We present a set of library routines that allow easily parallelized gr
aphics rendering routines that require no communication between each p
arallel task, such as ray-tracing, to be run efficiently in an environ
ment of distributed workstations. The presentation of the paper focuse
s on the problems encountered in implementing a distributed system und
er Unix and proposes solutions to each problem. Specifically, we discu
ss the challenges involved in overcoming the limits of communicating w
ith a large number of processes in Unix and in providing fault toleran
ce when using sockets. Technical aspects of the implementation and som
e additional problems that were encountered are discussed. Finally, we
compare the rendering times for a complex image with a renderer using
the library and show that the library routines are able to exploit mu
ch of the existing parallelism. The library is presented using a graph
ics application, though the concepts are generic enough to be of use i
n designing any distributed system under Unix.