Comment by mhandley

There's probably a distinction to be made between being obvious and having no surprises. I agree Python is relatively obvious, and I write a lot of small python scripts for data analysis because if I can figure out what I want to do, doing it in python is usually very simple. But for larger programs I care less about obvious, and more about no unpleasant surprises. Once the amount of code grows larger than I can hold in my head at once, I really want the language to help me avoid footguns. Python mostly isn't great for this due to its type system (with a few exceptions such as its big integer support avoiding most integer overflow surprises).

It probably isn't a popular opinion, but I actually think C++ does a fairly good job of this, so long as you're disciplined in how you use it. That discipline is not necessarily obvious, but I've acquired it over 30+ years, and now I find that I almost never have memory safety issues (I do use sanitisers here, but they rarely show up anything). Perhaps more importantly, C++'s fairly strong typing means I'm generally pretty confident refactoring code as requirements change without getting unpleasant surprises. Sure, C++ could do better in many ways, but it is pretty good at avoiding many of the unpleasant surprises I care about, at least for single-threaded code.