
Hi, I'm having problems with the xinerama support. I follow the installation steps and I run autoreconfigure. The xmonad binary is also linked with libXinerama.so. Howeverer when I create the first window it spans along the two monitors. Appart from that xmonad looks great :D Thanks.

On Sat, Jun 02, 2007 at 12:45:36AM +0200, Ruben Porras wrote:
Hi,
I'm having problems with the xinerama support. I follow the installation steps and I run autoreconfigure. The xmonad binary is also linked with libXinerama.so.
Howeverer when I create the first window it spans along the two monitors. Appart from that xmonad looks great :D
Hmm. The xmonad binary being linked to libXinerama.so would seem to indicate that X11-extras built with Xinerama support enabled. Weird. So let's try to figure this out, eh? Please give the versions that you are running of the following: xmonad X11-extras your X server What does this program output when you run it under X? import Graphics.X11.Xlib import Graphics.X11.Xinerama main = do d <- openDisplay "" xineramaQueryScreens d >>= print getScreenInfo d >>= print Jason Creighton

nahoo82:
Hi,
I'm having problems with the xinerama support. I follow the installation steps and I run autoreconfigure. The xmonad binary is also linked with libXinerama.so.
Can you check that X11-extras found and built with Xinerama support? Perhaps provide the configure logs for X11-extras.
Howeverer when I create the first window it spans along the two monitors. Appart from that xmonad looks great :D
Yes, a sure sign X11-extras didn't build with Xinerama. Usually this is due to the xinerama libs not being found by configure. -- Don

Am Samstag, den 02.06.2007, 13:44 +1000 schrieb Donald Bruce Stewart:
nahoo82:
Hi,
I'm having problems with the xinerama support. I follow the installation steps and I run autoreconfigure. The xmonad binary is also linked with libXinerama.so.
Can you check that X11-extras found and built with Xinerama support? Perhaps provide the configure logs for X11-extras.
I think so, this are the steps that I'm following $ runhaskell Setup.lhs configure ... checking X11/extensions/Xinerama.h usability... yes checking X11/extensions/Xinerama.h presence... yes checking for X11/extensions/Xinerama.h... yes configure: creating ./config.status config.status: creating X11-extras.buildinfo config.status: creating include/config.h config.status: include/config.h is unchanged $ runhaskell Setup.lhs build Preprocessing library X11-extras-0.2... Building X11-extras-0.2... /usr/bin/ar: creating dist/build/libHSX11-extras-0.2.a As a test: $ nm dist/build/libHSX11-extras-0.2.a | grep xinerama 000000a0 D X11zmextraszm0zi2_GraphicsziX11ziXinerama_xineramaIsActive_closure 00000c58 T X11zmextraszm0zi2_GraphicsziX11ziXinerama_xineramaIsActive_info 0000009c D X11zmextraszm0zi2_GraphicsziX11ziXinerama_xineramaQueryExtension_closure 00000c40 T X11zmextraszm0zi2_GraphicsziX11ziXinerama_xineramaQueryExtension_info 00000094 D X11zmextraszm0zi2_GraphicsziX11ziXinerama_xineramaQueryScreens_closure 00000c10 T X11zmextraszm0zi2_GraphicsziX11ziXinerama_xineramaQueryScreens_info 00000098 D X11zmextraszm0zi2_GraphicsziX11ziXinerama_xineramaQueryVersion_closure 00000c28 T X11zmextraszm0zi2_GraphicsziX11ziXinerama_xineramaQueryVersion_info $ runhaskell Setup.lhs install --user Installing: /usr/local/lib/X11-extras-0.2/ghc-6.6.1 & /usr/local/bin X11-extras-0.2... Registering X11-extras-0.2... Reading package info from ".installed-pkg-config" ... done. Saving old package config file... done. Writing new package config file... done. Attached is also the buildlog from xmonad. If it helps, I'm usisg xorg 1.3.0 and I'm activating the second monitor through the command xrandr --output VGA --left-of LVDS Thanks

On Sat, 02 Jun 2007 09:48:55 +0200
Ruben Porras
If it helps, I'm usisg xorg 1.3.0 and I'm activating the second monitor through the command
xrandr --output VGA --left-of LVDS
Hmm, this should work (I run the same command every day). Are you running xrandr before or after xmonad starts? Does xmonad pick up the right screen geometry after restart (mod-q, make sure xmonad is in your path)? Has your server loaded the Xinerama extension? Check with 'xdpyinfo | grep XINERAMA'. Cheers, Spencer Janssen PS - you should try running Jason's little diagnostic script from an earlier post in this thread

Am Samstag, den 02.06.2007, 03:08 -0500 schrieb Spencer Janssen:
On Sat, 02 Jun 2007 09:48:55 +0200 Ruben Porras
wrote: If it helps, I'm usisg xorg 1.3.0 and I'm activating the second monitor through the command
xrandr --output VGA --left-of LVDS
Hmm, this should work (I run the same command every day). Are you running xrandr before or after xmonad starts?
after xrandr. Im using the same .xsession file as for ion3, I'm only changing the last line, from exec ion3 to exec xmonad.
Does xmonad pick up the right screen geometry after restart (mod-q, make sure xmonad is in your path)?
the same problem persist, also after closing all the windows and openining a new xterm. Xmonad is in the path.
Has your server loaded the Xinerama extension? Check with 'xdpyinfo | grep XINERAMA'.
yes, it works well with ion3 + the xinerama module and with metacity.
PS - you should try running Jason's little diagnostic script from an earlier post in this thread
I cannot find the script in the archives, could someone mail it to me? Thanks.

nahoo82:
Am Samstag, den 02.06.2007, 03:08 -0500 schrieb Spencer Janssen:
On Sat, 02 Jun 2007 09:48:55 +0200 Ruben Porras
wrote: If it helps, I'm usisg xorg 1.3.0 and I'm activating the second monitor through the command
xrandr --output VGA --left-of LVDS
Hmm, this should work (I run the same command every day). Are you running xrandr before or after xmonad starts?
after xrandr. Im using the same .xsession file as for ion3, I'm only changing the last line, from exec ion3 to exec xmonad.
Does xmonad pick up the right screen geometry after restart (mod-q, make sure xmonad is in your path)?
the same problem persist, also after closing all the windows and openining a new xterm. Xmonad is in the path.
Has your server loaded the Xinerama extension? Check with 'xdpyinfo | grep XINERAMA'.
yes, it works well with ion3 + the xinerama module and with metacity.
PS - you should try running Jason's little diagnostic script from an earlier post in this thread
I cannot find the script in the archives, could someone mail it to me?
Here's the script, http://www.haskell.org/pipermail/xmonad/2007-June/000494.html -- Don

Am Samstag, den 02.06.2007, 03:08 -0500 schrieb Spencer Janssen:
On Sat, 02 Jun 2007 09:48:55 +0200 Ruben Porras
wrote:
PS - you should try running Jason's little diagnostic script from an earlier post in this thread
The output is: Nothing [Rectangle {rect_x = 0, rect_y = 0, rect_width = 2304, rect_height = 1024}] the version of x11-extras and xmonad, are the darcs from yesterday.

nahoo82:
Am Samstag, den 02.06.2007, 03:08 -0500 schrieb Spencer Janssen:
On Sat, 02 Jun 2007 09:48:55 +0200 Ruben Porras
wrote: PS - you should try running Jason's little diagnostic script from an earlier post in this thread
The output is:
Nothing [Rectangle {rect_x = 0, rect_y = 0, rect_width = 2304, rect_height = 1024}]
the version of x11-extras and xmonad, are the darcs from yesterday.
Ok. We're officially mystified. If X11-extras was correctly built with Xinerama, and Xinerama is enabled on your system, the above should have found a couple of screens -- its a simple call into X after all. Any further information about your system that might be relevant? -- Don

dons@cse.unsw.edu.au (Donald Bruce Stewart) writes: [...]
Ok. We're officially mystified. If X11-extras was correctly built with Xinerama, and Xinerama is enabled on your system, the above should have found a couple of screens -- its a simple call into X after all.
It's probably worth looking at what's actually in Graphics/X11 in the source for X11-extras. I managed somehow to get to the situation where everything was linked to libXinerama, but it was still using the dummy functions that didn't use it. (Though the nm output you gave also looks plausible, so probably this isn't it.) [...]

On Sun, Jun 03, 2007 at 01:06:30PM +0100, Bruce Stephens wrote:
dons@cse.unsw.edu.au (Donald Bruce Stewart) writes:
[...]
Ok. We're officially mystified. If X11-extras was correctly built with Xinerama, and Xinerama is enabled on your system, the above should have found a couple of screens -- its a simple call into X after all.
It's probably worth looking at what's actually in Graphics/X11 in the source for X11-extras. I managed somehow to get to the situation where everything was linked to libXinerama, but it was still using the dummy functions that didn't use it. (Though the nm output you gave also looks plausible, so probably this isn't it.)
If you managed to do that, it's definitely a bug. Are you able to reproduce this behavior? Jason Creighton

Jason Creighton
On Sun, Jun 03, 2007 at 01:06:30PM +0100, Bruce Stephens wrote:
[...]
It's probably worth looking at what's actually in Graphics/X11 in the source for X11-extras. I managed somehow to get to the situation where everything was linked to libXinerama, but it was still using the dummy functions that didn't use it. (Though the nm output you gave also looks plausible, so probably this isn't it.)
If you managed to do that, it's definitely a bug. Are you able to reproduce this behavior?
Almost certainly not. It was quite possibly user error: I was messing about quite a bit, trying to work out what was wrong. My guess is it won't happen to anyone following the instructions properly (in particular, running autoconf), or just using more recent tarballs or darcs downloads.
participants (5)
-
Bruce Stephens
-
dons@cse.unsw.edu.au
-
Jason Creighton
-
Ruben Porras
-
Spencer Janssen