Rewrites are risky, but sometimes necessary (EOL software with security risks, increasing costs for implementing new features within the existing codebase, code

Rewrites are risky, but sometimes necessary (EOL software with security risks, increasing costs for implementing new features within the existing codebase, code rot etc.). I think a lot of developers also lean in that direction because new code is easier to write than the old one is to read, understand and maintain, but attempting to create something that encapsulates and carries the old business logic over 1:1 can be very hard in practice! There will be problems regardless of whether you're trying to do a gradual rewrite and carry over API paths one by one, or do a big bang rewrite, where you replace the whole thing. I actually recently watched a conference talk called "Top 5 techniques for building the worst microservice system ever", which also touches upon the issues with rewrites: https://youtu.be/88_LUw1Wwe4?t=775 https://youtu.be/88_LUw1Wwe4?t=775 But as long as you're okay with the loss of some functionality or changes to some of it (hopefully for the better), then it should be a bit more doable.