- Arch-general - lists.archlinux.org

[arch-general] [signoff] kernel26 2.6.27.10-1
by Tobias Powalowski 22 Dec '08

22 Dec '08

Hi guys, new kernel adresses the following things: - bump to latest version Plan to avoid file conflict: - Move module-init-tools 2 weeks after .10 kernel moved to core. - Release a NEWS ITEM: Possible file confilct, while upgrading kernel26 2.6.27.10 and module-init-tools 3.5, please use 'pacman -Sf kernel26' to avoid warnings about existing module-init-tools files. any other suggestion? please signoff for both arches, greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa(a)archlinux.org

2 1

[arch-general] scons "update"
by Alessandro Doro 20 Dec '08

20 Dec '08

# pacman -Syu ... warning: scons: local (1.1.0.d20081104-1) is newer than extra (1.1.0-1) ... Is there a reason for reverting to the stable scons release or the packager intended to update to latest 1.1.0.d20081207? bye

3 5

[arch-general] Access arch-general ML archive before November 2007
by Grigorios Bouzakis 20 Dec '08

20 Dec '08

Does anyone know of a place were tha arch-general (at the time simply arch) ML archive is available before November 2007? I wanted to access it but realised it isnt available. I guess the name change had much to do with it, but i dont believe the previous years are lost. Greg

2 2

[arch-general] More simple udev blacklisting
by Michal Soltys 19 Dec '08

19 Dec '08

With reference to older discussions about blacklisting methods - modprobe supports -b (or in long form --use-blacklist) that will also block explicit loading by standard module name. Thus you could nicely avoid shell wrapper on fb stuff. The switch won't block implicit dependencies though (so if A requires B, B is blacklisted, then modprobe -b A will still load A and B). That shouldn't be a problem from udev's perspective though. Stock udev rules will use -b by default from the next release, peeking at the mailing list. FYI.

1 0

[arch-general] X crashes after upgrade to intel
by ert256 19 Dec '08

19 Dec '08

Hello After system upgrade, and configuring intel driver, X crashes width message : Backtrace: 0: /usr/bin/Xorg(xorg_backtrace+0x3b) [0x813017b] 1: /usr/bin/Xorg(xf86SigHandler+0x51) [0x80d8cb1] 2: [0xb7f8f420] 3: [0xb7f8f420] 4: /usr/bin/Xorg(ProcGetImage+0x5d0) [0x808a130] 5: /usr/bin/Xorg(Dispatch+0x34f) [0x808c20f] 6: /usr/bin/Xorg(main+0x47d) [0x8071c8d] 7: /lib/libc.so.6(__libc_start_main+0xe5) [0xb7b7e6c5] 8: /usr/bin/Xorg(FontFileCompleteXLFD+0x209) [0x8071071] Fatal server error: Caught signal 11. Server aborting Complete log is here :http://student.agh.edu.pl/~ert/xlog.txt Complete configuration is here: http://student.agh.edu.pl/~ert/config.txt Please help. -- Rafał (ert16) Trójniak m@il : ert256(a)gmail.com Jid : ert256(a)gmail.com GPG key-ID : DD681D47 749F E1DC A58F 9084 BBC0 797A 0691 53D6 DD68 1D47

2 2

[arch-general] Arch Linux automatic (prescripted) installer/deployment tool
by Dieter Plaetinck 18 Dec '08

18 Dec '08

Hi all, I've started working on a new procedure for AIF called 'automatic'. With automatic I don't mean a newbfriendly installer that automatically configures your xorg etc. I rather mean an installer that you pass a small configfile (disk layout, package list,...) so it can work non-interactively to setup your system exacty the way you want. I hope I/we can someday implement such automatic installation in two ways: - an environment that gets loaded through by pxe booting and can work without any user input at all. Best case scenario we can generate such an environment like we generate cd iso's, and we automatically start aif with the automatic procedure. A sufficient live network config will be done (the bios will use dhcp before pxe booting) and the config file for the procedure can be made available for example by the user himself who generated the pxe environment. (low prio for me personally but maybe someone is interested in this, it's something to keep in mind) - we use the installer cd with AIF on it, you boot it and start "aif -p automatic" yourself. 2 more manual things need to happen: * Network config (for that I plan to write a simple script that uses the aif libraries) * getting the config file for the automatic procedure After that, the automatic procedure can work independently. Given the modular, reusable code base we already have now, this should be fairly easy to implement (although I still need to work more on aif, the way pacman functions are exposed etc) Does anyone have thoughts, ideas, requirements, ...? Dieter

4 6

[arch-general] [arch-dev-public] Move Azureus to community?
by Grigorios Bouzakis 18 Dec '08

18 Dec '08

On Thu, Dec 18, 2008 at 11:25 PM, Aaron Griffin <aaronmgriffin(a)gmail.com> wrote: > http://bugs.archlinux.org/task/12480 > > Opinions? > Theres also bittorrent which hasnt been able to operate for at least a year. AFAIK it cant be updated either. http://bugs.archlinux.org/task/6764 http://bugs.archlinux.org/task/6758 Greg

1 0

[arch-general] wrong file permissions in /var/lib/pacman/local
by Hubert Grzeskowiak 18 Dec '08

18 Dec '08

hi there, can it be, that the some dirs written to /var/lib/pacman/local/* are not chmod'ed properly? sometimes i get errors because as user i don't have the permissions to do anything with it (it's set to drwx------). this only occurs on some packages (eg. the new nmap package). other dirs and files inside all (also the corrupted) directories seem okay (files: -rw-r--r--; dirs: drwxr-xr-x) i played around with umask and sudo (always used on de-/installations) with no result -everything was okay- so i conclude it's a per-package problem. is anybody else encountering this? thx in advance Hubert Grzekowiak

6 19

[arch-general] new AIF release
by Dieter Plaetinck 17 Dec '08

17 Dec '08

Hello Arch devs and users, here is my holidays present for you: AIF alpha-0.6 ! * Changes since alpha 0.5: - fully reworked disk subsystem - support for lvm and dm_crypt - Autoprepare looks exactly the same (ui-wise) - Manual mode UI is entirely reworked and has a general editor (see screens) - Unified, more flexible/reusable backend code for partitioning, making filesystems and mounting them. Uses structured file formats for defining partitions and filesystems - Rollback functionality: when something failed while processing, or user changes his mind afterwards, you can do a rollback (things like umount, lvremove etc are done automatically in the correct order) and start over/retry - fully decoupled UI (dia/cli): - there are now generic, transparent functions for asking dates,numbers,strings,checklist,.. etc with an implementation in both ncurses (dialog) and cli. - all code now uses these functions, making the codebase much cleaner. (look at the code for autoprepare for instance) - with a simple command line flag you can toggle between running the program in cli or dialog mode. - support for commandline args - /arch/aif -p interactive -d # runs the interactive profile with debugging enabled - /arch/aif -p interactive -i cli # runs the interactive profile in CLI mode. - updated howto for using AUR packages - replaced all refs to previous name (fifa) to aif. - enabled date/time setting - a massive amount of bugfixes - a lot of code cleanups * Screenshots - AIF manual filesystems, here showing a setup with /dev/sda3 containing a dm_crypt, containing /, and /dev/sda3 containing an LVM setup with 1 LV for /home http://users.edpnet.be/dieter/aif/aif-alpha_0.6-disks-overview-dia.png - Processing the filesystems. The installer automatically reorders everything http://users.edpnet.be/dieter/aif/aif-alpha_0.6-disks-processing-dia.png - a CLI shot where the partitions/filesystems of the autoprepare just have been processed http://users.edpnet.be/dieter/aif/aif-alpha_0.6-autoprepare_processed-cli.p… * Usability I tested the installer personally in virtualbox in different scenarios (ftp & cdrom based, autoprepare, manually with normal disk layout,/ on lvm, dm_crypt,.. ), and all produced working installations. Although I often need to fix bugs after testing a specific scenario, I don't re-test all scenarios each time. YMMV Note that the installer doesn't automatically add the lvm2/encrypt modules to mkinitcpio yet, and doesn't add the right string for lvm/dm_crypt in grubs menu.lst yet. There are some known issues (which are not critical, just avoid them or live with them). Here are the biggest ones (for more, see TODO) - don't remove lvm LV's if you have more then one in one lvm VG. So think before you create :-) - make the correct devicemapper devices directly because they are not cleaned up correctly if you remove them again - keyboard and timezone settings from installer don't go into target systems config - long selections (eg packages) are very cumbersome to do in CLI mode. - rollback of dm_crypt partitions doesn't work yet * How? http://github.com/Dieterbe/aif/tree/master/HOWTO Use package aif-git from AUR. Even though I tag releases in git, I don't build packages for them. I now use the master branch as "stable" branch where I pull in code from the experimental branch after it has been tested. So the master branch (package aif-git) always contains the latest, stable code. Use "/arch/aif -p interactive -d" (the interactive procedure is the port of /arch/setup) Have fun, Dieter

4 6

Re: [arch-general] [arch-dev-public] [signoff] pacman 3.2.1-2, pacman-mirrorlist
by Grigorios Bouzakis 17 Dec '08

17 Dec '08

On Sun, Dec 14, 2008 at 12:40 AM, Dan McGee <dpmcgee(a)gmail.com> wrote: > This is a dual signoff. I have split the pacman mirrorlist into its > own package because > 1) Pacman releases are not all that common (because its perfect > software, of course!) > 2) Our packaged mirrorlist often lags waaay behind the actual one in > SVN, because I find it a waste to rebuild pacman just for mirrorlist > changes. This will prevent that. > > I don't actually run pacman, so 2 signoffs per arch would be > wonderful. In addition, PLEASE let me know if anything goes awry > during a normal upgrade process. This could include mirrorlist not > being written to pacsave, mirrorlist disappearing, etc. > May i ask why pacman requires pacman-mirrorlist and not the other way around? Pacman can operate without a mirrorlist. Pacman-mirrorlist can not operate without pacman.. Greg

7 8