Replies

In my world, co-workers all have degrees in math/physics/chemistry/software from BA to PhD. As a working group, we are aging. The young guys are in their 40s. The upper end is under 75. We're not seeing a reasonable number of qualified applicants to hire and learn the details of systems we have been supporting since the early 80s. A fair bit of corporate memory is going to retire/die before it can be passed along.

I'm witnessing firsthand the regression in IT from the late 1990s/2000s. Things that were common knowledge are now unknown to supposed "experts". It's about compensation; companies want something but don't want to pay for it and think a system is self-maintaining.

I find myself annoyed about building systems since 1980 and having to spend time/$$$ getting "certificates" to make my government customer happy. I've been a UNIX kernel developer on UNIX1100 (Sperry/UNIVAC), Power6/32 (tahoe) UNISYS 7000, HPUX-7 (68040) and Linux (networking) since the early 90s. As a full time developer through the 80s/90s, the system admin chores were collateral duty. It was expected. Now it has become a "profession" by itself. The extra layers of security in networking have certainly become more complex. As a developer, you need to be aware of the intricacies to write correct implementations. The fine details are often OS specific, so you must refine your skillset depending on what your customer demands. I'm currently converting old Windows server apps to live in a Docker container with Spring Boot REST interfaces and deployable with Kubernetes. Other members of the team are converting Linux apps in the same direction. Composable microservices in the cloud. As part of the old guard who built it, it's my task to adapt the old code to the new world. Inside the Docker container is OSGi bundles. Forcing Spring Boot to run inside an OSGi was not obvious, but it now runs fine.