90k sounds pretty standard for inexperienced (although maybe not first job) devs in general for most markets. Throw in factors like experience or skills in low supply and that changes pretty fast.

I know that COBOL isn't going away anytime soon, but most companies have seen the writing on the wall for a long time. Anywhere that COBOL can be replaced with something more modern, it's already underway. Some places even have a surplus of COBOL devs because of it. But there are countless places where it can't be replaced, at least not reasonably.

The only way a COBOL dev is making $90k after 5 years is if there are very specific fringe benefits that make them not want to move along, or they are extremely naive about the market.

[–] dan@upvote.au 4 points 11 months ago

Anywhere that COBOL can be replaced with something more modern, it's already underw

Rewrites are extremely risky though, and some companies don't want to risk it. That COBOL code probably has 40 years worth of bug fixes and patches for every possible edge/corner case. A rewrite essentially restarts everything from scratch.

Do you know of a decent sized company that successfully migrated away from COBOL? I'd be interested in reading a whitepaper about how they did it, if such a thing exists.