Doomcoding: Defensive Coding and Design Against Indians and Outsourcers - Design, tooling and choices we make can make us vulnerable

[Demonslayer1776]

In the coming decades.

For someone starting out now, if they're any good at it there's still a good 15-20 years or so of juice left to squeeze, especially as pajeets won't go anywhere near something as arcane as COBOL and the old guys still coding probably wanted to retire 10 years ago but couldn't. That's plenty of time to build up skills in other languages / technologies, and make industry contacts so you can move onto something else once the final COBOL machine is switched off some time in the 2060s.

Maybe get a job today in one of the single worst dev positions possible with no outside relevance, and hamper any ability to modernize if you lack strong self motivation - self motivation that could be used on a low level language that actually has future prospects.

Anyone going out of their way to learn a specialized skill to try and capitalize on niche jobs has better options than COBOL. Imagine if that time went towards actually getting good at a language like C. Newbies may have written some basic C for a college course, but its quite rare for them to have strong fundamentals. It's in demand, used in many different situations, and even if you don't use it directly, it builds fundamentals and earns respect from people that are hiring.

The few COBOL positions that exist aren't good experience and won't build industry connections. The complexity is not born from fundamental problem solving, its born from obscur limitations and decades of technical debt. It's all about building bespoke knowledge of that specific implementation, which then locks you to that position without an offramp.