See my response to the sibling - my point is that optional types are useful as a design tool for writing more robust code that deals with expected missing data as per the design, but don't help at all with actual bugs that cause unexpected missing data at runtime.

Yes and no. Consider that there are functions where returning "no value" (null in our case) is a valid option. It is possible that the callers of those methods forget to check for a null response; leading to a NPE.

If the language differentiates between fields/results that can be null and those that cannot, and makes the caller specifically handle the null case, then that cause of NPE will longer happen.

I politely suggest you re-read my comments, as I have indeed considered that and it's exactly what I'm saying :-) I agree with what you're saying 100%, I perhaps haven't written clearly enough because I think you're thinking I'm arguing against optionals being useful. I'm not, I'm just saying they're not a solution to the bugs that cause 'real' NPEs - only design bugs where the code doesn't model the data correctly.

I'm not speaking to whether or not you consider optionals useful. Rather, I'm saying that calling a method that can return null without dealing with that case in the code is one of the most common reasons for a NPE. The calling code _is_ the actual error, and forcing the use of optionals removes the possibility of them.