Having done some research work in CS I'm not inclined to cast too many stones at the poor climate scientists just for having outrageously bad code. I was on a project with $N million in funding that had no source control, a concept I was totally ignorant of the time. We had one of our profs attempt to edit the sole copy of the code, running in production, during a customer demo. This is a "Physician, heal thyself" for the academic branch of our discipline.