This story is super interesting and I think can teach us some valuable lessons about refactoring and the price of truly understanding the domain in which the code operates.
The accompanying article is also a pleasant reading with a nice bit of background, and I really liked the motivation behind "layering" on top instead of rewriting from scratch.
I referenced a LOT of articles and case studies on legacy rewrites and ports in that first year. It is definitely a challenge that even with LLM still exists and I am happy to add to the body of knowledge surrounding the subject.
Thanks for bringing this story to HN!