[Xquartz-dev] How to revert to native X11 builtins from latest X11 release?
Jeremy Huddleston
jeremyhu at apple.com
Mon Jul 19 11:19:58 PDT 2010
/opt/X11/lib/X11/xinit/xinitrc.d/98-user.sh (and the corresponding SL one in /usr/...) are responsible for executing the scripts that you provide in ~/.xinitrc.d
What file did you edit? I do not know what you mean by "98-user-sh to reference /usr/X11 rather then /opt/X11/lib/xinitrc.d/98-usr.sh" since neither of the 98-user.sh files reference anything in /opt/X11 or /usr/X11:
$ cat /opt/X11/lib/X11/xinit/xinitrc.d/98-user.sh
if [ -d "${HOME}/.xinitrc.d" ] ; then
for f in "${HOME}"/.xinitrc.d/*.sh ; do
[ -x "$f" ] && . "$f"
done
unset f
fi
fink links its applications against X11 libraries in /usr/X11/lib ... my understanding from your earlier emails is that you changed something such that they would build against libraries in /opt/X11/lib ... is that correct or not? That was my understanding from this quote in one of the fink list emails:
> You'd also need to revert your /sw/bin/xinitrc.sh to have
>
> x_prefix="/usr/X11R6"
On Jul 19, 2010, at 10:58, Richard Miles wrote:
> Jeremy, I changed no links. I just changed 98-user-sh to reference /usr/X11 rather then /opt/X11/lib/xinitrc.d/98-usr.sh
> I changed /sw/etc/xinitrc-last-hook to reference /usr/X11/lib/xinitrc.d/98-user-sh.
>
> On Jul 19, 2010, at 9:35 AM, Jeremy Huddleston wrote:
>
>> Right. What makes your think that this has anything to do with my software? Additionally, did you rebuild your fink projects after changing which X11 you linked against?
>>
>>
>> On Jul 18, 2010, at 22:47, Richard Miles wrote:
>>
>>> Here are the processes running in gone when I just execute X11.app:
>>> 37619 ?? S 0:00.49 /sw/lib/gvfs/gvfsd
>>> 37667 ?? S 0:00.02 /sw/lib/gvfs/gvfs-hal-volume-monitor
>>> 37686 ?? S 0:00.66 /sw/lib/gvfs/gvfsd-trash --spawner :1.8 /org/gtk/gvf
>>> 37695 ?? S 0:00.05 /sw/lib/gvfs/gvfsd-burn --spawner :1.8 /org/gtk/gvfs
>>> 37814 ?? S 0:00.06 /sw/lib/gvfs/gvfsd-metadata
>>> 44609 ?? R 13:09.12 /Applications/Utilities/X11.app/Contents/MacOS/X11.b
>>> 44659 ?? S 0:00.02 /bin/sh /usr/X11/bin/startx
>>> 44707 ?? S 0:00.01 /usr/X11/bin/xinit /usr/X11/lib/X11/xinit/xinitrc --
>>> 44709 ?? S 0:00.01 /usr/X11/bin/X :6 -nolisten tcp -auth /Users/rmiles/
>>> 44799 ?? S 0:00.25 /sw/bin/gnome-session
>>> 44882 ?? Z 0:00.00 (metacity)
>>> 44887 ?? S 0:00.21 /sw/lib/gconf2/gconfd-2
>>> 44894 ?? Ss 0:00.56 /sw/lib/gnome-settings-daemon/gnome-settings-daemon
>>> 44897 ?? S 0:00.00 gnome-keyring-daemon --start
>>> 44898 ?? Ss 0:00.29 seahorse-daemon
>>> 44908 ?? S 0:00.82 gnome-panel --sm-config-prefix /gnome-panel-QGc9uJ/
>>> 44925 ?? S 0:00.96 nautilus --sm-client-id 1066dda4841f6df2e01277267988
>>> 44927 ?? Ss 0:00.07 /sw/lib/libbonobo2/bonobo-activation-server --ac-act
>>> 44928 ?? S 0:00.90 gnome-terminal --sm-client-id 10db21541bd3cada112790
>>> 44936 ?? S 0:00.14 /sw/lib/gnome-panel/wnck-applet --oaf-activate-iid=O
>>> 44938 ?? R 0:01.31 /sw/lib/gnome-applets/geyes_applet2 --oaf-activate-i
>>> 44940 ?? S 0:00.21 /sw/lib/gnome-applets/gweather-applet-2 --oaf-activa
>>> 44942 ?? S 0:00.18 /sw/lib/gnome-utils/gnome-dictionary-applet --oaf-ac
>>> 44944 ?? S 0:00.36 /sw/lib/glunarclock/glunarclock-applet-2 --oaf-activ
>>> 44964 ?? S 0:00.01 gnome-pty-helper
>>> 45036 ?? S 0:00.98 /sw/lib/gnome-panel/clock-applet --oaf-activate-iid=
>>> 45038 ?? S 0:00.06 /sw/lib/gnome-panel/notification-area-applet --oaf-a
>>> 45177 ?? Ss 0:00.03 /usr/libexec/taskgated -p -s
>>> 36991 s000 Ss 0:00.16 login -pf rmiles
>>> 37002 s000 S 0:00.33 -bash
>>> 45179 s000 R+ 0:00.00 ps ax
>>> 36992 s001 Ss 0:00.03 login -pfq rmiles /bin/bash --login
>>> 37001 s001 S+ 0:00.21 -bash --login
>>> 36994 s002 Ss 0:00.16 login -pf rmiles
>>> 37006 s002 S+ 0:00.20 -bash
>>> 44965 s003 Ss+ 0:00.20 -bash
>>> 44973 s004 Ss+ 0:00.20 -bash
>>>
>>> You can see that metacity is not in operation and does not interact with gnome wind borders.
>>>
>>>
>>> On Jul 18, 2010, at 9:50 PM, Jeremy Huddleston wrote:
>>>
>>>> Wow. This is all kinds of wrong.
>>>>
>>>> 1) Just put the script in ~/.xinitrc.d (don't have another script put it there)
>>>> 2) Make sure it is executable
>>>> 3) Don't use startx to start X11... just start it.
>>>>
>>>> If you want to have differnet "types" of X11, then you should conditionalize the script in ~/.xinitrc.d
>>>>
>>>> On Jul 18, 2010, at 21:31, Richard Miles wrote:
>>>>
>>>>> The script is :
>>>>> #!bin/sh
>>>>> . /sw/bin/init.sh
>>>>> export G_FILENAME_ENCODING=@locale
>>>>> icewm --gnome &
>>>>> #metacity &
>>>>> exec gnome-session
>>>>> This is put in ~/.xinitrc.d by this script:
>>>>> #!/bin/sh
>>>>> rm ~/.xinitrc.d/*
>>>>> #cp ~/bin/zz-metacity.sh ~/.xinitrc.d
>>>>> cp ~/bin/zzz-gnome-session.sh ~/.xinitrc.d
>>>>> cd
>>>>> wait 10
>>>>> startx&
>>>>> You say rebuild all my X11 fink packages? I am puzzled by this. Whay do I have to rebuild all X11 fink packages?
>>>>>
>>>>> On Jul 18, 2010, at 9:22 PM, Jeremy Huddleston wrote:
>>>>>
>>>>>>
>>>>>>
>>>>>> On Jul 18, 2010, at 20:56, Richard Miles wrote:
>>>>>>
>>>>>>> I changed it back to use /usr/X11R6 and it
>>>>>>
>>>>>> What is this "it" ? X11.app in SL and XQaurtz.app will both run scripts in ~/.xinitrc.d unless you have a ~/.xinitrc.
>>>>>>
>>>>>>> still will not run metacity in the starup script that is in ~/.xinitrc.d.
>>>>>>
>>>>>> The only thing that variable makes sense for is in building your packages. You should rebuild all your X11 fink packages whenever you change that variable.
>>>>>>
>>>>>>> This has always been the case.
>>>>>>
>>>>>> Do you have a ~/.xinitrc ?
>>>>>>
>>>>>>> Some update from apple did something so that my script to start gnome does not work anymore.
>>>>>>
>>>>>> I doubt that. We didn't change anything in 10.6.3 or 10.6.4
>>>>>>
>>>>>>> That is it will not start metacity using exec gnome-session.
>>>>>>
>>>>>> It is probably something in your own configuration that changed... especially since other fink users report it working for them in the email that you quoted.
>>>>>>
>>>>>>
>>>>>
>>>>> Richard E. Miles
>>>>> r.godzilla at comcast.net
>>>>>
>>>>>
>>>>>
>>>>
>>>
>>> Richard E. Miles
>>> r.godzilla at comcast.net
>>>
>>>
>>>
>>
>
> Richard E. Miles
> r.godzilla at comcast.net
>
>
>
More information about the Xquartz-dev
mailing list