4 ms·
From where I sit as a dev, it's less "I want shiny new tech on my resume" and more "we need to refactor / simplify everything for ease of maintainability". It'
by andrewfong 5y ago
From where I sit as a dev, it's less "I want shiny new tech on my resume" and more "we need to refactor / simplify everything for ease of maintainability".
It's a hard problem. Sometimes it's unnecessary but sometimes the tech debt is real. It's not uncommon to have one blessed path for doing most things but hey, here's this edge case that doesn't fit our mental model and is a pain to test and bloats our bundle and we keep getting paged for it and the engineers who built it left the company and didn't leave any documentation or tests, so let's just kill that feature.
And now users are pissed and don't understand, because from their perspective, they think the cost of leaving the feature alone is zero. But the truth is the cost of keeping it is quite high and there just isn't enough usage to justify engineering effort to reduce ongoing maintenance costs.
- swiftcoder 5y agoIt's much worse in ecosystems that are changing the core frameworks at a rapid pace (hi, React!). If you don't do the same thing Facebook does internally, and continually refactor your entire codebase onto the latest paradigms, you quickly end up unable to adopt new features and 3rd party libraries when you actually need to do so. If you look instead at a mature ecosystem, like, say, C++ or Java, you might reasonably maintain a codebase for a decade without needing to adopt new language features.