First Impressions

[Georg Wrede]

Anders F Björklund wrote:
> If you're willing to handle the "surrogates", then UTF-16 is a rather
> good trade-off between the default UTF-8 and wasteful UTF-32 formats ?
> A downside is that it is not "ascii-compatible" (has embedded NUL chars)
> and that it is endian-dependant unlike the more universal UTF-8 format.

Problem is, using 16-bit you sort-of get away with _almost_ all of it. 
But as a pay-back, the day your 16 bits don't suffice, you're in deep 
crap. And that day _will_ come.