[email protected] wrote:
> James Carlson wrote:
>
>> [email protected] writes:
>>
>>
>>> Yes. I have now read the GPL license and think it
>>> will in the long term be simpler to re-write my code to use no GPL bits.
>>> I don't expect this to take more than a week or so...
>>>
>>>
>> Sounds like a nifty and worthwhile contribution to me. Good luck!
>>
>>
>>
> I've given some thought to how to do this, and I just want to bounce some
> ideas off of people, partly to see if I am making sense, and partly to see
> if someone else is already doing something similar that I can use.
>
> In the current (linux and my Solaris code), much of the work is done in
> the kernel, including:
> - Determining which pen is in use, and which end of the pen (stylus
> or eraser)
> - Determining position of pen on the tablet
> - Determining when tablet/pen buttons/keys are pressed/released
>
> The kernel side takes this information and forms events (in a defined
> structure)
> that it passes up to the X library module. The X library module then
> does some
> more work and finally sends the events to interested parties.
>
> What I would like to do is make the kernel side of this much simpler.
> Instead
> of doing calculations, all the driver will do is pass up the raw data
> (and handle some
> ioctl calls for getting things like vendor and product ID, etc.). The
> library module side
> would then do all processing of the raw data, including the stuff
> currently done
> in the kernel driver.
>
> If this is done correctly, the amount of change to the kernel side will
> be extremely minimal,
> if/when Wacom comes out with a new tablet model. And the X library
> module could be made generic to handle "tablet" devices. The specific
> tablet vendor/product stuff could be added to the "generic" tablet xdrv.
>
A simple kernel driver seems like a fine idea, as long as you know that
you only want to support a specific userland application (the X
server). If you want to do something more sophisticated (such as have
the table "emulate" a mouse or something like that) then having a more
complex driver might be useful.
You should also have a look at the USB HID specification. Can this
device either act as a normal USB HID device, or emulate one? It might
be the case that our HID support needs to be more completely implemented
-- I don't know. But I do know that the *spec* for USB HID is
incredibly rich, supporting everything from keyboards and mice to
complex devices like VR gloves and helmet sensors! :-)
-- Garrett
> Any thoughts on this?
>
> thanks,
> max
>
> _______________________________________________
> driver-discuss mailing list
> [email protected]
> http://mail.opensolaris.org/mailman/listinfo/driver-discuss
>
_______________________________________________
driver-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/driver-discuss
