Drum input and voices
MuseScore treats drum staves different from others in terms of how note input is performed. This is unavoidable due the nature of how drum notation works - there are multiple notes that can be placed on each line or space with different shaped note heads to indicate different drums, etc. Also, drum input more or less demands use of multiple voices but in a unique way, where any given "note" is used almost exclusively in the same voice throughout the piece. But some aspects of how drum input works with respect to multiple voices seem confusing or even broken. I'd like to take this opportunity to collect ideas for how things could be improved.
First, here is the issue that started the discussion: #153686: Drum input palette: notes set to enter as voice 2 are entered into voice 1 in error. There are several different ways of entering drum notes - computer keyboard, double click icon on drum palette, single click icon on drum palette then click in score, MIDI, and Piano Keyboard. Of these, only the computer keyboard always takes advantage of the predefined voice for each drum note (as specified in the drumset definition for the staff). Single click icon sets the toolbar to the specified voice so that clicking the score will normally enter into its predefined voice, but you can change to a different voice before clicking in the score to override the default. MIDI and the Piano Keyboard ignore the predefined voice entirely and just use whatever is selected on the toolbar. Double click in the toolbar is the one that is kind of broken with respect to voices - it behaves in what appears to be a random manner (actually depends on what is already present in the measure).
Obviously I want to fix double click to be more consistent and predictable. The question is, what *should* the behavior be? More like keyboard input where we use the predefined voice? Or more like MIDI / Piano Keyboard entry where we use the toolbar?
And more broadly, is it really valuable to keep the idea of predefined voices, given that it only really applies to some input methods and we need to provide the ability to override this anyhow? I can see both advantages and disadvantages to the predefined voice. One advantage is, it really is the case that snare drum is probably always going to be in voice 1 and bass drum in voice 2 throughout a score, so having predefined voices can save a bit of voice switching during note input. One disadvantage is, it just feels different from regular note input. It also feels inconsistent between methods of drum input. Of course we *could* change this so that all methods of input use the predefined voice, but then the only way to override it would be using the single click icon / click in score method, which is a step backwards in usability to me.