Re: [idn] Why follow IDNA with UTF-8?

[Keith Moore]

> Just-use-ACE is actually the most succeptible to this particular problem,
> since old and new applications pass data into the system through the old
> APIs equally, without any distinctive markings. Was it nameprepped? nobody
> knows...

ACE is sufficiently ugly that it's difficult to imagine that it will be produced 
other than by calling the APIs that produce it - and those APIs will support 
nameprep.