"Unicode everywhere" is something I hear often, but don't really agree with, unless you mean "Unicode by default and preferred and easy". The real world still

"Unicode everywhere" is something I hear often, but don't really agree with, unless you mean "Unicode by default and preferred and easy". The real world still has and probably always will have a lot of text in other encodings - Latin-1, Shift-JIS, heck even things like GSM03.38. I like how Ruby handles it - every string has an encoding attached, and manipulations with non-compatible encodings will complain.