On Wednesday, 27 November 2013 at 12:46:38 UTC, bearophile wrote:
D+Phobos seem to fail most things (it produces BAFFLE):
I still think we're doing pretty good.
At least, we *handle* unicode at all (looking at you C++). And we
handle *true* unicode, not BMP style UCS (looking at you
Java/C#), with the options of storing said strings in any
encoding: UTF8 through UTF32, and the possibility to also have
ASCII.
We don't yet totally handle things like diacritics or ligatures,
but we are getting there.
As a whole, I find that D is incredibly "unicode correct enough"
out of the box, and with no extra effort involved.
