[Cooker] [Bug 6389] [mandrake_theme] provide a default background
http://qa.mandrakesoft.com/show_bug.cgi?id=6389 [EMAIL PROTECTED] changed: What|Removed |Added Status|NEW |RESOLVED Resolution||INVALID --- Additional Comments From [EMAIL PROTECTED] 2003-11-19 16:52 --- pre/post scripts on all packages that provide mandrake_theme should ensure that /usr/share/mdk/backgrounds/default.png exists and points the right background (Mandrake.png, Powerpack.png etc). Check in %pre of mandrake_theme: if [ ! -f /usr/share/mdk/backgrounds//default.png -o -L /usr/share/mdk/backgrounds//default.png ]; then rm -f /usr/share/mdk/backgrounds//default.png ln -s Mandrake.png /usr/share/mdk/backgrounds//default.png fi -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: RESOLVED creation_date: description: Mandrake_theme doesn't provide a default backaground across different Mandrake versions, or even within the same Mandrake version. Mandrake 9.1: There's only mandrake_desk which provides /usr/share/mdk/backgrounds/default.png. Mandrake 9.2 download: There's mandrake_desk which provides /usr/share/mdk/backgrounds/default-9.1.png, but no default.png. Then there's mandrake_theme which provides /usr/share/mdk/backgrounds/Mandrake.png Mandrake 9.2 PowerPack: Mandrake_theme is not installed by default, but there is mandrake_theme-powerpack, which provides mandrake_theme, and the file /usr/share/mdk/backgrounds/Powerpack.png So there is no consistency, and setting up a windowmanager to use a default background is impossible. This will result in windowmanager using their own backgrounds, which costs space and adds inconsistancy across desktops/wm's. It will be a good idea to rename mandrake_theme to mandrake_theme-download, so it's clear that it's not the same package as the mandrake_theme on the powerpack. It should still provide mandrake_theme. A solution would be to provide a default background. And I mean just the same filename, like Mandrake.png. I don't really care if the download edition would have a Download.png, while Mandrake.png becomes a symlink to one or the other, or if both packages contain a Mandrake.png, as long as it is possible to use the same filename for a background, and set mandrake_theme as a dependency. I would prefer a solution that is compatible as possible with the current release, 9.2, that's why I suggest Mandrake.png as default filename. Just to make an example of where the problem comes up: XFce4 (actually xfdesktop-4) has it's default background hardcoded, and I changed it from an xfce background to Mandrake.png. The problem is that when people install these packages on their powerpack, that it will not have a background, since mandrake_theme-powerpack is the default package installed.
Re: [Cooker] Updated heartbeat, and request
On Tuesday 18 November 2003 12:39 pm, Buchan Milne wrote:
Bret Baptist wrote:
This is the result from rpm --rebuild heartbeat-1.0.2-2mdk.src.rpm:
error: Bad package specification: %package -n %mklibname heartbeat 0
Easy fix (I think)
Add something like this to the top of the spec file:
%{!?mklibname: %define mklibname
%lib%{1}%{?2:%{2}}%{?3:_%{3}}%{-s:-static}%{-d:-devel}}
(that's one line)
This is what I use on samba, and it should still build on 8.0 even
(not that I have tried this particular feature on it recently ...)
If it's a couple of newer packages on 8.2, just add the macro to your
~/.rpmmacros (might need some tweaking).
No such luck Buchan, get this message now.
error: Bad package specification: %package -n %lib%{1} heartbeat 0
D'Oh, that one seems to be a new addition that wasn't in fact tested.
I've just tested on 8.2 in a chroot with:
%{!?lib: %global lib lib}
%{!?mklibname: %global mklibname(ds)
%lib%{1}%{?2:%{2}}%{?3:_%{3}}%{-s:-static}%{-d:-devel}}
(that's two lines ... in case the last one wrapped to two)
I only tested with:
rpm -q --specfile samba.spec --with system
with the two changed lines above (relative to the packages in cooker).
Otherwise, I can give the heartbeat SRPM a rebuild now since I have the
chroot available now ...
Regards,
Buchan
[Cooker] [Bug 5821] [kernel] kernel 2.4.22-9mdk-i686-up-4GB does not boot w/ACPI enabled
http://qa.mandrakesoft.com/show_bug.cgi?id=5821 --- Additional Comments From [EMAIL PROTECTED] 2003-11-18 08:58 --- Tomas: You can boot without acpi at the LILO prompt, hit ESC and type: linux acpi=off That should help you, but AFAIK it isn't related to this bug. It's better to get machines blacklisted for acpi so that there is one kernel only that supports all machines, with all features, and there is no sense distributing another 20MB of kernel if one setting can be changed. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEW creation_date: description: New cooker kernel freezes when ACPI is enabled. Hardware is HP omnibook 4500XE.
Re: [Cooker] Updated heartbeat, and request
On Tuesday 18 November 2003 3:32 pm, [EMAIL PROTECTED] wrote:
On Tuesday 18 November 2003 12:39 pm, Buchan Milne wrote:
Bret Baptist wrote:
D'Oh, that one seems to be a new addition that wasn't in fact tested.
I've just tested on 8.2 in a chroot with:
%{!?lib: %global lib lib}
%{!?mklibname: %global mklibname(ds)
%lib%{1}%{?2:%{2}}%{?3:_%{3}}%{-s:-static}%{-d:-devel}}
(that's two lines ... in case the last one wrapped to two)
OK we are making progress now. Only issues I am seeing is why would
these be build dependencies?
libgnome32-devel is needed by heartbeat-1.0.2-2mdk
libid3tag0-devel is needed by heartbeat-1.0.2-2mdk
I found the same issue now, I just --nodeps'ed it (libid3tag0-devel is not
in 8.2, I installed the rest painfully with my old 8.2 CDs which I seem to
have written on bad media ...) and it seems to be building OK.
I really can't think why I need half of GNOME installed to build this!
Maybe I will find out some time ;-).
Regards,
Buchan
Re: [Cooker] libtool problem while using %configure in kde rpm
Hi, I'm working on my second rpm for submission to mandrake contrib but I've encountered a problem. I'm trying to create a spec file for kscope 0.4 I've followed the standard .spec given in the wiki and have this snippet in my file; %prep %setup %build %configure %make %install rm -rf %buildroot %makeinstall However I get the following error message while trying to build the package; c++ -DHAVE_CONFIG_H -I. -I. -I.. -I/usr/lib/qt3//include -I/usr/X11R6/include -DQT_THREAD_SUPPORT -D_REENTRANT -O2 -O2 -fomit-frame-pointer -pipe -march=i586 -mcpu=pentiumpro -fno-exceptions -fno-check-new -c kscope_meta_unload.cpp /bin/sh ../libtool --mode=link --tag=CXX c++ -O2 -O2 -fomit-frame-pointer -pipe -march=i586 -mcpu=pentiumpro -fno-exceptions -fno-check-new -o kscope -L/usr/X11R6/lib -L/usr/lib/qt3//lib -L/usr/lib -R /usr/lib -R /usr/lib/qt3//lib -R /usr/X11R6/lib projectfilesdlg.o progressdlg.o kscopepixmaps.o calltreedlg.o scanprogressdlg.o dirscanner.o kscopeconfig.o preffrontend.o prefcolor.o preferencesdlg.o openprojectdlg.o newprojectdlg.o ctagslist.o ctagsfrontend.o frontend.o querypage.o querywidget.o cscopefrontend.o searchlist.o editorpage.o editormanager.o filelist.o projectmanager.o editortabs.o kscope.o main.o projectfileslayout.o calltreelayout.o scanprogresslayout.o prefcolorlayout.o preffrontendlayout.o openprojectlayout.o newprojectlayout.o preffrontend.moc.o frontend.moc.o prefcolor.moc.o querypage.moc.o projectmanager.moc.o calltreedlg.moc.o newprojectdlg.moc.o progressdlg.moc.o kscopeconfig.moc.o searchlist.moc.o editorpage.moc.o kscope.moc.o filelist.moc.o ctagsfrontend.moc.o scanprogressdlg.moc.o projectfilesdlg.moc.o querywidget.moc.o cscopefrontend.moc.o editormanager.moc.o editortabs.moc.o openprojectdlg.moc.o preferencesdlg.moc.o ctagslist.moc.o kscope_meta_unload.o -lktexteditor -lkparts -lkdeui -lkdecore -lqt-mt -lpng -lz -lm -lXext -lX11 -lresolv -lSM -lICE -lpthread -lresolv libtool: unrecognized option `--tag=CXX' Try `libtool --help' for more information. make[2]: *** [kscope] Error 1 make[2]: Leaving directory `/home/nick/rpm/BUILD/kscope-0.4/kscope' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/nick/rpm/BUILD/kscope-0.4' make: *** [all-recursive-am] Error 2 error: Bad exit status from /home/nick/rpm/tmp/rpm-tmp.99463 (%build) RPM build errors: Bad exit status from /home/nick/rpm/tmp/rpm-tmp.99463 (%build) The problem would appear to be caused by the fact that Mandrake ships with libtool 1.4.3 and kscope (and it would appear most kde packages) uses libtool 1.4a [EMAIL PROTECTED] nick]$ cd rpm/BUILD/kscope-0.4/admin/ [EMAIL PROTECTED] admin]$ ./ltconfig --version ltconfig (GNU libtool) 1.4a (1.641.2.206mm 2001/04/03 21:47:47) This would appear to mean that I cannot use %configure in my .spec file. Are there plans to upgrade mandrakes libtool to version 1.4a? How have others addressed this problem? And is there a standard Mandrake replacement for %configure in kde packages? AFAIK, the current work-around for this is to add this at the top of your spec file: %define __libtoolize true Everything else should work. Things like this probably need to be added to the wiki (along with the mklibname fix etc etc). Regards, Buchan
[Cooker] [Bug 6383] [drakxtools] New: diskdrake offers too too much space when resizing LVM volumes
http://qa.mandrakesoft.com/show_bug.cgi?id=6383 Summary: diskdrake offers too too much space when resizing LVM volumes Product: drakxtools Version: 9.3-5mdk Platform: PC OS/Version: All Status: UNCONFIRMED Severity: normal Priority: P2 Component: diskdrake AssignedTo: [EMAIL PROTECTED] ReportedBy: [EMAIL PROTECTED] I have one volume group: # vgdisplay --- Volume group --- VG Name default_volume VG Access read/write VG Status available/resizable VG # 0 MAX LV255 Cur LV2 Open LV 1 MAX LV Size 255.99 GB Max PV255 Cur PV2 Act PV2 VG Size 4.98 GB PE Size 4 MB Total PE 1276 Alloc PE / Size 1176 / 4.59 GB Free PE / Size 100 / 400 MB VG UUID OlTYL1-NxcV-VN6u-AtUC-05T3-8Rut-smYY3F When resizing this volume: # lvdisplay /dev/default_volume/2 --- Logical volume --- LV Name/dev/default_volume/2 VG Namedefault_volume LV Write Accessread/write LV Status available LV # 2 # open 1 LV Size608 MB Current LE 152 Allocated LE 152 Allocation next free Read ahead sectors 1024 Block device 58:1 Diskdrake offers to resize it to up to 5104 MB, even though the unassigned space on the VG is correctly indicated in diskdrake (400MB) as a grey block. Of course, resizing larger than the current volume size plus unallocated volume space (I tried 1200MB) fails (diskdrake shows a vextend failed error. (I actually encountered this on 9.1 with much larger volumes too). It's not too serious, but it would be nice if everything worked exactly correctly (everything else is really nice to use - diskdrake's LVM/XFS resize/ext2,3 resize support etc makes it trivial to manage data). -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is.
Re: Re[2]: [Cooker] 9.2 ISOs on the mirrors.
--- Galileo [EMAIL PROTECTED] wrote: O Ouch There is no need for it now since the LG released updates for their firmware and procedure how to fix the dead drives. It would be nice if Mandrake would release updated iso-s for club members or a howto on how to do it ourselves. btw does this mean that the updated kernel is not necessary ? Are there any changes in the updated kernel except the packet writing being left out ? There are other fixes such as the issue with Phillips webcams freezing Mandrake (see errata). But that certainly doesn't need a new ISO.
Re: [Cooker] Samba update failed?
Did a urpmi --auto-select tonight - possible samba update bomb? Here's the console output: [...] 1:drakxtools-newt ## 2:drakxtools ## 3:harddrake-ui ## 4:samba-common warning: /etc/samba/smb.conf created as /etc/samba/smb.conf.rpmnew ## 5:samba-server ## 6:samba-client ## 7:libsmbclient0 ## 8:harddrake ## 9:drakconf ## error: %preun(samba-client-2.2.8a-13mdk) scriptlet failed, exit status 1 Yes, this was an error in the preun script of the package you are upgrading from. Nothing to worry about, and fixed in both -14mdk and 3.0.1-0.pre3.2mk. If you alternatives (ie /usr/bin/smbclient) still exist and work, the upgrade should have been ok, if not I would want more details ... Regards, Buchan
Re: [Cooker] User says MDK 9.2 is a backward step and requests features
Hello guys, I give support at MandrakeExpert and there was an incident created by an user very dissatisfied with Mandrake Linux 9.2. If you guys want to take a closer look, the incident number is 72784. The user says after upgrading to Mandrake 9.2 from Mandrake 9.0 he's lost a lot of features and customization he'd made. In answer to him, I said I would forward his complains about mdk 9.2 to Cooker guys to try to improve the next version. What do you guys think about his opnions? Do you think these things can be changed? Below is the message transcript. Later, Alberto M. -- I just installed Mandrake 9.2 over my 9.0. I have lost the following functions, and I want them back. kpp - I had a button on my pannel to launch kpp to dial my isp. Now there is no function there. Some large KDE packages were split into smaller packages. This is a good thing, since now you can install kmail without having to install potential security risks (like pppd which is required by kppp), for instance in an office where no-one has a modem. You don't need the scanning libraries just to install a screenshot tool. Yes, unfortunately this causes a bit of pain for end-users installing software which disappears. Superuser file manager - There was a superuser file manager under the file manager section of my start menu. It is now gone. What is the function that it linked to so I can get it back and manipulate the files as root without haveing to mess around with an su in a terminal somewhere? No idea why this was removed, after we actually fixed them to work correctly (use filemanagement profile and work in non-KDE desktops). For now, drag/drop your Home icon to the desktop, right-click-properties-execute-run as-root. Kmail address book is gone - Kmail is now asking for an address book. I want the one I was using before. See (1). Under the configuration menu the login manager asked for the root password and let me do useful stuff. The new version doesn't. The Administration button was still there last time I looked. So far, I consider 9.2 a major step backward. I have to send this message from Windows. [after, he sent another one]: I have reloaded the packages for KPPP and KAddressBook to my system. KAddressBook did find all my e-mail addresses. Incidently, I also lost the game Freecell. My bigger problem is I can't connect to my ISP through my modem. The connections got lost when I upgraded. I did use the config tool to setup the internet connection, but none of my programs (Kmail, pan, galleon, mozilla, opera, etc.) can connect, and all report that they are unable to connect to the requested host. Finally, my SCSI devices have disappeared. My swap file is on the SCSI drive, so this is a potential problem. At least I have 512 meg of memory. What SCSI controller is it? Are any errors listed in dmesg? Please help. I have to use Windows whenever I want to go to the net, and I can't install my Win4Lin until I get the internet connection back up to get the latest modified kernel from Netraverse. No need for this, win4lin kernels are available in contrib (and thus PowerPack/ProSuite). Finally, can you go back to the old x-windows signon window. With the new one, you can only login for users that are on the list. The old one also allowed you to enter the user's name. I guess you have mdkkdm. Use Mandrake Control Center-Hardware-Login Manager (or similar) to change this. (This is a change from 9.0-9.1) I was hoping to change the window, or at least add root to the list by using the login manager from the desktop manager (don't you think this should be in the configuration manager too?), and selected the administrator button. The window appeard to allow me to enter the root password. The main window went blank and said loading. But then it blinked for a minute. I could see that the text boxes were active. Then almost instantly, it went blank again, and the window was as it is when the program starts. So I cannot change the login interface. I can't make out what the user is doing here. kcontrol-Administration-Login Mangager? I am sure this works. It seems the user has some X configuration issue. Try starting a terminal, and running: $ su # kcontrol I'm guessing it can't connect to the X server (no idea how he got it that way ...). Is there some way to get rid of the background on all the pseudo terminals on F1 through F6? Edit /etc/sysconfig/bootsplash (and make the obvious change) Can the startup and shutdown all be set to default to verbose mode? Add 'splash=verbose' or 'splash=0' to the append options in the bootloader. The new kde icons look like they were stollen from Windows XP. Any way to go back to the old ones? KDE Control Center-LooknFeel-Icons ? I can't remember what 9.0 had by default, maybe we didn't have Crystal back then? It does look to me as though you rushed this to the
Re: [Cooker] KDE 3.2 beta1 - KDM Session Type Selection
The strange thing is if I type in startx (ie don't use any form of autologin) It works perfectly For me it works fine if I choose Default. So I think it might have something to do with the user authentication problem No (and this is using LDAP even). I think there's still a problem with getting /etc/wmsession.d into kdm (since I don't think I see all the WMs I have installed). Regards, Buchan
Re: [Cooker] KDE 3.2 beta1 - KDM Session Type Selection
[EMAIL PROTECTED] wrote: For me it works fine if I choose Default. Well, it depends on what you want to get for Default. Well, at present it seems to mean KDE. If I go through the drakfw wizard and choose GNOME, that's what I expect to get the next time. Yes, I guess this is also broken, but I normally avoid drakfw myself. But that aside, the problem is that no matter what WM you choose explicitly from the (I assume) MDKKDM submenu, all you get is ICEWm. Maybe this doesn't happen if all you ever try to use is KDE. Well, actually after kde-3.2 made it onto my cooker box, I tried the KDE entry, and got IceWM, tried the GNOME entry, and got IceWM, tried the Default entry, and got KDE. My last login to X before the kde-3.2 upgrade was KDE, no idea if that makes a difference or not, but it works for me so I'm not going to change it in the near future. Regards, Buchan
[Cooker] [Bug 6273] [Installation] bootsplash can not be unmarked in package list
http://qa.mandrakesoft.com/show_bug.cgi?id=6273 [EMAIL PROTECTED] changed: What|Removed |Added Status|NEW |RESOLVED Resolution||INVALID --- Additional Comments From [EMAIL PROTECTED] 2003-11-13 12:20 --- bootsplash is required by the mandrake-theme package: $ rpm -qR mandrake_theme|grep bootsplash bootsplash = 2.0.2 mandrake_theme is required by mandrake_desk $ rpm -q --whatrequires mandrake_theme mandrake_desk-9.2-8mdk mandrake_desk is required by any window manager: $ rpm -q --whatrequires mandrake_desk blackbox-0.65.0-1mdk WindowMaker-0.80.2-5mdk libgnome2-2.4.0-3mdk icewm-1.2.13-0.3.1mdk gdm-2.4.4.5-1mdk xscreensaver-4.14-3mdk gnome-session-2.4.1-1mdk kdebase-progs-3.1.93-14mdk If your real issue is turning off bootsplash, just change the 'splash=silent' to 'splash=0'. Alternatively, maybe you want to just see all the boot messages all the time, and you can use 'splash=verbose'. If you really want to file a bug, file it on mandrake_theme, that it should not require bootsplash. But, since Mandrake wants to have bootsplash by default, I don't think it will be confirmed as a bug (since it's probably a feature considering it was intentionally added). You could also make your own package which provides mandrake_theme, but does not require bootsplash (if this is for a larger installation). Either way, installation is doing the right thing according to the rpm dependencies. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: RESOLVED creation_date: description: Well it can be unmarked but the warning follows that almost all KDE, Gnome etc. is going to be excluded from installation as well. The following packages are going to be removed list is much longer than warning window can show (BTW this warning window is not scrollable, as anyone can see trying to unmark bootsplash, so you even canot see what else is NOT going to be installed... ;-) You might understand that not everyone wants to close boot screen text messages with some much less informative bootsplash. If somebody wants to partially blind himself, that's OK for him... I don't want to. For me the easiest way NOT to see this bootsplash at all has been in 9.1 simply to not install it. Unmark during the installation. Now with 9.2 it seems impossible because simple unmarking of bootsplash in Graphical Environment - KDE Workstation - bootsplash seems to imply mandatory unmarking whole graphical interface set with it. Something is wrong with this. I want to get rid of bootsplash but still use KDE and whole lot of Graphic.
[Cooker] [Bug 6273] [Installation] bootsplash can not be unmarked in package list
http://qa.mandrakesoft.com/show_bug.cgi?id=6273 --- Additional Comments From [EMAIL PROTECTED] 2003-11-13 17:17 --- Mandatory installation(!) of any non-vital part of Mandrake GPL-ed download distribution You can always 'rpm -e --nodeps bootsplash' if you want. is something definitely worth to be discussed as well. Bugzilla is the wrong place, try cooker. I do not consider such a bit Orwellian thing to be feature. How about the desktop backgrounds? Maybe you should request to have them removed too, and we can have a blank screen instead. We could remove the X patch too, so that instead of a blue background when starting X, we could have the more conventional black/white diagonal hatching and hurt our eyes looking at it until we log in. What additional _mandatory_ parts there are for Mandrake 9.2? Any list of them? Sure, just diff the output from 'urpmq -d basessytem|sort' and 'urpmq -d mandrake_desk|sort' You may find things like 'menu' and 'mdk-menu-messages' forced upon you. Actually, since the initscripts package actually calls the /sbin/splash binary, maybe initscripts should depend on bootsplash instead? Then we can have it on servers too ;-). -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: RESOLVED creation_date: description: Well it can be unmarked but the warning follows that almost all KDE, Gnome etc. is going to be excluded from installation as well. The following packages are going to be removed list is much longer than warning window can show (BTW this warning window is not scrollable, as anyone can see trying to unmark bootsplash, so you even canot see what else is NOT going to be installed... ;-) You might understand that not everyone wants to close boot screen text messages with some much less informative bootsplash. If somebody wants to partially blind himself, that's OK for him... I don't want to. For me the easiest way NOT to see this bootsplash at all has been in 9.1 simply to not install it. Unmark during the installation. Now with 9.2 it seems impossible because simple unmarking of bootsplash in Graphical Environment - KDE Workstation - bootsplash seems to imply mandatory unmarking whole graphical interface set with it. Something is wrong with this. I want to get rid of bootsplash but still use KDE and whole lot of Graphic.
[Cooker] [Bug 5673] [kernel] SIS 746FX - lost interrupts
http://qa.mandrakesoft.com/show_bug.cgi?id=5673 --- Additional Comments From [EMAIL PROTECTED] 2003-11-12 18:18 --- Have you tried booting with 'acpi=off'? This fixes problems for me on the machines I have seen with this problem. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEW creation_date: description: Mainboard: ASRock K7S8X, SIS 746FX chipset, Bios 1.90 Sept, 15, 2003 I tried to boot up from 9.2rc2 CD 1. It detected my IDE-chipset as SIS 5513: SIS 962/963 MuIOL UDMA/133 Then it got problems while detecting my HDDs: hda: attached ide-disk driver hda: lost interrupt hda: lost interrupt hda: lost interrupt hda: host protected ares=J1 hda: lost interrupt hda: xxx sectors etc. (harddisk properly detected) from the first lost interrupt to the detection of my first harddisk two minutes elapsed. As it tried to detect my hdc and started again with hdc: lost interrupt I stopped the installation. Maybe a workaround for this could be to use the kernel - parameter noapic, but this would be a *workaround* only. On my current 9.1 installation with kernel 2.4.21 I also get APIC-errors in /var/log/messages even when starting with the noapic - option :-( Any help would be appreciated! Thank you!
[Cooker] [Bug 5359] [XFree86] xf86cfg crashes in LM 9.2rc1/9.2final
http://qa.mandrakesoft.com/show_bug.cgi?id=5359 --- Additional Comments From [EMAIL PROTECTED] 2003-11-10 12:15 --- Works for me. AFAIK, xf86cfg (without -text option) only runs in X. Did you try this in X or not? -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: This is a new problem. Run xf86cfg (as root or non-root user) and it starts up but then crashes with the message: X Error of failed request: BadName (named color or font does not exist) Major opcode of failed request: 45 (X_OpenFont) Serial number of failed request: 813 Current serial number in output stream: 817 However, xf86cfg -textmode works fine.
[Cooker] [Bug 6313] [rpm] conditional BuildRequires not picked up in query
http://qa.mandrakesoft.com/show_bug.cgi?id=6313
--- Additional Comments From [EMAIL PROTECTED] 2003-11-10 12:17 ---
But the same method should still work:
[EMAIL PROTECTED] rpm]$ rpm -qRp XFree86-4.3-24mdk.src.rpm
zlib-devel
flex
bison
groff
pam-devel
ncurses-devel
perl
libpng-devel
freetype2-devel
fontconfig-devel = 2.1-4mdk
rpmlib(CompressedFileNames) = 3.0.4-1
[EMAIL PROTECTED] rpm]$ rpm -qRp XFree86-4.3-24mdk.src.rpm |wc -l
11
[EMAIL PROTECTED] rpm]$ rpm -ivh XFree86-4.3-24mdk.src.rpm
1:XFree86### [100%]
[EMAIL PROTECTED] rpm]$ rpm -bs --nodeps mdk/SPECS/XFree86.spec
Wrote: /home/bgmilne/rpm/mdk/SRPMS.mdk10.0/XFree86-4.3-24mdk.src.rpm
[EMAIL PROTECTED] rpm]$ rpm -qRp
/home/bgmilne/rpm/mdk/SRPMS.mdk10.0/XFree86-4.3-24mdk.src.rpm
zlib-devel
flex
bison
groff
pam-devel
ncurses-devel
perl
libpng-devel
freetype2-devel
fontconfig-devel = 2.1-4mdk
Glide_V3-DRI-devel = cvs-2mdk
rpmlib(CompressedFileNames) = 3.0.4-1
[EMAIL PROTECTED] rpm]$ rpm -qRp
/home/bgmilne/rpm/mdk/SRPMS.mdk10.0/XFree86-4.3-24mdk.src.rpm|wc -l
12
If you could query buildrequires from the spec file (something like 'rpm -qR
--specfile XFree86.spec' but for buildrequires), then the useless creation of
the SRPM could be avoided.
--
Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug, or are watching someone who is.
--- Reminder: ---
assigned_to: [EMAIL PROTECTED]
status: NEW
creation_date:
description:
When a package has conditional BuildRequires (like ImageMagick) these are not
picked up when querying the package:
Normal:
[EMAIL PROTECTED] i586]$ rpm -qpR
/mirrors/cooker/SRPMS/ImageMagick-5.5.7.12-1mdk.src.rpm | wc -l
15
--with plf:
[EMAIL PROTECTED] i586]$ rpm -qpR
/mirrors/cooker/SRPMS/ImageMagick-5.5.7.12-1mdk.src.rpm --with plf | wc -l
15
[EMAIL PROTECTED] i586]$ rpm --with plf -qpR
/mirrors/cooker/SRPMS/ImageMagick-5.5.7.12-1mdk.src.rpm | wc -l
15
[EMAIL PROTECTED] i586]$ rpm -qpR --with plf
/mirrors/cooker/SRPMS/ImageMagick-5.5.7.12-1mdk.src.rpm | wc -l
15
From the ImageMagick.spec file:
%if %build_plf
%define enablelzw 1
%define enablejasper1
%define enablefpx 1
%endif
%if %{enablefpx}
BuildRequires: libfpx1-devel
%endif
%if %{enablejasper}
BuildRequires: libjasper-devel
%endif
So, one would expect to see a number larger than 15 when queried with the --with
plf option.
[Cooker] [Bug 5359] [XFree86] xf86cfg crashes in LM 9.2rc1/9.2final
http://qa.mandrakesoft.com/show_bug.cgi?id=5359 --- Additional Comments From [EMAIL PROTECTED] 2003-10-11 18:00 --- -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [ndeb] wrote: only have I have XFree86-100dpi-fonts and XFree86-75dpi-fonts installed. I might help to run via strace without any changes to your system, so see where it crashes. Regards, Buchan - -- |--Another happy Mandrake Club member--| Buchan MilneMechanical Engineer, Network Manager Cellphone * Work+27 82 472 2231 * +27 21 8828820x202 Stellenbosch Automotive Engineering http://www.cae.co.za GPG Key http://ranger.dnsalias.com/bgmilne.asc 1024D/60D204A7 2919 E232 5610 A038 87B1 72D6 AC92 BA50 60D2 04A7 -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.3 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQE/r8NvrJK6UGDSBKcRAmg4AKC/bnLaw3q7UcOYbP5/c/VU2YJ0WwCfdZxh kI/PruFF5zAvg61k+qRETag= =RVdz -END PGP SIGNATURE- -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: This is a new problem. Run xf86cfg (as root or non-root user) and it starts up but then crashes with the message: X Error of failed request: BadName (named color or font does not exist) Major opcode of failed request: 45 (X_OpenFont) Serial number of failed request: 813 Current serial number in output stream: 817 However, xf86cfg -textmode works fine.
Re: [Cooker] Trying to find packages in cooker.
Hello, list. I'm new to this list, and I am interested in helping debug, test, and work with the cooker release of Mandrake. However, I'm having problems finding a couple of packages that I absolutely require in order to work with Linux. I have a visual problem seeing the screen, but in Gnome 2.4 there is the atk libraries, gnopernicus screen reader, festival tts, gnome-mag magnifier, as well as other Gnome accessibility packages. All of GNOME-2.4 is available in Mandrake 9.2 and cooker, including (AFAIK) all accessibility software in GNOME-2.4. Since I am very new to cooker I can't seam to tell if anyone has created rpms for these very necessary packages for those of us without perfect vision. If these are not included in cooker could someone please take some time to include all the gnome accessibility libraries and accessibility programs in Mandrake. Thanks, and looking forward to your responce. $ urpmq -d -r gnopernicus|egrep (fest|atk|gnope|gnome-mag) festival-1.4.3-2mdk festvox-kallpc16k-1.4.3-1mdk|festvox-kallpc8k-1.4.3-1mdk|festvox-kedlpc16k-1.4.3-1mdk|festvox-kedlpc8k-1.4.3-1mdk gnome-mag-0.10.3-1mdk gnopernicus-0.7.1-1mdk libatk1.0_0-1.4.1-1mdk libgnome-mag1-0.10.3-1mdk So, all that you need to do is install gnopernicus, which will pull in atk (if you don't have it already), festival, a festival voice of your choice, and gnome-mag: # urpmi gnopernicus Regards, Buchan
[Cooker] More kde conflicts
file /usr/bin/umbrello from install of kdesdk-3.1.93-3mdk conflicts with file from package umbrello-1.1.1-3mdk (kdesdk should probably provide/obsolete umbrello?) A lot of these occur if using 'urpmi --auto-select', there should be conflicts in kdepim-common for kdenetwork-common 3.1.93 etc. file /usr/bin/kgpgcertmanager from install of kdepim-common-3.1.93-6mdk conflicts with file from package kdenetwork-common-3.1.3-37mdk file /usr/lib/libkdenetwork.la from install of libkdepim2-common-3.1.93-6mdk conflicts with file from package libkdenetwork2-common-3.1.3-37mdk file /usr/lib/libkdenetwork.so.2 from install of libkdepim2-common-3.1.93-6mdk conflicts with file from package libkdenetwork2-common-3.1.3-37mdk file /usr/lib/libmimelib.la from install of libkdepim2-common-3.1.93-6mdk conflicts with file from package libkdenetwork2-common-3.1.3-37mdk file /usr/lib/libmimelib.so.1.0.1 from install of libkdepim2-common-3.1.93-6mdk conflicts with file from package libkdenetwork2-common-3.1.3-37mdk (I have now done 'rpm -Uvh /var/cache/urpmi/rpms/*.rpm' on this cooker box, but we have another which will probably show the problems ...). Then, I also get this: file /usr/share/mimelnk/text/x-fortran.desktop from install of kdevelop-3.0.0-0.alpha7.6mdk conflicts with file from package mandrake-mime-0.4-5mdk file /usr/share/mimelnk/text/x-pascal.desktop conflicts between attempted installs of kdevelop-3.0.0-0.alpha7.6mdk and kdelibs-common-3.1.93-11mdk (maybe there is a newer pacakge, our mirror is a bit behind) Regards, Buchan
Re: [Cooker] Danger: Magazine with 9.2 with problematic kernel
sionii wrote: The LINUX FORMAT magazine carried a very prominent warning about LG on their website following the publication of their issue with 9.2 as a cover disc. And when was the last time you checked a magazine website before installing a package included on a CD or DVD? Most people won't go to the website until after they experience the problem. This is not going to save more than a handful of those who could be affected by this problem. For those who think that 9.2 was in a fit state to be released and that those who have criticised were over-reacting - have a look at Linux Format's Help forum. It is being swamped with installation support requests for 9.2. It is _very_ bad news for linux (and Mandrake) that the latest release of the distro that has the best reputation for being user-friendly is going to leave so many people stranded or even damaged at install time. We are already seeing sweeping generalisations along the lines of linux isn't ready for primetime, because people assume Mandrake = linux. This sort of damage to reputations could take years to undo. If you are talking about only the LG issue, I think your point is moot. 1)LG will replace or repair the devices 2)An exploit of this is possible on Windows, and I am sure we will see one soon, and then I guess people will say Windows isn't ready for the desktop? If you have any *real* reasons why your claim that 9.2 was not in a fit state to be released, please give me your bug numbers. AFAIK, the only major as-yet unfixed problem on 9.2 is the disappearing menus issue. And any people who are running 9.2 now are either: -Club members (and there has been enough coverage on fixing and working arounf the issues -installing from FTP (these aren't newbies then) -getting the ISOs via some other means, and can't really complain until official ISOs are out (which will address the LG issue).
Re: [Cooker] Problems with new kde 3.1.93
Le Saturday 08 November 2003 12:57, Crispin Boylan a écrit : Hi The same ;-) I've just updated the packages to kde 3.1.93 on my system and now I seem to have lost anti-aliased fonts, Whilst the Use anti-aliasing for fonts option is checked, I've exactly the same feeling. In konqueror, fixed and serif fonts have a rendering I've never seen before. along with icons on the menu items (icons on the hierarchies are ok, but the actual programs seem to have no icon with them). It seems god for me (cooker from this morning). I've also lost the ability to use the k control centre, all that appears is a Network menu with nothing in it! The problem is known and Laurent is working on it. I Use Configuration - KDE menu entries instead, but I two System entries (the first one with Crypto and Passwords, and the second with Date, Fonts and Paths) and no way to configure kdm. This would work: # kcmshell crypto # kcmshell kdm For anymore, see the /usr/lib/kde3/kcm_(\w+).so files and run kcmshell $1, some should work, some won't. (caveat: I haven't tried this with KDE-3.2 on cooker since it hasn't made it to my cooker due to the fact that new packages are uploaded every day just before our mirror syncs :-(). Regards, Buchan
[Cooker] [Bug 5669] [Installation] not possible to select mirror for updates at end of install
http://qa.mandrakesoft.com/show_bug.cgi?id=5669 --- Additional Comments From [EMAIL PROTECTED] 2003-11-08 13:36 --- Simon, I don't think the /sbin/dhclient-script is the immediate problem (it is optional to dhcp, but is used to set routes etc), the immediate problem is this: socket: Protocol not available - make sure CONFIG_PACKET (Packet socket) and CONFIG_FILTER (Socket Filtering) are enabled in your kernel configuration! This means that the af_packet module is not loaded, which is required to do a DHCP REQUEST etc. I do network installs quite often, but can't test this since DrakX doesn't allow a custom update source to be used, we have an internal updates mirror and no access to any official mirrors. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: Since mandrake 9.1 I've not been able to select a mirror for updates at the end of the installation process (even though this step is announced). This used to work in 9.0. This problem is still valid in 9.2rc2
Re: [Cooker] FHS 2.3 (fwd)
Bryan, I think you managed to drop the cc, I don't think Chris is cc'ed. One reason the proposal came up for /srv is that the fhs is commonly asked by sysadmins for guidance as to where they should put data for services. On linux at least, /var/www and /home/www is common, though neither seem to fit into the definitions for those areas. Is it so hard to broaden the definition of /var then? IMHO the definition doesn't get any broader anyway. In the case where you have a large volume of data in /var/www, you most likely have a database supporting that data in /var/lib/mysql, and you would have planned ahead for this and created a large /var, or partitioned it better to allow quota use etc. There is a distinction between the types of data we have in /var though. We have data that could be used by multiple pieces of software (/var/www by any web server, /var/spool/mail by different MTAs), then we have data that is specific to the software (/var/spool/postfix, /var/spool/imap, /var/lib/urpmi, /var/lib/mysql etc etc). The structure under /srv has been left undefined, with some suggestions as a non-normative rationale. Being that the structure under /srv is left undefined we still end up with the same mess. /srv/database/postgresql or /srv/postgresql or /src/disk1/postgresql Yes, I see no advantage to just moving most of /var into /srv. This does not solve a problem: /srv will still be a mess from distro to distro. Or, even worse, some distros may ignore /srv sticking to the previous LSB/FHS (undefined) waiting for clarification or changes in future versions of the FHS. So we'll end up with both /var /home /srv depending on the flavor of linux. And symlink mess between them, some symlinks in /var pointing to /srv/X, some symlinks in /var/lib pointing to /srv/Y etc. No, the rationale states that data of interest to a specific user should go in that user's home directory. Here is the full text of the rationale: This main purpose of specifying this is so that users may find the location of the data files for particular service, and so that services which require a single tree for readonly data, writable data and scripts (such as cgi scripts) can be reasonably placed. IMHO, scripts don't belong in /var (for samba we place supporting scripts in shell and perl in /usr/share/samba or /usr/share/samba3 which is IMHO a more sane location), but neither do they belong in some other location. We use /var/lib/samba for other non-user-specific data that should be available by default, so we have /var/lib/samba/printers for Windows printer drivers, /var/lib/samba/profiles for user profiles (which shouldn't really be in $HOME) and /var/lib/samba/netlogon for login scripts etc. I don't see the sense in placing this all in /srv. On existing servers, this will only mean more headaches for admins who must either symlink everything (and rpm doesn't deal well with changing a filesystem entry from a symlink to a directory and vice-versa so the admin has more headaches ahead with this method) or repartitioning (since / is likely too small to hold /srv too) or remounting /var/ (or some directory below it) under /src (which will likely require a reboot or at least dropping to single). And I don't see any benefit in this. Data that is only of interest to a specific user should go in that users home directory. The methodology used to name subdirectories of /srv is unspecified as there is currently no consensus on how this should be done. One method for structuring data under /srv is by protocol, eg. ftp, rsync, www, and cvs. On large systems it can be useful to structure /srv by administrative context, such as /srv/physics/www, /srv/compsci/cvs, etc. This setup will differ from host to host. Therefore, no program should rely on these locations. This assumes data is arranged and stored in seprate places based off of services. rsync, cvs, ftp, and www can easily share the same target data. Yup, we don't have FHS police :-) Admins do make local changes. In this case its more about setting up defaults for applications/packages to place or access their data. Then please keep it out of / . If this is just to be a dumping ground for apps to leave their default junk we don't need to clutter up / with another dir do we? /var seems to have always been the ugly stepchild of unwanted crap... I'm fine with adding more to it... ;) My issue is that I really don't like typing extra keys unless really necessary. One of the nice things about linux used to be that you could hit two keys per directory level and get quite far. Now I am going to have to type: /sb[TAB]/ifco[TAB] instead of /s[TAB]/ifco[TAB]. Please don't pollute / any more (one of the reasons I hate Windows is polluted root directories due to lack of structure by most users ...). The /media entry doesn't bother me... but I'd rather see them redefine /mnt to be what most use it for anyway: exactly
Re: [Cooker] Re: Versioning for multiple releases: a modest suggestion
On Fri, 7 Nov 2003 01:09, Buchan Milne wrote:
This would need
some rpm changes, so that a package built on 9.2 would have an
epoch of 92 and one build on 10.0 would have epoch 100. But this
would be a problem with packages that already has an epoch, as you
can use only integer numbers for the epoch tag.
It could also cause some problems for Requires/BuildRequires with
epoch values? Any BuildRequires/Requires with an existing Epoch value
would need to be bumped also, which makes for some additional
'macro-isation' for the packager, and makes it difficult to automate
this.
How big can the epoch be?
Existing, or absolute maximum?
$ rpm -qp --qf %{EPOCH}\n libarts1-1.1.3-7mdk.i586.rpm
3001
:-(
But, 30092001 is still bigger than 30091001
If you used release * 100 + existing, would
that work? So if a package was up to epoch 42, you would get 9242 for
the current release and 10042 for the same package aimed at 10.0.
Yes, this is the idea.
Next question, how about switching to even release numbers for stable
releases and odd for Cooker? That would make the epoch 9342 for the
pre-10.0 Cooker, 10042 for the 10.0 release, 10142 for the post-10.0
Cooker, 10242 for the next release (10.2, there is no 10.1), presuming
that no version bumps happened in between times.
I had wondered about possibly having an approach closer to what FreeBSD
has, where we have a 9.x branch and a 10.x branch, and after a certain
point, end-users should be happy to test with it, and at some point goes
into maintenance.
It might be better to combind both ideas, but this doesn't have too large
an impact on using the epochs or not (although it may be an advantage of
managing the epochs well).
This would violate tradition in that the releases would go 9.2, 10.0,
10.2, 10.4, 11.0,... but it would be understandable to anyone used to
the kernel versioning system and ordinary end-users shouldn't notice
much or care much either.
If we have room for one more digit, we could use that for the alpha,
beta and RC releases; so 93042 is the Cooker version of the package,
93142 is the alpha, betas run from 93242 to as high as 93542 then RCs
from 93642 up to (heaven forbid) a potential 93942, then 100042 for the
10.0 final.
I don't think the epoch should be abused like that. Sub-release numbers
(ie 0.alpha24.3mdk) are fine for this.
How much of this could we do without confusing people too much? How
complex would it be to set up the corresponding RPM macros?
I don't think rpm macros will be a good solution, since any buildrequires
with epoch values will need to be adjusted, so either we're going to be
doing shell arithmetic in rpm macros for the epoch values, or rpm needs to
be appropriately hacked ...
Regards,
Buchan
Re: [Cooker] syslog-ng as default?
On 2003-11-02(Sun) 12:58:24 +0100, Oden Eriksson wrote: I've been using it for a while as well and don't have any problem. I'm all for it, though I'd imagine we will be hearing Where's the old syslogd? WHAT'S GOING ON? in mdk 10 :-) Abel Tough, 10.0 is the first release in a new series, there will be a lot of new things (kernel-2.6, samba-3.0.x, KDE-3.2 etc) for people to get used to, they won't notice syslog for a while ...
Re: [Cooker] Want to do a magazine CD with 9.2 plus LG-proof kernel
On Fri, 7 Nov 2003 04:48, Pascal Terjan wrote: I'm pretty sure the magazine didn't even bother to tell Mandrakesoft they'll distibute the 9.2... OK... I'm planning to include a slightly-less-than-single-ISO version of 9.2 plus the updates and an LG-proof kernel on the cover of an Australian magazine. It will be aimed at a single purpose: installing a secure, database-backed Mandrake Linux webserver/mailserver/gateway machine for developers to use either for real or as a staging area or an intranet server/testbed. You might want to consider a live CD version ... (if it's prime purpose isn't for installation, but more for demo). What are the rules Mandrake would like me to follow? For example, if I include a PLF package or the NVidia/ATI drivers, what must I change? The distro's name? Licence terms? What can I do with the advertising screens? Are Mandrake happy for me to add extra images to the sequence, one to loudly remind the audience that this is not an official Mandrake distribution - it's been built for the magazine - one to advertise Linux contacts in Australia, and so on? You probably want to replace mandrake_theme at least. IMHO bootsplash-themes should have subpackages, which preferably provide mandrake_theme if they have the backgrounds and lilo splash screen. What else? Well, if we get an answer on this, can we get it on the Wiki. This really needs to be documented, considering Fedora is now a better option for this (all RH artwork removed, all fedora artwork in fedora* packages) Regards, Buchan
[Cooker] [Bug 6271] [kdebase] Konqueror has no option to change font as a file manager.
http://qa.mandrakesoft.com/show_bug.cgi?id=6271 --- Additional Comments From [EMAIL PROTECTED] 2003-11-03 16:51 --- BTW, try KDE Control Center-Components-File Manager It may be that the removal of the option was intentional, it may not be, but since the KDE team doesn't care to document what is a feature (such as double click settings not working?), we have no way of knowing. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: Open konqueror as a file manager. Settings-ConfigureKonqueror. There is no appearance option. KDE developers say it is mandrake's fault. http://bugs.kde.org/show_bug.cgi?id=67021 What have you done to kdebase? :-)
Re: [Cooker] Re: Konqueror bug
On Mon, 3 Nov 2003 00:35, Jan Ciger wrote: Leon Brooks wrote: | Mine breaks totally arbitrarily, http: or file: or fish: makes no | difference. Always seems to be text/html. Konqueror also dies | silently (ie, no bomb message) a lot more often than under 9.1. I filled a bug for this, please vote: http://qa.mandrakesoft.com/show_bug.cgi?id=6275 Click vote for this bug, get: Internal Server Error The server encountered an internal error or misconfiguration and was unable to complete your request. Please contact the server administrator, [EMAIL PROTECTED] and inform them of the time the error occurred, and anything you might have done that may have caused the error. More information about this error may be available in the server error log. Apache-AdvancedExtranetServer/2.0.47 (Mandrake Linux/4mdk) mod_perl/1.99_09 Perl/v5.8.1 mod_ssl/2.0.47 OpenSSL/0.9.7b Server at qa.mandrakesoft.com Port 80 I am sure there is a bug for this, maybe we should vote for it ;-)
Re: [Cooker] Re: [CHRPM] nss_ldap-211-1mdk
On 31-Oct-03 at 17:32, Buchan Milne ([EMAIL PROTECTED]) wrote: But, I don't see the problem: [EMAIL PROTECTED] bgmilne]$ grep ^passwd /etc/nsswitch.conf passwd: files ldap [EMAIL PROTECTED] bgmilne]$ wc -l /etc/passwd 38 /etc/passwd Same for me here, so far. [EMAIL PROTECTED] bgmilne]$ getent passwd|wc -l 187 Here I get the same symptom as before. # getent passwd|wc -l getent: relocation error: /lib/libnss_ldap.so.2: undefined symbol: dbopen 0 # getent passwd root root:x:0:0:root:/root:/bin/bash # getent passwd etutest1 getent: relocation error: /lib/libnss_ldap.so.2: undefined symbol: dbopen May be you have a ldap.conf file that does not trigger the call to dbopen? - could you send me the relevant part of your ldap.conf? Here is mine: # egrep -v '^#|^$' /etc/ldap.conf host myhost.unige.ch base ou=people,dc=unige,dc=ch ldap_version 3 scope sub pam_filter objectclass=posixAccount pam_login_attribute unigeChStudentUid pam_member_attribute gid pam_password clear nss_base_passwd ou=People,dc=unige,dc=ch?sub nss_base_shadow ou=People,dc=unige,dc=ch?sub ssl on sslpath /etc/ssl/certs/cert7.db Maybe it is this ^^^ ? (I don't see the point in wanting to verify the SSL cert against the commercial CAs when I use my own CA cery anyway, which is available and configured) nss_map_attribute uid unigeChStudentUid pam_template_login_attribute unigeChStudentUid That configuration did work with Mandrake 9.0. Here is mine, we have production machines running with configs like this, Mandrake 9.0 through 9.2: [EMAIL PROTECTED] bgmilne]$ egrep -v '^#|^$' /etc/ldap.conf host bgmilne.cae.co.za base dc=cae,dc=co,dc=za ldap_version 3 scope one pam_filter objectclass=posixaccount pam_login_attribute uid pam_password md5 nss_base_passwd ou=People,dc=cae,dc=co,dc=za nss_base_shadow ou=People,dc=cae,dc=co,dc=za nss_base_group ou=Group,dc=cae,dc=co,dc=za ssl start_tls tls_cacertfile /etc/ssl/ca.crt tls_checkpeer yes TLS_CACERT /etc/ssl/ca.crt Or maybe dbopen is provided by another library loaded at runtime not in the dependancies shown by ldd. It would be strange though: You would expect all symbols of a dynamic libraries to be resolved within the dependancies. [ Florin: is it OK that dbopen is not defined in the dynamic libraries? ] - Any suggestion on where to look now to figure this out? Maybe try without the certdb, since this seems to be about the only thing that has anything to do with libdb (besides libsasl)? Regards, Buchan
[Cooker] [Bug 6255] [kdeutils] kdiskfree has some problem detecting partitions
http://qa.mandrakesoft.com/show_bug.cgi?id=6255 --- Additional Comments From [EMAIL PROTECTED] 2003-10-30 11:32 --- Probably related to bug 4321 -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: It seems kdiskfree detects each partition twice... and only shows sizes for one of them. See the picture I am going to post. Note: I never edited /etc/fstab manually, but I mounted some partitions with diskdrake, during the installation of mandrake 9.2.
Re: [Cooker] Gnome Ghostview does not allow changing printer settings
Thanks, this is what I have tried to find out. I hope this kind of functionality could be integrated in future to printerdrake. (or configurations/gnome menu) Yes, a GUI for alternatives would be nice (and also something to check the existence of alternatives to catch problems with migrating to alternatives - see smbclient problems now and gcc problems back in 9.0 IIRC), but this is not printerdrake specific. Maybe we need a drakalternatives or something.
Re: [Cooker] Xboard fixes
Now, gnuchess application is in %_gamesbindir, I just fixed xboard, until I can't upload it, my rpms are here: http://compil.mandrake.org/~nanardon/xboard-4.2.6-6mdk.src.rpm http://compil.mandrake.org/~nanardon/xboard-4.2.6-6mdk.i586.rpm Hmm, someone restarted apache on klama? The one time it would have been useful (last week) it wasn't running ... /me wonders about abusing %post to start apache on klama ;-)
[Cooker] [Bug 6253] [kdebase] Cannot set konqueror to double click mode.
http://qa.mandrakesoft.com/show_bug.cgi?id=6253 [EMAIL PROTECTED] changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution||DUPLICATE --- Additional Comments From [EMAIL PROTECTED] 2003-10-29 14:02 --- *** This bug has been marked as a duplicate of 982 *** -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: RESOLVED creation_date: description: Fresh install of mandrake 9.2. Set konqueror's view mode to detailed list view. Notice it is in single click mode. I want to change that. I go to kde control center, peripherals, mouse. bug #1) The radio button says double click, but konqueror is in single click mode. bug #2) I try to select single click, then again double. konqueror is still in single click mode. I'm surprised this has not been fixed since mandrake 9.1. Is there a particular reason?
[Cooker] [Bug 982] [kdebase] file views in konqueror in detailed mode don't respect double click settings
http://qa.mandrakesoft.com/show_bug.cgi?id=982 [EMAIL PROTECTED] changed: What|Removed |Added CC||[EMAIL PROTECTED] --- Additional Comments From [EMAIL PROTECTED] 2003-10-29 14:02 --- *** Bug 6253 has been marked as a duplicate of this bug. *** -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: ASSIGNED creation_date: description: When KDE has been set to use double-click to open files/direcories, konqueror will still open files/directories on single-click when using 'detailed' view. When icon view is used, the behaviour is correct.
[Cooker] [Bug 5372] [pcmcia-cs] pcmcia service hangup
http://qa.mandrakesoft.com/show_bug.cgi?id=5372 --- Additional Comments From [EMAIL PROTECTED] 2003-10-29 14:09 --- KNOPPIX basically disables all the features that Mandrake has enabled at boot time (compare /proc/cmdline). Have you tried booting with the 'acpi=off apm=off noapic nolapic' options, which is basically what KNOPPIX uses, to see if one of those features is the problem? -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEW creation_date: description: Hi, I have recently installed mdk9.2rc1 on the laptop Compaq Presario 2100. The install was done without problems but on the startup the system just hangs on te Starting PCMCIA services. I have tried to disable the service by booting in single user mode, however in single user mode the keyboard simply doesn't work. I was able to resolve by bootting from the install cd and starting the install stage(on rescue keyboard didn't worked) and then switching to the console and mount the old root and disable pcmcia from the startup. The system booted fine without the pcmcia services. Note: I don't have any PCMCIA card on the sockets. Thanks
[Cooker] [Bug 6228] [fontconfig] Some sites have jagged fonts
http://qa.mandrakesoft.com/show_bug.cgi?id=6228 --- Additional Comments From [EMAIL PROTECTED] 2003-10-28 12:30 --- Fonts are anti-aliased by default in everything except GTK-1.x applications since Mandrake 9.1 in default installations. Either: 1)Try with a different user account or remove your ~/.fonts.conf (I think) 2)Tell us which version of freetype2 you have installed: $ rpm -q freetype2 Fonts look fine for me and have for a long time. Note, there are some other options available for users of LCD monitors, like sub-pixel hinting, which can be set in the KDE Control Center or with 'gnome-font-properties' Maybe it would help to attach a small screenshot of an affected site (otherwise it is difficult to compare fonts). -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: No matter what browser I use, and what fonts I use in the kde/gnome control center, when I go to some sites I always get nonantialiased, jagged fonts. For example the mandrake club. This is very ugly: just look at the italic fonts. Also non italic fonts are not ok: small fonts do look well, but medium and large fonts don't. I would like an option in the mandrake control center to say always use antialiased fonts for WEB-BROWSING. (On the other hand, I still want to use the default font in konsole, which is not antialiasing). Debian's fontconfig prompts you with that question. Better yet would be: always use antialiased fonts if the font size is over XXX.
Re: [Cooker] ldap + samba3 + pam_mount
Hello there, I'm trying to install a long awaited (by myself, because I only have 3 boxes) full ldap based authentification, with smb as network filesystem for both Linux and Windows worstation (I want to throw away nfs and anyway I have to use samba for my windows box, stupid ?). Be aware that NFS is currently the best generic (ie exlucing AFS and Coda) unix-to-unix file sharing system available, and NFSv4 should probably provide for the only reasons you would want to use smb over NFS. At present it is not possible to run GNOME or KDE on a SMB/CIFS-mounted home directory, even with a samba server running on a unix machine with unix extensions available (or at least it was last time I tested which was just before cifs went into the Mandrake kernel). The ldap based authentification and samba-ldap are fine working. Last step is therefore using pam_mount for home directory (+ few other shares). Are you using group mapping? If so, is it working (I have problems using the Windows User Manager for Domains under certain circumstances, but I have a bug open on it ...). Here is what I faced, could you please tell me if I'm wrong pam_mount is able to pass password to the stack, but not to keep it back from the stack - pam_mount must be stacked first in the auth section. No, this is not necessary. I don't know why, but if I stack pam_mount after pam_ldap in the session section, mount operation will abort with the following message : pam_mount: unable to open /var/run/pam_mount/foo pam_mount: received order to close things (...) su: unbind.c:40: ldap_unbind_ext: Assertion `( (ld)-ld_options.ldo_valid ==0x2 )' failed. (...) Maybe you need the use_first_pass option in your pam_mount line if it's after a real authenticating module. If I toggle pam_ldap pam_mount, things are ok, but I still have the following warning message (last line, /var/run/pam_mount is root owned): pam_mount: pam_mount: checking to see if //192.168.1.12/foo is already mounted at /home/foo pam_mount: creating mount /home/foo pam_mount: checking for encrypted filesystem key configuration pam_mount: about to start building mount command pam_mount: mount type is SMBMOUNT pam_mount: waiting for homedir mount pam_mount: command: /bin/mount mount -t smbfs //192.168.1.12/foo /home/foo -o username=foo,uid=foo,gid=foo,dmask=0700 Note taht you will have no chance at all to run KDE or GNOME on smbfs, with cifs you may get a bit further (relative symlinks actually work in some circumstances on cifs, but not at all on smbfs). Most of the other WMs work OK (I have tested at least fluxbox and WindowMaker). pam_mount: unable to open /var/run/pam_mount/toto Never seen that. The current cooker package works fine for me. [EMAIL PROTECTED] /]$ Does anyone has yet faced (and understood) the above errors/warnings (I've found nothing on it googling) ? Thanks in advance, Sébastien. PS : It's on an up to date 9.2 : pam_mount-0.9.2-3mdk pam_ldap-164-2mdk samba3-common-3.0.0-2mdk samba3-client-3.0.0-2mdk samba3-server-3.0.0-2mdk The last pam_mount version is the 9.4, I'll compile it and see if things are going to another way. I don't think that's the problem. The pam.d/system-auth used was -- authrequired /lib/security/pam_env.so authrequired /lib/security/pam_mount.so authsufficient/lib/security/pam_unix.so nullok use_first_pass authsufficient/lib/security/pam_ldap.so use_first_pass authrequired /lib/security/pam_deny.so account required /lib/security/pam_unix.so account sufficient/lib/security/pam_ldap.so passwordrequired /lib/security/pam_cracklib.so retry=3 minlen=2 dcredit=0 ucredit=0 passwordsufficient/lib/security/pam_unix.so nullok use_authtok md5 shadow passwordsufficient/lib/security/pam_ldap.so use_authtok passwordrequired /lib/security/pam_deny.so session required /lib/security/pam_limits.so session required /lib/security/pam_unix.so session optional /lib/security/pam_mount.so session optional /lib/security/pam_ldap.so -- Some comments: 1)I don't think it is useful putting pam_mount in system-auth, since I don't see any value having your smb share mounted when you read your mail via an IMAP on such a machine, or when you connect to a samba printer (if you use 'obey pam restrictions = yes') etc. Also, I have had some problems using pam_mount in system-auth (maybe it doesn't work too well with pam_stack) in the past. 2)I don't know if you want to make the pam_mount entry required or requisite, since it may make it impossible to do certain things if your samba server is unavailable (although I guess there could be scenarios where you don't want users to be able to log in in such a case). I use pam_mount by adding the
Re: [Cooker] Boot-up speed
Hi, - Sysv is already mostly parallelized. It takes 26 seconds which is better than I expected. Actually, as I understand, on this single-cpu machine the overhead of context switching is so high that xinetd and cups would probably start faster if they were not parallelized with X. - Idea: sshd is the only service that uses xinetd. Standalone sshd would probably boot faster? This computer connects to the Internet using a modem, I don't even own a network card. Therefore I probably don't need sshd at all. Isn't xinetd startup of ssh disabled by default, in favour of starting as a daemon? - While I'm at it, I also don't need netfs, nfslock (thus portmap) If you want to use NFS you need them all. If you don't use NFS, 'urpme nfs-utils-client' or whatever it is. AFAIK you need portmap for fam to work? or tmdns. Isn't it possible to disable these services if no NIC is detected during installation? Also since this is a desktop pc i don't need apmd. And you don't want to be able to use the sleep/wake buttons on your keyboard ;-). Can't this also be detected? - I just can't understand why kde takes so long to start!??!? Unfortunately, gnome isn't much better. The loading session part in kde can be accellerated by disabling KOrganizer (if someone wants to use it, he/she can find it in the menus) I don't think korganiser has a huge impact, and it's point is to remind you, not to tell you the things you want to be reminded of when you remember to run it, and it can easily be disabled. - Idea: postpone cups and scannerdrake startup for 60 seconds to give desktop a chance to start and avoid unnecessary context switching. Yay, then Mandrake samba print servers will now only be usable 2 minutes after having fully booted ... since we HUP samba a minute after starting it to give CUPS time to pick up all the printers (I haven't removed this since samba-2.2.8a can't update the printers automatically as it is supposed to - it worked in 2.2.7a though ...). It's very unlikely that someone would like to print or scan something in the first 60 seconds? Maybe on a desktop, and I guess it's not too bad on a server unless there's an MCSE driving it ... The same trick could be used for crond and atd I believe. What do you think? Should I apply these optimizations to my Pc? will anything burn up (except the lg cd-rom) ;) I wonder if the 23s to start KDE can be reduced significantly by pre-linking? BTW, Windows (even Windows 2000) shuts down very quickly (ie usually 5 seconds on my laptop), and I think the margin is greater on shutdown ... Regards, Buchan
Re: [Cooker] Fwd: [plug] Mandrake Bug announce warning rant ahead
On Tue, 28 Oct 2003 22:45, Rolf Pedersen wrote: Perhaps, at least in the current economic environment, it would be unwise to commit further, limited resources in the uncompensated service of those who merely just don't want to pay for software and (...) don't want to steal it either. I understand your point, but that's not Scott's issue. Scott's issue is that he _is_ a fully subscribed Club member and is on the appropriate announce lists MandrakeClub was the place this issue was first brought to light (in the MandrakeClub forum), about 19 October. I reported it on the maintainers list when I saw two different reports. IIRC that was 20 Oct. Since the only people who officially have access to the ISOs up to now are in MandrakeClub, this is the logical place to look. I also don't think it would have been responsible for the immediate action to be a story on MandrakeClub's front page without at least some verification/investigation, and that occured on Oct 20 and 21 on the maintainers list, whereafter it moved to cooker. 3 days after the 2nd report on MandrakeClub (IIRC) there was a note on the errata page (which really should be the place *everyone* looks before installing - and BTW we actually discussed that on maintainers before the LG issue and hopefully in future users will be urged to read the errata by the splash screen or similar). Just over a week after the 2nd report on MandrakeClub and we have kernels and full advisories out. I believe a lot of the threads about the LG drives were CC'ed to the expert list? I really don't think it could have been handled any better with current resources. but still found out about LG's faux pas first from other sources. Yes, I do find it strange LG didn't notify their customers who all paid hard cash for their hardware of this vulnerability in their hardware earlier, and urge them to do a firmware update. I wonder how long it will before (now that this vulnerability is more public) until there is a Windows worm which sends FLUSH_CACHE to LG cd-rom drives ... Regards, Buchan
Re: [Cooker] Kernel and Samba
Hi, My Samba box stalls after some usage, mapped drives disappear and users can't write or read from drives. The stalls happen randomly. I'm running 2.4.19-16mdksmp and Samba 2.2.7a-9.2mdk. May I ask is this a kernel bug or Samba bug? Does anyone know a fix for it? I checked the memory from BIOS, they didn't report any errors. BIOS memory check is (mostly) useless. Use memtest86 or similar. Also, your kernel seems a bit behind, I think current on 9.0 is 2.4.19-32mdk (though I don't think this should make that much difference). /var/log/kernel/warnings Oct 27 09:19:12 smbserver kernel: xfs_force_shutdown(md(9,5),0x8) called from line 1039 of file xfs_trans.c. Return address = 0xe08ae312 Oct 27 09:19:12 smbserver kernel: Corruption of in-memory data detected. Shutting down filesystem: md(9,5) Oct 27 09:19:12 smbserver kernel: Please umount the filesystem, and rectify the problem(s) This seems to point quite strongly to either hardware (most likely memory) or kernel (xfs driver or md driver, it seems you are running software raid?) If the kernel has problems with a filesystem, there's nothing much samba can do about it ... /var/log/kernel/errors -- Oct 27 10:36:44 smbserver kernel: Unknown bridge resource 2: assuming transparent Oct 27 10:36:44 smbserver kernel: PCI: Unable to handle 64-bit address space for Oct 27 10:36:44 smbserver kernel: PCI: Unable to handle 64-bit address space for Oct 27 10:36:44 smbserver kernel: Unknown bridge resource 2: assuming transparent Oct 27 10:36:44 smbserver kernel: PCI: Device 00:1f.1 not available because of resource collisions You need to give some more information on the hardware on this machine, but something does not look right ... what's in /proc/interrupts ? /var/log/samba/log.winbindd --- [2003/10/27 10:37:23, 0] nsswitch/winbindd.c:process_loop(626) process_loop: Invalid request size (1701996389) sent, should be (1304) Some winbind users have reported winbind in 2.2.8a works a bit better, you can find packages on the samba FTP mirrors for all supported releases (hmm, except ldap-enabled packages for 9.2 ... I must do this still ...). BTW: [EMAIL PROTECTED] bgmilne]$ cat /etc/mandrake-release Mandrake Linux release 9.0 (dolphin) for i586 [EMAIL PROTECTED] bgmilne]$ uname -r 2.4.19-32mdksecure [EMAIL PROTECTED] bgmilne]$ uptime 11:26pm up 165 days, 6:00, 1 user, load average: 0.01, 0.02, 0.00 [EMAIL PROTECTED] bgmilne]$ rpm -qa 'samba*' samba3-common-3.0.0-1.1mdk samba-server-ldap-2.2.8a-2mdk samba3-server-3.0.0-1.1mdk samba-common-ldap-2.2.8a-2mdk samba-client-2.2.8a-2mdk (This is our production samba domain controller for about 60 concurrent users, with 100GB of data shared on XFS over LVM on hardware RAID - never have problems with it). Regards, Buchan
Re: [Cooker] Re: LG Drives
On Sunday 26 Oct 2003 01:03, Greg Meyer wrote: On Saturday 25 October 2003 07:20 pm, Juan Quintela wrote: marc == Marc Guise [EMAIL PROTECTED] writes: marc I read the post about Mdk 9.2 and LG on mandrakeusers.org. I have cd-rw drive, marc model HL-DT-ST GCE 8400b and I have Mdk 9.2rc2 running. There are no problems marc with my LG drive 21mdk just updated (vdanen should be doing the official update) fixes that problem. Only LG plain CD-ROMS are affected. Later, Juan. PD. Yep, whoeved decided at LG that reusing for UPLOAD_FIRMAWARE command FLUSH_CACHE comand should be shoot. Twice. SO it turns out to be a firmware bug after all that. I really hope you guys don't take the heat for this in the court of public opinion. Yes, it is a firmware bug, and yes, the LG drives are responsible for this. But, it is Mandrakes own fault that this happens. I don't agree. If you take beta / heavy modified kernels instead of kernels that have been tested by the entire linux community, you can expect things like these to happen. AFAIK (I may be wrong), SuSE has been shipping with this patch for quite a while (AFAIK it is their patch), and there have been some reports of SuSE doing the same thing to LG drives ... And, if no-one found this problem before, in a well-known patch which is supposed to be merged into 2.6.x, how was it going to be found? True, strange that this bug was able to tunnel trough all RCs, but this once again proves that it's better to use the entire linux community as testers instead of a few beta testers, i.e. use only stock kernels. OK, so you want us to ship with a totally vanilla kernel? No XFS? OSS only (no ALSA)? No software suspend? No ACLs? No cloop? No CIFS? BTW, there is a vanilla kernel in contrib. How many people actually run it? Mandrake is not a linux distribution known for stability, let's please take stability as top priority for Mandrake 10.0. My servers tell me otherwise (some machines with clients I haven't seen in over a year because they haven't ever had problems with them). They wouldn't be running Mandrake without ACLs/XFS (which is one reason they don't run Redhat BTW, it's too much work to get a real samba server running on Redhat and a few other distros which have kernels without the features I need). This way Mandrake will get the good name that belongs to such a cool distribution. A distro with no features, that's cool. Not! Regards, Buchan
Re: [Cooker] Club RPMs?
Okay, I give up. Where is the @[EMAIL PROTECTED] GD urpmi line for commercial club rpms? The mirror script is just a list of ftps for cooker contrib. There is no script at all. I even tried the urpmi.setup, which immediately dies on my cooker box. Please, it shouldn't be this hard to find the fsckin' urpmi setup for club. Someone at Mdk pull their head out and put this somewhere quasi-prominent. It is THE primary reason for me to pay for club. You've tried: # drakclub ? (I haven't yet, on 28.8 dialup at home, and our local mirror doesn't have a full 9.2 yet)
Re: [Cooker] newer gideon/kdevelop3 packages
áéåí øàùåï, 26 áàå÷èåáø 2003, 18:47, ðëúá òì éãé [EMAIL PROTECTED]:
Please change your release to be a value which rpm will consider to be
less than 1, otherwise unsuspecting users who install your package
will be stranded with your package when a real kdevelop package is
released. This is a non-trivial issue, and a similar issue is
currently causing users of Tex's KDE packages to have problems when
upgrading to 9.2. Please apply the patch below before you distribute
this to too many people.
i see. it looks like a really ugly hack, but i understand why it's
needed.
It's not really a hack. It's policy (if you care to read about it in the
Manadrake RPM Howto).
patched locally, in the next release it will be available.
Thanks.
Also, this is bad practice (yes, I know it is used by KDE spec file
templates, but that is no excuse):
find . -type d | egrep kdevelep | sed
'1,2d;s,^\.,\%attr(-\,root\,root) %dir,'
$RPM_BUILD_DIR/file.list.kdevelop
because you will own many directories which you should not own (/usr,
/usr/bin, /usr/lib, etc etc). But you can fix those yourself.
so, what is the best remedy for that?
making the file list by my own?
The file list can be made automatically, but you cannot use a simple 'find
%{buildroot} -type d |sed -e s,%{buildroot},,g to choose which
directories you own. Since there aren't that many directories, yes, paste
your directory list into the spec file and take out any directories that
should be owned by another package (ie only leave those directories which
have the package name as the last component or something like that). This
might do:
find %{buildroot} -type d |sed -e s,%{buildroot},,g|grep '%{name}$'
Regards,
Buchan
Re: [Cooker] newer gideon/kdevelop3 packages
todo for next version:
* fix the icon problem (the app does not have an icon for some reason) *
compile also the dot program. apparently I am missing some
functionality.
Here are the links:
http://iglu.org.il/pub/Hebrew/diego/kdevelop/
http://iglu.org.il/pub/Hebrew/diego/kdevelop/kdevelop-3.0-cvs_20031026.i586.rpm
http://iglu.org.il/pub/Hebrew/diego/kdevelop/kdevelop3cvs.spec
Please change your release to be a value which rpm will consider to be
less than 1, otherwise unsuspecting users who install your package will be
stranded with your package when a real kdevelop package is released. This
is a non-trivial issue, and a similar issue is currently causing users of
Tex's KDE packages to have problems when upgrading to 9.2. Please apply
the patch below before you distribute this to too many people.
Also, this is bad practice (yes, I know it is used by KDE spec file
templates, but that is no excuse):
find . -type d | egrep kdevelep | sed '1,2d;s,^\.,\%attr(-\,root\,root)
%dir,' $RPM_BUILD_DIR/file.list.kdevelop
because you will own many directories which you should not own (/usr,
/usr/bin, /usr/lib, etc etc). But you can fix those yourself.
Regards,
Buchan
--- downloads/kdevelop3cvs.spec 2003-10-26 15:37:31.0 +0200
+++ downloads/kdevelop3cvs-fixed.spec 2003-10-26 18:39:29.0 +0200
@@ -19,7 +19,7 @@
Name: kdevelop
Summary: Integrated Development Environment for C++/C
Version: 3.0
-Release: %_date
+Release: 0.%{_date}
#Source1: kdevelop-%{version}_for_KDE_3.1.tar.bz2
#Source1: c_cpp_reference-2.0.2.tar.bz2
Re: Re[2]: [Cooker] rpm and missing gpg key gpg-pubkey-22458a98-3969e7de
bccz Because it's the key for security updates, and it should be imported for bccz you by urpmi.addmedia when you setup a security update source. Didn't know that. I have a local rsync mirror of security updates which are then available to the rest of the machines in my network. No problem. I just have to sync base dir also. Maybe next time you can give us information like this? You seemed to imply you were running urpmi.addmedia against an official updates mirror.
Re: Re[2]: [Cooker] 9.2 Updates a total mess
bccz Maybe you have an old page cached, but at least all the NVidia kernel bccz packages have been updated. I hope the ATI and winmodem drivers could also I know how to check for a page in cache. It wasn't it. bccz BTW, this is better than 9.1, where Mandrakesoft did not update any of the bccz proprietary kernel modules, and Club contributors did it all. True. bccz I don't think so. I think the current problem is not enough resources (ie bccz more people sending patches, maintaining packages would help). Well that's mandrakesofts problem. It's your problem just as mich (if you have issues). If they don't have enough money to hire people they should ask for more volunteers. Development is totally open. If you want to help, you can. There is no need to ask for developers, because the policy is clear. Volunteers are welcome and always have been. bccz If you're going to use shorewall, get Florin's updated packages. Shorewall is a nice but when you have to build a firewall for network which has 150+ users and 15+ classes of users that won't do. Your point? bccz I haven't seen the menu problem on my cooker box, so I don't agree. And I bccz think the fact that many bugs have been fixed by updates so fast is a good bccz thing. Well if you haven't seen it a loot of others have. But that doesn't help fix it. Mandrakesoft has problems reproducing it too (and 'just install a package' isn't a reproducible test case), so it's difficult for them to solve it. Don't get me wrong updates are always good but they also show that 9.2 wasn't tested enough. Or that bugs took longer to solve than the time available before they had to meet their commitments. Of course, Mandrake didn't go out telling people not to test this release ... bccz Most people are just clueless, and run Redhat because everyone else does, bccz Instead of dissing a distro you have significant investment in, maybe you bccz could actually help to improve it? Many of us do, and that is one reason bccz why it *is* actually worthwhile. O now you really got me going. Well let see. I few years ago I needed postfix with mysql support. postfix in mandrake didn't have it. I downloaded srpm and found that mysql was in but switch in spec file wasn't activated. It took quite a while for a new package to appear. Even one security update was released but the switch wasn't activated. Once more i had to build it myself. Have you considered that maybe Mandrakesoft doesn't want postfix to depend on mysql by default, but *does* want to make it easy for someone to build the package with mysql support All you need to do is: $ rpm --rebuild --with mysql postfix*.src.rpm Luckily, ldap support is on by default. What's the big deal? I had to avoid using automatic update procedure since it would break my mail server. Yes, this needs to be solved in some way (but building postfix with mysql isn't the total solution, rather we need to deal with custom packages better). Anyway, you could just have added postfix to your /etc/urpmi/skip.list. Next: The same time I needed cyrus-imapd but no mandrake didn't have it. Strange it was in 7.0/7.1. let me check. I pull from the closet my old powerpacks and see that cyrus imapd was inside. Ok again the same thing. I make a cyrus-imapd rpm. Test it for a few days. everything is ok. After 2 years it's still working happily. After some time I clean that spec file and send it cooker list. Just wonted to help out. Nobody cares. 2 years later cyrus-imapd still is not in standard set. No, it's in contrib, along with courier and binc and dovecot. There was resistance to having two IMAP servers in a desktop distro. IMHO, we should replace UW, and have Cyrus in main (reclaiming the space sendmail wastes ;-)). I wasted all of my votes on it as so many others did but nobody listens to us. So, then why is cyrus-imapd in contrib now? Why are there people who are working on making it easier to set it up? The newest msn chat client or some other crap is probably more important. Actually, it's not (as the gaim users will probably complain). Next: a couple of posts concerning bugs missing functionality etc. all passed by nobody cared. Please, post your bug numbers, otherwise there is no reason to give any credence to reports like this. Next: mandrakeclub been a member since it was made. It wasn't worth the money but still I'm a member. Why ? because I want to help. It looks like nobody cares. How so? bccz Oh yes, of course you should always read the errata before installing. And bccz maybe someone needs to add the menu issue to it for now. How is errata being built. Look for yourself: http://qa.mandrakesoft.com/twiki/bin/view/Main/Mandrake92, look at the older revisions of that page to see the history of the errata, and you will see a number were up before ISOs were released. Some user has complained that that is not working. Possibly in a bug report on
Re: Re[4]: [Cooker] 9.2 Updates a total mess
bccz Development is totally open. If you want to help, you can. There
is no bccz need to ask for developers, because the policy is clear.
Volunteers are bccz welcome and always have been.
???
Don't have to ask developers ? If I want something to be changed in ML i
will post a message to cooker list.
Sometimes it is more effective to post a patch, and/or file a bug.
Sometimes not.
If other people want the same thing
then it is up to the developers to make the change. After all they are
the ones who have the control over the final product.
But you seem to infer that only Mandrakesoft people are developers, and
exclude yourself from possibly being a developer.
bccz Your point?
My point ? If you want to do it wright you do it yourself.
I do. Do you?
bccz Or that bugs took longer to solve than the time available before
they had bccz to meet their commitments.
bccz bccz Of course, Mandrake didn't go out telling people not to test
this release ... maybe the best solution would be a kind of a fork to
create a
perfect ml 9.2 (9.2.1) apply all of the updates over the next month or
so and post a howto for people who wont to build their own distrib.
isos.
The question is, would the bugs be found in 9.2, or would people think
9.2 is just an RC, I will wait for 9.2.1, and then you end up in the
same situation?
There have been a question on cooker list concerning this but I haven't
seen a god answer for it. I'm currently truing to do the same thing
myself.
bccz Have you considered that maybe Mandrakesoft doesn't want postfix
to depend bccz on mysql by default, but *does* want to make it easy for
someone to build bccz the package with mysql support
No. It was a mistake. The mistake got corrected a few releases later.
Anyway postfix doesn't depend on mysql it only has a module in it
built to support mysql.
Which depends on MySQL, and can't be packaged seperately (AFAIK).
From:
http://cvs.mandrakesoft.com/cgi-bin/cvsweb.cgi/SPECS/postfix/postfix.spec?rev=1.63content-type=text/x-cvsweb-markup
%define with_MYSQL 0
[...]
%if %{with_MYSQL}
Requires: MySQL, MySQL-client
BuildRequires: MySQL, MySQL-client, MySQL-devel
%endif
bccz All you need to do is:
bccz $ rpm --rebuild --with mysql postfix*.src.rpm
Yeah i know how to build rpm.
bccz No, it's in contrib, along with courier and binc and dovecot.
There was bccz resistance to having two IMAP servers in a desktop
distro. IMHO, we should bccz replace UW, and have Cyrus in main
(reclaiming the space sendmail wastes bccz ;-)).
Yeah I know it's in the contrib but we voted for it to be in the main.
And maybe if can be done for 10.0, and we can put UW in contrib? But you
have people upgrade from 9.1 to 9.2 and lose support for the IMAP server
they were using.
We are making progress, but it is slow, but it doesn't help if you
complain ...
bccz So, then why is cyrus-imapd in contrib now? Why are there people
who are bccz working on making it easier to set it up?
Yes I know that it is in the contrib thanks to the votes.
Look the main problem here is that some of us need some new features
which are not so common. If added they would not make any difference to
the most people. Most wouldn't even notice them but it would make a
great difference to some of us since it would be supported by
mandrake.
Example cyrus-sasl doesn't have a support for mysql or postgre.
If added this wouldn't make any difference to most of the people.
Most of the people don't even use it.
But it would make a great difference to me since the next time a
security patch is released I wouldn't have to go try the same process of
patching building and testing. I don't have enough resources to keep all
of the old mandrake releases installed so I could build a few rpm when a
bug is found.
So, send your patches. If Florin won't take them directly (he often
doesn't manage to look at cyrus stuff), one of the Luca's will take a look
at it, or I will.
I needed LDAP support in samba-2.2.x, and since it couldn't be available
at run-time, I made it available at compile time. In samba3, xml and mysql
are optional at runtime, so I have provided the additional packages (and
hope to do the same for LDAP so the normal samba packages don't
necessarily need to depend on ldap).
Things *do* happen when the people who need them contribute ...
Maybe a moderated feature request list should be built for mandrake
developers as a reference point.
http://qa.mandrakesoft.com/wiki
I meant to summarise some of the authentication-related aspects already,
but haven't had time.
New features are always welcome as long they don't affect stability.
bccz Please, post your bug numbers, otherwise there is no reason to
give any bccz credence to reports like this.
Already wasted to much time writing this. Don't have the time to go try
cooker archive.
OK, waste our time then ...
bccz Possibly in a bug report on the RC or similar. How else do you
make an bccz
Re: [Cooker] kernel-source bug in 9.2
Hi Not sure if this is the right place to ask/report ?? Is this a bug? Trying to compile a kernel with gcc-2.96 using: Well, this is not supported AFAIK. The default compiler is gcc-3.x, and that is what is supported for compiling kernels. Do I need to report this as a bug to MandrakeSoft Bugzilla? Please don't, this isn't our bug I think ... I need this kernel compiled with gcc-2.96 for my satellite internet connection card driver. Get your hardware vendor to provide an update binary object for your kernel module, or use an ancient kernel. Regards, Buchan
Re: [Cooker] rpm and missing gpg key gpg-pubkey-22458a98-3969e7de
G it looks like that I need this key :gpg-pubkey-22458a98-3969e7de G So where can i find it ? Newer mind already found it on http://www.mandrakesecure.net/en/cks/search.cgi?stype=keyid_4bkeyid_4b=22458A98 Why wasn't this key imported the first time ? Because it's the key for security updates, and it should be imported for you by urpmi.addmedia when you setup a security update source. See: ftp://mandrake.contactel.cz/Mandrake/updates/9.2/base/pubkey The reason the keys differ is that they are fundamentally different packages. This also has the advantage that you can assign certain keys to certain urpmi media, so for instance if packages signed with the updates key appear in main, they will not be verified. This gives you an extra level of control about ensuring only the rigth packages get installed. Regards, Buchan
Re: [Cooker] msec fscked me over. good
On Saturday 25 October 2003 06:14 pm, illogic-al wrote: up until this morning i was happily at msec 4. i changed one of the default settings and now i can't su to root from anywhere. Either in X or at console. Help. I don't wanna reinstall. I have 9.2 btw but once again man prevails over machine! or man prevails over sh ... uh crappy that is, software in this case. Somehow msec decided to change the suid bit on /bin/su Well, it didn't decide that by itself, since su does not feature in the default permissions: $ grep su /usr/share/msec/perm.? /usr/share/msec/perm.0:/var/lock/subsys root.root 755 /usr/share/msec/perm.1:/var/lock/subsys root.root 755 /usr/share/msec/perm.2:/var/lock/subsys root.root 755 /usr/share/msec/perm.3:/var/lock/subsys root.root 755 /usr/share/msec/perm.4:/var/lock/subsys root.adm 750 /usr/share/msec/perm.5:/var/lock/subsys root.root 700 i followed the instructions here after some judicious googling: http://www.ornl.gov/cts/archives/mailing-lists/tru64-unix-managers/2003/05/msg00025.html and now i'm back in business. Lesson for the day, DO NOT allow msec to run cron jobs I think you're going to have to look a bit further for that lesson. All my servers run msec 4 with some small customisations, and it's just great IMHO (with msec running from cron). Regards, Buchan
Re: [Cooker] Error buildin Openoffice.org1.1 0.rc4.2mdk
On Friday, October 24, 2003 07:52 am, R N dev wrote: Can anyone help me? Angelo = Building project readlicense_oo = /usr/src/RPM/BUILD/oo_1.1rc4_src/readlicense_oo /usr/src/RPM/BUILD/oo_1.1rc4_src/readlicense_oo/html /usr/src/RPM/BUILD/oo_1.1rc4_src/readlicense_oo/docs/readme - java -classpath /usr/src/RPM/BUILD/oo_1.1rc4_src/external/common/xt.jar:/usr/src/RPM/ BUILD/oo_1.1rc4_src/external/common/parser.jar -Dcom.jclark.xsl.sax.parser=com.sun.xml.parser.Parser com.jclark.xsl.sax.Driver readme.xrm ../readme.xsl os1=LINUX gui1=UNX cp1=INTEL com1=GCC lang1=en-US type=html file=../../unxlngi4.pro/misc/01/readme01.html FIXME: Emulate java runtime with gij here dmake: Error code 1, while making '../../unxlngi4.pro/misc/01/readme01.html' ---* TG_SLO.MK *--- ERROR: Error 65280 occurred while making /usr/src/RPM/BUILD/oo_1.1rc4_src/readlicense_oo/docs/readme dmake: Error code 1, while making 'build_all' ---* TG_SLO.MK *--- error: Bad exit status from /var/tmp/rpm-tmp.88932 (%build) I can't help, but I'd be extremely interested in the solution. I posted this problem last month in Mandrake Expert (because I was building for 9.1 rather than cooker). The only responses were two off-list emails asking if I'd figured it out. OK, well, it ovbiously looks like it's dying in some java/gcj/gij section. Unfortunately this stuff is a mission to work with (trust me, I rebuilt OO.o-1.0.1 for 8.2), since it depends on the features available in gcj etc etc and normally needs the latest features. Anyway, you will need to give a lot more detail on the system you are running this on for anyone to have a chance to help you. What Mandrake release is it? Do you have a jre installed? Do you have gcc-gcj installed? (and which version for those of us who can't remember what it was on the release you are running) Which 'java' binary is first in the path? Aare you compiling with any switches to rpm (ie --with switch)? Regards, Buchan
Re: [Cooker] Re: LG Drives
On Saturday 25 October 2003 07:20 pm, Juan Quintela wrote: marc == Marc Guise [EMAIL PROTECTED] writes: marc I read the post about Mdk 9.2 and LG on mandrakeusers.org. I have cd-rw drive, marc model HL-DT-ST GCE 8400b and I have Mdk 9.2rc2 running. There are no problems marc with my LG drive 21mdk just updated (vdanen should be doing the official update) fixes that problem. Only LG plain CD-ROMS are affected. Later, Juan. PD. Yep, whoeved decided at LG that reusing for UPLOAD_FIRMAWARE command FLUSH_CACHE comand should be shoot. Twice. SO it turns out to be a firmware bug after all that. I really hope you guys don't take the heat for this in the court of public opinion. /. posted a really badly researched post on this (with a bad title too), but it seems public opinion is that hardware shouldn't be vulnerable. It wouldn't hurt though to have an official position sent to /., I think they should reasonably update the summary to indicate that this is a hardware issue that would also affect other users, and hopefully push the availability of updates and how to use them. Regards, Buchan
Re: [Cooker] network icon on deskop?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Brad Felmey wrote: | I will smile and politely pat you on the shoulder, understanding | perfectly that you obviously have zero experience migrating an | enterprise environment to Linux. Well, yes, you are right, but what I do not agree with is putting a link to smb:/ on the desktop, unless it is made more reliable. It works great for me (at least, it shows all the workgroups the windows machines show, even if we can't access them anyway since we have firewalled ourselves off from the worms long ago ...). It does not work properly for me (I am current on Cooker) - I am seeing some shares from our Windows server but not others, Have you authenticated correctly to the server (setting in KDE Control Center-Networking-SMB Shares I think)? some directories seem to be empty even though they are not and many other strange things like this. Just recently, I have had to hit refresh when entering some directories. This wasn't necessary just a few weeks ago ... but I assume if can be restored to work the way it did before). Anyway, it seems your samba browsing is somewhat broken (or you are in an AD-only network without a WINS server or something like that). Also, it should be better when we can compile kio_smb against libsmbclient again (as soon as cooker is open and I can put samba3 in main). Either use LinNeighborhood (which works fine for me) or nothing for now, because a half-broken solution is worse than none. LinNeighborhood works worse for me (WINS support doesn't really work correctly I think). Regards, Buchan
Re: [Cooker] network icon on deskop?
Le jeu 23/10/2003 à 23:04, [EMAIL PROTECTED] a écrit : I love you , u know it ? ;) Don't tell my girlfriend ;-) 100% agree. Lindows do it too. it's true that lisa ( and gnome-vfs2 ) should be improved so that we could have the ability to display everything with a good organisation. I will see if I have the time to look at some of this ... but I suspect not :-(. But at this time, with what we have, we should do the best and there it's putting this icon on the desktop. That's simple ! Exactly! Note : nautilus need to be patch so that it doesn't display this icon ( cf Home Trash case ) Actually, I think it should work fine? I tested an icon with smb:/ URL, and it displayed the workgroups in Nautilus when opening it. Just, I think gnome-vfs2 isn't quite up to it quite yet, Nautilus doesn't know what to do with the SMB shares (asks which application to use to open it ...). Of cousre, lan:/ does not work, which is one of the reasons for going for smb:/ for now instead of lan:/. Regards, Buchan
Re: [Cooker] Boot Linux Faster
Hello, I don't know if any one else saw this article but it seemed very good to me. http://www-106.ibm.com/developerworks/linux/library/l-boot.html A chap at IBM Hursley, in the UK, has come up with a way to make Linux boot faster using make. And there have been other efforts too, such as serel (http://www.fast-boot.org IIRC). However, the speed advances aren't that exceptional, we could get at least 50% of the speed increase by not running depmod on every boot (which most of these methods avoid anyway ). I have not got around to trying it out. However, given the speed that XP now starts up compared to most Linux setups I would think anything to get the boot times down must be good. I responded to this e-mail asking why everything needs to be started at boot time. For example it takes a huge amount of time to get CUPS up and running. Why not defer this until the first print job is required ? Well, in any network, CUPS would need to be running to allow the printers to be visible to other machines (whether via CUPS or Samba). Also, how do you start the service up for the user when they want to print? BTW, XP starts most services (look for spoolss.exe in the task manager before you print something) automatically, but I notice that (at least under Windows 2000), some things are only started up when a user logs in, and Windows 2000 Server takes an age to Start network services (although this could be due to some excessive time spent on DNS lookups for it's domain if the internet is not available ;-)). It appears that extension of his idea could be to work out which services really need to be started at boot time and which might be deferred. First step is to not run depmod on every boot, which will save 10-15 seconds for everyone with minimal changes. Regards, Buchan
Re: [Cooker] network icon on deskop?
On Thursday 23 October 2003 04:04 pm, [EMAIL PROTECTED] wrote: (sorry, but it's just depressing to see a feature which worked great in 9.0 trashed, when there is only one other distro with the same feature, and they get rave reviews for it ... and I don't feel confident enough in it being working sufficiently in 9.1 and 9.2 to actually do any work advertising it - even if it was responsible for getting 10 desktops into an enterprise windows network) yes can you say corel linux way back when had this it was great, xandros has it now as well as lindows, oh and dont forget about ark linux which does this just fine also. Hmm, AFAIK it's only Xandros which can actually do full domain membership as we do with Winbind. Most other solutions are hacks (like RH's use of pam_smb which is only about a quarter of a solution). We are more than a little behind on this if so many others can have it working just fine and in the case of corel that was literally years ago. IIRC, Corel didn't have domain membership working. And their fancy file manager is proprietary. It is a testament though that bero with ark linux has had it working for about a year now also. Anybody here remember when bero joined mandrake to help setup the i586 builds besides me? I answered one of Bero's questions on samba last week (and it seems they're implementing something like diskdrake's smb mount support), so I don't really think we are that much behind Ark anyway. And to however above that said this is a bad idea you have no idea what you are talking about. This is a sorely needed feature. And there are a lot more (but some of them won't be feasible for a while, such as full Kerberos integration etc to make this all seamless in the right environment). On a side note I love admining my boxen with kio_fish it is so nice to edit files from the gui and then just restart the deamon from a terminal. /me only trusts vi ;-). But, I hope libConf gets taken seriously for 10.0 it has a lot of potential, and I wish I had the time to actually start working on the front-end bits for samba and openldap ... Regards, Buchan
Re: [Cooker] Some wishes for Mandrake 10.0...
| Never call a ntp server with cron! This is the best way to overload a ntp | server if several people do that. I think that the original idea was not to have non-stop synchronized clock (aka ntpd) but just the possibility to e.g. once a day synchronize clock with some external server. For desktop machines, this is usually Change your thinking to a large scale sysadmin. Set up an NTP service on one of your boxen and then configure all of your clients to sync in whatever way you want to your local master. To avoid the server overload (100+ machines syncing their clock at the same time), the script could be smart and randomize the time when it No need. Again, set up your own NTP server, make all the local machines sync to that one local master. Of course, some solution for them to find the NTP server automatically (DHCP/DNS) would be nice (like _ldap._tcp kerberos-style SRV records do for nss_ldap/pam_ldap etc). Regards, Buchan
Re: [Cooker] 9.2 Updates a total mess
Next : On this page http://www.mandrakesecure.net/en/advisories/advisory.php?name=MDKA-2003:020 Mandrake says that all of the mandrakeclub commercial drivers are available. Quote: New commercial drivers for this kernel are available at Mandrake Club. Now my English is a bit rusty but this should mean that they are there wright NOW ! I don't see them here : http://rpms.mandrakeclub.com/rpms/MandrakeClub/comm/9.2/i586/ByDate.html Maybe you have an old page cached, but at least all the NVidia kernel packages have been updated. I hope the ATI and winmodem drivers could also be updated, if not I will do at least the ltmodem drivers (but if Mandrakesoft isn't going to take care of this they could at least make the SRPMS available to contributors - my NVIDIA and ltmodem packages have been done mostly from scratch). BTW, this is better than 9.1, where Mandrakesoft did not update any of the proprietary kernel modules, and Club contributors did it all. Also, it would be nice if we could be warned in advance for some updates, as if I had known .18mdk would be the kernel update, I could have had win4lin kernel updates out at the same time, instead I built packages yesterday (remotely) and haven't had a chance to test them yet ... although Netraverse has released some but their packages have some issues . After updating my system KDE lost its desktop and also menus. Deleting a user and adding it again didn't work. I wasn't even able to run rxvt since alt+f2 didn't work. Luckily there was menudrake in the menus so i restored the original menus. You people need a reality check. I don't think so. I think the current problem is not enough resources (ie more people sending patches, maintaining packages would help). And some really heavy quality assuring since this is really unprofessional. I don't remember something like this happening before. Yes, Redhat didn't manage to stuff rpm up this badly before ... Now desktop issues don't bother me really but if such obvious things are happening what is going to happen with nonobvious. I'm currently building a new firewall/ruter using 9.2 and I'm starting to question myself should i do it with 9.2. I can't afford such stupid things to happen on a central firewall in my network. If you're going to use shorewall, get Florin's updated packages. Now how do you think this looks in the eyes of a desktop user ? It shoes that Mandrake is really unprofessional. I haven't seen the menu problem on my cooker box, so I don't agree. And I think the fact that many bugs have been fixed by updates so fast is a good thing. It's no wonder that people look at me weird when a say that i use Mandrake for all of my servers. Most people are just clueless, and run Redhat because everyone else does, Debian because they think they don't need newer software releases, or Gentoo because they think that cool compile flags will really make their firewall run faster. Maybe I should try some other distribution but since i have been using Mandrake since 7.0 or so it would be really sad. Instead of dissing a distro you have significant investment in, maybe you could actually help to improve it? Many of us do, and that is one reason why it *is* actually worthwhile. Oh yes, of course you should always read the errata before installing. And maybe someone needs to add the menu issue to it for now. BTW, doesn't logging out and back in fix it? update-menus should be run from /etc/X11/xinit.d I think?
Re: [Cooker] Re: Huge List of Updates
ef2 == ef2 [EMAIL PROTECTED] writes: ef2 If you want that more people test the beta versions, you need to consider ef2 the human psychology : the more the figure will be big, the more people will ef2 want to try the distribution. I even suggested after 9.1 to lie about the ef2 actual release date. Maybe this would not be really useful though. I am thinking that if the solution is to apply psycology, we will have to told people that we are going to have only 1 RC and then doing 3, then people will begin to test RC1 instead of beggining testing in RC3 :p And then what do you do for the next release? Start at RC3, up to RC5? The only solution is to make people who are going to report bugs do it by RC1 at the latest. Maybe there should be a policy on bugs ... bugzilla should not allow new bug submissions from when RC2 is released ;-). And maybe only bugs that were reported before RC1 can be voted on?
Re: [Cooker] network icon on deskop?
On 10.23, Brad Felmey wrote: On Thu, 2003-10-23 at 11:40, Buchan Milne wrote: So, when can we add an icon on the desktop for smb:/ in KDE and GNOME by default, called something like Local Network (I have one on my desktop - - actually called Windows Network, and it works great). I concur. This is exactly what I've been doing with my users. Just put a konq link on the desktop called Windows Network, and point it to smb:/. I also added it to the personal toolbars in Konq, so it's easy to get there from wherever. It's seamless to them, and really went a long way toward their happiness in Linux. I disagree here. Why is everybody so found on making Linux look like Windows ? How about the fact that migrating an enterprise network means catering to 95% of users who have never seen anything but Windows? Of course, this issue would be much less pressing if there were some way to control desktop shortcuts globally (at least pre-desktop) without having to script it all. Linux is linux, it is great, and works its way. Why everybody thinks that what Winblow$ does is sliced bread wrt user interfaces/interaction ? So, it is a good idea to hide *all* the functionality of the OS? Many people have problems working with the Windows machines (make that 95% of the machines they access) on their network. I would like to spend less time answering but where are the shares on the network? which could be achieved by this. That is the never ending problem. You can make linux _look_ like windows, but never _behave_ like windows (unless you fill it up with bugs ;). I didn't ask to be able to right-click on the icon and choose properties. I just think that it should be readily apparent to newbies (who are otherwise clueless about ssh, scp, kio_fish, NFS, rsync, smbclient, smbmount etc etc etc) that Linux is viable on a Windows network (or other network for that matter). The user comparing it will always say 'mmm, almost, but not the same, this is a windows-wannabe'. But that's better than it can't even network with my office lan at all. BTW, if this ever comes to mdk, plz think that a network neighborhood is not a 'windows' nbh, you can also see nfs shares... Learn from osx, the network browse window shows AppleTalk zones and SMB zones, you do not distinguish them. Well, at present lisa is so depressing that I can't really suggest lan:/ over smb:/. If lisa is improved (to show something like this: LAN Browser: -All machines (current content of lan:/, showing all protocols) -Windows-compatible (smb:/) -Netware-compatible (as yet non-existent ncp:/) -Unix-compatible (as yet non-existent nfs:/ or autofs:/) -Directory (a better ldap:/, with working AD and NDS support) ) then I would agree. But, in a network with more than one subnet and/or workgroup, lan:/ is not able to do as well (for smb) as smb:/ (trust me). Since there is no precedent on unix for even needing to show an NFS browser on the desktop, I don't think it's such a high priority (since autofs can take care of it anyway, but can't do the same for most smb shares). We could have had some cred for shipping with an smb:/ icon (since it is trivial to do), and do plan on the rest for later. Or we could just wait for everyone else to have it first. Innovation, here we come. And maybe someone can actually make drakconnect work sufficiently for winbind authentication setup to work during installation for people other than RHCEs and LPI cert holders. (sorry, but it's just depressing to see a feature which worked great in 9.0 trashed, when there is only one other distro with the same feature, and they get rave reviews for it ... and I don't feel confident enough in it being working sufficiently in 9.1 and 9.2 to actually do any work advertising it - even if it was responsible for getting 10 desktops into an enterprise windows network) Regards, Buchan
[Cooker] [Bug 4002] [galaxy-kde] using galaxy the color of the buttons doesn't appear
http://qa.mandrakesoft.com/show_bug.cgi?id=4002 [EMAIL PROTECTED] changed: What|Removed |Added Status|UNCONFIRMED |NEW Ever Confirmed||1 --- Additional Comments From [EMAIL PROTECTED] 2003-22-10 12:29 --- The bug can also be seen with kcalc, by setting colours for the buttons (if this is not the default). When switching to kcalc when it is behind another window, you will see the buttons have a background colour assigned correctly, but the Galaxy theme then paints over them. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEW creation_date: description: use Kalzium to test this. Kalzium shows a quimical table with colors in all quimical components. with galaxy theme this doesn't work (the colors are missing), with any other theme this works fine.
[Cooker] [Bug 6195] [Installation] New: Users should be encouraged to visit errata page before installing
http://qa.mandrakesoft.com/show_bug.cgi?id=6195 Product: Installation Component: isolinux Summary: Users should be encouraged to visit errata page before installing Product: Installation Version: 1.810 Platform: PC OS/Version: All Status: UNCONFIRMED Severity: normal Priority: P2 Component: isolinux AssignedTo: [EMAIL PROTECTED] ReportedBy: [EMAIL PROTECTED] Many users have issues during or just after installation (ie first boot), which could be avoided if they were aware of the errata before starting the installation. Ideally, we want some way to ensure that the user has seen the errata list before even initiating an installation (web browser in stage1 ? ;-)). Probably the easiest highly visible solution would be to include a message in the isolinux/syslinux boot screen to that effect. Additionally, users who had previously used proprietary drivers may have issues (especially users with drivers which are necessary for their internet connection) may be unaware that they should obtain new drivers before installation. The message should be quite strongly worded (ie Before considering installation, please see the errata for this release at ... . Additionally, if your hardware requires proprietary drivers, please ensure you have access to them before continuing.), but maybe a bit more brief ... -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is.
Re: [Cooker] When can we finally get X to fall back to XFdrake??
Le Mardi 21 Octobre 2003 20:42, Buchan Milne a écrit : A long time ago, X used to fall back to running a configuration tool. This may have been long ago, when XFdrake wstill had a link for XF86configure (or whatever Redhat used to call their tool). Why, if a user makes a mistake, do they end up with a console login, when they could have XFdrake instead to help them fix the problem? See the comments in the OSNews review to see why I am asking this again. Windows95 even had a safe mode which would help you reoconfigure your display settings. 8 years later, and Mandrake doesn't. I know it's not the right place but: I can't find how to run windows in console mode Recovery console in Windows 2000 and Windows XP. But, there are very few times you actually need it, because you can ... or even failsafe. Hit F5 or F8 during boot (since Windows 98, maybe even Windows 95). On all versions of Windows since then, that will bring you up in VGA mode and load no drivers which aren't necessary. I have never seen a Windows system that will not get into safe mode. In Windows NT4, this is the VGA Mode boot option in the boot loader. Do you known how to do? The question is, can Mandrake also provide a solution for users who don't know all the console tools (whether MS has them or not is irrelevant, since you would never need them) can still recover from choosing the wrong display settings, or needing to choose a different display driver? Note, I don't say hardware configuration under Windows is the greatest (I have spent lots of time fighting hardware support in Windows, like about 2 hours of work to get it to work with the IDE controller on a different M/B when setting up hardware profiles for different machines on the same disk), but surely we can do at least as well in ensuring newbies don't end up with: localhost login: and believe the FUD about linux being too difficult for the average user to use. Regards, Buchan
[Cooker] [Bug 6162] [Installation] Advanced LDAP authentication Option fails
http://qa.mandrakesoft.com/show_bug.cgi?id=6162 --- Additional Comments From [EMAIL PROTECTED] 2003-20-10 13:13 --- You can now access the authentication setup by running: # drakauth (please list the packages it installs ... since they should have been installed during the authentication setup). My tests with 9.2rc2 worked well (I don't have 9.2 final yet). Could you upload the relevant contents of /root/drakx ? -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: I specified my Mandrake 9.1 LDAP authentication server durring the installation and the client machine doesn't appear to have been configured for LDAP authentication. This is a most important component for enterprise deployment.
[Cooker] [Bug 6180] [samba3-common] Segmentation fault net3 groupmap list - with ldaps only
http://qa.mandrakesoft.com/show_bug.cgi?id=6180 --- Additional Comments From [EMAIL PROTECTED] 2003-20-10 18:38 --- Can you please confirm exactly which packages you are using (you noted samba-3.0.1pre1 but I have not made packages for Mandrake 9.1, using the cooker packages on Mandrake 9.1 *will* cause problems as cooker/9.2 run openldap-2.1.x, Mandrake 9.1 shipped with openldap-2.0.x, and there are some incompatabilities between the two versions in the libraries). $ rpm -qa 'samba3*' should do. If you are running non-standard versions of any other related software, please note those too ... Also, can yo confirm whether or not you have correct certificate infrastructure in place (ie you have a CA certificate available and defined in /etc/ldap.conf on the samba server, and the LDAP server has a valid certificate, with the CN value matching it's hostname as set in the passdb backend line, signed by the CA certificate)? I will test on my test samba3/LDAP network later today, but I think I had SSL or TLS working ... -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: I've configured samba3 (smb.conf) with ldaps to connect our LDAP Server. When executing the command net3 groupmap list I always get a Segmentation fault. gdb -c core.22231 Program terminated with signal 11, Segmentation fault. #0 0x40286e38 in ?? () If I change the line: passdb backend = ldapsam:ldaps://mse.foo.com to passdb backend = ldapsam:ldap://mse.foo.com everything works. I'm able to reproduce this error on two different servers with ML 9.1. I can make a connection to a share on the samba server when ldaps is activated - so ldaps seems to work.
[Cooker] [Bug 308] [drakconf] Soundcard not set up properly
http://qa.mandrakesoft.com/show_bug.cgi?id=308 --- Additional Comments From [EMAIL PROTECTED] 2003-20-10 19:06 --- Reports from the thinkpad mailing list say that the Thinkpad models 380XD, 600/600E, and 770X/Z should all use cs4232. I would guess the snd-cs4232 should also work (can someone with a 600/600E or similar test?). Thierry, it seems the information on some of these cards may be incorrect (but the 600X works correctly out-the-box). -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEEDINFO creation_date: description: The soundcard is identified as this: cs46xx: Cirrus Logic|Cirrus CS4610/1 CrystalClearSoundFusion Audio [M ULTIMEDIA_AUDIO] (vendor:1013 device:6001 subv:1014subd:1010) The soundcard installed in the PC is a CS4239 so there could be some compatability problems here? or should the CS46XX module support the CS4239 card? Here are the output from lspcidrake -v agpgart : Intel Corporation|440BX/ZX - 82443BX/ZX Host bridge [BRIDGE_HOST] (vendor:8086 device:7190) unknown : Intel Corporation|440BX/ZX - 82443BX/ZX AGP bridge [BRIDGE_PCI] (vendor:8086 device:7191) yenta_socket: Texas Instruments|PCI1251 PC Card Controller [BRIDGE_CARDBUS] (vendor:104c device:ac1d subv:4000 subd:) yenta_socket: Texas Instruments|PCI1251 PC Card Controller [BRIDGE_CARDBUS] (vendor:104c device:ac1d subv:4800 subd:) unknown : IBM|20H2999 PCI Docking Bridge [BRIDGE_PCI] (vendor:1014 device:0095) cs46xx : Cirrus Logic|Cirrus CS4610/1 CrystalClear SoundFusion Audio [MULTIMEDIA_AUDIO] (vendor:1013 device:6001 subv:1014 subd:1010) unknown : Intel Corporation|82371AB PIIX4 ISA [BRIDGE_OTHER] (vendor:8086 device:7110) unknown : Intel Corporation|82371AB PIIX4 IDE [STORAGE_IDE] (vendor:8086 device:7111) usb-uhci: Intel Corporation|82371AB PIIX4 USB [SERIAL_USB] (vendor:8086 device:7112) sonypi : Intel Corporation|82371AB PIIX4 ACPI - Bus Master IDE Controller [BRIDGE_OTHER] (vendor:8086 device:7113) Card:NeoMagic (laptop/notebook): Neomagic Corporation|[MagicGraph 256AV] [DISPLAY_VGA] (vendor:10c8 device:0005 subv:1014 subd:00dd) eepro100: Intel Corporation|82559 [Ethernet Pro 100] [NETWORK_ETHERNET] (vendor:8086 device:1229 subv:1014 subd:00d8) yenta_socket: Texas Instruments|PCI1250 PC card CardBus Controller [BRIDGE_CARDBUS] (vendor:104c device:ac16 subv:9400 subd:) yenta_socket: Texas Instruments|PCI1250 PC card CardBus Controller [BRIDGE_CARDBUS] (vendor:104c device:ac16 subv:9c00 subd:) unknown : Virtual|Hub [] (vendor: device:) usb-storage : Shuttle Technology Inc.|CD-RW Devicee [] (vendor:04e6 device:0101) System: IBM Thinkpad 600E, 124Mb, 30GB HD, Docking
[Cooker] [Bug 6180] [samba3-common] Segmentation fault net3 groupmap list - with ldaps only
http://qa.mandrakesoft.com/show_bug.cgi?id=6180 [EMAIL PROTECTED] changed: What|Removed |Added Status|UNCONFIRMED |NEEDINFO --- Additional Comments From [EMAIL PROTECTED] 2003-21-10 01:50 --- No, your setup looks good. I have a simliar setup running on Mandrake 9.0 with (packages built on the machine): # rpm -qa 'samba3*' samba3-passdb-xml-3.0.1-0.pre1.1mdk samba3-doc-3.0.1-0.pre1.1mdk samba3-server-3.0.1-0.pre1.1mdk samba3-common-3.0.1-0.pre1.1mdk and Mandrake 9.1 running the same packages (built on 9.1) But this works here: # net3 groupmap list bgmilne (S-1-5-21-2244014245-3637982190-3323613867-2003) - bgmilne milne (S-1-5-21-2244014245-3637982190-3323613867-2009) - milne home (S-1-5-21-2244014245-3637982190-3323613867-3001) - home Domain Admins (S-1-5-21-2244014245-3637982190-3323613867-512) - adm Domain Users (S-1-5-21-2244014245-3637982190-3323613867-513) - users Domain Guests (S-1-5-21-2244014245-3637982190-3323613867-132069) - nogroup Machine accounts (S-1-5-21-2244014245-3637982190-3323613867-1843) - machines root (S-1-5-21-2244014245-3637982190-3323613867-1001) - root However, when running with TLS/SSL against a different host, ie with samba on 9.1 and openldap on 9.0, I get the segmentation fault. However, I also get the segmentation fault running an ldapsearch: Search against local host: [EMAIL PROTECTED] bgmilne]$ ldapsearch -x -h bgmilne.home.control.co.za -ZZ -LLL (uid=bgmilne) dn dn: uid=bgmilne,ou=People,dc=home,dc=control,dc=co,dc=za [EMAIL PROTECTED] bgmilne]$ ldapsearch -x -h fw.home.control.co.za -ZZ -LLL (uid=bgmilne) dn Segmentation fault It works the other way around: [EMAIL PROTECTED] bgmilne]$ ldapsearch -x -h bgmilne.home.control.co.za -ZZ -LLL (uid=bgmilne) dn dn: uid=bgmilne,ou=People,dc=home,dc=control,dc=co,dc=za Running net3 groupmap list with samba on 9.0 and openldap on 9.1 also gives no problems. I believe the problem is with the openldap/ssl libraries on 9.1 (as problems also exist with other software using libldap2, such as gq, when running against a remote server with SSL/TLS). I have one 9.1 box with working ldap/TLS to another machine, but that is running: $ rpm -qa '*ldap*' libldap2-devel-2.1.22-1mdk libsasl2-plug-ldapdb-2.1.13-2.2mdk php-ldap-4.3.0-3mdk libldap2-2.1.22-1mdk openldap-2.1.22-1mdk openldap-servers-2.1.22-1mdk nss_ldap-207-1.1mdk perl-ldap-0.27.01-1mdk openldap-clients-2.1.22-1mdk pam_ldap-164-1.1mdk (of course with all relevant software rebuilt against openldap and sasl). At present the suggestion is to run an LDAP slave on any samba server anyway (for performance reasons). If you are getting sefaults against the local host, that is more serious. Of course, this makes failover more difficult. Anyway, I think this bug is specific to Mandrake 9.1, and in openldap/libldap. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEEDINFO creation_date: description: I've configured samba3 (smb.conf) with ldaps to connect our LDAP Server. When executing the command net3 groupmap list I always get a Segmentation fault. gdb -c core.22231 Program terminated with signal 11, Segmentation fault. #0 0x40286e38 in ?? () If I change the line: passdb backend = ldapsam:ldaps://mse.foo.com to passdb backend = ldapsam:ldap://mse.foo.com everything works. I'm able to reproduce this error on two different servers with ML 9.1. I can make a connection to a share on the samba server when ldaps is activated - so ldaps seems to work.
Re: [Cooker] The JPackage Project.
Svetoslav Slavtchev wrote: The new 9.2 contribs jpackage directory, specifically: rsync rsync://ftp.sunet.se:873/Mandrake/9.2/contrib/jpackage/i586/ contains 415 noarch RPMs of the form: adaptx-0.9.6-1jpp.noarch.rpm placed there by the JPackage (David Walluck) project. David Walluck is just one of the contributors, Guillaume Rousse is also a large contributor. What is this JPackage project? J[ava related] Packages :-) How clever you are! (NOT!) ??? You asked the question. It's really polite to be sarcastic to someone who answered it (NOT!). Does it have a web site and mailing lists? have you tried http://jpackage.org ? Yes - it appears that the new 9.2/contribs/jpackage tree is a list of some 415 available Java applications, which is a great idea! But nowhere from MandrakeSoft do I see any indication whether 9.2 as installed includes Java support and if not what are the recommended RPMs to add Java support. Mandrake has shipped with Java support for a long time. Just not in the download edition. you will notice from the comparison page (http://www.mandrakesoft.com/products/92/comparison) that only ProSuite ships with JPackage, and I would guess it ships with jpackage-compatible java packages. However, it doesn't seem the packages on Club are suitable (http://rpms.mandrakeclub.com/rpms/MandrakeClub/comm/9.2/i586/j2re-1.4.2_01-1mdk.i586.html). Anyway, JPackage packages really want to be used with JPackage-compatible Java RPMs, so you may want to try rebuilding one of the .nosrc.rpm's (I hope this isn't necessary on ProSuite). You will also notice (on the same page) all the commercial releases ship with j2re/j2sdk packages. What is the modus operandum necessary for us to make use of appropriate ones of these RPMs to achieve Java operability in Mandrake 9.2? Either use the Java packages from the Club, or from one of the boxed sets, and then use urpmi Whats' it all about, please? Is this directory a new medium (NOT ?media?! ugh!) for urpmi? If so, what do we install? Whatever you want. Maybe you have heard of projects like JEdit, Eclipse, tomcat, ant etc. If you want to install them, you most likely already know what they are. check the url I suspect the answer should have been ¨Yes¨, but none of the jpackage RPMs include any summary text in their RPM, so how are we to know what each is and whether we might need it? Are you sure you checked them all??? I think you must have missed at least a few ... it looks like on average they have about 3 lines of desciption each: Of course, if you want better descriptions, I am sure JPackage will take patches ... Regards, Buchan
Re: [Cooker] The JPackage Project.
Svetoslav Slavtchev wrote:
The new 9.2 contribs jpackage directory, specifically:
rsync rsync://ftp.sunet.se:873/Mandrake/9.2/contrib/jpackage/i586/
contains 415 noarch RPMs of the form:
adaptx-0.9.6-1jpp.noarch.rpm
placed there by the JPackage (David Walluck) project.
David Walluck is just one of the contributors, Guillaume Rousse is also a
large contributor.
What is this JPackage project?
J[ava related] Packages :-)
How clever you are! (NOT!)
??? You asked the question. It's really polite to be sarcastic to someone
who answered it (NOT!).
Does it have a web site and mailing lists?
have you tried http://jpackage.org ?
Yes - it appears that the new 9.2/contribs/jpackage tree is a list of
some 415 available Java applications, which is a great idea!
But nowhere from MandrakeSoft do I see any indication whether 9.2 as
installed includes Java support and if not what are the recommended RPMs
to add Java support.
Mandrake has shipped with Java support for a long time. Just not in the
download edition. you will notice from the comparison page
(http://www.mandrakesoft.com/products/92/comparison) that only ProSuite
ships with JPackage, and I would guess it ships with jpackage-compatible
java packages. However, it doesn't seem the packages on Club are suitable
(http://rpms.mandrakeclub.com/rpms/MandrakeClub/comm/9.2/i586/j2re-1.4.2_01-1mdk.i586.html).
Anyway, JPackage packages really want to be used with JPackage-compatible
Java RPMs, so you may want to try rebuilding one of the .nosrc.rpm's (I
hope this isn't necessary on ProSuite).
You will also notice (on the same page) all the commercial releases ship
with j2re/j2sdk packages.
What is the modus operandum necessary for us to make use of
appropriate ones of these RPMs to achieve Java operability in
Mandrake 9.2?
Either use the Java packages from the Club, or from one of the boxed sets,
and then use urpmi
Whats' it all about, please?
Is this directory a new medium (NOT ?media?! ugh!) for urpmi? If
so, what do we install?
Whatever you want. Maybe you have heard of projects like JEdit, Eclipse,
tomcat, ant etc. If you want to install them, you most likely already know
what they are.
check the url
I suspect the answer should have been ¨Yes¨, but none of the jpackage
RPMs include any summary text in their RPM, so how are we to know what
each is and whether we might need it?
Are you sure you checked them all??? I think you must have missed at least
a few ... it looks like on average they have about 3 lines of desciption
each:
$ rpm -qp --qf %{DESCRIPTION}\n contrib/jpackage/RPMS/*.rpm|wc -l
1136
$ ls contrib/jpackage/RPMS/*.rpm|wc -l
389
Of course, if you want better descriptions, I am sure JPackage will take
patches ...
Regards,
Buchan
[Cooker] [Bug 5244] [lisa] Lisa bombards the net with packets
http://qa.mandrakesoft.com/show_bug.cgi?id=5244 --- Additional Comments From [EMAIL PROTECTED] 2003-19-10 13:43 --- Zeb, I can't fix the package (only Laurent can upload KDE packages), I merely provided the fixed script. There is nothing more I an do. Use the script here to fix it. I will take this issue up on the maintainers list, maybe we can have an errata at least if Laurent won't fix it. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: ASSIGNED creation_date: description: I have the LAN at 10.0.0.0/8. tcpdump shows that the lisa tries to do an ARP request for probably all addresses in the network and literally bombards the net with packets. I don't know what this might cause on a large network. To add to this it continues these probes even after the LAN interface goes down. This can also result in a costly problem. The following happened to me: - I configured my notebook for GPRS over bluetooth - I removed the LAN card from my notebook so it does not interfere with the experiments - I brought a PPP connection up - The defaultroute was set to ppp0 - The lisa continued to send packets - They all travelled to the phone, generating paid traffic :-( The distribution is Mandrake 9.2 Beta 2 Proposed resolution: don't enable the server by default and document this problem somewhere
Re: [Cooker] 9.2 disasters list (continuing)
Buchan Milne wrote: Come on Ron, please don't be so dramatic on every problem you have, until you can actually provide a useful bug report. The degree of dramaticness has nothing to do with whether there is a reproducible bug report. And neither does it: -contribute any information of value that anyone can use to debug the problem -give us any reason to respect the next Stop the presses bug report from you. This has occurrred three times under 9.1 and now once under 9.2 on this one machine, which is otherwise 100% long term totally stable. So, it's but a 9.2 disaster, but rather something you have configured on your machine. With about 6 production desktops that have run 9.1 and 9.2 for significant amounts of time, plus a few production servers (running 9.0 and 9.1) I have never seen anything like this, and it doesn't seem as if anyone else has. I need some kind of daemon that will watch all the time for /root/*.n file creation, but that is beyond my skill level. Well, you could start by looking at the modification times, and if you were using the machine at the time, remember what you were doing. If not, you could look in the logs to see what ran then. 9.2 is not yet installed on my other desktop machine, but will be soon unless any further disasters pop up. OK, so a disaster isn't something that prevents you from using the machine as a full-time desktop? So, in future we should classify anything from you with urgent, disaster, critical etc as minor irritation then? Sorry, but I can't take bug reports like this seriously ... maybe you should try a different list (since this isn't the 9.2 support list).
Re: [Cooker] printer packages
Le Samedi 18 Octobre 2003 11:43, Buchan Milne a écrit : No significant advantage over CUPS (AFAIR). So would you prefer cups for a printer not shared (for both security and size reasons) ? I have no such printers ;-). Till probably has some other reasons ...
Re: [Cooker] 9.2 disasters list (continuing)
lördagen den 18 oktober 2003 15.36 skrev Greg Meyer: It won't be a problem if I *buy* the powerpack, so as far as Mandrake is conserned this is no problem. I only wish I had known about this _before_ downloding it, then I would not have spent my time on it. What about the download edition, is the kde-i18n-sv package in there? Yes, but then you don't get kernel-source. Seems we can no longer get a full distribution on 3 CDs? Time to go to 4 for download?
Re: [Cooker] No XFS for you, sucker
This is a big bummer and the waste of a morning... I bought a fancy new machine (one of those E-Cube mini sytems... it's very neat, and glows BLUE!). I installed my freshly downloaded powerpack, with XFS root and XFS home. If you are going to use an XFS root, you need a /boot which is not XFS. This has something to do with the compatablity SGI wants to keep with the on-disk format used by Irix XFS, which is incompatible with what LILO wants to put there on any partition containing a kernel you want to boot. I was under the impression drakx warned you about this. At the lilo section of drakx, I asked for lilo to be installed to MBR. I got this message: Fatal error, filesystem would be destroyed by LILO boot sector /dev/hda. But then it just goes on with the lilo setup. No way to go back. After the lilo step, I got the same error again, and no bootloader was installed. Grrr. So I reran drakx, and this time chose 'upgrade' instead of 'install', hoping I could change the configuration to install bootloader to root. But no, I was not asked that question, it just kept trying to install to MBR. So I reran drakx, doing another complete install. This time I chose to install the bootloader to root. The error: You cannot install the bootloader to an XFS partition. Awesome. Once again, instead of going back and letting me select MBR, it goes on into the LILO setup forever. I don't think that would help. Either, make a small ext3 root filesystem, and use XFS for data partitions (this is my suggestion, since it's about the safest), or make an ext2/3 /boot partition. BTW, on servers, I typically install with a small root ext3 filesystem, /usr and /tmp ext3, /var and data partitions on LVM, usually XFS. This avoids the root on LVM issues, but still gets you the majority of the LVM benefit (resizing partitions which can vary greatly in required capacity). Regards, Buchan
Re: [Cooker] No XFS for you, sucker
[EMAIL PROTECTED] escribió: This is a big bummer and the waste of a morning... I bought a fancy new machine (one of those E-Cube mini sytems... it's very neat, and glows BLUE!). I installed my freshly downloaded powerpack, with XFS root and XFS home. If you are going to use an XFS root, you need a /boot which is not XFS. This has something to do with the compatablity SGI wants to keep with the on-disk format used by Irix XFS, which is incompatible with what LILO wants to put there on any partition containing a kernel you want to boot. I was under the impression drakx warned you about this. Well, my 9.1 system is completely xfs and lilo here has no problem in installing the bootloader on the mbr. Don't tell me that the lilo that comes with 9.2 cannot do that, it'd mean I cannot upgrade. OK, well, I wasn't absolutely sure about the MBR, but it *should* work on the MBR according to: http://oss.sgi.com/projects/xfs/faq.html#lilowork I am not sure why Austin's installation of LILO on the MBR failed. You should be ok ... I think ;-). Regards, Buchan
[Cooker] Re: bug with upgrading samba3
hello, there is a small problem i have been incurring for a while while upgrading samba3 (i always rebuild samba3 --with system). immediatly after upgrade [EMAIL PROTECTED] ~ $ smbclient -L moskow protocol negotiation failed zsh: 26932 exit 1 smbclient -L moskow and in syslog: Oct 18 20:21:19 Moskow smbd[19939]: [2003/10/18 20:21:19, 0] tdb/tdbutil.c:tdb_log(724) Oct 18 20:21:19 Moskow smbd[19939]: tdb(/var/cache/samba/messages.tdb): tdb_reopen: open failed (No such file or directory) Oct 18 20:21:19 Moskow smbd[19939]: [2003/10/18 20:21:19, 0] smbd/server.c:open_sockets_smbd(410) Oct 18 20:21:19 Moskow smbd[19939]: tdb_reopen_all failed. [EMAIL PROTECTED] /etc/squid # ll /var/cache/samba/messages.tdb ls: /var/cache/samba/messages.tdb: No such file or directory %post is postinstall scriptlet (using /bin/sh): /usr/share/rpm-helper/add-service samba $1 smb #/usr/share/rpm-helper/add-service samba $1 wrepld # Add a unix group for samba machine accounts groupadd -frg 421 machines # Migrate tdb's from /var/lock/samba (taken from official samba spec file): for i in /var/lock/samba/*.tdb do if [ -f $i ]; then newname=`echo $i | sed -e's|var\/lock\/samba|var\/cache\/samba|'` echo Moving $i to $newname mv $i $newname fi done # Remove the transient tdb files (modified from version in off. samba spec: for TDB in brlock unexpected locking messages; do if [ -e /var/cache/samba/$TDB.tdb ]; then rm -f /var/cache/samba/$TDB.tdb; fi; done if [ -d /var/lock/samba ]; then rm -rf /var/lock/samba fi So we remove /var/cache/samba/messages.tdb after restarting samba server this is not nice (and i should have reported this way earlier, sorry) I've actually been meaning to remove the whole offending piece. Remind me never to take anything from the official samba spec file again Will fix for the next 3.0.1 prerelease. Regards, Buchan
Re: [Cooker] provides, conflicts, obsoletes
each time i need to use one of the above i am confused again when too use what. I think this info is not in rpm howto yet, i would add it, if i didn't feel unsure about what is best (I use things like described below, but i might be wrong). I see Conflicts and provides are in the wiki, but Obsoletes is not (for that matter, don't see prereqs either). Perhaps someone can give a thorough explanation and i will add it to wiki? Or someone who does know can add it, whatever. Perhaps it should be case described, like: - new package has function of another old package as well (I assume this obsoletes other old package). Obsoletes also means installing the Obsoleting package will remote the Obsolete package But should it provide old package as well? Yes, since something may have depended on the old package. BTW, rpmlint will compain if you don't provide an obsoletes. - new package does not work anymore with another old package (conflicts) - provides is only used to set aliasses for the name Mostly (besides the Obsoletes case). Next rpm issue that needs clarification is alternatives (they are explained briefly, but nothing is said about what you need to do when introducing obsoletes for new binaries - the trap I fell into). Regards, Buchan
Re: [Cooker] printer packages
[EMAIL PROTECTED] wrote : Le Samedi 18 Octobre 2003 11:43, Buchan Milne a écrit : No significant advantage over CUPS (AFAIR). So would you prefer cups for a printer not shared (for both security and size reasons) ? I have no such printers ;-). unlike many people (such as standalone desktops) ;-) Huh? What is that? ;-) (sorry, but no place I have been recently has less than 2 computers, and if you have 2 computers and a printer ...). Till probably has some other reasons ... I wonder which ones. I found this script: /etc/dynamic/scripts/functions.script check_activated $0 MODE=660 OWNER=lp.sys if [ -x /usr/sbin/update-alternatives ]; then TYPE=`/usr/sbin/update-alternatives --display lpr 2 /dev/null | grep currently | cut -f 6 -d ' '` case $TYPE in *lpd) MODE=660; OWNER=lp.lp;; *cups) MODE=660; OWNER=lp.sys;; *pdq) MODE=666;; # should be done via pam_console esac fi ... Is pdq less secure than others for this ? In some ways yes, in others, no. PDQ doesn't have a spooler, so all users need to be able to print. Currently it looks like this is just done by makeing the print devices world-writeable. As the comment says, pam_console (see /etc/security/console.perms) would be better, which is how a lot of other devices are managed. Of course, PDQ doesn't have a daemon running with elevated priveleges (like CUPS does). So, with PDQ you are allowing any logged in user to send any data they like to the machine, with CUPS you are allowing any remote user (unless you do some work securing it) to send data to port 631 on your machine. Worst case scenario for PDQ is someone sends data to your printer that does it some harm (or uses your paper). With CUPS, they could again elevated priveleges, *and* waste all your paper ;-). For a single-user, non-networked machine, PDQ is better. But, for a network, CUPS is just so much easier (easier than Windows printing since you just plug a machine in the network and it finds all printers, and you don't have to worry about printer drivers etc). Regards, Buchan
[Cooker] [Bug 982] [kdebase] file views in konqueror in detailed mode don't respect double click settings
http://qa.mandrakesoft.com/show_bug.cgi?id=982 --- Additional Comments From [EMAIL PROTECTED] 2003-17-10 11:54 --- Simon, the patch Tex removed is of course a Mandrakesoft patch, so the question is really whether it is intentional or not (since Laurent has been made aware that the problem is caused by the patch). We can only assume that this is intentional, but it would be nice to have an official answer to know if it is worthwhile bothering with this bug or whether we should just resort to rebuilding kdebase without the patch on our own. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: ASSIGNED creation_date: description: When KDE has been set to use double-click to open files/direcories, konqueror will still open files/directories on single-click when using 'detailed' view. When icon view is used, the behaviour is correct.
[Cooker] [Bug 5309] [gdm] GDM does not show NIS accounts in the browser
http://qa.mandrakesoft.com/show_bug.cgi?id=5309 --- Additional Comments From [EMAIL PROTECTED] 2003-17-10 20:57 --- Fred, it doesn't change matters, but it may be a feature, since it seems it is only showing users with a valid shell. However, kdm shows all users, and it is perfectly feasible to use a linux machine without a valid shell (if the admin doesn't want you to get lost ;-)). So, I think there should be an option to remove the requirement for a valid shell (if there isn't one). However, I don't think this is a Mandrake bug ... so I think we need a report from Edward Tandi on a stable system ... -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEW creation_date: description: I use the 'Happy Gnome with Browser' configuration for the Graphical Greeter. The Browser part allows a user to click on their user icon to select the account to log in to. Unfortunately, ulike kdm, it does not show NIS accounts in the browser. If you type the user's account name in manually however, it _is_ possible to log in.
[Cooker] [Bug 6149] [Installation] getAndSaveFile error when upgrading from Mandrake 9.0
http://qa.mandrakesoft.com/show_bug.cgi?id=6149 --- Additional Comments From [EMAIL PROTECTED] 2003-16-10 13:09 --- /dev/ram3 has always been 30MB, as you can see in the boot disk images (9.0 and 9.1 both have this in the append line in syslinux.cfg: append initrd=network.rdz ramdisk_size=32000 root=/dev/ram3 vga=788). Are you sure you aren't running out of disk space on the target partitions? You could always make your own boot disk and bump up the ram disk size if you really think that is the issue. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: I'm upgrading from Mandrake 9.0 using Mandrake 9.2 PowerPack club CDs. In the installer, after I selecct Upgrade Mandrake 9.0 release (dolphin), I get the following error: An error occured getAndSaveFile: No space left on device. on the 3rd virtual console it says: * starting step 'choosePackages' * trying to read hdlist1.cz for medium 1 * getFile Mandrake/base/hdlist1.cz: * warning: getAndSaveFile: No space left on device at /usr/bin/perl-install/install_any.pm line 144, $F chunk 221. * not XSetInputFocus since already done and not on top
Re: [Cooker] Discovery Pack
Hi all. Aha, now I know what the surprise was about... Discovery Pack. Good for Mandrake(Soft) I guess, financially speaking... Where does that leave odd morons like me? ProSuite DVD and CD's-only? That's what I took ... (wasn't sure if Workstation DVD had jpackage, plus my laptop doesn't have a DVD drive ... yet ...)
Re: [Cooker] megaraid driver
I am having some trouble with my dell poweredge 1600sc. I have a megaraid controller in it. After installing mandrake 9.1 and mandrake 9.2 I have periodic lock-ups. After using the 2.6 test kernel all seems to be well. For some reason it seems that the default 9.x kernels use the old v1 megaraid drivers and the 2.6 test kernels use the v2 driver. How can I switch over to the megaraid 2 drivers using the 2.4.22 kernels or can someone upgrade the driver so the default megaraid driver in the 2.4.22 kernel is the latest v2 driver Hmmm, our 1600SC is running fine (under pretty heavy load), we have the Perc IDE RAID in it, we're running 2.4.21-0.22mdk on it (I think ... we might have updated to 2.4.21-0.25mdk but we needed XFS before 0.25mdk was available), which is Danny's kernel from Club. Our 2500 (with the Perc SCSI RAID) is doing fine too, it seems it is about 145 days since we did the last kernel update ... Anyway, if you can give more detail about exactly what card and driver is being used, I can take a look (I know a few others that run either 9.0 or 9.1 on 1600SCs with no problems too ...). Regards, Buchan
[Cooker] [Bug 6118] [xinitrc] How about fast user switching?
http://qa.mandrakesoft.com/show_bug.cgi?id=6118 [EMAIL PROTECTED] changed: What|Removed |Added Severity|normal |enhancement Component|mdkkdm |packaging Product|mdkkdm |xinitrc Version|9.2-9mdk|2.4.4-76mdk --- Additional Comments From [EMAIL PROTECTED] 2003-13-10 14:25 --- Add something like this to /etc/X11/xdm/Xservers: :1 local reserve /usr/X11R6/bin/X :1 vt8 :2 local reserve /usr/X11R6/bin/X :2 vt9 :3 local reserve /usr/X11R6/bin/X :3 vt10 :4 local reserve /usr/X11R6/bin/X :4 vt11 :5 local reserve /usr/X11R6/bin/X :5 vt12 Now, you will have 5 reserve X11 servers, which will be started on demand by kdm. Just log out, restart the dm service, log back in, and you should see a new menu entry in your KDE menu, called Start New Session. This has been available for a long time, and works quite well. Additionally, SuSE ships with this by default (and BTW this is one of two reasons some local users have for using SuSE over Mandrake). -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEW creation_date: description: Hello, It would be nice to have a user switch feature which does not force me to shut down programs. For example if my mother must check mail, I don't want to log off (and shut down xmule). Currently I have to lock the session CTRL+ALT+f2 startx -- :1 But this does not work fully: I have many resource locking problems, e.g. the sound does not work, and the tv card doesn't either. Also, it is not apparent enough. 99% of the users don't know it is possible. Windows XP does this since years, and I heard SUSE 8.2 does it too. cheers!
[Cooker] [Bug 6074] [samba-client] Mandrake Control Center cannot setup Samba Mount Points
http://qa.mandrakesoft.com/show_bug.cgi?id=6074 [EMAIL PROTECTED] changed: What|Removed |Added Status|NEW |RESOLVED Resolution||DUPLICATE --- Additional Comments From [EMAIL PROTECTED] 2003-13-10 14:32 --- Please read the proposed errata list on the Wiki (http://qa.mandrakesoft.com/twiki/bin/view/Main/Mandrake92) until the official errata list is out. *** This bug has been marked as a duplicate of 5889 *** -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: RESOLVED creation_date: description: In the Mandrake Control Center, when I click on Setup Samba Mount Points it loads that configuration module, and then claims that the samba-client package needs to be installed. It then asks if I want to install it. I click on yes and it goes to a dialog that says Mandatory samba-client package not installed. I opened up a terminal and su'd to root and typed urpmi samba-client. Urpmi responded with Everyting already installed. So, 1) MCC does not know samba-client is installed 2) Despite asking about it, MCC will not install the package. I think this may be because the binary of the samba client name changed. In 9.1 I believe it was smbmount. I don't remember exactly what gets installed in 9.2, but I think the name changed. So the tool might need to be updated for this. Just perahps a place to start looking. I marked this as critical, because the networking aspects of the distribution have always been touted and I think it's important that this gets fixed.
[Cooker] [Bug 5889] [samba-client] During upgrade from 9.1 to 9.2RC2, alternative names aren't updated
http://qa.mandrakesoft.com/show_bug.cgi?id=5889 --- Additional Comments From [EMAIL PROTECTED] 2003-13-10 14:32 --- *** Bug 6074 has been marked as a duplicate of this bug. *** -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: RESOLVED creation_date: description: I was upgrading from Mandrake 9.1 to 9.2RC2 using the CD install method, and afterwords, all samba client references had 2 at the end (e.g. smbclient2, and no smbclient).
[Cooker] [Bug 5326] [kernel] kernel fails to load usb
http://qa.mandrakesoft.com/show_bug.cgi?id=5326 --- Additional Comments From [EMAIL PROTECTED] 2003-13-10 15:23 --- Answering Kevin: 1)Please add your comments to bugs that are more applicable, there are a number of bug reports on the Philips Webcam issue (including workarounds), and AFAIK, this bug report has nothing to do with it. 2)The bug isn't in the kernel, but in the ALSA sound drivers (which will be in-kernel in 2.6). Most sound cards are better supported with ALSA than with OSS (the sound system that ships with the 2.4 kernel). 3)No distribution (at least of the commercial distros) ships with stock kernels as the default kernel, since at present there are so many features which don't work as well without 3rd-party patches 4)A vanilla 2.4 kernel is available in contrib (but you have to be careful to not use features it does not have ...) 5)There are a number of mailing lists, including [EMAIL PROTECTED], [EMAIL PROTECTED] etc. Please see the relevant webpage on the Mandrake web site. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEW creation_date: description: On my acer aspire 1700+ with MDK 9.2rc1 the usb ports don´t work. I had this problem with MDK 9.1 too.lsusb returns a nice 0x0 even with my webcam and speedtouch modem plugged in. Adding pci usb ports has the same result. I know these ports are from VIA,and they work perfectly with suse live cd 8.2 (That uses kernel 2.4.20). I think it could be a bus system problem, as in MDK 9.1 sound did't work even if it installed and loaded the drivers, and this works with the suse live cd too, I´m installing KDE and will report on sound as soon as I can. Please try and fix this, without usb I loose my modem, mywebcam, my printer and my scanner. I can send more info on the chips, just let me know how to get it (Cat something in proc, run a script,whatever)
[Cooker] [Bug 6123] [XFree86-SVGA] X Hangs on Logout or CTL+ALT+Fx
http://qa.mandrakesoft.com/show_bug.cgi?id=6123 --- Additional Comments From [EMAIL PROTECTED] 2003-13-10 15:29 --- Are you booting with framebuffer (high resolution text console) or not? On some cards, problems like this occur when using framebuffer. Try booting with 'vga=0' to see if it makes a difference. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: I am running a purchased 9.2 Beta version so I'm not sure on the cooker version. Vanilla KDE desktop on a Toshiba Satellite 1000-S157 with Intel 830 graphics. Logging out of KDE or pressing CTL+ALT+Fx for another console gives black screen. CTL+ALT+F7 does nothing. I think it's just X hanging because CTL+ALT+DEL still initiates a reboot. I tried creating a new basic XF86Config with xf86config -- no help. Tried turning off hardware video acceleration -- no help. I did the same install on a desktop with an nVidia card and it works fine so it looks like it's specific to my graphics chip. I found similar problems reported on previous 9.x versions but never found the resolution. Any help greatly appreciated. Thanks, Lee
[Cooker] [Bug 6112] [initscripts] service network restart should not disable non ON_BOOT=yes interfaces
http://qa.mandrakesoft.com/show_bug.cgi?id=6112 --- Additional Comments From [EMAIL PROTECTED] 2003-09-10 16:51 --- The second solution would be better, since it would also solve the problem where PCMCIA devices (which are now correctly not started at boot, so they are started by PCMCIA instead, avoiding the error message before PCMCIA is started) are not restarted by 'service network restart'. When restarting the network, one expects all devices that were running before to be running after ... -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: description: = service network restart shut down net interfaces that do not have ON_BOOT=yes in / etc/sysconfig/network-scripts/icfg-INTF but does not restart them due to the ON_BOOT filter: if LANGUAGE=C LC_ALL=C egrep -q -L ^ONBOOT=['\]?[Nn][Oo]['\]? ifcfg-$i; then # this loads the module, to preserve ordering is_available $i continue fi suggestion: = make start() and stop() out of case/esac code and make restart state call start();stop();; then: - either add a parameter to start() that tell it we're not at boot and really want all interfaces to come up - or get the list of active interfaces before stoping them and pass it again to start()
Re: [Cooker] missing all.img in cooker
Hi! I would like if there was an all.img with all the boot stuff available in cooker. There was one at some time in the beta stage, but now it is gone. I think you are looking for all.rdz? ftp://mandrake.redbox.cz/Mandrake-devel/cooker/i586/isolinux/alt0/all.rdz It needs a kernel too: ftp://mandrake.redbox.cz/Mandrake-devel/cooker/i586/isolinux/alt0/vmlinuz Could I use such an image from a small hd partition to boot from? Yes, and I think this is covered in the CookerHowto on the Wiki. What is the best method to boot of disk? Save it in my home partition which normally survives between installs? If you have a local cooker mirror, you can boot it straight from there. I have a seperate /boot on my desktop, so I put things there. On my laptop, I stuck it and my lilo.conf in /home (along with a script to reinstall lilo, since I did this a few times). Save it on my windows partition, which also survives, and is bootable? Is there some setup that would be easy for a newbie? Download to ypur windows partition and then boot from there? (may be only a win98 thing, XP may not allow that) Yes, this should be possible with loadlin (but this seems not to be in /dosutils any more), but not for 2k/XP, and probably not NT either. I don't think they can boot linux directly :-(. Regards, Buchan * Please click on http://www.cae.co.za/disclaimer.htm to read our e-mail disclaimer or send an e-mail to [EMAIL PROTECTED] for a copy. *
[Cooker] [Bug 5945] [mingetty] /etc/securetty is empty (no console root login)
http://qa.mandrakesoft.com/show_bug.cgi?id=5945 --- Additional Comments From [EMAIL PROTECTED] 2003-07-10 14:09 --- Well, cooker is always newer ... But, I haven't noticed this (console logins as root work fine) on my clean install of 9.2rc2, and this hasn't been reported by anyone else or confirmed in the 2 weeks since you reported this. BTW, /etc/securetty is owned by the setup package. You could take a look in /root/drakx to see the logs from the installation, to see if you can see why this happened. If you don't know what to look for, attach the *.log files to this bug report. Also, check if there are any /etc/securetty.rpm* files. -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: NEEDINFO creation_date: description: See summary, this is what the system looked like after installing MDK-9.2-rc2. ...and that ends my list of bug reports. It's a darn nice system, with just a few perks. I hope the separation of the bug reports helps you to distribute and resolve them in time for curtains-up ;-)
[Cooker] [Bug 6085] [kernel-source] error running make xconfig in the kernel source directory
http://qa.mandrakesoft.com/show_bug.cgi?id=6085 [EMAIL PROTECTED] changed: What|Removed |Added Summary|error running make xconfig |error running make xconfig |in the kernel source|in the kernel source |directory |directory --- Additional Comments From [EMAIL PROTECTED] 2003-06-10 12:27 --- Did you run 'make mrproper' first as indicated in /usr/share/doc/kernel-source-2.4.22/README.Mandrake ? -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: Trying to recompile the kernel (due to an ACPI problem with a notebook Toshiba) I get an error 2; it's impossible to continue. Similar error I get with the kde config tool (tou have found a bug in the tool or the config file is wrong) Toshiba Satellite 5200-801 updgraded from mdk 9.1 to 9.2 RC2
[Cooker] [Bug 5889] [samba-client] During upgrade from 9.1 to 9.2RC2, alternative names aren't updated
http://qa.mandrakesoft.com/show_bug.cgi?id=5889 --- Additional Comments From [EMAIL PROTECTED] 2003-06-10 13:20 --- -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [simon] wrote: You mean by the installation program? Yes, it might be an idea to run something like: # pushd /etc/alternatives; for i in * ;do update-alternatives - --auto;done;popd (or something more intelligent which checks for the existence of the link first) But, I am not sure if this would adversely affect user's previous choices (if a more intelligent version is not possible). And, it wouldn't help users doing an upgrade via 'urpmi --auto-select - --auto'. Regards, Buchan - -- |--Another happy Mandrake Club member--| Buchan MilneMechanical Engineer, Network Manager Cellphone * Work+27 82 472 2231 * +27 21 8828820x202 Stellenbosch Automotive Engineering http://www.cae.co.za GPG Key http://ranger.dnsalias.com/bgmilne.asc 1024D/60D204A7 2919 E232 5610 A038 87B1 72D6 AC92 BA50 60D2 04A7 -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.2 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQE/gU22rJK6UGDSBKcRAgW0AJ9w9P2w5W+GhjmwRK5vNp8cPslPYgCfcjeT FBz9kVLx+xTXuUNCXU8yZAY= =W/WZ -END PGP SIGNATURE- * Please click on http://www.cae.co.za/disclaimer.htm to read our e-mail disclaimer or send an e-mail to [EMAIL PROTECTED] for a copy. * -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: RESOLVED creation_date: description: I was upgrading from Mandrake 9.1 to 9.2RC2 using the CD install method, and afterwords, all samba client references had 2 at the end (e.g. smbclient2, and no smbclient).
[Cooker] [Bug 6087] [gnome-terminal] should depend on libvte4
http://qa.mandrakesoft.com/show_bug.cgi?id=6087 --- Additional Comments From [EMAIL PROTECTED] 2003-06-10 13:59 --- $ rpm -qR gnome-terminal|grep vte libvte.so.4 $ urpmq -p libvte.so.4 libvte4 You should run 'urpmi --auto-select', and not just install individual packages from cooker (since Fred doesn't track the versions required by GNOME packages in the requirements) -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: I could install newest version of gnome-terminal but it failed to run. in .xsession-errors, I had /usr/bin/gnome-terminal: relocation error: /usr/bin/gnome-terminal: undefined symbol: vte_terminal_set_scroll_background After installing libvte4, it works.
Re: [Cooker] errata 9.2
Gwenole Beauchesne [EMAIL PROTECTED] writes: Hi, Error scenario: when doing a network install, on recent machines, the network card seems to not respond (no DHCP answer received for example) Why: problem with APIC Solution: type F1 then linux noapic when booting BTW, I saw that occuring on some instances of nvidia|via based motherboards. It is currently workarounded on AMD64 with disabling APIC there by default. Could you please precise the type of those recent machines? On plain x86, it should be even easier to blacklist those. Happened for me with an Asus Terminator K7 DDR[1] which has an ASUS A7S motherboard owning a SiS 740 chipset (I think). How should I provide identification information? I would like to know that too ... I have seen it myself on a few Gigabyte Socket A boards with VIA chipsets (like GA-7VTXH IIRC). Regards, Buchan * Please click on http://www.cae.co.za/disclaimer.htm to read our e-mail disclaimer or send an e-mail to [EMAIL PROTECTED] for a copy. *
[Cooker] [Bug 5991] [drakxtools] Drakconnect sets something wrong, kde slow opening windows.
http://qa.mandrakesoft.com/show_bug.cgi?id=5991 --- Additional Comments From [EMAIL PROTECTED] 2003-07-10 01:28 --- But, the question here is what the desired behaviour is. -Does the user want the hostname provided by the DHCP server? If so, then tmdns can solve this problem (well, I am not convinced tmdns is the right solution for anything, but it seems to be preferred by drakconnect), as long as the hostname the user entered for zeroconf name is *not* set manually in tmdns.conf -Does the user want the hostname they set when running drakconnect? If so, then tmdns can solve this (to a certain extent, it's still not the snappiest way of discovering your own IP address ...), as long as the name tmdns claims is the same as the hostname. So, in both cases, tmdns.conf should *not* have the hostname hardcoded, since in the first case it is harmful, and in the second case it is redundant (since if it is not set, tmdns will use `hostname`). The major problem here though IMHO is setting the hostname in tmdns.conf, since in this scenario tmdns *could* have helped if it's config file modified by drakconnect. Anyway, I think we need documentation on how drakconnect should work, otherwise it's almost impossible to tell bugs from features ... -- Configure bugmail: http://qa.mandrakesoft.com/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are on the CC list for the bug, or are watching someone who is. --- Reminder: --- assigned_to: [EMAIL PROTECTED] status: UNCONFIRMED creation_date: description: Hello. I have a router and I use drakconnect to set up my adsl connection with dhcp. THE PROBLEM: Drakconnect apparently sets up something wrong, since kde windows take about 5 seconds to open, and gnome complains that /etc/hosts is wrong. On the other hand, if I choose a static ip assignment and manual DNS assignment (as opposed to dhcp), windows open instantly. The difference is huge. POSSIBLE REASONS: I'm no expert, but I am going to report what others told me. :-) 1) Some kde developer told me it must be a problem not in dhcp, but in dns. unfortunately, drakconnect doesn't allow me to specify dns if I choose dhcp. 2) Someone else told me that the hostname does not resolve because the /etc/hosts file produced by drakconnect is broken, and I should manually specify a fully qualified hostname in /etc/hosts, together with dhcp. He told me that dhcp allows the hostname to be specified OR assigned automatically. But it seems to me that drakconnect does not allow a hostname to be specified with dhcp. Maybe this is the problem. I hope I' ve been of help. cheers!
Re: [Cooker] Bootsplash dependencies
With previous releases of Mandrake, I've always deselected the bootsplash package during install. With 9.2 RC2, that's impossible, as bootsplash is required by a long list of packages, including icewm, galeon, gedit and evolution. Actually, bootsplash is only required by mandrake_theme, which is required by mandrake_desk. Is the bootsplash requirement really necessary? For mandrake_theme to work, yes. But, the idea was that there could be more then one package providing mandrake_theme. (In other words, is it safe to do a rpm -e --nodeps botsplash after the install?) Yes, but otherwise you may prefer to look at some of the other themes in the bootsplash-themes package in contrib, or just add something like 'splash=0' to your bootloaders append line. Regards, Buchan * Please click on http://www.cae.co.za/disclaimer.htm to read our e-mail disclaimer or send an e-mail to [EMAIL PROTECTED] for a copy. *
Re: [Cooker] gaim 0.69 and 0.70 are out ...
On Sunday 05 Oct 2003 00:57, Adam Williamson wrote: On Sat, 2003-10-04 at 19:09, Austin wrote: On 10/04/2003 05:23:35 PM, Adam Williamson wrote: 9.2 is finalised and gone to duplication. It's way too late. You have to learn to let go, man. :D True, but I think the question was: can it be provided in the online updates That's not how I read the post I replied to. It seemed to want 0.70 in 9.2 proper. That was also my idea, but ok. IMHO Gaim will someday post a 9.2 version of .70 is so very not te way to deal with this. Well, I am sure there will be a package in MandrakeClub, and probably very soon on the free MandrakeClub mirrors, which are quite easy to install from. I think users really should be able to upgrade during the install procedure. Upgrades aren't provided for stable releases. Security and bugfix updates are. If this *really* qualifies as a bugfix release, then it will be tested a bit in cooker, and then put up on the mirrors. But consider samba3 for a moment. You can't print to a Windows2003 server, or connect to it with smbclient with samba-2.2.x. We have had samba3 packaged in contrib since before 9.1 was released, and we(I) have done a lot of testing of samba3, and 3.0.0 final was released 2 weeks ago. But it stays in contrib, since it was too late for the release schedule (we were already in version freeze). So, users in networks with Windows2003 servers need to install samba3-client from contrib, which will at least allow them to print and connect with smbclient (they must use mount.cifs which is in main to mount network shares) with minimal effort (by using alternatives). Those of you who depend on gaim could have made similar arrangements ... Making your way trough sourceforge to find the correct update is not done for newbies that try Mandrake as Redmond OS replacement, and Mandrake attracts a lot of those. Sure, but it's just way too late for the 9.2 release itself, and a package does at least need some testing (what if a vulnerability has been introduced in the hasty implementation?), and most likely won't be available for 9.2 until at least a few days after cooker reopens. There is nothing that can be done now besides testing packages and/or convincing the maintainer that this is worthy of a bugfix update (ie file a bug). Regards, Buchan * Please click on http://www.cae.co.za/disclaimer.htm to read our e-mail disclaimer or send an e-mail to [EMAIL PROTECTED] for a copy. *
Re: [Cooker] Mandrake 9.2 update policy (was: gaim 0.69 and 0.70 are out )
On Sunday 05 Oct 2003 14:55, Thomas Backlund wrote: Jos Hulzink kirjoitti viestissään (lähetysaika Sunnuntai 05 Lokakuu 2003 15:30): This is not a matter of depending on gaim, this is a matter of willing to prevent reviews from writing Mandrake 9.2 is no good, it comes with broken software Mandrake has a nice update feature during install, let's use it, instead of defining what a bugfix is. What is better ? A few extra kB of updates, or a broken distribution ? The problem that arises here is the bigger picture... You say a few extra kB... But once you open the bugfix/security updates system to this, where does it end...?? People will start saying: ... but you updated that package ..., and this package adds or modifies this or that feature that is needed... if we don't have it I consider it's broken... and we can't have a broken distrib... Now we're comparing OpenOffice with Gaim... What these bugs are for Gaim, is oowriter and oocalc won't start for OpenOffice. (Which actually happened to me in RC2, but happened to be fixed in time). I.e. buggy enough to make the package unusable for most people. Oh, come on. Rename ~/.openoffice with your favourite file management tool, and then try again. Or run it from the console to see what the error is. I think Gwenole had problems reproducing it, and I don't remember anyone providing a real test case. so where do you draw the line...? and since all packages always gets updated with more features, or feature enchancements... To be honest: for me, the sky is the limit, but it's easy talking with a cable connection. But to draw a line: most packages have a distinction between bugfix releases and feature releases. Ok, of course with Gaim we got the exception to this rule... Exactly, gaim would be a feature release, not a bug release = it should not go in updates IMHO, but in unsupported/MandrakeClub. today gaim, tomorrow OpenOffice.org final..., and then ... and suddenly the updates mirrors are not a few kB more... instead you will be calculating the updates in MBs, then in GBs For some reason I immedately must think about the famous Redmond OS service packs, sometimes 100 MB in size. You happily download those, for they fix bugs. So IMHO here comes the power of an Club membership..., join the club and you will most likely get the downloads there... and all are happy... (you get what you pay for... remember...) You say, become a club member. Ok, fair. Problem is, when I Mr newbe user download Mandrake, I see a buggy distro, and no updates. Or, I see that using proprietary, controlled communication protocols is a bad idea (Gaim). A new user shouldn't see any problems with OpenOffice.org (AFAIK the problem you mention only happens on upgrade). In the first place, I see nowhere that many problems are solved if I pay (and I'm reluctant to pay, for Mandrake is buggy, and who tells me that my problems are gone after I paid ?), in the second place, after I paid, the manual how to get the club updates in, is practically non existent (I found something deep in a forum once). What do you mean by buggy And *real* bug can be filed at bugs.mandrakelinux.com, and will be solved. For feature updates to software, go to MandrakeClub. And, BTW, the Mandrake firsttime wizard should set the Club source up for you, and if you bought a box, you will get a month membership, so you can join immediately (with the product code on the box). It should be painless. Users are willing to pay for something good, not for something bad that mabye gets better after they paid. To quote you: join the club and you will most likely get the downloads there. I don't see whether that is true, so I should just pay and hope ? Sorry, that's not the way customers think in 2003. Sure, but we aren't responsible for other peoples instant messaging networks. If they *really* want to play change-the-protocol, that's their problem. What I am really discussing is the way users see MandrakeSoft, that almost bankrupt company with a rather cool distro that needs all money it can get. The problem with any linux distribution is that the Download Edition is in fact your advertisement. And bad advertisement doesn't sell. And selling is essential for MandrakeSoft. I don't see how this is relevant. I don't use gaim (instant messaging doesn't work through our corporate firewall, and if I really wanted instant messaging I would use jabber or something open), so I think 9.2 really is a good release. Maybe -but that's too late now- it is an option to add a feature during the install period: Enter MandrakeClub membership ID / Passwd (or registration code) to get less critical updates. You obviously haven't run the firsttime wizard, otherwise you would have seen this, username/password option or product code for setting up acecss to the MandrakeClub urpmi media. I can live with paying for some updates, if
Re: [Cooker] [Bug 5859] [initscripts] Improve startup speed (solution in this report)
http://qa.mandrakesoft.com/show_bug.cgi?id=5859 --- Additional Comments From [EMAIL PROTECTED] 2003-05-10 12:39 --- For me network start takes REALY long when not connected to then net: eth0 1 minute 20 seconds (8139too) Have you tried enabling network hotplugging for this card? It is capable of it, and in that case you will never wait for it to come up. eth1 1 minute 40 seconds (wlan orinoco) This is with default setup from Drak on Fujitsu E7010 (P4 notebook). I am not sure if wireless cards work that well with ifplugd, but this could certainly be fixed by using network profiles in the Mandrake control center. I've tested and it's working for me ! It takes me less than an hour to configure it ... It would be great if the init script could be replace by this one... All informations are on this page : http://www-106.ibm.com/developerworks/linux/library/l-boot.html Sure, but it has nothing to do with the problem you have described (long delay bringing up network connections), since you would still have to wait for network services to be brought up if you are not plugged in, whereas enabling network hotplugging (should work out-the-box for 8139too-based cards at least) *does* deal with this. Please provide the contents of your /etc/sysconfig/network-scripts/ifcfg-eth0 and /etc/sysconfig/network-scripts/ifcfg-eth1 if you are unsure about how to check for network hotplugging. * Please click on http://www.cae.co.za/disclaimer.htm to read our e-mail disclaimer or send an e-mail to [EMAIL PROTECTED] for a copy. *
