bug#53452: 29.0.50; C-<arrow> selection broken by translate-upper-case-key-bindings

[Eli Zaretskii]

> From: Robert Pluim <rpluim@gmail.com>
> Cc: spacibba@aol.com,  53452@debbugs.gnu.org
> Date: Tue, 25 Jan 2022 11:19:33 +0100
> 
> >>>>> On Mon, 24 Jan 2022 15:39:28 +0200, Eli Zaretskii <eliz@gnu.org> said:
>     >> Why not?
> 
>     Eli> Because people will probably  want it applied to any key.  There's
>     Eli> nothing special about letters, except that "upper-case" implicitly
>     Eli> refers to letters.
> 
> Let's start again:
> 
> Currently, S-<right> produces S-<right>, but is translated to <right>,
> and C-S-a is translated to C-a
> 
> If you set translate-upper-case-key-bindings to nil, then S-<right>
> produces S-<right>, and C-S-a produces C-S-a
> 
> So the effect is as if the option were called
> 'translate-shift-bindings', but that effect is not called out
> anywhere.
> 
> So either we fix the behaviour, or the documentation. Would it be that
> hard to restrict this to something like [a-z] or !function key? Or we
> could always add yet-another-user-option.

My point was that the behavior is probably okay, and the documentation
should be amended if it could confuse.