Re: [lmi] patch: compilation fix for wx 2.9

[Greg Chicares]

On 2010-05-15 15:01Z, Vadim Zeitlin wrote:
> 
> And to test whether the string is convertible to a char* in the current
> locale encoding without loss of information you need to perform the
> conversion and check that the result is not empty.

That's the approach I followed in revision 4952, committed 20100515T2041Z.

It's at least as simple as the other approaches discussed, and it doesn't
remove existing support for anyone's language. I understand that such
support is locale-dependent, tenuous, fragile, accidental, brittle, etc.,
but I judge it more culturally respectful to erect no new barriers.

>  Of course, dealing with the problem of representing Spanish characters in
> a string using Chinese locale is exactly the kind of nightmare that Unicode
> frees us from.

Someday we'll upgrade to a boost version that provides boost::wpath,
and then these problems will be easy to solve.