[moomoo11]

Going to extremes never works like ONLY bug fixes or address tech debt for X time. Or only shipping features even if the foundation is shit.

The best option is to accept software is soft and malleable.

If it prints money but it’s some legacy code that smells like shit and has bugs, just write a new clean interface on top and build to it. Over time you’ll end up replacing the underlying system while shipping new features and fixing bugs as they pop up. Oh and the money keeps printing.

It it doesn’t print money and the code is shit, the project should just be scrapped. Who cares.

If people constantly keep building shit code and never improve, just quit because that company probably sucks.

Be extreme by being practical.