Make acecad driver work with 2.6.x kernels

[Giuseppe Bilotta]

<posted & mailed>

Hello all,

this is would be my first contribution to the X.org project so please have a
little patience :)

I own a Medi at com-branded Acecad tablet, and I've had lots of problems making
it work in Linux, with and without X. I've recently found that my non-X
problems were due to some system misconfigurations due to upgrades from
older distribution versions.

Yet making the tablet work under X was still impossible: it wouldn't work at
all with the acecad input module, and barely work as a pointer with the
evdev driver.

I've finally narrowed down the issue to two (known) issues in the respective
drivers, specifically

https://bugs.freedesktop.org/show_bug.cgi?id=3995

for the acecad driver and

https://bugs.freedesktop.org/show_bug.cgi?id=2758

on the evdev side.

Since there was a patch offered for the evdev driver already, and I assume
it's only a matter of time before it gets merged closing the issue (even
though it has been over a year since the last significant comment), I've
decided to spend a couple of hours fixing the acecad problem, if possible
without breaking the driver for older kernel versions.

As it happens, I have a patch ready now that allows me to use the tablet
(movement, pressure, buttons) without any apparent problems, even though
there are some minor issues that I'd like to iron out, in particular:

* it's quite difficult to make the cursor move up to the right or bottom
screen borders, as it usually stops somewhat early

* the absolute movement mapping is hard-coded to screen 0, and I would like
to make that configurable at runtime.

So here are my questions:

* should I begin by submitting the patch as-is, attaching it to the
aforementioned issue 3995, or should I wait until the other issues are
ironed out?

* is somebody willing to spend some time tutoring me (I'm a TOTAL newbie
when it comes to X.org and kernel input driver coding) on how to fix those
two issues and possibly implement other (run-time configurable via the
xinput interface) features?

* is there any chance of these patches getting in mainline?

Thanks all in advance,

-- 
Giuseppe "Oblomov" Bilotta