RE: Character folding in the pretest

[Drew Adams]

> I think it is clear that people want various different character
> folding rules.  The differences depend partly on what language the
> text is in, partly on whether the user actually speaks that language,
> and partly on personal preference.
> 
> Rather than arguing for an a-priori rule, we should let users show us
> what they actually like, and then try to find general patterns in
> those preferences so that we can make general defaults that users
> tend to like.

+1