Re: [hatari-devel] Enhanced keymap support

[Eero Tamminen <oak@xxxxxxxxxxxxxx>]

Hi,

On 3.10.2022 18.53, Thomas Huth wrote:
> Am Sun, 2 Oct 2022 15:57:18 +0200
> schrieb Yves Le Berre <yves.le_berre@xxxxxxxxx>:
> > Due to Thomas recent changes :
> >
> > - Reduce our dependencies to screen.h,
> > - Replace strlcpy() with a better implementation.
> >
> > I adapted enhanced_keymap patches by rebasing and grouping the patches
> > by filenames rather by added/modified function.
>
> Please don't squash patches together in such a big patch. That makes the
> changes very hard to review.
>
> Eero, you once started looking into this problem, do you still plan to
> take care of this? Or don't you have time for this anymore?

* I'm still not convinced that using SDL scancode indexing (instead of 
earlier ST scancode indexing) is better

  - Last time I tried the patch set, it was still occasionally 
triggering one of the added asserts after using AltGr shortcuts (which 
eats some key events).  Those extra asserts are needed because of switch 
to SDL scancodes

* I concluded that symbolic keymapping issues should be handled first [1]

However, I kind of run out of steam (motivation) after starting on 
symbolic keymapping stuff, as I had spent so *much* time on this patch 
series (my wife was complaining about it too).


[1] It's ridiculous that people need to write keymap files to use their 
keyboards with Hatari. Proper symbolic keymapping should be able to 
solve need for keymaps for most people.

Making it work for every TOS / keyboard layout version is just a lot of 
work, as you need separate builtin mapping for each layout:
	http://tho-otto.de/keyboards/

After that, it can be evaluated whether (and how much) the current patch 
series is still needed.


	- Eero