how to change repeat speed of tilt wheel

[riow]

Hi there, I'm trying this software. May I know how to change the repeat speed of left and right tilt wheel? Changing of the numbers in "fixup (de-bounce) tilt wheel auto-repeat" doesn't take any effect.

[phil]

Sorry, AFAIK, the tilt wheel repeat speed is determined by your hardware and can not be changed.
All XMBC can do is listen to the incoming messages (from the mouse) and re-map them.

The tilt wheel repeatedly sends a "TILT LEFT or TILT RIGHT" message while the user holds the tile wheel. This is completely different from a normal mouse button (which has a specific DOWN (pressed) and UP (released) message). Therefore it makes it impossible to know if a user still has the tilt wheel pressed (held) down, other than to look for the absence of tilt messages.

The purpose of the fixup/de-bounce options for the tilt wheel are to attempt to determine the difference between a "normal" repeat message (tilt held down) and a new message (tilt pressed, released, pressed). This was to allow it to work with simulated keystrokes and not repeat the key while the tilt was held down, but to still send a new key when the tile was released and pressed again in quick succession.

The intention is not to effect the hardware repeat rate, but to attempt to determine the difference between repeats (at a fixed interval) and genuine presses and an irregular user controlled interval. It works to do that but only in that very specific scenario and only if the user does not click at the same rate the autorepeat is at!

Thanks,
Phil

[em27]

I see... I was wondering the same thing.

So basically what the fixup does is even if the mouse sends messages at lets say 100ms intervals, if I set it at 500ms, only one in five messages will arive, every 500ms?

I'm using the tilt for my right click menu, since the button seems to have died. It's pretty comfortable to operate especially with one finger, the only problem is that the menu repops several times and sometimes it ends in a closed state after a "tilt click".

[phil]

The fixup is supposed to debounce it completely, so you just get the one message. The two times are related to the initial repeat speed and the following speeds (some mice send the first repeat quicker than the following ones). But I know it does not always work as intended, normally because of different mice.

If set to 100, Essentially it is supposed to ignore messages that come in less than 100ms after the previous message. If it is not doing that, then let me know. It might be worth enabling debug logging as this should give you enough timing information to determine what values to use (and it should also log any de-bounce actions etc.).

Thanks,
Phil