Consider a system that uses pure demand paging.
a. when a process first starts execution, how would you characterize the page-fault rate?
b. once the working set for a process is loaded into memory, how would you characterize the page-fault rate?