[PATCH weston 1/3] Introduce pointer lock interface

[Bill Spitzak]

On 09/23/2014 10:34 PM, Peter Hutterer wrote:

>> I figured the fixed type would be used, and wrap around. At 100 units/per
>> inch the 24.8 fixed type would wrap to where the sign changed after the
>> mouse was moved in the same direction for 1.3 miles. I think clients can
>> ignore this possibility, or if they really want to it should be easy to
>> detect and correct it.
>
> fwiw you'd be hard-pressed to find a mouse with less than 400dpi these days,
> 800 and 1000dpi are common, ignoring specific gaming mice with higher
> resolutions.

Those numbers are not fixed-point, they are integers and can use all 31 
bits (or 32 depending on how you treat negative wrap around). A 1000dpi 
mouse would wrap from positive to negative (2^31) after 33.8 miles.

The compositor/libinput has to convert to display units (ie pixels when 
there is no output scale) and uses 24.8 fixed point numbers for this. 
Unless the mouse dpi exceeds 256 per display unit (ie about 25,600 dpi) 
this is higher resolution than the mouse and in fact this is the 
limiting factor. And as I said for a 100dpi screen it is 1.3 miles, the 
mouse dpi is irrelevant.