experimental status of C++17

[Jonathan Wakely]

On Tue, 19 Mar 2019 at 15:53, Florian-Bogdan Tudoran wrote:
> Supposing one sticks to standard C++, does one need to worry about
> functionality breaking in the future? Seems unlikely, as the standard is
> already published but then how should we read that statement?

The problem isn't that a published standard might change, it's that
our implementation is incomplete and therefore doesn't necessarily
follow the standard correctly, and so might need to change. API
changes should only happen where we are not conforming and need to
change to follow the standard, but we reserve the right to make ABI
changes while support is experimental. As a hypothetical example, we
might realise that our implementation of std::variant has a bug and
needs an extra bool member added to it in GCC 9+. That would change
the size of std::variant objects, and so would make C++17 code
compiled with GCC 9 incompatible with C++17 code compiled with GCC 8.

I'm not aware of any such problems at the moment, but if we need to
make such changes then we will do so.