The goal of research in debugging is to help the user to understand
what's happening in the execution of the program, usually in order
to fix something that's flawed. The two major directions we've
looked at are providing information about optimized programs, and
interrupting programs at particular execution states. Based on your
experience as a programmer, and on the material we've covered
concerning the state-of-the-art, what do you see as feasible
improvements that might be made to debuggers for these purposes.
(You need not restrict your consideration to your cadaver and its
debugger).
What additional research needs to be done? Be sure to relate your
answer to concrete examples from the papers and from debuggers you
have used.