The use of lines of code metrics for productivity and quality studies [is] to be regarded as professional malpractice starting in 1995.