Tristan Brindle

Wait, MSVC has a concepts implementation?

Yeah, this is a bug with AppleClang -- for some reason they removed the runtime parts of `std::experimental::bad_optional_access` from their version of LLVM in 7.3 but kept the header. I...

Hmmm, this is a bit of a headache. It turns out I was wrong, and this isn't fixed in XCode 8. Furthermore, it looks like it occurs even if using...

The above commit forces STX_NO_STD_OPTIONAL when using libc++

Agreed. I tried doing it, so there were four versions of most tests (two for `variant`); one with C++11, one with C++11 and `-DSTX_NO_BLAH` (via CMake's `target_compile_definitions()`), and the same...

As you can see from the commit, I've tried again to test all four possibilities (C++11/14, with/without overrides) using some CMake hackery. To address the issue of test_optional.cpp and test_any.cpp...

No problem filing bug reports! :) Just no promises that I'll be able to fix them... This is a known (to me) issue that I've run into myself from time...

I'll need to think about this one -- I don't think allocation should be necessary if we're willing to forego constexpr flatten in C++17 mode, but it would mean adding...

Agreed, `byte` should be in the `detail` namespace

There's `contract_violation()` and (sometimes) `struct contract_violation_error` as well as `byte`... are there any others? `size()`, `data()` etc are already in the `detail` namespace.