Detect Intentional Input Binding for PTT

[justin-lovell]

The number of active bindings must be no more than two so that users do not inadvertently switch radios when a neighbouring hat switch is toggled for a brief millisecond.

This is an implementation as per the algorithm suggested in issue: #272

However, this PR would not resolve the illustrated scenario keybindings as the modifier would be superseded through the logic of the following lines:

DCS-SimpleRadioStandalone/DCS-SR-Client/Input/InputDeviceManager.cs

Lines 612 to 638 in e8b71ad

	if (previousBind.ModifierDevice == null && bindState.ModifierDevice != null)
	{
	//set previous bind to off if previous bind Main == main or modifier of bindstate
	if (previousBind.MainDevice.IsSameBind(bindState.MainDevice))
	{
	previousBind.IsActive = false;
	break;
	}
	if (previousBind.MainDevice.IsSameBind(bindState.ModifierDevice))
	{
	previousBind.IsActive = false;
	break;
	}
	}
	else if (previousBind.ModifierDevice != null && bindState.ModifierDevice == null)
	{
	if (previousBind.MainDevice.IsSameBind(bindState.MainDevice))
	{
	bindState.IsActive = false;
	break;
	}
	if (previousBind.ModifierDevice.IsSameBind(bindState.MainDevice))
	{
	bindState.IsActive = false;
	break;
	}
	}

To resolve the original issue completely, it is suggested that the PTT, radio switching, and radio tuning should be elevated up to a state machine.

This PR would resolve my specific use case where the neighbouring hats are being triggered for a fraction which causes the radio to switch rapidly during transmission.

[justin-lovell]

Closing PR since no activity nor comments

[ciribob]

Apologies apparently I've missed this...

Please reopen if you want too, and I'll have a look for the next release