Legacy COBOL systems (hell, a lot of legacy code in general) is a perfect example of the "if it's not broken, don't fix it" mantra. There's an incredible amount

Legacy COBOL systems (hell, a lot of legacy code in general) is a perfect example of the "if it's not broken, don't fix it" mantra. There's an incredible amount of work involved in rewriting legacy code, and when you're talking about doing so in a constantly updating, critical context like financial transactions, it kind of highlights the need to take things slow and carefully. My experience with old legacy systems (20+ year old C in particular) is limited, but it was enough to recognize that just ripping it all out is an awfully ignorant approach compared to a gradual migration of what needs to be replaced. It's also an incredibly expensive one, especially since it's a recurring situation. No matter how clean and clear you think your code is, or how thoroughly you've commented everything, in a decade or two, someone is probably going to be groaning about it being legacy code.