- Arch-dev-public - lists.archlinux.org

[arch-dev-public] problems uploading a package
by Roman Kyrylych 25 May '08

25 May '08

[roman@home trunk]$ extrapkg gtkspell-2.0.13-1-i686.pkg.tar.gz 100% 24KB 24.0KB/s 00:00 File integrity okay. ===> Uploaded /home/roman/packages/gtkspell-2.0.13-1-i686.pkg.tar.gz ===> Commited with "upgpkg: gtkspell 2.0.13-1" message D ../repos/extra-i686/gtkspell-2.0.11-enchant.patch CG ../repos/extra-i686/PKGBUILD property 'svnmerge-integrated' set on '../repos/extra-i686' ~/scm-repos/svn-packages/gtkspell ~/scm-repos/svn-packages/gtkspell/trunk svn: Commit failed (details follow): svn: Aborting commit: '/home/roman/scm-repos/svn-packages/gtkspell/repos/extra-i686/PKGBUILD' remains in conflict ~/scm-repos/svn-packages/gtkspell/trunk ===> Tagged for extra-i686 [roman@home trunk]$ As can bee seen from arch-commits - commit to trunk/ was successful, but committing to repos/ fails. Any ideas why this happens? -- Roman Kyrylych (Роман Кирилич)

2 2

[arch-dev-public] [signoff] ntfs-3g 1.0.2506-1
by Thomas Bächler 25 May '08

25 May '08

Forgot to send the signoff message yesterday: Another update, please somebody sign off this time. For the users reading this: As very few developers use ntfs-3g, please do a short functionality test (mounting, reading/writing a few files) and sign off if it works, so this can go to core fast.

3 2

[arch-dev-public] kernel26 moved, announcement draft
by Thomas Bächler 24 May '08

24 May '08

I'm sorry I didn't think about it earlier, but it was time for the kernel to be moved, so I did it. So this is (perhaps a little late), a draft for the announcement: Linux 2.6.25 moves to core It took much longer than usual, I'm sorry for that, but kernel26 can now be moved to core. Here are the changes: - Configuration has been reviewed completely, some more drivers have been enabled, configurations for both architectures are now in sync - IO accounting has been enabled, this makes it possible to use iotop - Cleaned up the kernel patches, everything is now in one big -ARCH patch, for details look at http://projects.archlinux.org/?p=linux-2.6-ARCH.git;a=summary Problems: - The USB subsystem can now only be accessed by GPL kernel modules. Therefore, the slmodem USB module does no longer work. I compiled the slmodem package without the slusb.ko module, I have no idea if this is still useful. - martian no longer compiles, I have no idea why and didn't find a fix for it. - Some users report that their iwl4965 is broken in 2.6.25, while others say it works fine. See these two bug reports: http://bugs.archlinux.org/task/10389, http://bugs.archlinux.org/task/10442 I advise all affected users to keep a copy of the old kernel26 and module packages or not to update at all.

4 4

[arch-dev-public] introducing qt-4.4
by Pierre Schmitz 24 May '08

24 May '08

Hi all, I just uploaded a first version of qt-4.4 to [testing]. This is more than a minor upstream update, so please have a look at your qt4-apps. I have updated lyx to 1.5.5 to make it work with new qt. For PSI I had to use a svn-snapshot. But I think there'll be a final release of 0.12 quite soon. I haven't reviewed the configuration of the qt package yet; so I am open to any suggestions. Pierre -- archlinux.de

1 0

[arch-dev-public] On vacation until 23rd of May
by Daniel Isenmann 23 May '08

23 May '08

Hi, I will be on vacation until 23rd of May. There will be no computer with internet access. Feel free to update my packages if it is necessary... Have a good time.... Bye, Daniel

2 2

[arch-dev-public] Moving the DB scripts
by Aaron Griffin 22 May '08

22 May '08

Hey all, I'm going to move the new db scripts live. In an hour or so. I still need to generalize Pierre's testing2core script and commit it, then I will push them live. Once I do that, it is going to break the website updates for a short period - until Eliott can get to the web changes (we haven't been able to coordinate properly, so rather than wasting time, I'm doing this move now). So, if you guys could hold off doing repo related activities in about an hour or so, that'd be great. One fun addition: /arch/db-remove No more need for package files to remove a package. Syntax is: db-remove pkgname repo arch Cheers, Aaron

7 28

[arch-dev-public] Installer release was: [signoff] kernel26 2.6.25.4-1...
by Aaron Griffin 22 May '08

22 May '08

On Thu, May 22, 2008 at 4:24 PM, Simo Leone <simo(a)archlinux.org> wrote: > On Thu, May 22, 2008 at 04:20:48PM -0500, Simo Leone wrote: >> >> Signed off kernel, mkinitcpio, and catalyst - i686 >> >> Let's do this thing. >> > > Oh and one more thing, the new installer release is only blocked on the > kernel. They'll be getting out to mirrors within 12 hours of the kernel > getting pushed. So I guess that makes today a big day :) Quick question - does the installer still use packages.txt? Those files are way broken.

1 0

[arch-dev-public] [signoff] kernel26 2.6.25.4-1, mkinitcpio 0.5.19-1 and 2.6.25 modules
by Thomas Bächler 22 May '08

22 May '08

The new mkinitcpio works around the custom dsdt bug, thus I think we can move to core now. Please sign off on all you can. I didn't include support for more than 4 CPUs now, because it would have required yet another rebuild of all modules (I already had to rebuild them all for 2.6.25.3 due to configuration changes). More CPUs are not that critical, so they can wait until 2.6.26. martian is still broken and I have no idea how to fix it. That means it will be dropped (or just kept lying around until there is a fix), I doubt that many people actually need it.

6 6

[arch-dev-public] [signoff] initscripts 2008.05-1
by Thomas Bächler 22 May '08

22 May '08

I think it is time to release this, here are the changes: - Remove usbfs special handling from rc.sysinit, for explanation see http://localhost/git/?p=arch/initscripts/.git;a=commit;h=f3e064ec822b3dae19… - Make USEDIRECTISA default to no - this is a change in behaviour compared to what we had in the past, but it makes more sense like this - Remove the initrd hacks, this has all been solved more elegantly with initramfs - Fix some bug in the network daemon - Remove the screen blanking setting from rc.sysinit, we don't usually hardcode such things - Fix some weird udev issues, hopefully (I have no idea why the fix works, but it has been confirmed) - Some random cosmetics/fixes in code and output If you want to know it all, look at the git repository.

4 4

[arch-dev-public] [signoff] libtool 2.2.2-1
by Andreas Radke 18 May '08

18 May '08

regression fixes we almost all had in our last pkg. -Andy EWS - list of user-visible changes between releases of GNU Libtool New in 2.2.2: 2008-04-01: CVS version 2.2.1a, Libtool team: * New features: - In compile mode, compiler output occurs in the user locale. This feature has been present in 1.5.26 but not in 2.2. * Changes in supported systems or compilers: - Initial shared library support for AmigaOS4 on powerpc. * Bug fixes: - Fix 2.2 regression in libltdl that causes memory corruption upon repeated `lt_dlinit(); lt_dlexit()'. - Fix 2.2 regression in libltdl that skipped the dlopen loader if the system also supports other loaders (e.g., Cygwin, HP-UX). - Fix 2.2 regression in that `libtool --mode=execute CMD ARGS' does not transform ARGS that do not look like shell or C wrappers of libtool programs. - Fix 2.2 regression that kept cross-compiling to w32 from working. - Several testsuite issues have been fixed, thanks to user feedback.

4 10