I’m specifically interested in Czech keyboard layout. Instead having numbers in the top row, we use it to house out many characters with diacritics: ě, š, č, ř and so forth. Right now the results of pressing those keys are not very useful and also inconsistent:
Holding keys “š”, “ž”, “ý” and the rest of the row only sends a text action with the correct letter. Keys “ě”, “č” and “ř” send one text action with the correct letter and then repeatedly normal action “key_e”, “key_c” or “key_r” respectively, so any action bound to these letters without diacritics would also happen when pressing the key with the accented version.
Since I won’t be needing to keep track of whether these particular keys are held and am only interested in the initial press, I think I can use the text action for most of these and come up with some ugly workaround to detect whether a text action “ě” was sent in the same frame as the “e” key was pressed, but I’d rather find a better solution, also because of future projects.