This article describes the failings of an autograding system in use in a computing science class in Berkeley. The use of autograders in computer sciences is a natural development, as programs can be tested by debuggers and efficiency algorithms to determine not only whether they run at all, but also how well they run. This article records what happens when they fail, but it's more useful reading when looked at from the context of what should happen when they run successfully (which is what they usually do).
Today: 0 Total: 19 [Direct link] [Share]
