Feature request: improved output chord handling

[jtroo]

Is your feature request related to a problem? Please describe.

There have been several instances of users asking about a way to do symbol layers and they have an issue where kanata does not gracefully un-mod an output chord such as S-1 or RA-a for the user.

The typical solution to this is to use macro, but that means key repeat no longer works. There are probably ways to do better.

Describe the solution you'd like.

Un-mod an output chord once a different key is pressed.

This will make it so that for key repeat, the modifier key(s) in an output chord will remain pressed e.g. with lsft in S-1), but as soon as a different key is pressed, the modifier key(s) go away. If the original output chord key remains pressed, the modifier key will not come back. The original modifiers are no longer important since the original key will not repeat anyway.

Describe alternatives you've considered.

Make a macro repeat on key repeat.

This solution is less nice because it still relies on macro, which is less discoverable. There is already the repeatable macro action, but this does not have the same timing as native OS key repeat.

Additional context

No response

[gerhard-h]

the modifier auto-release could really be helpfull for new users (the solution)

for the alternativ solution I'm personally very happy with the current onetime-macro and use it a lot,
but maybe os-auto-repeat is a more intuitiv default behavior.

But in combitation with dead keys I belive autorepeat would do more harm than help e.g.
DOWN ´
TAP a
TAP a
gives áa or ´´áa,...