There are lots of perfectly usable, productive, and profitable applications written in cobol, still chugging along, making, in some cases, $millions for their o

There are lots of perfectly usable, productive, and profitable applications written in cobol, still chugging along, making, in some cases, $millions for their owners. Often these are government organizations or large enterprises who have no incentive to change, and lots of reasons not too. One of the biggest reasons to stick with legacy cobol applications is that, often times, these applications grew organically over decades to encompass business process for entire industries. There was never one single design document or even goal. And hence the functionality of the systems are poorly documented and only really understood by the cadre of users who use it from day to day. Ie they are incredibly expensive or even impossible to replace. Another point that's never made in discussions about cobol is how incredibly "safe" it is, from a security perspective. Cobol is a lot like (an incredibly verbose) assembler, all storage is statically allocated before the program even starts. In cobol there in no runtime stack, and no dynamic memory allocation. So no chance of buffer overflow attacks or trashing the stack.