Comment # 26 on bug 1209594 from Stefan Dirsch

(In reply to Luca Billi from comment #24)
> Sorry to start sounding like a broken record but I'd like to reiterate that
> set-keymap and set-x11-keymap are not orthogonal but they overstep each
> other:
> they BOTH change all XKB* variables *and* KEYMAP.

That's  how mapping between Linux console and X11 keyboard is currently.

> Unless the toggle "--no-convert" is used that is.

Well, but we need the convert for YaST (from Linux console to X11 at least)

> Namely, if a user has configured KEYMAP != XKBLAYOUT, no sequence of
> set-keymap / set-x11-keymap will ever be able to reproduce that configuration
> and the user settings will be inevitably overidden.
> 
> That is not my case but I'm just pointing out a limitation of the current
> approach.

Well, who wants that. Different Linux console keyboard and defaultl X11
keyboard
setting?

> Taking a step back, I don't know about the big picture but I'm wondering
> about the role of the script "keytable".

The main purpose of this script is to finish what YaST fails to do during
installation.

* Wed Oct 18 2017 sndirsch@suse.com
- reintroduced /etc/X11/xdm/keytable script and the hook-up in
  /usr/lib/X11/display-manager to setup X11 keymaps, since YaST
  is not able to use localectl already during installation due to
  dbus and appropriate systemd services not running :-(
  (bsc#1046436)

I already killed it before, but I needed to revive it. :-(

> If it is to just provide sensible defaults in case some settings are
> missing, 
> maybe this logic might do the trick:
> 
> if KEYMAP and XKBLAYOUT are defined 
>   do nothing (and preserve user settings)
> 
> else if KEYMAP is defined but not XKBLAYOUT
>   set-keymap KEYMAP (which will also implicitly set XKB*)
> 
> else
>   do nothing (current behavior)
> fi

Yes, that may work. Haven't tested this though.