mchristianl

Unfortunately that makes no difference. I was able to get the ```c++ float h(const A * a){ … } ``` version running with two additional ```c++ jlcxx::static_type_mapping::set_julia_type((jl_datatype_t*)jlcxx::julia_type("ARef")); jlcxx::static_type_mapping::set_reference_type((jl_datatype_t*)jlcxx::julia_type("ARef")); ``` and...

I got it working, but seemingly not in the most performant/unboxed way. By leaving all `IsBits`, `IsImmutable` out and using a mutable struct in Julia: ```C++ … float f(A a){...

PS: we have `Ptr = T ∀ T. JuliaReferenceMapping< [b] T& > = StaticIf ∀ T. JuliaReferenceMapping< [c] T*& > = T*& ∀ T. JuliaReferenceMapping< [d] const T& > =...

> just pass by value, but this was a mistake during my research about this, I found a quote from Chapter 3.7 of Design and Evolution of C++: > The...

Okay. I will follow the new branch then. Thank you for developing that further! So _dynamic_ is the wording for Julia-corresponding-Julia-types via the Julia-C-API in C++ and _static_ is for...

> this would be limited to structs that are POD in C++, right Well, not at all. Although it is more tricky to produce _some_ compatible layout, the important trait...

Nah, I take parts of that back. Julia structs seem to follow different alignment rules than C/C++ structs

That's strange. I got the notification email, so this should be the correct identifier at least

About the boolean-valuedness of decidable orders `_

Hi @ayberkt, The [latest progress](https://github.com/mchristianl/synthetic-reals#news--current-development-status) that I made was in proving [Cubical.HITs.Rationals.QuoQ](https://mchristianl.github.io/synthetic-reals/html/Cubical.HITs.Rationals.QuoQ.html) to be an instance of a [LinearlyOrderedField](https://mchristianl.github.io/synthetic-reals/html/Number.Bundles2.html#LinearlyOrderedField) in [Number.Instances.QuoQ](https://mchristianl.github.io/synthetic-reals/html/Number.Instances.QuoQ.Instance.html#%E2%84%9Abundle). This is just a preparation for a definition of...