The test framework should return an error instead of panicking if the VM execution of a program fails. The reason for this is that it is easier to test returned errors than a panicked execution. So when we e.g. make a test to verify that a u64 multiplication with overflow crashes the VM, we want to be sure why execution was stopped. For this, errors are better than panics.
The test framework should return an error instead of panicking if the VM execution of a program fails. The reason for this is that it is easier to test returned errors than a panicked execution. So when we e.g. make a test to verify that a
u64
multiplication with overflow crashes the VM, we want to be sure why execution was stopped. For this, errors are better than panics.