areas are reviewed. The main type of network environments in which real-time computing applications have been realized are the LAN environments. Newly emerging network environments such as sensor networks at one end and optical grids at the other end add additional dimensions to the challenges posed to the research community. Those additional dimensions are also briefly reviewed.