On Sun, May 16, 2010 at 11:25 AM, walt <w41ter@gmail.com> wrote:

On 05/16/2010 08:43 AM, Kevin O'Gorman wrote:


On Sun, May 16, 2010 at 12:11 AM, Mick <michaelkintzios@gmail.com <mailto:michaelkintzios@gmail.com>> wrote:




* *On Sunday 16 May 2010 02:24:10 Kevin O'Gorman wrote:

* * > On Wed, May 12, 2010 at 3:33 PM, Dale <rdalek1967@gmail.com <mailto:rdalek1967@gmail.com>> wrote:


* * > > Mine has xf86-* drivers as well. *OP, do you have your setting in

* * > > make.conf correctly? *Mine looks like this:

* * > >

* * > > INPUT_DEVICES="keyboard mouse evdev"

* * > >

* * > > I do NOT use hal so your settings may need to be different but you do

* * > > need the line tho.

* * > >

* * > > I have INPUT_DEVICES="evdev", and adding either of the others makes X go

* * >

* * > back to not starting at all.



* *That's right, you will also then need to install the appropriate x86 driver;

* *e.g. x11-drivers/xf86-input-mouse





You mean like this, the way it's always been? *Or is there something more specific I have to do?

treat src # eix x11-drivers/xf86-input-mouse

[i] x11-drivers/xf86-input-mouse

* * *Available versions: *1.5.0{tbz2} {debug}

* * *Installed versions: *1.5.0{tbz2}(09:38:05 PM 05/11/2010)(-debug)

* * *Homepage: http://xorg.freedesktop.org/

* * *Description: * * * * X.Org driver for mouse input devices



treat src #



BTW, the most recent boot started X without the mouse working, but these two lines appear in /var/log/Xorg.0.log



line 44-47: (==) |-->Input Device "evdev"

* * * * * * (==) |-->Input Device "<default keyboard>"

* * * * * * (==) The core pointer device wasn't specified explicitly in the layout.

* * * * * * * * * * Using the first mouse device.





line 457: * (==) MACH64(0): Silken mouse enabled



(MACH64 is my video card)

(My mouse is a Microsoft optical with a USB cord that I use with a PS/2 adapter and a KVM switch, which works with Live disks. I have no idea what Silken is)





These are the only lines with the word "mouse" in them.




I just did the experiment of building xorg-server with the hal useflag *off*, and

found that neither keyboard nor mouse worked until I restored the two InputDevice

sections that I commented out when I switched to evdev+hal:



Section "ServerLayout"

* *Identifier * * "X.org Configured"

* *Screen * * *0 *"Screen0" 0 0

* *Option * * * * "AIGLX" * * * *"false"

* *InputDevice * *"Mouse0" "CorePointer" * * *<------ restored these two lines

* *InputDevice * *"Keyboard0" "CoreKeyboard" *<-------

EndSection



Section "InputDevice"

* *Identifier * * "Keyboard0"

* *Driver * * * * "evdev"

* *Option * * * * "Device" "/dev/input/event3"

EndSection



Section "InputDevice"

* *Identifier * * "Mouse0"

* *Driver * * * * "evdev"

* *Option * * * * "Protocol" "auto"

* *Option * * * * "Device" "/dev/input/event4"

* *Option * * * * "Emulate3Buttons" "True"

EndSection



Note those "Device" entries. *I found those devices in /dev/input/by-path/:



$ls -l /dev/input/by-path/

lrwxrwxrwx 1 root root * 9 2010-05-16 10:40 platform-i8042-serio-0-event-kbd -> ../event3

lrwxrwxrwx 1 root root * 9 2010-05-16 10:40 platform-i8042-serio-1-event-mouse -> ../event4

lrwxrwxrwx 1 root root * 9 2010-05-16 10:40 platform-i8042-serio-1-mouse -> ../mouse0



"evdev" uses "event-devices", hence the name. *I don't use the mouse0 device at

all but I'm guessing I would if I used the "mouse" driver instead of "evdev".



Starting with xorg-server-1.8 the mouse and keyboard Inputdevice sections are no

longer needed (not sure about synaptics, though), because the server uses evdev

automatically (no manual configuration like cited above) and ignores hal completely.



Using the evdev driver alone, and xorg-server built without hal, I get this:



(**) Option "CorePointer"

(**) Mouse0: always reports core events

(**) Mouse0: Device: "/dev/input/event4"

(II) Mouse0: Found 9 mouse buttons

(II) Mouse0: Found scroll wheel(s)

(II) Mouse0: Found relative axes

(II) Mouse0: Found x and y relative axes

(II) Mouse0: Configuring as mouse

(**) Option "Emulate3Buttons" "True"

(II) Mouse0: Forcing middle mouse button emulation on.

(II) XINPUT: Adding extended input device "Mouse0" (type: MOUSE)

(**) Mouse0: (accel) keeping acceleration scheme 1

(**) Mouse0: (accel) acceleration profile 0

(II) Mouse0: initialized for relative axes.

(**) Option "CoreKeyboard"

(**) Keyboard0: always reports core events

(**) Keyboard0: Device: "/dev/input/event3"

(II) Keyboard0: Found keys

(II) Keyboard0: Configuring as keyboard

(II) XINPUT: Adding extended input device "Keyboard0" (type: KEYBOARD)

(**) Option "xkb_rules" "evdev"

(**) Option "xkb_model" "evdev"

(**) Option "xkb_layout" "us"








GENIUS.* It worked!** That which has been dead for 2 weeks is risen!

Many thanks.
--
Kevin O'Gorman, PhD

On 05/16/2010 04:39 PM, Kevin O'Gorman wrote:



On Sun, May 16, 2010 at 11:25 AM, walt <w41ter@gmail.com wrote:



I just did the experiment of building xorg-server with the hal useflag *off*, and
found that neither keyboard nor mouse worked until I restored the two InputDevice
sections that I commented out when I switched to evdev+hal:

Section "ServerLayout"
Identifier "X.org Configured"
Screen 0 "Screen0" 0 0
Option "AIGLX" "false"
InputDevice "Mouse0" "CorePointer" <------ restored these two lines
InputDevice "Keyboard0" "CoreKeyboard" <-------
EndSection

Section "InputDevice"
Identifier "Keyboard0"
Driver "evdev"
Option "Device" "/dev/input/event3"
EndSection

Section "InputDevice"
Identifier "Mouse0"
Driver "evdev"
Option "Protocol" "auto"
Option "Device" "/dev/input/event4"
Option "Emulate3Buttons" "True"
EndSection



GENIUS. It worked! That which has been dead for 2 weeks is risen!


Bless you, my son (I always thought the deadline is the third day.)

I could give you that information only after hours of frustrating research
and trial-and-error hacking, most of which I've already forgotten, so it's
a good thing that I commented those lines out instead of deleting them.

Happily for all of us, the xorg devs are well aware of the frustrating user
experience with all of this configuration nonsense, and are making very good
progress towards automating it.