Re: [idn] Fast nameprep vs. slow nameprep

[Patrik Fältström <paf@cisco.com>]

At 19.57 +0000 01-01-27, D. J. Bernstein wrote:
>I'm explicitly assuming that we _won't_ allow names needing nameprep in
>configuration files. Bad names disappear as soon as they leave the
>keyboard. That's fast nameprep. The changes are isolated inside the
>keyboard interface.
>
>You're implicitly assuming that we _will_ allow names needing nameprep
>in configuration files. That's slow nameprep. It's horribly expensive.
>Hundreds of programs will need to learn about nameprep.

No, you misunderstand me. We talk about exactly the same thing. We 
need something between the keyboard and the config file which do the 
nameprep, so never non-nameprepped things end up in configuration 
files.

That's the whole idea with IDNA. Do nameprep as "close" to the 
keyboard as possible (and the reverse when displaying).

   paf