Re: Fwd: Re: touchscreen calibration again

Hi Lau,

What's up?

Do you think you will have time this week to work on the raw touchscreen
data that we sent you?

That's a quite critical issue for us, and if you think you can't take
care of this in the next days (or the next week-end), I think I will
implement a brute-force temporary workaround.

The idea is to use a simple look-up table based on the 120 points from
one of our harvest files. When it gets a point from the touchscreen
device, the h2200_lcd driver would look up the real screen coordinates
from the 2 closest measured x and y coordinates, and would perform a
linear interpolation (or extrapolation if the point is outside the
measured area).

That would consume a bit of RAM, but not much CPU. So, it won't hurt to
use this in the beginning instead of a better solution as the one you
may propose.

    Cheers,

    Michael.

-- 
Michael Opdenacker
http://opdenacker.org/