I read this often, and since people tend to do overengineering, this isn't bad advice. However: Some things can only be reduced so much. Rebuilding the product

I read this often, and since people tend to do overengineering, this isn't bad advice. However: Some things can only be reduced so much. Rebuilding the product we ship at $dayjob, and which I enjoy working on, with a comparable feature set would probably take me ~8 years fulltime, assuming I know the whole stuff as well as I think and I don't hit major blockers (e.g. I don't do UI stuff, but most work is in the backend anyway); there is virtually no FOSS to accelerate this. A lot of the features can be dropped to get to a MVP, but I suspect I'd still be working on that for 6 month to a year, fulltime, and I will not be as easily extensible as the current incarnation. So that's 1000 to 2000h with a huge amount of technical debt and magnitudes inferior to the "competition"/$dayjob. Could I still pull that off if I wanted to? Yes. Would I enjoy it that way? Hell no.