Re: Character folding in the pretest

[Eli Zaretskii]

> From: Óscar Fuentes <address@hidden>
> Date: Thu, 04 Feb 2016 21:52:08 +0100
> 
> Eli Zaretskii <address@hidden> writes:
> 
> > You misunderstood.  Decomposition is just a tool that is used to
> > search for equivalent character sequences.
> 
> Equivalent in the Unicode sense, right?

Equivalent in the following sense: if the text includes ñ (these are 2
separate characters, they are just combined for display), then
searching for either n or ñ (a single character in both cases) should
find that 2-character sequence.

This follows the "canonical equivalence", described in more detail
here:

  http://unicode.org/reports/tr10/#Canonical_Equivalence

> If my understanding is correct now (the feature is some Unicode thing
> and not about how characters are used by people) I insist on defaulting
> to off, unless we renounce to make Emacs amenable to those who use a
> text editor for natural languages.

It _is_ about how characters are used, see above.

And you don't need to insist, you can just turn it off in your
sessions.  You have heard at least 2 people whose opinions are to the
contrary, for various valid reasons.