Suppose that you intentionally put 90 "fake" errors into a large software system just before the system was subjected to a large system testing effort. this testing effort uncovers 80 errors in the system, of which 30 were "fake" and 50 were not. clearly, 60 of the fake errors were not uncovered. how many non-fake errors in addition to the 50 already found would you expect to be in the system? realize that i'm only asking for an estimate here based on straightforward (and probably over-simplified) assumptions about the errors and their discovery.