bug#25747: 25.1.91; hideshow binding C-c @ C-h overrides help

[Drew Adams <drew.adams@oracle.com>]

> I also think that "C-c @ C-h" is a very unfortunate keybinding, even if
> it is mnemonic for "hide".

I agree.  There are other *hide* commands that don't use `C-h'
in their bindings.  There are even other hideshow hide commands
that don't use it:

C-c @ C-M-h
C-c @ C-l
C-c @ C-t
C-c @ C-d

There's anyway redundancy: C-c @ C-h is the same as C-c @ C-d.
(And C-c @ C-M-h is the same as C-c @ C-t.)  Are all of the
default bindings really necessary/useful?

(If you think it's important to have something that fits with
the `C-c @ C-M-h' of `hs-hide-all' then use `C-c @ M-h'.)

My guess is that the author just didn't know about using `C-h'
after a prefix arg.  And the more exceptions to that working,
the more people will do likewise: mistakenly (in ignorance)
bind `C-h' after a prefix key, preventing it from that use.

> But it seems much too late to do anything about that now.

Why is it too late?  Just drop `C-c @ C-h' - that's my suggestion.