The case for boring technology

Every stack has a novelty budget, whether anyone names it or not — a limited amount of "let's try the new thing" the project can absorb before the risk stops being interesting and starts being a liability. Spend it on the part of the product that's actually new. Everywhere else, pick the boring option and move on.

Boring means documented, means the error messages are googleable, means the person debugging it at 11pm six months from now — probably me — isn't also relearning a framework. None of that is glamorous. All of it is why the interesting part of the project got built at all.