Syntax, Logic, and Run-Time Errors

A program is a set of instructions that takes input and is supposed to produce a correct output. Three kinds of error can break that chain, and they show up at three different moments. A syntax error is a grammar mistake: a missing colon, an unmatched parenthesis, a misspelled keyword. The language's rules are violated, so the translator checks the code before running anything and refuses to start. No statement executes, not even the first one.

A run-time error waits until the program is already running, then hits a step it cannot perform: dividing by zero, asking for the fifth item in a list of three, opening a file that doesn't exist. The program crashes partway through, usually with an error message pointing at the exact line. A logic error is the sneaky one. Nothing is grammatically wrong and nothing crashes, so the program runs start to finish and prints an answer confidently. The answer is just wrong, because the instructions themselves encode the wrong idea. Example: averaging 8 numbers by dividing the sum by 10 instead of 8. It runs fine. It lies fine too.

This is why a clean run with no error messages proves nothing about correctness. Error messages only catch syntax and run-time problems, the ones the language itself can detect. Logic errors are caught only by checking output against known correct answers, which is the whole reason testing with sample data matters once a program runs without crashing.