twitter thread - https://twitter.com/allenholub/status/1496326760719216643 Not my article and I don't even necessarily fully agree - but I thought this might spark some interesting discussion on HN.

Seems to be a difficult way to go once you're past the early stages of a project. I can't imagine a team at a say Series B/C startup pausing long enough to test their codebase thoroughly, if they were not already doing it to some reasonable degree.

Tweet below and the second below it, call out non-eng users and priority as two good reasons to have a tracker and I agree with that https://twitter.com/mekkaokereke/status/1497267436315312128?...

I love ideas like this; radical flip-the-script concepts that you basically can't incrementally apply but I do believe it worked for the author, and can work for others.

Stuff we could never have done from the beginning, but can basically undo the patches we've applied to fix issues that no longer apply.