Tom Honermann wrote:
> This history is why I emphatically assert that there isn't anything broken with
> this programming model. It has limitations, it has sharp edges, and we can all
> rejoice that support for other programming models has emerged, but that
> doesn't change the fact that there is a lot of code written to this model that
> is still being maintained.

Do we agree that this existing code doesn't do this

std::cout << u8"Привет!" << std::endl;

?

Do we agree that, assuming the narrow literal encoding can represent Cyrillic,
e.g. it's Windows-1251, it's eminently reasonable to expect the above to do
the same thing as

std::cout << "Привет!" << std::endl;

?

If the answer to both these questions is "yes", where do the objections against
using the narrow literal encoding as the intermediate (a strategy that produces
a Unicode-based model when the narrow literal encoding is UTF-8, and degrades
gracefully when it isn't) come from? What's the source of disagreement?