| Just a counterpoint to the comments who talk positively about the COBOL idea: When it comes to COBOL, the main problem isn't actually the programming language itself, it's everything else around it. Most companies who need COBOL have a horrible tech culture & situation and are stuck with a pile of shitty, undocumented COBOL dating back decades. If you're going to be doing COBOL, your problem won't be COBOL, it will be the archeology that you will have to do to reverse-engineer the existing system, the pressure you'll be under to deliver something fast under those constraints and the blame you'll take when it inevitably explodes (forget about CI/testing or a staging environment). COBOL is not just a programming language but an entire ecosystem completely different from mainstream computing. Forget SQL DBs, forget UNIX terminals - you'll need to learn how to use & develop for mainframes - it's a closed ecosystem guarded by IBM which means there aren't many free resources available to learn from or get help, and it is so different that the skills you'll learn won't be transferrable to modern, non-mainframe development. Furthermore, all of the above doesn't actually pay that well. I'm sure you've heard all the media attention about COBOL and how systems are failing all over the place because there are no developers to maintain them - well that's a lie, the problem isn't that there are no developers to maintain them, it's that there are no developers to maintain them at the price the company is willing to pay. Finally, COBOL is mostly still used by governments or large companies which would be very hard to get into due to your origin. |
Many companies that use COBOL/RPG use IBM i (formerly known as AS/400) computers, and you can get hands-on practice for free in Pub400 (https://pub400.com). The salaries are decent but not great, and the platform only changes a bit with each OS release, so you can build up mastery without burning out.