When working on introducing our stdlib, I found myself removing a lot of Python backend code just to get things not failing at runtime. The output is of course far from correct. It's just painful to work on it now, with the current level of mess 🤷 😆 now I see your pain @SerVB. That's why I think we should really invest into cleaning things up and understanding everything that happens in our Python backend. I'm shelving the stdlib thing and will focus first on removing commented and other dead code, e. g. I suspect some lowerings are not needed (either at all or not yet) in our case. It may take us like 3 months, but I see it as an investment into future development speed.