A. Jiang

> Is this a resolvable problem? I don't think this is solvable, unless we change the standard wording to restrict ADL for `swap` to enumeration and class types (like `ranges::swap`).

Currently no implementation accepts this example due to ADL in the comparison of iterators ([Godbolt link](https://godbolt.org/z/hfdobM4cY)). However, the standard doesn't seem to state that this is UB/ill-formed. ```C++ #include struct...

If VS can be updated to use Clang 18 before C++23 ABI lockdown, it seems possible to use `[[msvc::no_unique_address]]` for `in_value_result`, `out_value_result`, and `expected`.

See also #3857. ~Perhaps we can handle both issues together.~ Oh, two issues are about opposite directions.

It seems that this is only the resolution of LWG-2524 and should be handled as other LWG issues.

It seems that we should pass `_Transform_op` to `_Apply_inclusive_predecessor` (with its signature changed) and use `_Transform_op` in it. We should also pass the beginning "input" forward iterator. Perhaps we should...

WG21-P0571R2 is supposed to resolve the issue of `GENERALIZED_SUM`, although the current revision is still less than ideal. (Not yet rebased onto _Mandates_; only implicit conversion is mandatory, but explicit...

We have a number of issues on difference types. I guess there're also remain issues in `` ~, ``,~ and ``. CC @hewillk. Edit: `deque` is likely to be fixed...

I tried to implement the suggestion from @davemcincork in https://github.com/microsoft/STL/issues/3254#issuecomment-1991532148. ```diff diff --git a/stl/src/syserror_import_lib.cpp b/stl/src/syserror_import_lib.cpp index e2f2d2b7..a70424ba 100644 --- a/stl/src/syserror_import_lib.cpp +++ b/stl/src/syserror_import_lib.cpp @@ -46,11 +46,48 @@ extern "C" { if...

> I _think_ your proposal will leak. As in my suggestion, you're calling `FormatMessageA` with the `FORMAT_MESSAGE_ALLOCATE_BUFFER` which causes it to allocate the buffer. But unlike my suggestion, you don't...