Re: Glidix re-write
Posted: Sat Nov 02, 2019 9:17 am
Whenever people talk about rewriting their code, I advise them to reconsider that. Rewrites are rarely productive. It is usually much faster (and less error-prone) to start with some semi-working project and refactor the bad parts than it is to rewrite the project from scratch. If I look at my own projects, I notice that some parts of them have been refactored a lot. Managarm's memory management has been completely refactored at least 2 times. If I look at my stats on GitHub, I notice that I tend to refactor *a lot* of code (e.g., in managarm: 207,750 ++ 154,902 --, not including third party code). LAI has also been completely refactored at least once. Still, it is much more productive to refactor existing components than it is to rewrite them from scratch.
Hint: once you have rewritten your code, you will still not be satisfied with all aspects of its architecture. The latter is just a state that you (or any other software developer) will ever reach for non-toy projects.
Hint: once you have rewritten your code, you will still not be satisfied with all aspects of its architecture. The latter is just a state that you (or any other software developer) will ever reach for non-toy projects.
