November 2015 - Arch-general - lists.archlinux.org

[arch-general] Policy about packages and file capabilities
by Damjan Georgievski 16 Nov '15

16 Nov '15

What's the policy about capabilities for executables in Arch packages? I'm asking since in my setup I'm running wpa_supplicant as the 'nobody' user, but I let it keep the NET_ADMIN and NET_RAW capabilities (excerpt from the .service file): User=nobody SupplementaryGroups=rfkill CapabilityBoundingSet=CAP_NET_ADMIN CAP_NET_RAW The executable needs to have those capabilities on the executable in order to keep them when switching to user `nobody`, so I set them with: setcap 'cap_net_raw=ep cap_net_admin=ep' /usr/bin/wpa_supplicant I'd suggest (and send a patch) to add these capabilities in the next update to the package if that's acceptable. PS. wpa_supplicant is still not updated from version 2.3 which doesn't work with the wifi interface in the 4addr mode, bridged and with the -b option of wpa_supplicant. 2.4 and 2.5 work in that scenario. so I hope an update is due. -- damjan

2 3

[arch-general] gsl 2.1
by Scott Lawrence 16 Nov '15

16 Nov '15

Hi all, It seems that the latest officially stable version of gsl (gnu scientific library) is 2.1. The repository is at 1.16, and I don't see anything more recent in [testing] or AUR. Is there an intention to update this package straight to 2.1, or to introduce a gsl2 package? I've no problem with putting this on AUR myself, but I don't want to step on toes / waste time. -- Scott Lawrence

4 4

[arch-general] Thunar sftp connection not working - access denied
by Friedrich Strohmaier 15 Nov '15

15 Nov '15

Hi Folks, since some time I cannot get thunar connect my remote servers folders via sftp. Error dialog: Failed to open "/my/remote/path on my.server" access denied ssh access works as expected. I feel like it stopped working after I got this arch announce: https://www.archlinux.org/news/d-bus-now-launches-user-buses/ System: recent arch x86_64 xfce4 4.12 I didn't find anything useful by net search, so my question here: Did anyone else get hit by this? Is there likely a relation to changes in D-Bus. Any hint to nail down this one appreciated. -- Friedrich

3 8

[arch-general] Line Drawing Characters
by David N Murray 14 Nov '15

14 Nov '15

Hi all, I've built several machines by installing Arch from scratch. Recently, I built a Linode VM using their Arch image, dated 2015-08. From there, I did a normal `pacman -Syu` to update the system. My problem is probably trivial, annoying, and I haven't been able to figure it out. I'm hoping someone can give me a hint on where to look. I've been through the post-install docs and can't figure out what might be causing this. The machine I'm on is running xfce and I'm using xfce4-terminal 0.6.3. If I do something like `systemctl status`, I see output that is mostly ascii characters: * on the first line next to the name of the host, pipe (|) next to each line in the CGroup. If I ssh into another local machine, I see the same thing. My problem is on the Linode box. When I ssh there and run the same command, I'm seeing garbage where I see pipes next to each line in the CGroup output. I don't know if this is going to paste properly, but I'll try: Local: CGroup: / |-init.scope Linode: CGroup: / ââinit.scope On the linode box, those 2 characters actually occupy 6 characters in the output and the last 5 characters are each drawn in a box and overwrite the first character of the process name. All the characters appear as "garbage." What's really strange is that the output from `top` is fine, and that appears to use some graphics and color. Has anyone seen something like this? All sessions have the same TERM variable (xterm). This is an area of any Linux I've never spent much time. It always just worked out-of-the-box (for any distro I've ever used). The only thing I can think of at this point is to try and diff the two systems' /etc directories and see if I can track it down that way. Maybe that's next. TIA, Dave

3 3

[arch-general] Xorg 1.18.0 enters [testing].
by Laurent Carlier 14 Nov '15

14 Nov '15

Xorg 1.18.0 is entering [testing] with the following changes: - You can now choose between xf86-input-evdev and xf86-input-libinput. - xf86-input-aiptek will not be updated and will be removed when xorg-1.18.0 is moved to [extra] Caveat: NVIDIA drivers are not yet compatible with 1.18.0. You can block the upgrade by appending --ignoregroup=xorg to the pacman command or adding 'xorg' to IgnoreGroup in pacman.conf -- Laurent Carlier http://www.archlinux.org

2 1

Re: [arch-general] [arch-announce] Xorg 1.18.0 enters [testing]
by Leonid Isaev 14 Nov '15

14 Nov '15

On Fri, Nov 13, 2015 at 01:01:09PM -0000, Arch Linux: Recent news updates: Laurent Carlier wrote: > Xorg 1.18.0 is entering [testing] with the following changes: > > * You can now choose between `xf86-input-evdev` and `xf86-input-libinput`. > > * `xf86-input-aiptek` will not be updated and will be removed when xorg-1.18.0 > is moved to [extra] > > Caveat: > > NVIDIA drivers are not yet compatible with 1.18.0. You can block the upgrade by > appending --ignoregroup=xorg to the pacman command or adding 'xorg' to > IgnoreGroup in pacman.conf > > URL: https://www.archlinux.org/news/xorg-1180-enters-testing/ Hmm, on my machine Xorg now runs as root, and I had to create /etc/X11/Xwrapper.config to enfoce the old behavior of dropping suid privs... Has anyone else noticed this? Thanks, -- Leonid Isaev GPG fingerprints: DA92 034D B4A8 EC51 7EA6 20DF 9291 EE8A 043C B8C4 C0DF 20D0 C075 C3F1 E1BE 775A A7AE F6CB 164B 5A6D

1 0

[arch-general] systemd-networkd and netctl with multiple interfaces
by Bennett Piater 13 Nov '15

13 Nov '15

Hello! I installed Arch on my new Thinkpad T450s over the weekend. Everything works well, but I have a question: I use systemd-networkd to manage my network interfaces and netctl for the connections. I set everything up according to (this)[0] and (this)[1] to get automatic activation of wifi via netctl-auto and netctl-ifplugd. My question is as follows: I use i3wm, and i3status shows *both* ethernet and wifi as connected if I plug in the cable while having a wifi connection. What does this mean exactly, and how is my traffic routed? Thanks in advance for clearing that up. Cheers, Bennett [0]: https://wiki.archlinux.org/index.php/Systemd-networkd#Wired_and_wireless_ad… [1]: https://wiki.archlinux.org/index.php/Netctl#Automatic_switching_of_profiles -- GPG fingerprint: 871F 1047 7DB3 DDED 5FC4 47B2 26C7 E577 EF96 7808

8 11

Re: [arch-general] QT applicacion dies on plain fluxbox, but not on fluxbox on top of LXDE
by Javier Vasquez 12 Nov '15

12 Nov '15

> On Wed, Nov 11, 2015 at 10:21 AM, Javier Vasquez <j.e.vasquez.v(a)gmail.com> wrote: > ... > > In case more information is required, well besides bein using Arch for > x86-64, up to date, with fluxbox version 1.3.7-1, attached go some > outputs from strace, gdb, and coredump. I actually can't make much > sense out of them... OK, attachments failed, so linking to gist files with the outputs: https://gist.github.com/je-vv/89c4910a740afcac89a7#file-hpmyroom_coredump-t… https://gist.github.com/je-vv/7595e5949870db15d6d0#file-hpmyroom_gdb-txt https://gist.github.com/je-vv/a3a7b9d95b8d0eca9078#file-hpmyroom_strace-txt Thanks, -- Javier

1 1

Re: [arch-general] systemd-networkd and netctl with multiple interfaces
by Tinu Weber 11 Nov '15

11 Nov '15

On Wed, Nov 11, 2015 at 16:06:18 +0100, Bennett Piater wrote: > They have different metrics as per the example from the wiki. > > However, no wiki article or manpage that I encountered explained what > exactly the metric does. Could you explain that to me? > > Cheers, > Bennett > > > If there are multiple default routes with the same metric, this might > > cause issues, as packets might be sent randomly through one of the two > > interfaces (thus with different IP addresses), leading to packet drops. > > > > If the metric value is different, this shouldn't be a problem, though. The idea is that if none of the routes match the destination, the default route is used. If there multiple of them... well, nobody can tell you what happens. Usually the kernel chooses the route with the lower metric (to "break the tie"), but I cannot tell you exactly what it signifies. I'd suggest to try to find out why netctl makes it possible to have multiple default routes (however, as I don't use netctl, I'm not able to answer this), as this is usually a bad idea. By the way, I'll redirect this discussion back to the mailing list, maybe someone can answer this. Best, Tinu

1 0

[arch-general] QT applicacion dies on plain fluxbox, but not on fluxbox on top of LXDE
by Javier Vasquez 11 Nov '15

11 Nov '15

Hi, I've been using plain fluxbox, rather than any DE, for several years now. I use dunst for libnotify server BTW... However now a days there's an application "hpmyroom" [1], based on QT for linux, which dies any time there's an interactive popup. For example deleting a buddy will pop up a dialog to ask whether you want to confirm or not. Or when someone sends a buddy request, a dialog pops up to confirm whether to accept it or not. When using plain fluxbox, "hpmyroom" immediately dies with a segmentation fault, when such popups are suppose to happen (nothing pops up actually). Other guys using Arch were not having problems, but they were using DEs like XFCE, Gnome, KDE and others. So I tried LXDE, but using fluxbox instead of openbox. To my surprise, that worked out. Same fluxbox, just on top LXDE. Problem is that I don't really want to be running a DE. I was exploring what would LXDE be providing that I'm not using on plain fluxbox with no DE, but couldn't find what it might be. As mentioned, I was already using a libnotify server, dunst. I changed it for others like "xfce4-notifyd" and "notification-daemon", which made no difference. So I discarded a libnotify issue. Then the other thing that came to mind, though more relunctantly was graphics polkit. So I called "lxpolkit" (part of lxsession from LXDE) on the .xinitrc, prior to calling "startfluxbox", but that didn't help either. So, as I suspected initially, that didn't seem like it. Arch handles dbus through xinit, when calling "startx" or using "xdm", but then just in case I tried manually: systemctl --user import-environment DISPLAY XAUTHORITY dbus-update-activation-environment DISPLAY XAUTHORITY As expected, that didn't help a bit... Anyone with some clues as to what I might be missing to keep using plain fluxbox? -- Javier [1] https://www.myroom.hpe.com

1 1