Re: libstdc++ breakage (__throw unresolved)

2001-04-25 Thread Michel Daenzer
Daniel Jacobowitz wrote:
> 
> On Tue, Apr 24, 2001 at 08:16:47PM -0700, Matt Brubeck wrote:
> > On Wed, 25 Apr 2001, Michel [iso-8859-1] Dänzer wrote:
> >
> > > I just did a minor upgrade involving libstdc++2.10-glibc2.2 and in the
> > > process update-menus failed:
> >
> > I had this problem also, and want to warn everyone very strongly to place
> > this package on hold. Do not upgrade to 2.95.4-0.01. It breaks package
> > installation and removal, and caused several packages on my system to be
> > left in a broken state. It wasn't too difficult to recover and downgrade,
> > but definitely an unwanted hassle.
> >
> > > I rebuilt menu and it works again. But it seems a lot of binaries
> > > linked against /usr/lib/libstdc++-libc6.2-2.so.3 fail in the same way
> > > so I wonder what would be the best way to fix this?
> >
> > Has anyone contacted the maintainer? I assume this problem only occurs on
> > powerpc, since I didn't see anything on the other Debian lists.
> 
> First I'd heard of it.  I know where the problem is; we need to get a
> fixed GCC uploaded with DWARF EH turned on again.  I'm working on it.

I got 2.95.4.ds1-0.010424 from incoming and it works fine so far. Thanks!


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: undefined symbol: __throw

2001-04-25 Thread Michel Daenzer
Damien GUIHAL wrote:
> 
> Le Wed, 25 Apr 2001 10:42:31 +0200
> [EMAIL PROTECTED] (Stefan Haller) a ecrit:
> 
> > I was trying to upgrade my testing installation today, and I got lots of
> > errors of the following kind:
> >
> >   Setting up eterm (0.9.0-9) ...
> >   update-menus: error while loading shared libraries: update-menus:
> >   undefined symbol: __throw
> >   dpkg: error processing eterm (--configure):
> >subprocess post-installation script returned error exit status 127
> >
> > Now everything is in a somewhat half-installed state, which doesn't feel
> > good...  Any idea what I am missing?
>
> I did have the problem too, I'm running Debian Testing, and that's the
> paskage menu which semms to be corrupted. I corrected the problem by
> compiling my own menu package (apt-get source -b menu) with the new
> gcc-2.95.4

That's not a fix but a workaround, the bug is with gcc really. (See the other
thread about this)

libstdc++2.10-glibc2.2 2.95.4-0.010407 should be avoided, fixed gcc packages
should be in unstable soon.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: RFC: What should be done about kernel-image's bootloader questions

2001-06-07 Thread Michel Daenzer
Ethan Benson wrote:
> 
> On Fri, Jun 08, 2001 at 02:14:51AM +0200, Michel D?nzer wrote:
> >
> > Thanks for the pointer. AFAICS it's for the user who installs a
> > kernel-image package though; is there a way to influence this with
> > kernel-package or in my source package that I'm not seeing?
> 
> you can install a /etc/kernel-img.conf file to begin with, but if your
> talking about the kernel-image package itself then no.
> 
> manoj's answer to these situtations is the user, or the package owning
> the bootloader should install an appropriate /etc/kernel-img.conf.

As the bootloader lives in AmigaOS entirely it doesn't make sense to create a
package for it.


> > That wouldn't be a problem, but it asks the exact same questions as it
> > would on an i386 box. I wasn't keen to find out what happens when it tries
> > to run lilo so I said no to all of them...
> 
> i don't see how this is possible, i removed lilo from my x86 boxes
> awhile ago when i installed grub, as soon as i did all the boot floppy
> and lilo questions in kernel-image installation went away.  they
> install quietly now no questions asked (this is packages build with
> potato's kernel-package, i took no other action to disable these questions).
> 
> so you must have somehow installed the lilo package on your apus for
> this to happen, or else kernel-package has become severely
> braindamaged recently.
> 
> since lilo is arch: i386 short of unpacking it manually i don't see
> how you installed it.

[EMAIL PROTECTED]> locate lilo  ~
/usr/lib/menu/default/lilo
/usr/share/vim/vim57/syntax/lilo.vim


No lilo on this Pismo, where I built and installed the kernel-image-2.4.4-apus
package.

I see the preinst sets do_boot{floppy,loader} to "Yes" and defaults the
bootloader to lilo; maybe I could change at least the former two using
kernel-pkg.conf during the build. I'll see.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: MOL

2001-06-08 Thread Michel Daenzer
Jens Schmalzing wrote:

> > It also does not allow me to browse the internet (once I have
> > established a ppp connection to my isp).
> 
> This will work if you enable network address translation and IP
> forwarding.  I use
> 
>  iptables -t nat -A POSTROUTING -j SNAT --to $IP -s $MOLNET -o $DEVICE
>  echo 1 > /proc/sys/net/ipv4/ip_forward
> 
> where $IP and $DEVICE are the address and networking device,
> respectively, that Linux uses for connecting to the outside world, and
> $MOLNET is the network used between Linux and mol (192.168.1.0/24 for
> you).

Just my $0.02: apt-get install ipmasq does this automatically.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: Powerbook G3 1999 xfree 4.0.2 accel?

2001-06-08 Thread Michel Daenzer
Michael Schmitz wrote:
> 
> > I read somewhere that BootX did not like XFree86-4 before 2.4.x for
> > the Lombard. Don't know if that is true after todays work.
> 
> That's still true unless XFree86 changed their mind about resolving PCI
> resource conflicts on PPC regardless of the UseFBDev option.

UseFBDev is still a driver specific option (does the atimisc driver even have
it? Its maintainer doesn't like fbdev too much...), so the upper layer doesn't
care about it. It does have a 'framebuffer' mode now where it doesn't mess
with PCI directly if all devices are claimed by drivers which support it, but
AFAIK only the fbdev driver supports that right now.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: IR on TiBook?

2001-06-08 Thread Michel Daenzer
Wilhelm Fitzpatrick wrote:
> 
> Recently there was a thread about using the IR port on Pismo's, saying
> that there is needed kernel support that is missing.  Does this affect the
> TiBook as well?  That is has anybody got IRDA support working with the
> TiBook?  Or do I need to wait?

Most of the TiBook hardware is the same as a Pismo, so I wouldn't expect it to
behave differently there.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: RFC: What should be done about kernel-image's bootloader questions

2001-06-08 Thread Michel Daenzer
Ethan Benson wrote:
> 
> On Fri, Jun 08, 2001 at 03:07:35AM +0200, Michel Daenzer wrote:
> > As the bootloader lives in AmigaOS entirely it doesn't make sense to
> > create a package for it.
> 
> nor to add anything about it to kernel-package...

It already has to handle the subarch, do you think this would be a big effort?


> you would be better off to make a real bootloader not relying on
> proprietary OSes, or teach the current one to read the kernels off the
> ext2 root filesystem.

AFAIK Roman Zippel had implemented a LILO like loader at one point but it got
lost. :(


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: KDE2/libqt2 errors on sid.

2001-06-08 Thread Michel Daenzer
Mark Nellemann wrote:

> I have been running KDE2 until a recent (1-2 weeks) apt-get
> update+upgrade. When I start any KDE related program now, I get this error:
> 
> [EMAIL PROTECTED]:~$ startkde
> kdeinit: error while loading shared libraries: /usr/lib/libqt.so.2:
> undefined symbol: __pure_virtual
> knotify: error while loading shared libraries: /usr/lib/libqt.so.2:
> undefined symbol: __pure_virtual
> /usr/bin/kdeinit: error while loading shared libraries:
> /usr/lib/libqt.so.2: undefined symbol: __pure_virtual
> 
> Am I doing something wrong or is this a known problem?

I'd say neither.

[EMAIL PROTECTED]> ldd /usr/lib/libqt.so.2|cut -f 3 -d ' '|xargs grep -l
__pure_virtual
/usr/lib/libGLU.so.1

This library is in xlibmesa3. Please try reverting that package to an older
version.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: postgresql?

2001-06-09 Thread Michel Daenzer
Michael Schmitz wrote:

> I've created a snapshot of my chroot for Michel and others, it's at
> ftp.biophys.uni-duesseldorf.de:/pub/linux/powerpc/debian/buildd/. A quick
> setup guide for using the chroot is there as well.

Thanks, I'll give it a try when I have more bandwidth again...


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: Save trackpad settings during sleep

2001-06-09 Thread Michel Daenzer
Tom Rini wrote:

> # Save the settings.
> /sbin/trackpad show | cut -d \  -f 5- > /var/run/trackpad
> # Make sure we write to disk
> sync
  
Is this really necessary BTW? If we wake up normally, it should be there
whether it's on the disk or not, and if we don't wake up, it doesn't matter.


Am I missing something?


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: KDE2/libqt2 errors on sid.

2001-06-10 Thread Michel Daenzer
Mark Nellemann wrote:
> 
> The package (xlibmesa3) wasn't installed at all! KDE is working again
> after I installed it.

Which package providing libgl1 did you have installed then? mesag3?

Anyway, this should probably be reported to the libqt2 maintainer, preferrably
via a bug report.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: X screen wraps around

2001-06-10 Thread Michel Daenzer
Alan Macdougall wrote:
> 
> On 8/6/01 at 1:40 AM, Michel DSnzer <[EMAIL PROTECTED]> wrote:
> 
> > Duncan Sands wrote:
> 
> > > If I use more bpp then the screen image is
> > > displaced to the right, and wraps around onto
> > > the left of the screen.  Having the left of a
> > > window on the right of the screen and the
> > > right of the window on the left of the screen
> > > makes my head hurt...  However, having only
> > > 8bpp is a pain too. Can anyone suggest how I
> > > can have more colours and less head pain?
> >
> > Help fix the bug. ;^)
> 
> I had similar problems on my 7600 w/. 15" Apple Multisync AV monitor
> having gotten XFree86 4.03 / fbdev to work - the wrap (only about 1.5 cm
> though) at what I thought was 8 bpp, and a really terrible pinkish hued
> display at what I thought was 16 bpp.
> 
> This with kernel args in BootX of "video=controlfb:vmode:16,cmode:16".
> 
> Turns out (after reading some of the debian-x list) that changing the
> "Default Depth" in the "Screen" section of the XF86Config-4 file from 16
> to 15 fixes *both* the wrap problem and the colour problem.  (So I've
> now learned that bpp , depth.)

A lot of framebuffer devices are broken such that they only work correctly in
either depth 15 or 16, not both.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: keymaps

2001-06-10 Thread Michel Daenzer
Lenticchia Emiliano wrote:
> 
> i have an amiga ppc with italian keyboard, i need to configure the layout in
> shell and in X server, i have try some change and the result is:
> in shell the ò à è ì ecc key not work
> in mc only the ù § and \ | not work is exchanged
> in x some programs work like mc and other like the shell...
> 
> is a great confusion, the rigth step to make all work fine is 

Fix the kernel keymap (edit /usr/share/keymaps/amiga/amiga-it.kmap.gz) and
submit it to the console-data maintainer. X might work as well then if you
disable Xkb.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



Re: Powerbook G3 1999 xfree 4.0.2 accel?

2001-06-11 Thread Michel Daenzer
Matthew Bernhardt wrote:
> 
> Well, I got X to work with the ati driver finally. I used 4.0.3-3 for
> everythiong except xserver-xfree86, which is at 4.0.2-7.
> 
> The only problem now is that a 'startx' brings a blank screen that dies
> after about 10 seconds. There is 'NO' error, just a die. I thought it was
> all of X until I just hit a 'X' in shell and it brought up everything other
> than a window manager. (I even exported a nice gnome hack on solaris I've
> been working on. Looks great.)
> 
> Is there any way to get an extended log for startx/xinit?

X server output is in /var/log/XFree86.0.log, client stuff in
~/.xsession-errors and/or ~/.gnome-errors . Black screen can hardly be a
client problem.


-- 
Earthling Michel Dänzer (MrCooper)\   Debian GNU/Linux (powerpc) developer
CS student, Free Software enthusiast   \XFree86 and DRI project member



emacs21: build failure related to -z nocombreloc

2001-10-29 Thread Michel Daenzer
Package: emacs21
Version: 21.1-2
Severity: serious


gcc   -R/usr/X11R6/lib   -z nocombreloc  -g   -o temacs
dispnew.o frame.o scroll.o xdisp.o xmenu.o window.o charset.o coding.o
category.o ccl.ocm.o term.o xfaces.o xterm.o xfns.o xselect.o xrdb.o
fontset.o   emacs.o keyboard.o macros.o keymap.o sysdep.o   buffer.o
filelock.o insdel.o marker.ominibuf.o fileio.o dired.o filemode.o
cmds.o casetab.o casefiddle.o indent.o search.o regex.o undo.o  alloc.o
data.o doc.o editfns.o callint.oeval.o floatfns.o fns.o print.o
lread.o abbrev.o syntax.o unexelf.o  mocklisp.o bytecode.o
process.o callproc.oregion-cache.o sound.o atimer.o doprnt.o
strftime.o intervals.o textprop.o composite.o md5.o  terminfo.o
lastfile.o vm-limit.o  widget.o   ../lwlib/liblw.a -L/usr/X11R6/lib
-lXaw -lXmu  -lXt -lSM -lICE -lXext -ltiff  -ljpeg  -lpng -lz -lm  -lungif
-lXpm  -lX11  -lncurses -lm  -lgcc -lc -lgcc
/usr/lib/crtn.o
Running gcc with ppc ld script.
gcc: nocombreloc: No such file or directory
make[3]: *** [temacs] Error 1

(full log at http://voltaire.debian.org/buildlogs/emacs21)


There is a patch which should fix this at
http://savannah.gnu.org/patch/download.php?id=46 .


-- System Information
Debian Release: testing/unstable
Architecture: powerpc
Kernel: Linux pismo 2.4.13-ben0 #1 Sat Oct 27 17:05:53 CEST 2001 ppc
Locale: LANG=en_US, LC_CTYPE=



Re: xterm not working on Dan's X packages ....

1998-12-01 Thread Michel Daenzer
Sven LUTHER wrote:

> btw, the current base has a kbd-dataxxx-10, but only a kbd_xxx-9 which depends
> on kbd-data (=xxx-9), so it is not useable. I didn't have the latest kbd 
> source so i compiled the -7 version of them.

I have installed kbd*-9 with kbd-data*-10, works without problems. You just
have to ignore the dependencies, best done by installing directly with dpkg.

Hope this helps.

Michel

-- 
Why drink & drive when you can smoke and fly?
__________
Michel Daenzer///   [EMAIL PROTECTED]
 __  /// AmigaOS/Linux(Debian/PPC)
Student of computer science at the   \\\///   Team *AMIGA*  ICQ #: 5675698
Swiss Federal Institute of Technology \\\/   IRC: CoOpER / dALEb33


Problems compiling some KDE stuff

1998-12-07 Thread Michel Daenzer
Hi

I have successfully compiled the basic KDE stuff, it works great and also *is*
great IMO. I recommend using Qt 1.33 because 1.4 and later causes all apps to
crash when you mark text.

Now I have encountered problems compiling some advanced stuff:

- kuser (in kdeadmin): Aborts with a complaint that my glibc (2.0.100) has an
error concerning shadow passwords. Sorry, don't have the exact output at hand.
It advises to redo the configure with --with-pam and then remake, but that
doesn't help. Anyone knows what this means and how to fix it?

- kfloppy (kdeutils) doesn't compile. Too bad I don't remember the error, just
wanted to mention it :)

Oh and I use the latest gcc/binutils packages.

Michel

-- 
Why drink & drive when you can smoke and fly?
__________
Michel Daenzer///   [EMAIL PROTECTED]
 __  /// AmigaOS/Linux(Debian/PPC)
Student of computer science at the   \\\///   Team *AMIGA*  ICQ #: 5675698
Swiss Federal Institute of Technology \\\/   IRC: CoOpER / dALEb33


Re: X startup hang

1999-01-21 Thread Michel Daenzer
Sven LUTHER wrote:
 
> it is possible that this is fbdev related, michel, did it work for you ?

Still not Friday ;-)
I'll report on the weekend, although I can't imagine they don't work for me if
they do for you, Sven.

Michel 

-- 
Why drink & drive when you can smoke and fly?
__________
Michel Daenzer///   [EMAIL PROTECTED]
 __  /// AmigaOS/Linux(Debian/PPC)
Student of computer science at the   \\\///   Team *AMIGA*  ICQ #: 5675698
Swiss Federal Institute of Technology \\\/   IRC: CoOpER / dALEb33