nanogui: NanoCal
Subject:
RE: [nanogui] NanoCal
From:
"Gabriel Goldstein" ####@####.####
Date:
20 Dec 2002 07:54:41 -0000
Message-Id: <000c01c2a7fc$0a8dd660$6401000a@zorak>
I'm new to this whole open source thing. I have a lot of stuff. I've
even implemented the api call to the engine. I don't know how I would
submit anything? Do you want it or do you think you have enough stuff?
I have the api implemented (from the way I understood it to work) and
the linearization routines borrowed from a magazine article that solves
for the determinates.
Thanks,
Gabriel
-----Original Message-----
From: Jordan Crouse ####@####.####
Sent: Thursday, December 19, 2002 10:04 AM
To: ####@####.####
Subject: Re: [nanogui] NanoCal
I am currently working on getting this all integrated in to a new patch.
I have
most of the code written, but I haven't had a chance to test it on a
touch
screen based device yet.
Basically, I have borrowed a page from our friends at PicoGUI, and
reasoned
that there is no reason for each individual mouse driver to have to
handle
calibration. Some drivers handle calibration in the kernel, and for
those we
need to make a special exception, but like I have mentioned before, that
is no
longer the accepted way to handle calibration, especially as we move
headlong
into the brave new world of the input subsystem.
So I have implemented a series of "input filters" that sit in the input
data path, and modify the incoming mouse data by rotating and
translating as
required.
I am using Alex's API for bringing in the data from userland, and I am
planning
on bringing in the calibration routines from PicoGUI, which through a
weird
twist of fate were originally borrowed from Bradley LaRonde's tpcal in
Microwindows. The PicoGUI folks have added a few fixed point routines,
and we
now have up to 4 different translation filters we can choose.
This will of course, add more size to the engine, but it will make the
input calibration much more stable, and it will make it much easier for
us to
add new mouse drivers (which in turn will be much smaller).
Anyway, there is no substitute for actually having working code on hand,
so give
me a few days, and I'll post something as soon as I can get the bugs
worked out.
Jordan
