Crystal Durham

I'll keep it in mind. I'm mostly concerned about people accidentally stumbling into cases that use nonconst methods, potentially even in a non threadsafe manner (thus correctly nonconst per cxx's...

The warning is actually emitted from the `.cxx` file, so it might be prudent to `#pragma warning( default )` in the generated code for compilation as well.

If you aren't going to get to it soon @adetaylor, I could potentially take a good stab at implementing the above design this coming week.

(Had some free time, decided to take a look at this again.) > @adetaylor > Design proposal: > > - [...] > - But then we call a new function...

#908 attempts to implement this via keeping a new mapping of types to atoms, which is used by downstream passes instead of `Atom::from`. It needs significant testing to ensure that...

Unfortunately, the C++ code I'm trying to interface with is too complicated for autocxx/bindgen to process cleanly at the moment, so while I'd love to use autocxx, it's not possible...

Reported https://github.com/google/autocxx/issues/479. Unfortunately the code isn't reproducible but it is freely obtainable, so perhaps someone able to track down where autocxx is panicking can reproduce the panic and minimize the...

> I don't know what the actual risks are, since the struct is still standard layout It mostly seems to be that while UDTs with constructors are still _StandardLayoutType_, the...

See also context in https://github.com/fitzgen/bumpalo/issues/128 the test seems to be a poor nondeterministic test to begin with.

Makes sense. I just sort of spitballed a rough design expecting further refinement. I'm glad to see this is available now!