linuxconsole-dev

Mostly successful with Backstreet Ruby but lost VT1

[Sam <ukh...@sn...>]

Hi all,

I just spent the whole day getting Backstreet Ruby working on my system,
mostly successfully.  I'll report my experience then ask for help
regaining my virtual terminal #1 (or virtual console; I'm not clear on
the terminology.  I mean the thing I used to get when I pressed
Ctrl-Alt-F1).

I'm running Debian Woody and my project is a digital picture frame
(read: "unwanted flat panel display gotten for cheap") near my
computer.  The display is now in a nice wooden frame, hanging on the
wall, so it doesn't look like a computer screen, and it's attached to my
PC by a 15' VGA cable to a cheap PCI GeForce2.  When I boot my computer
the picture frame display comes up with its own X server then one of my
digital photographs is chosen at random and displayed.  It's totally
independent of gdm and whether I'm logged in or not (thank you,
Backstreet Ruby).

I followed the directions in the XFree Local Multi-User HOWTO, which was
very helpful.  Thank you, Svetoslav Slavtchev.

I first tried using Andreas Schuldei's binary kernel package
(kernel-image-2.4.20-backstreet-ruby-p4_10.00.Custom_i386.deb) but when
I booted up my PS/2 keyboard didn't work.  I then tried to patch a
Debian 2.4.21 kernel but the patch failed, so I downloaded a vanilla
2.4.21 kernel which patched successfully.  When configuring the kernel I
tried to set "Input Devices/Keyboards/AT keyboard support" to Y so I
wouldn't have to deal with loading the module (as recommended by the
instructions).  The kernel configurator wouldn't let me; it insisted on
it being set to 'M'.  I tried booting this kernel and it failed in the
same way (no keyboard).  I then reconfigured the kernel and turned on
"Prompt for development and/or incomplete code/drivers".  Once that was
on I was allowed to set "AT keyboard support" to 'Y' instead of 'M'. 
Booting the patched kernel now works with my keyboard.  I have to add
"dumbcon=1" to my kernel command line (maybe because I've got only one
keyboard plugged in).

I've got an AGP GeForce3 and a PCI GeForce2 in my system and I'm using
nVidia's binary drivers.  I first tried the kernel-only solution but my
system crashed with bizarre patterns on my screens.  Fine, so I switched
to Andreas Schuldei's binaries for the patched XFree86 3.0 and these
worked, but only if I probe my GeForce2 first, i.e.:

        /usr/X11R6/bin/X -probeonly -layout Small
        /usr/X11R6/bin/X :1 -layout Small -prefbusid 2:7:0 vt8 &
        
after which I load gdm, which starts the other X server on the AGP
card.  It seems odd that I have to probe the PCI card right before
starting the PCI card (this isn't about which card comes first), but
that's what worked.

I made some small changes to a program called "qiv" to behave itself
better when it displays the photograph (such as hiding the mouse
pointer).  Yay, open source!

Here's the part where I need help.  Now when I press Ctrl-Alt-F1, the
main display switches back to the text-only console BUT it's "greyed
out"!  Both the foreground text and background look grey--the text is
darker than it should be and the background is lighter.  I can read text
on it but I cannot type anything to it.  This alarms me because I like
dropping out of X once and a while and getting a command line.  Any
ideas?

During my experimenting I did try having gdm launch both X servers as
explained in the HOWTO.  When I did this, instead of launching one
myself, I was able to use VT1 but suffered the garbage font problem. 
Typing "consolechars -d" fixed it immediately.  I don't care about this,
because I don't want gdm to launch both servers, but I'm noting it
because I didn't have the "greyed-out" VT1 problem in that setup.

A big thanks to those people who write the code and documentation which
makes this possible.

Sam

listing bruby debs with apt-get.org

[hugo v. <hug...@ca...>]

Svetljo showed me a deb depository for the 4.3.0
X patched stuff. Looks like a lot of work went
into that. Should that not be listed with
http://www.apt-get.org/ ?
I got there because I was unsure about the
sources list entry:
deb /http://www.schuldei.org/debian/bruby/ ./ 

but that's how you do it alright.
Hugo.


The following signature is automatically
added by the web email service I use,
which is beyond my control:
=======================================
Help the planet each day! It's free and easy:
http://www.Care2.com/dailyaction/

Re: (no subject)

[Svetoslav S. <ga...@st...>]

Quoting hugo vanwoerkom <hug...@ca...>:

> I have a few questions on running unpatched X
> versus running patched X with the ruby patched
> 2.4.21 kernel:
> 
> It says in the "Quick XFree" documentation:
> 
> Set up it with command:
> $ echo "1" > /proc/bus/pci/hackvideo
>  Ordinarily XFree86 for i386 architecture use
> native built in XFree86 functions to steer PCI.
> Please add necessary lines in /etc/X11/XF86Config
> file, which tell XFree86 to use Linux /proc
> filesystem interface:
> Section "ServerFlags"
> Option "PciOsConfig" "1"
> EndSection
>  After this XFree86 is ready for multiple
> independed heads without recopiling from source.
> Of course i recommend usage of "2" (idea by Kim
> "kill" Lilliestiena) XFree86 patch.
> 
> 1. I cannot find any info on that serverflags
> option, is it described anywhere?

i didn't look in to it, but may be you can find smth on http://www.xfree.org

> 2. Do you have to initialize
> /proc/bus/pci/hackvideo every time with a "1"

yes, you have to.
you can add it to some init script.

> before starting X?
> 3. Why is patch "2" preferred to using an
> unpatched X (in my case Debian X 4.2.1)?
>      I cannot run the patched binary from
> http://startx.times.lv/XFree86-patched.bz2 
> because I get an "rejected from local host"     
>    error and I suspect that it is because it is
> not compiled with gcc 3.3 while the X that I have is.

3.1 because it prevent lock-ups, and work better for the "majority" of video
cards. if you check the video compatibilty list in my HOWTO,
for most( all AFAIK) cards in "works, but with some glitches" 
patch "2" resolves the problems.

http://tldp.org/HOWTO/XFree-Local-multi-user-HOWTO/app_vid_comp.html

3.2 you can try the binaries from Andreas
(but i think they are compiled with older gcc-3.x, may be sid for 2 months ago)

http://www.schuldei.org/debian/bruby, or as apt repository 
"deb http://www.schuldei.org/debian/bruby ./ ".

best,

svetljo

(no subject)

[hugo v. <hug...@ca...>]

I have a few questions on running unpatched X
versus running patched X with the ruby patched
2.4.21 kernel:

It says in the "Quick XFree" documentation:

Set up it with command:
$ echo "1" > /proc/bus/pci/hackvideo
 Ordinarily XFree86 for i386 architecture use
native built in XFree86 functions to steer PCI.
Please add necessary lines in /etc/X11/XF86Config
file, which tell XFree86 to use Linux /proc
filesystem interface:
Section "ServerFlags"
Option "PciOsConfig" "1"
EndSection
 After this XFree86 is ready for multiple
independed heads without recopiling from source.
Of course i recommend usage of "2" (idea by Kim
"kill" Lilliestiena) XFree86 patch.

1. I cannot find any info on that serverflags
option, is it described anywhere?
2. Do you have to initialize
/proc/bus/pci/hackvideo every time with a "1"
before starting X?
3. Why is patch "2" preferred to using an
unpatched X (in my case Debian X 4.2.1)?
     I cannot run the patched binary from
http://startx.times.lv/XFree86-patched.bz2 
because I get an "rejected from local host"     
   error and I suspect that it is because it is
not compiled with gcc 3.3 while the X that I have is.

Thanks!

Hugo. 


The following signature is automatically
added by the web email service I use,
which is beyond my control:
=======================================
Help the planet each day! It's free and easy:
http://www.Care2.com/dailyaction/

Re: Q. about linux 2.4.20 w. bruby II

[<Aiv...@un...>]

>I forgot to say:
>I am under the impression because: "Since
>15-MAY-2003 kernel patch You need not to
>recompile XFree86." that I can run with the 4.2.0
>Debian Xfree86, that is what this is.

This is a political decision.

Kernel recompiling is relative easy and stock Linux
kernel runs fine with any distro.

But distro specific xf86 recompling is horrible, may
be unsuccessful (Mandrake 9.1). Stock xf86 binaries
mostly is incompatible with distro other programs.
Distro must be older than xf86 release.

As a result i push PCI filter into kernel.
This one oppress xf86 folk, because xf86 has nice
assembler based PCI steering code for i386 arch.
Fortunately linux PCI steering via /proc works
flawless and we can use unpatched xf86 binaries.

PCI steering is quite hard code and i hope nobody
touch them in distro specific xf86 release.
So this is safe.

Unsafe things is closed source xf86 drivers of chipset.
In case NVidia xf86 driver load single kernel module
which manage all videocards on the system. Thanks god
it works, but how long.

And without patch of xf86 most AGP servers restart lead
system hang up. Why?

Aivils Stoss

Re: Q. about linux 2.4.20 w. ruby backstreet patch

[Andreas S. <an...@sc...>]

* hugo vanwoerkom (hug...@ca...) [030709 02:44]:
> The nearest I can figure, since I made no changes
> to the kernel or the patch, it has to be my
> distro. 

i am running on debian (sid), too. i can compfort you: it is not
the distro`s problem, either.

your description does not help to identify the problem, it is too
high level.

My guess is that you forgot to load the modules you needed.

Q. about linux 2.4.20 w. bruby II

[hugo v. <hug...@ca...>]

I forgot to say:
I am under the impression because: "Since
15-MAY-2003 kernel patch You need not to
recompile XFree86." that I can run with the 4.2.0
Debian Xfree86, that is what this is.
Thanks again.
Hugo Vanwoerkom


The following signature is automatically
added by the web email service I use,
which is beyond my control:
=======================================
Help the planet each day! It's free and easy:
http://www.Care2.com/dailyaction/

Q. about linux 2.4.20 w. ruby backstreet patch

[hugo v. <hug...@ca...>]

I also asked this directly of Aivils Stoss, but
may be it is more appropriate here:

I am trying the bruby patch on a Debian "sarge"
(=next official Debian system) system. It has a
package ("kernel-package") that makes building a
new kernel-image absolutely painless.
Unfortunately the patch fails on a Debian
Kernel-source of 2.4.20. I have not investigated
why, but the reason no doubt is that Debian
patches the kernel too. I don't know exactly with
what and why.

So I downloaded the 2.4.20 kernel from the kernel
ftp and of course the patch applies perfectly.
I did not make any changes to the kernel other
than the patch and compiled it.

It boots fine. There are no messages of note.
However, X does not run, complaining about the
mouse device.
If I include mousedev.o it starts, pointing X at
/dev/psaux, but the mousecursor won't move.
X will fail when I point it at /dev/input/mouse0.
The failure is "can't open <...> device does not
exist".

The nearest I can figure, since I made no changes
to the kernel or the patch, it has to be my
distro. and/or my implementation of the patch.
The /dev/input dir. already existed, with its
entries mouse0.. etc.. The link between
/dev/mouse and /dev/input/mice also exists.

To eliminate more, I ran the plain Linux 2.4.20
kernel without the patch and that also runs X
fine, either with or without mousedev, pointing X
at /dev/psaux.

I am looking for a suggestion on where to look
next on getting this great idea to run.
Unfortunately I only have Debian distros, 3 of them.

I also put the /etc/hotplug scripts in the dir.
but I don't think that has anything to do with it.

BTW, mouse is PS/2, kbd is AT, video card is
NVidia TNT AGP with 32MB. Nvidia driver is
installed and runs fine on a non-bruby system.

Thanks!

Hugo Vanwoerkom



The following signature is automatically
added by the web email service I use,
which is beyond my control:
=======================================
Help the planet each day! It's free and easy:
http://www.Care2.com/dailyaction/

SAK v. ruby

[Brad M. <bmi...@ue...>]

hi

has anyone tried out the secure attention key? does it kill every x 
server on your multihead box?

i was just thinking it would be nice to make an app that provided the 
equivalent of "fast user switching" but something like that should work 
with SAK and should also know about multihead (if nothing else, to tell 
you to go sit in a different seat if you want to continue with a current 
session that's on another head)

brad

Re: Compiling Backstreet Ruby with Linux 2.4.21

[Svetoslav S. <ga...@st...>]

Quoting Gunter Ohrner <G.O...@po...>:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Hi!
> 
> I just tried to compile a bruby-2.4.21-20030615.diff.bz2 patched 
> 2.1.21-vanilla and always get the following error during linking:
> 
> ld -m elf_i386 -T /var/fspool/daten1/src/linux-2.4.21/arch/i386/vmlinux.lds
> -e 
> stext arch/i386/kernel/head.o arch/i386/kernel/init_task.o init/main.o 
> init/version.o init/do_mounts.o \
>         --start-group \
>         arch/i386/kernel/kernel.o arch/i386/mm/mm.o kernel/kernel.o mm/mm.o
> 
> fs/fs.o ipc/ipc.o \
>          drivers/char/char.o drivers/block/block.o drivers/misc/misc.o 
> drivers/net/net.o drivers/ide/idedriver.o drivers/cdrom/driver.o 
> drivers/sound/sounddrivers.o drivers/pci/driver.o drivers/video/video.o 
> drivers/media/media.o drivers/input/serio/seriodrv.o \
>         net/network.o \
>         /var/fspool/daten1/src/linux-2.4.21/arch/i386/lib/lib.a 
> /var/fspool/daten1/src/linux-2.4.21/lib/lib.a 
> /var/fspool/daten1/src/linux-2.4.21/arch/i386/lib/lib.a \
>         --end-group \
>         -o vmlinux
> drivers/char/char.o(.text+0x12055): In function `kd_nosound':
> : undefined reference to `input_event'
> drivers/char/char.o(.text+0x1206e): In function `kd_nosound':
> : undefined reference to `input_event'
> drivers/char/char.o(.text+0x12cc4): In function `kbd_bh':
> : undefined reference to `input_event'
> drivers/char/char.o(.text+0x12cd9): In function `kbd_bh':
> : undefined reference to `input_event'
> drivers/char/char.o(.text+0x12cf2): In function `kbd_bh':
> : undefined reference to `input_event'
> drivers/char/char.o(.text+0x13228): In function `kbd_connect':
> : undefined reference to `input_open_device'
> drivers/char/char.o(.text+0x1325d): In function `kbd_disconnect':
> : undefined reference to `input_close_device'
> drivers/char/char.o(.text+0x13386): In function `write_kbd_phys':
> : undefined reference to `input_find_handle'
> drivers/char/char.o(.text.init+0xa7a): In function `kbd_init':
> : undefined reference to `input_register_handler'
> make[1]: *** [vmlinux] Error 1
> make[1]: Leaving directory `/var/fspool/daten1/src/linux-2.4.21'
> make: *** [stamp-build] Error 2
> 
> Excerpt of my .config:
> 
> # Input device support
> #
> CONFIG_INPUT=m

this one should be :
CONFIG_INPUT=y

this is one of the most hated topics on lkml about subjects like:
2.5.X do not compile, 2.5.X doesn't boot ( did you set VT_CONSOLE=y )....
they were tring to figure out how to set it to =y and hide most of the option
behind smth like "Advanced Input Configuration" in order to have smth similar
to the 2.4 "look and feel"
but no solution yet.

best,

svetljo

ruby 2.4.21 2.5.70

[<Aiv...@un...>]

Hi all!

     Seems 2.4.21 works for me and 2.5.70 to
http://startx.times.lv/bruby-2.4.20-20030615.diff.bz2
http://startx.times.lv/ruby-2.5.70-20030615.diff.bz2

     You can use unpatched xf86 for locale Xservers.
Write into XF86Config
Section "ServerFlags"
Option "PciOsConfig" "1"
EndSection

and
echo "1">/proc/bus/pci/hackvideo

This command order kernel filter xf86 PCI-steering.
xf86 cannot disable neighbour video adapter.
This one works for me anytime with any X servers set.

AGP server restat/stop may lead system hang up :(
Only patched xf86 works nice.

Current xf86 patch hide video adapters and alllow
manipulation with one video adapter per Xserver only
http://startx.times.lv/xf86-430-prefbusid.diff

This work weird with Nvidia closed source xf86 drivers, but
it works. Tested Nvidia 1.0-4349. 1-st must be started Xserver
on 0:11:0 (PCI tag) then 0:9:0 and then 1:0:0 adapter without
adapter hidind.
If user will switch to text-mode-console 1-st must be started
xf86 on primary video device - 1:0:0 (save VGA fonts)
Nvidia kernel module remeber started adapters and then
allow start/stop in any order. I doe
X -probeonly -layout X2
X -probeonly -layout X1
startx -- :0 -layout X0 -prefbusid 1:0:0 vt7
startx -- :1 -layout X1 -prefbusid 0:9:0 vt8
startx -- :2 -layout X2 -prefbusid 0:11:0 vt16

This commandeset allow me start all 3 X servers in right order.
Unfortunately VGA fonts are allways corrupted and not allways
correct restored. I have not preform deeply VGA font save/restore
analizyng in xf86. Result - sometimes text-mode-console works.
I recieve recommendation to do "consolechars -d" after switch to
text mode , but this command works for only after all X servers stopping.


Aivils Stoss

Re: is Ruby dead ? lkml: Real multi-user linux

[Svetoslav S. <ga...@st...>]

Quoting James Simmons <jsi...@in...>:

> 
> > I was wonderin what does this post by James Simmons means.
> > 
> > James did you abondon ruby in order to bring something comersial ?
> > there are no changes in ruby for a more then 4 monts,
> > and ruby is hardly useable/ compilable, and you are talking about
> > a production status?
> 
> The 2.5.X version is broken. Instead I have beening working on making
> the fbdev layer stable. The api I developed while working on ruby. The 
> 2.4.X version is quite stable. Alot of the framebuffer drivers could be 
> dropped onto the 2.4.X ruby kernel and it would work. We do have a few 
> changes in the framebuffer api that happened but that wouldn't take much 
> to port to ruby 2.4.X.

ruby 2.4.X ?
does it exist?
there is nothing about ruby-2.4 on the linuxconsole web site, mailing list.
are there sources available anywhere?
does framebuffer console work in ruby-2.4

> > did i misunderstood something, or you have droped ruby for something
> comersial?
> 
> No. I'm still working on ruby. All my 2.5.X is geared toward ruby.

where can one get your latest work?
linuxconsole's cvs and bitkeeper have no changes since more then 4 months,
there where no diffs posted to the mailing list too.

Svetljo 

Re: is Ruby dead ? lkml: Real multi-user linux

[James S. <jsi...@in...>]

> I was wonderin what does this post by James Simmons means.
> 
> James did you abondon ruby in order to bring something comersial ?
> there are no changes in ruby for a more then 4 monts,
> and ruby is hardly useable/ compilable, and you are talking about
> a production status?

The 2.5.X version is broken. Instead I have beening working on making
the fbdev layer stable. The api I developed while working on ruby. The 
2.4.X version is quite stable. Alot of the framebuffer drivers could be 
dropped onto the 2.4.X ruby kernel and it would work. We do have a few 
changes in the framebuffer api that happened but that wouldn't take much 
to port to ruby 2.4.X.

> did i misunderstood something, or you have droped ruby for something comersial?

No. I'm still working on ruby. All my 2.5.X is geared toward ruby.
 

is Ruby dead ? lkml: Real multi-user linux

[Svetoslav S. <ga...@st...>]

I was wonderin what does this post by James Simmons means.


James did you abondon ruby in order to bring something comersial ?
there are no changes in ruby for a more then 4 monts,
and ruby is hardly useable/ compilable, and you are talking about
a production status?

did i misunderstood something, or you have droped ruby for something comersial?

if i missed smth please excuse me, but please give some sort of explanation.

best 

Svetljo

<!---sight very sad --->
http://marc.theaimsgroup.com/?t=105546067000002&r=1&w=2

<snip>
> solution, but these projects were orphaned. 

Not true. Your project has just finished it first stable 
release. Its just we are going commerical with this product.
We even have a modified distro to handle such a system. 
Yes you have to do special modifications to make 
multi-desktop systems work. 

> My motivation is simply a priva
..............

   Now that we have a working product we can go to the 
next stage. Finding customers to sell it to. This weekend
I'm getting together with a few sales guys to discuss this.
   As for the potential of this. Its large. With Gnome/KDE
and OpenOffice running for 3,4 or more people at the same
time out of one box will be a enormous cost reducer for 
any business. You coudl even do Wine to run MS apps but I 
don't know what the legal ramifactions of that would be.
As a business we wouldn't do it out of fear of being sued.
    The next stage will be non PC boards supporting more than
one graphics display output. Every now and then you see such a
board. I seen a 8 graphics chip board with 8 video outputs. 
Embedded devices that open like a wallet with a display on each 
side. Multiuser game consoles would be really cool. Its going to 
be a fun future.

<snip>
-- 

Re: joystick whitout axis

[Brad H. <bh...@bi...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Sat, 7 Jun 2003 01:21 am, Juan Pablo wrote:
> > If the device doesn't have an absolute axis, would it be easier to use
> > the input layer and event output? What do you need in userspace?
> >
> > Brad
>
> In userspace there's no problem with the joystick having a dummy axis, I
> use it that way at the moment. The driver is for connecting a nintendo's
> powerpad via the parallel port, and use it as a joystick with 12 buttons
> and no axis, as I said there's no problem with having an axis that's never
> used. BTW do you want me to send you the driver for inclusion in the
> kernel?
Don't send it to me. It needs to go to Vojtech.

Brad
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQE+4TYrW6pHgIdAuOMRAmwoAKCYAXqd6wH/Ieq4Zmyj8uRgqvZCfACcCB7/
Xcf/1/yfGb9qEsWXrMNN0p0=
=+6Ps
-----END PGP SIGNATURE-----

Re: joystick whitout axis

[Juan P. <jp...@ma...>]

> If the device doesn't have an absolute axis, would it be easier to use the 
> input layer and event output? What do you need in userspace?
> 
> Brad

In userspace there's no problem with the joystick having a dummy axis, I use it that way at the moment.
The driver is for connecting a nintendo's powerpad via the parallel port, and use it as a joystick with 12 buttons and no axis, as I said there's no problem with having an axis that's never used.
BTW do you want me to send you the driver for inclusion in the kernel?

Re: joystick whitout axis

[Brad H. <bh...@bi...>]

=2D----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 6 Jun 2003 05:21 am, Juan Pablo Sousa wrote:
> hi,
>
> is it posible to make a driver for a joystick that only has buttons (a
> mat)? I tried with the 2.4 kernel series and didn't succeed, is it
> designed taht way or was I doing something wrong. The quick solution was
> to initialize 2 axis for the joystick that weren't used.
I take it that you want to provide a low level driver that talks to the inp=
ut=20
layer, and outputs via the normal joydev driver to userspace. If so, then=20
there is no way. You must have at least one absolute axis - see below for=20
what the criteria is for a device to be associated with the joydev output:

See below for the description of what the=20
static struct input_device_id joydev_ids[] =3D {
        {
                .flags =3D INPUT_DEVICE_ID_MATCH_EVBIT |=20
INPUT_DEVICE_ID_MATCH_ABSBIT,
                .evbit =3D { BIT(EV_ABS) },
                .absbit =3D { BIT(ABS_X) },
        },
        {
                .flags =3D INPUT_DEVICE_ID_MATCH_EVBIT |=20
INPUT_DEVICE_ID_MATCH_ABSBIT,
                .evbit =3D { BIT(EV_ABS) },
                .absbit =3D { BIT(ABS_WHEEL) },
        },
        {
                .flags =3D INPUT_DEVICE_ID_MATCH_EVBIT |=20
INPUT_DEVICE_ID_MATCH_ABSBIT,
                .evbit =3D { BIT(EV_ABS) },
                .absbit =3D { BIT(ABS_THROTTLE) },
        },
        { },    /* Terminating entry */
};

If the device doesn't have an absolute axis, would it be easier to use the=
=20
input layer and event output? What do you need in userspace?

Brad
=2D----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQE+4DwYW6pHgIdAuOMRArzlAJsFAhz4TLd8xVBRFdWkTvtWuv7+xgCdFVWi
RA1fjGEOoeKotcCJRjtx2SI=3D
=3DPTJC
=2D----END PGP SIGNATURE-----

joystick whitout axis

[Juan P. S. <jp...@ma...>]

hi,

is it posible to make a driver for a joystick that only has buttons (a 
mat)? I tried with the 2.4 kernel series and didn't succeed, is it 
designed taht way or was I doing something wrong. The quick solution was 
to initialize 2 axis for the joystick that weren't used.

Re: latest ruby sources

[Svetoslav S. <ga...@st...>]

Quoting Rogelio Serrano <ker...@ho...>:

> where can we download the latest linuxconsole sources?
> TIA
> 

cvs : pserver:ano...@cv...:/cvsroot/linuxconsole co ruby
BitKeeper: http://linuxconsole.bkbits.net:8080/stable

but i think both of them are at least 2-3 months old :(

Aivils has a patch against 2.5.66 ( see http://startx.times.lv/ -> download )

best,

svetljo

latest ruby sources

[Rogelio S. <ker...@ho...>]

where can we download the latest linuxconsole sources?
TIA

_________________________________________________________________
Protect your PC - get McAfee.com VirusScan Online 
http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

powerpad kernel module

[Juan P. <jp...@ma...>]

hi,

I hacked Vojtech's gamecon.c module to connect a Nintendo powerpad and
it worked fine with 2.4.x kernels, have never tried 2.5
There's a catch tough: the powerpad uses two data wires, and gamecon is
designed with one wire per controller in mind, so I think it would be
easier to have a diferent module for it. The problem will be if someone
wants a powerpad and other gamepads in the same parallel port.

Do I send the file to the list? (~11Kb)

XFree Local multi-user HOWTO updated & published on tldp.org

[Svetoslav S. <ga...@st...>]

So the first official version is available,
and i'd like to ask you for comments, improvments, ideas :)

http://www.tldp.org/HOWTO/XFree-Local-multi-user-HOWTO/

Andreas: 
The kernel-images (except the p4) and .config are still missing. Comments?
And i think i lost somehow your mails about the German translation of Aivils web
site.( :-) Comments? )

best,

svetljo
-- 

Re: (BUGREPORT)XF-4.3 + event support patches

[Svetoslav S. <ga...@st...>]

Quoting Svetoslav Slavtchev <ga...@st...>:

> Quoting "Zephaniah E. Hull" <wa...@ba...>:
> > Ok, I'd like you to try something for the hotplugging problem, put the
> > actual echo to the fifo in the background, & should do the trick nicely.
> 
> so you mean:
> not -> .... {VAL_SEP}${ACTION}" >> ${dev} ' &
> but -> .... {VAL_SEP}${ACTION}" >> ${dev} & '
> is this correct?
> 
> no changes under backstreet ruby, will check with latest BK in an
> hour(compile
> time :) )

:( 
same with the older 2.5.69 (current BK seems brocken :( )
 
svetljo

Re: (BUGREPORT)XF-4.3 + event support patches

[Svetoslav S. <ga...@st...>]

Quoting Juan Pablo Sousa <jp...@ma...>:

> Svetoslav Slavtchev wrote:
> 
> >the X server detects the devices on start/restart, but repluging a mouse
> results
> >in loosing the mouse until restart of the X server.
> >( still have to check what hapens if keyboard is replugged )
> >
> >  
> >
> I've noticed that switching to console and back into X makes X find 
> repluged mice, I use gpm so maybe that has something to do with it.

I can confirm this when gpm is not used.
But doesn't X re-initialize the devices when one switches from VT to X ?
and this is really not an option for some cases when one can not switch to VT:
- X running on dumbcon (using bruby)
- loosing VT console with starting X (not happy owner of Matrox card :( )
( - what if keyboard is replugged, how would one switch to VT and back ? )
 
svetljo

Re: (BUGREPORT)XF-4.3 + event support patches

[Svetoslav S. <ga...@st...>]

Quoting "Zephaniah E. Hull" <wa...@ba...>:

> On Fri, May 09, 2003 at 01:10:02PM +0200, Svetoslav Slavtchev wrote:
> > Hi,
> > 
> > Quoting Svetoslav Slavtchev <ga...@st...>:
> > > Hi again,
> > > 
> > > some more info .
> > > 
> > > this is under both Mandrake-9.1 and Debian Sid.
> > 
> > both problems exist under linux-2.5.69 too :(
> > 
> > (of course only the first X server was started)
> > 
> > some one with ideas for debuging ?
> 
> Ok, I'd like you to try something for the hotplugging problem, put the
> actual echo to the fifo in the background, & should do the trick nicely.

so you mean:
not -> .... {VAL_SEP}${ACTION}" >> ${dev} ' &
but -> .... {VAL_SEP}${ACTION}" >> ${dev} & '
is this correct?

no changes under backstreet ruby, will check with latest BK in an hour(compile
time :) )

> That is about the only thing that comes to mind, the X side /should/
> work, and does for me. :/

the X server detects the devices on start/restart, but repluging a mouse results
in loosing the mouse until restart of the X server.
( still have to check what hapens if keyboard is replugged )


best,
 
svetljo