- Arch-general - lists.archlinux.org

[arch-general] Cannot install bacula-client error unknown public key
by Maykel Franco 06 Aug '15

06 Aug '15

Hi, I try install bacula-client in my archlinux but I get this error: -> Found bacula-fd.service ==> Validating source files with sha256sums... bacula-7.0.5.tar.gz ... Passed bacula-7.0.5.tar.gz.sig ... Skipped bacula-fd.service ... Passed ==> Verifying source file signatures with gpg... bacula-7.0.5.tar.gz ... FAILED (unknown public key 9E98BF3210A792AD) ==> ERROR: One or more PGP signatures could not be verified! ==> ERROR: Makepkg was unable to build bacula-client. ==> Restart building bacula-client ? [y/N] How can resolve it? Thanks in advanced.

3 3

[arch-general] Additional mouse (Lenovo N700) gesture binding
by Oon-Ee Ng 06 Aug '15

06 Aug '15

I just got the above bluetooth and/or wireless (dual-mode) mouse for my work, as the presenter laser looked pretty useful. 3 basic buttons (1, 2, 3) worked out of the box. So did the touch-scrolling (up/down). So I'm looking for advise for the remaining 3 inputs supported. This mouse has a 'Windows key' button and supports left/right swipe gestures (targeted at the new Windows side-screen gestures). Firstly, regarding the gestures, here's what I obtain from xinput test. I've also listed which keys these correspond to. Left swipe:- key press 22 Backspace key press 133 Super_L (left Windows key) key press 37 Control_L key release 22 Backspace key release 133 Super_L (left Windows key) key release 37 Control_L Right swipe:- key press 54 c key press 133 Super_L (left Windows key) key release 54 c key release 133 Super_L (left Windows key) This seems to correspond to pressing 3/2 key combos and releasing them in the same order. This seems to cause problems with xbindkeys and sxhkd (the only two key-binding daemons I've used before) as using the Super-c combo expects the Windows key to be pressed first, followed by the c key being pressed and released. How can I bind gestures to the above? Is it possible, for example, to bind keys to a specific input device (so for example I can bind to the letter c but only from the mouse, not from my keyboard)? Or something involving xmodmap? Thanks for your advise/help.

3 8

[arch-general] Since last systemd update dhcpcd no longer updates resolv.conf
by Manuel Reimer 04 Aug '15

04 Aug '15

Hello, since the last systemd update I have to "systemctl restart dhcpcd@..." every time I boot my system. What do I have to do to get sure that dhcpcd updates /etc/resolv.conf on the first time? Thanks in advance Manuel

3 2

[arch-general] systemd-networkd, bridging, and STP
by Zach La Celle 03 Aug '15

03 Aug '15

I'm trying to set up a server network configuration in Arch using systemd-networkd tools. The networkd configuration files that I have are not doing what I think they should be doing. My setup is this: 4 physical ports -> bond0 -> vlans 2-8 -> a few on bridges, a few on their own. First I'll post the two errors I'm seeing, then I'll post my configs below to help make sense of it. First error: systemd-networkd[315]: [/etc/systemd/network/10-create-bond0.netdev:7] Failed to parse bond transmit hash policy, ignoring: "802.3ad" Second error, is that I can't get br-devel bridge to use STP: # brctl show bridge name bridge id STP enabled interfaces br-devel 8000.4e565e92019a no bond0.3 bond0.6 I have a [Bridge] section in the .network file that sets the bridge address/network settings, but it doesn't seem to have any effect. __Config Files__ Right now, I have a bond set up: _10-create-bond0.netdev_ [NetDev] Name=bond0 Kind=bond [Bond] Mode=802.3ad TransmitHashPolicy=802.3ad LACPTransmitRate=fast MIIMonitorSec=0.1 UpDelaySec=0.2 DownDelaySec=0.8 I also have a bridge set up: _20-create-brdevel.netdev_ [NetDev] Name=br-devel Kind=bridge I then create some VLANs: _20-create-vlan3.netdev_ [NetDev] Name=bond0.3 Kind=vlan [VLAN] Id=3 _20-create-vlan6.netdev_ [NetDev] Name=bond0.6 Kind=vlan [VLAN] Id=6 Next, I set up the VLAN networks: _30-setup-vlans.network_ [Match] Name=bond0 [Network] VLAN=bond0.3 VLAN=bond0.6 Next, I add the interfaces to the bridge: _40-add-bond0.3-brdevel.network_ [Match] Name=bond0.3 [Network] Bridge=br-devel _40-add-bond0.6-brdevel.network_ [Match] Name=bond0.6 [Network] Bridge=br-devel Finally, I set up the br-devel bridge: _50-brdevel.network_ [Match] Name=br-devel [Network] Description="Development Bridge" [Address] Address=172.18.1.50/16 Broadcast=172.18.255.255 [Bridge] UnicastFlood=on HairPin=true UseBPDU=yes FastLeave=off AllowPortToBeRoot=on

1 0

[arch-general] Anyone using virtualbox-5.0.0 yet?
by walt 02 Aug '15

02 Aug '15

Arch updated vbox to 5.0.0 on 22 July, so I'm wondering if anyone is actively using it, and if you've had any problems with it. I certainly had a catastrophic problem when I tried it on my everyday gentoo desktop machine recently. If you Archers have been using it without any problems then I have a whole lot of research to do...

8 9

[arch-general] Update to OpenCV 2.4.11?
by André Apitzsch 29 Jul '15

29 Jul '15

Hi everybody, OpenCV 2.4.11 has been released about five months ago. Arch still ships 2.4.10. An Update to the newer version would simplify transition to 3.0. See http://code.opencv.org/projects/opencv/wiki/ChangeLog#30-rc What's holding back an update to 2.4.11? Regards, André

3 3

[arch-general] [extra] festival packge (Orphan)
by Tobias Markus 28 Jul '15

28 Jul '15

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi list, I hope this is the right place to post. The festival package (in [extra]) has been an orphan and out-of-date for a while, 2.1 is packaged, 2.4 is available. I fixed some of the errors coming up when trying to build festival 2.4 in the festival-patched-hts package (AUR) [1], and it works quite well now. The PKGBUILD can be modified for the non-HTS-patched version quite easily, only removing the relevant patches, and I kept modifications to a necessary minimum (to make it obvious I don't have any bad intentions...). Because of this, I wonder if some package maintainer could pick the actual festival package up and apply my changes to make it work with 2.4. I encountered a build error that is very likely an upstream bug [2], but I don't have time to report it right now. A workaround until the issue is resolved upstream is included. Aside from that, I changed the MD5 hashes to SHA-256 hashes and removed some parts of the some of the compatibility patches that were made obselete by upstream changes. Please do not hestitate to ask me further questions regarding the changes involved. Thank you very much in advance, Best regards Tobias [1] https://aur4.archlinux.org/packages/festival-patched-hts/ [2] festival/src/modules/clustergen/HTS_vocoder_me.cc includes a C file (not a header!), namely festival/src/modules/hts_engine/HTS_vocoder.c, causing linkage errors (multiple definitions of symbol). Workaround is a header file and some uncontroversial non-breaking changes. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iEYEARECAAYFAlW314IACgkQAO6N0EYmC9b0jgCgugOF+IgM7l7qfzIP+QmEETHA 4+YAmwUwmghQEkSXf3F/0Y4mdC8CBaTk =wg7q -----END PGP SIGNATURE-----

1 0

[arch-general] kernel package 4.1.3 still in testing
by David Kaylor 28 Jul '15

28 Jul '15

Anyone have any idea why 4.1.3 is being held back in testing, nearly a week now? Seems unusual for a .3 or newer, although I realize it might not be for a x.x.1 or .2. Anyone using the testing repo having major issues? No bugs shown against it, and nothing obvious in testing or kernel/hardware forums. Upstream, not sure about.

1 0

[arch-general] Possibility of including xorg input patch
by Kirill Churin 25 Jul '15

25 Jul '15

Can this patch https://bugs.freedesktop.org/attachment.cgi?id=33142 be applied to Arch's xorg? It solves famous problem https://bugs.freedesktop.org/show_bug.cgi?id=865 with broken keyboard input for users who use multiple keyboard layouts without any noticeable side-effects. It's known this patch has been applied in Ubuntu. -- Kirill Churin Jabber: reflexing(a)reflexing.ru, ICQ: 8163230, Skype on demand.

5 4

[arch-general] Signing kernel modules
by Damjan Georgievski 25 Jul '15

25 Jul '15

Since some time ago, the Linux kernel has had support for cryptographically signed modules, i.e. the kernel can be configured to only load properly signed modules. https://www.kernel.org/doc/Documentation/module-signing.txt I wouldn't go to lengths explaining the benefits of it, I was just wondering if Arch has the infrastructure to sign the modules - since developers already sign the kernel package itself [*], and in that sense we (the users) already trust them to that level. I'm not suggesting that CONFIG_MODULE_SIG_FORCE is enabled by default - that would break 3rd party modules, but people could optionally use the "enforcemodulesig=1" kernel command line option if they want to. Thoughts? [*] though packages are gpg signed, and modules use x.509 -- damjan

2 1