Send release key signal for keys not activated by macro

[soraxas]

Hi sorry to bother you again! Do you know if it's possible to send a key-up event for keys presses that is not emitted by the same macro?

In the docs, releaseKey it stated that

... This does not affect scancodes emited by different keyboard actions

which make sense. My use case is that I want to bind D to vim's cut to end-of-line, which I had set it up like

ifShift tapKeySeq S-end C-x

However, since the macro key D requires me to hold shift, the cut action becomes ctrl+shift+x which does not work (only ends up highlighting the line). I've tried ifShift tapKeySeq S-end S- C-x to add an empty shift press but it still doesn't work. Thanks for any advice on this!

[kareltucek]

Hi sorry to bother you again!

No trouble. I will try to provide you with information in real time, but leave feature requests / pull requests for another time...

Do you know if it's possible to send a key-up event for keys presses that is not emitted by the same macro?

Directly and generally, it is not possible - resulting usb report state is sum of us usb_repot_updater's usb repots and macro state reports - each macro slot/state has its own set of reports, which it uses to process tap/hold/release/press events. These are alive for as long as the macro is active, and then freed.

For more advanced features (for stateful vim-like things), you can use registers (search for setReg) to signal between macros.

Regarding the shift issue, there is suppressMods modifier, which should suppress regular scancode action modifiers. (This will not work if the shift is triggered/produced by a macro.) (Usually, folks want to bind shift/nonshift scancodes independently - yours is just a more advanced usecase of the same thing.) (Search the docs for suppressMods to see the generic examples.)

[soraxas]

No trouble. I will try to provide you with information in real time, but leave feature requests / pull requests for another time...

You are too kind!

Regarding the shift issue, there is suppressMods modifier, which should suppress regular scancode action modifiers.

Thanks for the pointer!

Actually I've lied: previously, I had followed the guide and mapped my shift key to double tap to caps

holdKey leftShift
ifDoubletap tapKey capsLock

... which, of course it doesn't work with ifShift, so my setup was actually

ifKeyActive 85 tapKeySeq S-end C-x

(where keyid 85 is leftShift).

supressMods doesn't work with mods-by-macro. But I don't care too much about double-shift-to-caps so I've switch my shift back to a normal key, and suppressMods works wonder. Thanks :)

---

I guess suppressMods doesn't work across macro because each macro holds its own bitfield of USB report (to avoid pollution across macro?), and therefore, a releaseKey event from macroA has no effect on macroB when they are bitwise-AND?

[kareltucek]

I guess suppressMods doesn't work across macro because each macro holds its own bitfield of USB report (to avoid pollution across macro?), and therefore, a releaseKey event from macroA has no effect on macroB when they are bitwise-AND?

Much more prosaic reason - because in context of a macro, you usually want to override non-macro behaviour, and therefore actually want to apply macro-induced-shifts irrespectively of applied suppressMods.

E.g., in case of ifShift suppressMods tapKey C-w, you want that control present.

Anyway, this is another small technical debt - if you can come up with sensible specification which will make things work better and not break too many current usecases, then a PR is welcome.

[kareltucek]

Anyway, this is another small technical debt - if you can come up with sensible specification which will make things work better and not break too many current usecases, then a PR is welcome.

I have refactored modifier handling some time ago into a concept of Input, Output and Sticky modifiers, so the behaviour should now be clearly defined, and controllable, so I consider it solved. Let me know if the manual is unclear or if you have any other remarks!