[potatolicious]

Honest question: but isn't that just a broken implementation (and a very obvious brokenness at that)? It seems to me there's a big difference between someone not coding to the standard, and the standard making your taks impossible.

[codeka]

The same could be said of UTF-16 implementations that don't support surrogate pairs.

[tlrobinson]

But that's not true for UCS-2, which can't represent certain code points.

On the other hand, UTF-8 is ASCII compatible and more efficient for text that's primarily ASCII.

[dsl]

If most implementations are broken, it becomes the standard.