I sympathise, but there is a flip-side. Code bases, get full of half-arsed framekworks and automation tools that nominally solve all your problems, but don't,

I sympathise, but there is a flip-side. Code bases, get full of half-arsed framekworks and automation tools that nominally solve all your problems, but don't, either because they address the wrong issue, or because bugs prevent them from doing even the nominal job. Yet these half-working things can become the required internal standard that teams have to adopt against their wishes. They will not be forgiving when the maintainers start ignoring bug reports. Now this is fundamentally a management failure: pushing a solution without knowing its limits or counting the cost of backing it up and compensating the maintainers. But as engineers we tend to be complicit, because when we make a new shiny thing, we also want to push it and don't want to tell mangement that it will really cost three times more than they were hoping.