Re: [hatari-devel] Preparing for next release 2.4

[Eero Tamminen <oak@xxxxxxxxxxxxxx>]

Hi,

On 7.5.2022 12.47, Thomas Huth wrote:
> Am Sat, 7 May 2022 10:48:18 +0200 (CEST)
> schrieb Yves Le Berre <yves.le_berre@xxxxxxxxx>:
> > Will the patch series for enhanced keymap be added to 2.4 release ?

Sorry, no.  I haven't *properly* looked at the patches you provided 
(months ago) on top of my patch series, and there are few other concerns 
(see below).


FYI: 2 years of work-from-home sapped me of desire to continue sitting 
longer times at the same chair / table at home, doing hobby coding. 
This may change once corona is properly over and work happens elsewhere 
than at home, but for now you unfortunately need to carry the torch for 
this change.


> I thought Eero has reworked the keymap handling quite a bit in the past
> months? Is there still anything missing there?

Several things:

* Looking at extra patches Yves provided (in "Enhanced keymap support" 
mail thread)

* Somebody else (e.g. you) reviewing the patch series I did

* Being 100% convinced that mapping host key(s) to (potentially 
multiple) Atari keys / modifiers, and identifying key releases by host 
scancode, does not cause issues [1]

* Starting to look into symbolic key mappings, but that being left 
*badly* unfinished.  I wanted to get that first to fairly complete 
state, to see whether it would be better solution to mapping all user's 
host keys to Atari keys, than the patch series Yves is interested about

(Properly working i.e. automatic symbolic key mapping would be better 
than users needing to write complicated key mapping files for scancodes.)


	- Eero

[1] E.g. losing keys due to IKBD buffer overflow, or user already 
pressing ALT, which release needs to be simulated to complete ALT-XXX 
sequence.  E.g. one of the asserts I added on key release, sometimes 
triggers when I use fast-forward shortcut (AltGr-X).