|
From: | Tacvek |
Subject: | Re: [Enigma-devel] TiltStick, USB based tilt sensor |
Date: | Tue, 5 Feb 2008 11:45:54 -0500 |
----- Original Message ----- From: "Till Harbaum / Lists" <address@hidden>
To: <address@hidden> Sent: Sunday, February 03, 2008 2:27 PM Subject: Re: [Enigma-devel] TiltStick, USB based tilt sensor
Hi, hmm, no feedback. Please let me explain why i am bothering you with this. The sdl layer supports joysticks and the tiltstick is a joystick. Thus everything should be a no brainer.Unfortunately there's some magic in the linux joystick driver (i never checkedif windows does the same) that extends the "death zone" of any joystick somewhat. This means that the range around zero of each axis is extended a little bit to make sure that a slightly miss-calibrated joystick doesn't permanently report tiny movements. Thus the joystick has to moved a significant amount before starting to report values != 0.
I would expect the nullzone to be fully user configurable. Having magic hardcoded into the kernel that cannot be turned off at runtime does not sound like standard procedure to me. Are you sure you cannot specify the null zone in the kernel level joystick calabration tool? (jscal, part of the joystick package). Since you seem to have hardware level calibration, in theory the software calibration could be set to just output the raw input values, with no nullzone.
Is that not correct?I'm pretty sure that can be done, but I'm having trouble seeing how this is done by just reading the driver source.
[Prev in Thread] | Current Thread | [Next in Thread] |