Jay Krell

@mclow I squashed and rebased, ok? (Is squash really required? A lot of repositories are configured to do that upon commit, not optionally. Imho that is 99% the right policy....

Why is the compiler option available? We believe we have code that depends on the old behavior. And we do get compilation errors where when we upgrade the compiler. (I...

Honestly I am a bit unsure, but this where we are.

Notice how the json motivation kinda came and went. The users eventually fixed their code. Maybe we will too eventually. I just figure, it is fairly harmless to support here....

Our code is really not portable, very Windows-specific, fairly hopelessly so, and likely will remain that way for a very long time. This is not the biggest problem by far....

> whose exception specification will change from compiler to compiler Isn't it the other way around? This tries to make the exception specification always the same?

> Not unless we use it everywhere in boost. Yeah, I kinda admitted that. My PRs only cover what I hit in our code. Selfish. Or a starting point. :)...

Right..there is a sort of transitiveness of noexcept(false) https://docs.microsoft.com/en-us/cpp/build/reference/zc-implicitnoexcept-implicit-exception-specifiers: " For user-defined destructors, the implicit exception specifier is noexcept(true) unless a contained member class or base class has a destructor...

https://ci.appveyor.com/project/jeking3/format-bhjc4/builds/41225712/job/p0rc2y68a1xtdh6q ``` \boost/format/feed_args.hpp(99): error C2280: 'std::basic_ostream &std::operator

Rude/lazy question: Was CI working before my change? Can we PR a nop/whitespace/comment change and see how it does?