Re: dbus problem

2005-10-25 Thread Dan Nicholson
On 10/24/05, Dennis J Perkins [EMAIL PROTECTED] wrote:
 I've found that gdm doesn't call .xinitrc for some reason, so
 dbus-launch never runs.

This is true, gdm does not behave the same way as issuing startx from
the console.  But you can get it to do what you want.  I have dbus
running right now from gdm.

This is what I can gather about how gdm starts up, but have a look at
the documentation for a fuller discussion
http://www.gnome.org/projects/gdm/docs/gdm.html.  Almost everything
resides in /etc/gnome/gdm, and the rest resides in
$GNOME_PREFIX/share/gdm and $GNOME_PREFIX/share/xsessions.  The main
difference here between what you want, is that gdm (and xdm I think)
are starting an Xsession.  You can make it do the same thing.

gdm starts up after running the Default script in /etc/gnome/gdm/Init.
 Then the user authenticates.  It runs the Default script in
/etc/gnome/gdm/PreSession.  Then it runs the /etc/gnome/gdm/Xsession
script.  Then it runs /etc/gnome/gdm/PostLogin/Default.  You don't
need to touch any of the Init/Pre/Post scripts, but Xsession is the
crux, and where you do most of the customization.

Xsession initializes /etc/profile, loads system and user Xresources,
keyboard setup, etc.  This is where you put stuff you want to run when
the session is going.  I have beagle, and it has a user run daemon,
beagled.  I kick this off in Xsession and it runs along until I log
out.  You could run an xterm or xclock or whatever.  I just added this
(way more than necessary):
# Run beagled if it's available
beagledaemon=`gdmwhich beagled`
if [ -n $beagledaemon ]  [ -x $beagledaemon ]; then
echo $0: Starting beagle daemon: $beagledaemon
beagled
fi

This has to go before the eval exec $command statement.  That
statement is the main session manager.  $command is parsed from one of
two files on the base installation.  If you selected Default session,
it uses the file
$GNOME_PREFIX/share/gdm/BuiltInSessions/default.desktop.  If you
selected Gnome session, the file is
$GNOME_PREFIX/share/xsessions/gnome.desktop.

$command is what is on the Exec= line in *.desktop.  By default, this
says Exec=gnome-session for gnome.desktop.  To make this use dbus,
change it to Exec=dbus-launch --sh-syntax --exit-with-session
gnome-session.  Now gdm runs the same way that my .xinitrc file does. 
There's also a way to have a Custom session with ~/.xsession, but I
haven't bothered with that yet.

One last thing.  $GNOME_PREFIX/share/xsessions is also where you would
put any other sessions you might want like kde or xfce.  This
directory is SessionDesktopDir in gdm.conf.

Sorry about being long-winded, but I thought it would be most helpful
if I explained the whole process as best as I understand it instead of
just saying change this line in file blah.

--
Dan
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: Question about libc5 libc6 incompatibility

2005-10-25 Thread Luca Dionisi
I was just wondering: have I said something wrong in my previous
post in this thread?
Why no answers from anyone in ten days?

Luca
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: dbus problem

2005-10-25 Thread Simon Geard
On Mon, 2005-10-24 at 21:11 -0600, Dennis J Perkins wrote:
 I've found that gdm doesn't call .xinitrc for some reason, so
 dbus-launch never runs.
 

Ah, you're using GDM to start Gnome - I'd missed that. In that case, try
the attached patch, which modifies GDM's Xsession file to run
dbus-launch as a wrapper around whatever command it's running (usually
gnome-session). That's what I use, and I find DBUS_SESSION_BUS_ADDRESS
is set in any xterm I open.

Simon.
diff -Naur gdm-2.8.0.5/config/Xsession.in gdm-2.8.0.5-patch/config/Xsession.in
--- gdm-2.8.0.5/config/Xsession.in	2005-07-19 19:53:41.0 +1200
+++ gdm-2.8.0.5-patch/config/Xsession.in	2005-10-12 22:33:22.0 +1300
@@ -224,6 +224,14 @@
   fi
 fi
 
+# add dbus-launch if found
+dbuslaunch=`gdmwhich dbus-launch`
+if [ -n $dbuslaunch ]  [ -x $dbuslaunch ]  [ -z $DBUS_SESSION_BUS_ADDRESS ]; then
+command=$dbuslaunch --exit-with-session $command
+elif [ -z $dbuslaunch ] ; then
+echo $0: dbus-launch not found!
+fi
+
 # add ssh-agent if found
 sshagent=`gdmwhich ssh-agent`
 if [ -n $sshagent ]  [ -x $sshagent ]  [ -z $SSH_AUTH_SOCK ]; then


signature.asc
Description: This is a digitally signed message part
-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: Question about libc5 libc6 incompatibility

2005-10-25 Thread Declan Moriarty
Recently, Somebody Somewhere wrote these words
 I was just wondering: have I said something wrong in my previous
 post in this thread?
 Why no answers from anyone in ten days?
 

Ok. Take as much from the list below as fits.

/begin nag

This list works by people volunteering personal time to help others
with their problems. Nobody _owes_ you an answer. The idea is, you
contribute, and garner some goodwill, so people will then contribute to
help you when you are stuck. It's personal time they are contri8buting. People 
don't bother when

1. Your question is clearly answered in lfs documentation (book, FAQ).

2. You are making many queries, and then demand of those who reply that
they should make things work with minimal effort on your part. In
effect, you are not making as much effort to solve your problems as you
want others to do for you. It's called ungratefulness.

3. Even the most elementary search with google would solve your problems
immediately. Try libc5 libc6 compatability as a google for instance on
this current problem. It's called laziness.

4. You are doing nothing for anyone else. All take, but no give. It's
called selfishness.

5. You don't provide the information necessary to solve your problems
and people have to extract it from you by queries. You don't mention
what you have tried, or else you tried nothing.

/end nag

If anything fitted, that's what you did wrong.
-- 

With best Regards,


Declan Moriarty.
-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: Printing or postscript problem.

2005-10-25 Thread David Lyne

Hi


You can do the same thing in cups with
lpadmin -p jerrypc -m laserjet.ppd.gz -v socket://[ip_address].
The only trick is finding the correct ppd for your printer. The conversion
problem is pretty much the same for cups or lprng.


Cheers for your reply jerry.

But I think my issue is with the conversion, NOT with the network link 
to the printer!


The ppd that I was using on lfs was newer than the ppd that I had 
installed on slackware, so I might try using the older version and see 
what happens.


Cheers
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: Question about libc5 libc6 incompatibility

2005-10-25 Thread Ken Moffat

On Tue, 25 Oct 2005, Luca Dionisi wrote:


I was just wondering: have I said something wrong in my previous
post in this thread?
Why no answers from anyone in ten days?

Luca


 Probably because *nobody* here knows anything about libc5.

Ken
--
 das eine Mal als Tragödie, das andere Mal als Farce
-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Re: Question about libc5 libc6 incompatibility

2005-10-25 Thread Rainer Peter Feller
On Tue, 2005-10-25 at 11:39 +0100, Ken Moffat wrote:
 On Tue, 25 Oct 2005, Luca Dionisi wrote:
 
  I was just wondering: have I said something wrong in my previous
  post in this thread?
  Why no answers from anyone in ten days?
 
  Luca
 
   Probably because *nobody* here knows anything about libc5.
 
 Ken
or nobody wants to admit that s/he is that old that s/he knows anything
about libc5

*searching in news.linuxfromscratch:blfs-support libc5*

o.k. found your question

As much as I guess:

ah no seems to be the answer for that, well for most cases.
it is a well discussed subject (even in lfs) and the (initial) reason
for the /tools dir
Yea old B/LFSers do you remember the struggle with the so called
newglibc? Yes this was a real PITA!!!

so if you want to run an old bin with an old extra lib you should be
prepared to compile much more to get this working.
But I think that this does not belong to blfs-support, because this has
nothing to do with beeding edge new software :-)

Try your luck at lfs-chat ...

-- 
  H
CUH Rainer Peter Feller
  H

-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: Question about libc5 libc6 incompatibility

2005-10-25 Thread Luca Dionisi
On 10/25/05, Rainer Peter Feller
[EMAIL PROTECTED] wrote:
 On Tue, 2005-10-25 at 11:39 +0100, Ken Moffat wrote:
Probably because *nobody* here knows anything about libc5.
 or nobody wants to admit that s/he is that old that s/he knows anything
 about libc5

Grazie
Luca
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Printing or postscript problem - Solved

2005-10-25 Thread David Lyne

Hi again


I had this with a canon bjc4000.

Everyone told me I had the wrong ppd file. I never solved it.
I was printing locally, so I had other options. I went away and never
came back.


Well, in my case it was the 'wrong' ppd file that was the issue. I 
noticed that on Slackware (where everything works well) the ppd was an 
older version, so I tried again with cups using the older one and hey 
presto, it worked! I don't know why new ppd which comes straight from 
the manufacturer's web site didn't work, but there you are ',8o|.


So, my thanks again for your time; and my apologies that the question 
had such a simple solution which I should have thought to try before 
bothering you all.


Dave
8o)
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: Cant rebuild GCC 3.4.4 on BLFS 6.1

2005-10-25 Thread Jonas Norlander
On Mon, Oct 24, 2005 at 08:32:41PM +0200, channelzero wrote:
 Am Montag 24 Oktober 2005 18:01 schrieb Jonas Norlander:
  Any idea why I got this error in make check?
 
 Do you have dejagnu package installed?
 

Yes it is installed.

[EMAIL PROTECTED]:~# runtest --version
WARNING: Couldn't find the global config file.
Expect version is   5.43.0
Tcl version is  8.4
Framework version is1.4.4

I'm going to rebuild tcl, expect and dejagnu just to be sure
the testtools is ok and report back
-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: Cant rebuild GCC 3.4.4 on BLFS 6.1

2005-10-25 Thread Jonas Norlander
On Tue, Oct 25, 2005 at 02:45:16PM +0200, Jonas Norlander wrote:
 On Mon, Oct 24, 2005 at 08:32:41PM +0200, channelzero wrote:
  Am Montag 24 Oktober 2005 18:01 schrieb Jonas Norlander:
   Any idea why I got this error in make check?
  
  Do you have dejagnu package installed?
  
 
 Yes it is installed.
 
 [EMAIL PROTECTED]:~# runtest --version
 WARNING: Couldn't find the global config file.
 Expect version is   5.43.0
 Tcl version is  8.4
 Framework version is1.4.4
 
 I'm going to rebuild tcl, expect and dejagnu just to be sure
 the testtools is ok and report back

Have rebuild them now and all check passed. Trying to build gcc
again

-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: transcode config can't find lzo headers

2005-10-25 Thread Dan Nicholson
On 10/25/05, Simon Scheiwiller [EMAIL PROTECTED] wrote:
 the configure script for transcode 0.6.14 looks for the lzo headers in
 /usr/include. when i pass the option --with-lzo-includes=/usr/include/lzo,
 it looks for the headers in /usr/include/lzo/include.
 is there a patch for transcode for that? or can someone tell me where to
 look for the failure?

You could try altering that default value.  Look at the configure
script and search for with-lzo-includes.  See what variables it sets.

Or just try this:
sed -i.orig 's|\(w_lzo_i\)/include|\1|' configure
or if that looks ugly to you:
sed -i 's|w_lzo_i/include|w_lzo_i|' configure

--
Dan
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: transcode config can't find lzo headers

2005-10-25 Thread Dan Nicholson
On 10/25/05, Randy McMurchy [EMAIL PROTECTED] wrote:
 Changing the defaults may not work however. The problem is because
 I believe Transcode wants LZO-1 and the OP has LZO-2 installed.

 Sometimes, just getting configure to see the changes isn't enough.

 There is a reason the LZO maintainer uses a different path for the
 include files. Probably because the -1 and -2 versions are not
 compatible.

Ahh.  I've never used LZO or even knew there were multiple versions of
it.  Good thing you're around Randy!

--
Dan
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: transcode config can't find lzo headers

2005-10-25 Thread Simon Scheiwiller
Thus spoke Dan Nicholson:
 On 10/25/05, Randy McMurchy [EMAIL PROTECTED] wrote:
 Changing the defaults may not work however. The problem is because
 I believe Transcode wants LZO-1 and the OP has LZO-2 installed.

 Sometimes, just getting configure to see the changes isn't enough.

 There is a reason the LZO maintainer uses a different path for the
 include files. Probably because the -1 and -2 versions are not
 compatible.

 Ahh.  I've never used LZO or even knew there were multiple versions of
 it.  Good thing you're around Randy!


It didn't work anyway, so I'll just do without for now. but just as I
start compiling, there are more problems coming up, so I guess there will
be a new transcode thread pretty soon...


-- 
simon scheiwiller
bollstrasse 21
ch-8405 winterthur
+41 78 624 16 73

--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: transcode config can't find lzo headers

2005-10-25 Thread Randy McMurchy
Dan Nicholson wrote these words on 10/25/05 17:28 CST:

 Ahh.  I've never used LZO or even knew there were multiple versions of
 it.

I ran into this last time I installed Transcode. Another alternative
for the OP is to install the most recent version of Transcode. They
are up to 1.0.1 as of now. You can find it at:

http://www.transcoding.org/cgi-bin/transcode

It may now support LZO-2.

-- 
Randy

rmlscsi: [GNU ld version 2.15.94.0.2 20041220] [gcc (GCC) 3.4.3]
[GNU C Library stable release version 2.3.4] [Linux 2.6.10 i686]
17:30:00 up 31 days, 1:54, 3 users, load average: 0.18, 0.25, 0.45
-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


transcode-0.6.14 fails to compile

2005-10-25 Thread Simon Scheiwiller
More problems here:

first, I get that error:

 gcc -DHAVE_CONFIG_H -I. -I. -I.. -Wall -D_LARGEFILE_SOURCE
-D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2 -MT bitstream.lo -MD
-MP -MF .deps/bitstream.Tpo -c bitstream.c  -fPIC -DPIC -o
.libs/bitstream.o
In file included from bitstream.c:36:
bitstream.h:36: warning: type qualifiers ignored on function return type
bitstream.c: In function 'bitstream_fill_current':
bitstream.c:100: error: invalid lvalue in increment
make[2]: *** [bitstream.lo] Error 1
make[2]: Leaving directory `/sources/blfs/multimedia/transcode-0.6.14/libac3'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/sources/blfs/multimedia/transcode-0.6.14'
make: *** [all] Error 2

I go in the libac3 directory and change line 100 in bitstream.c into what
it was till transcode 0.6.12 - and it works
the original 0.6.14 line is:
current_word = *((uint_32*)buffer_start)++;
i remove the cast and change it back to the 0.6.12 version:
current_word = *(buffer_start)++;

but then further ahead i get this error:

gcc -DHAVE_CONFIG_H -I. -I. -I.. -DMOD_PATH=\/usr/lib/transcode\
-D_REENTRANT -I.. -I../src -I/usr/include -I../libac3 -I../avilib
-I/usr/include/avifile-0.7 -I/usr/include -I/usr/include -I/usr/include
-I/usr/include -I/usr/include/libxml2 -I../libvo -I../libmpeg2
-I/usr/include/lqt -DUSE_XIO -I../libxio -Wall -D_LARGEFILE_SOURCE
-D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2 -MT a52_decore.lo -MD
-MP -MF .deps/a52_decore.Tpo -c a52_decore.c  -fPIC -DPIC -o
.libs/a52_decore.o
a52_decore.c:52: error: static declaration of 'verbose' follows non-static
declaration
../src/transcode.h:582: error: previous declaration of 'verbose' was here
a52_decore.c: In function 'a52_decore':
a52_decore.c:131: warning: pointer targets in passing argument 2 of
'p_read' differ in signedness
a52_decore.c:155: warning: pointer targets in passing argument 2 of
'p_read' differ in signedness
a52_decore.c:176: warning: pointer targets in passing argument 2 of
'p_read' differ in signedness
a52_decore.c:243: warning: pointer targets in passing argument 2 of
'p_write' differ in signedness
make[3]: *** [a52_decore.lo] Error 1
make[3]: Leaving directory `/sources/blfs/multimedia/transcode-0.6.14/import'
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory `/sources/blfs/multimedia/transcode-0.6.14/import'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/sources/blfs/multimedia/transcode-0.6.14'
make: *** [all] Error 2

i run configure again without liba52, but then i get the same error again
with dvd_reader.lo.

Simon

-- 
simon scheiwiller
bollstrasse 21
ch-8405 winterthur
+41 78 624 16 73

--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: transcode config can't find lzo headers

2005-10-25 Thread Dan Nicholson
On 10/25/05, Randy McMurchy [EMAIL PROTECTED] wrote:
 Dan Nicholson wrote these words on 10/25/05 17:28 CST:
 I ran into this last time I installed Transcode. Another alternative
 for the OP is to install the most recent version of Transcode. They
 are up to 1.0.1 as of now. You can find it at:

 http://www.transcoding.org/cgi-bin/transcode

 It may now support LZO-2.

I installed it last night.  It works nice (sans LZO as I said).

--
Dan
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


Re: transcode-0.6.14 fails to compile

2005-10-25 Thread Simon Scheiwiller

Thus spoke Dan Nicholson:
 Step up to 1.0.1.  It built cleanly with no patching, and it found all
 my multitude of plugins.  With the exception of libpostproc (from
 MPlayer).  Here's the configure statement I used:
 ./configure --prefix=/usr --enable-netstrem --enable-libpostproc \
 --enable-freetype2 --enable-ogg --enable-vorbis \
 --enable-theora --enable-libquicktime --enable-a52 \
 --enable-libmpeg3 --enable-libxml2 --enable-mjpegtools \
 --enable-sdl --enable-gtk --enable-libfame \
 --enable-imagemagick \
 --with-libpostproc-includes=/usr/include/postproc

configure with lzo doesn't work here either, it finds lzo1x.h (when i pass
--with-lzo-includes=/usr/includes/lzo to configure), but can't link to
-llzo, so i guess it wants lzo 1, too.

but at least the rest compiles cleanly :-)

Simon


-- 
simon scheiwiller
bollstrasse 21
ch-8405 winterthur
+41 78 624 16 73

--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


sudo vulnerability

2005-10-25 Thread Archaic
After digging around, I couldn't find any info other than what debian
gave. I've weeded out their extra junk and made an LFS-conforming patch
for anyone who uses sudo. I've also emailed the sudo list for their take
on this vulnerability and it's proposed fix.


-- 
Archaic

Want control, education, and security from your operating system?
Hardened Linux From Scratch
http://www.linuxfromscratch.org/hlfs

Submitted By: Archaic (archaic -aT- linuxfromscratch -DoT- org)
Date: 2005-10-25
Initial Package Version: 1.6.8p9
Origin: http://ftp.debian.org/debian/pool/main/s/sudo/sudo_1.6.8p9-3.diff.gz
Upstream Status: Sent
Description: (CVE-2005-2959) Tavis Ormandy noticed that sudo does not clean the
 environment sufficiently. The SHELLOPTS and PS4 variables are
 dangerous and are still passed through to the program running as
 privileged user. This can result in the execution of arbitrary
 commands as privileged user when a bash script is executed.

diff -Naur sudo-1.6.8p9.orig/env.c sudo-1.6.8p9/env.c
--- sudo-1.6.8p9.orig/env.c 2005-02-06 15:37:01.0 +
+++ sudo-1.6.8p9/env.c  2005-10-25 22:55:45.0 +
@@ -89,6 +89,8 @@
 static const char *initial_badenv_table[] = {
 IFS,
 CDPATH,
+SHELLOPTS,
+PS4,
 LOCALDOMAIN,
 RES_OPTIONS,
 HOSTALIASES,
-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page