sdbbs's questions

I've recently discovered, that when I open a new text file (extension .txt) in nano, there is actually a limited syntax coloring:

That is, if a line starts with number sign/hash, it is colored.

I think this is great, and I really like this (for a txt file, I really don't need any other syntax color, but one marking comment, with the same comment syntax as in bash), so I was wondering how/where it was defined.

According to How to set nano default syntax highlighting for files with no extension?, syntax coloring files for nano are in /usr/share/nano/ directory, and are "imported" via /etc/nanorc file.

So, I tried checking:

$ grep -r txt /usr/share/nano/
/usr/share/nano/cmake.nanorc:syntax cmake "(CMakeLists\.txt|\.cmake)$"

... but the only reference to .txt is from cmake, and I don't think this is what sets the coloring for test.txt.

So, how does nano determine the syntax coloring for a test.txt file?

Well, I thought this would have been easier, but I simply cannot find whether what I want is doable.

I want to mount a network Samba share from the command line in Ubuntu 20.04 (MATE) - preferably without any changes in files or saving settings somewhere; just a command line.

So first I tried:

sudo mount -t cifs -o user=MYSMBUSER //192.0.2.1/myshare /tmp/myshare

This works great - in the sense of it mounting, listing files, and providing read-only access to files; but it does not allow writing files (permission denied).

That is seemingly because of:

https://unix.stackexchange.com/questions/68079/mount-cifs-network-drive-write-permissions-and-chown

You are mounting the CIFS share as root (because you used sudo), so you cannot write as normal user. If your Linux Distribution and its kernel are recent enough that you could mount the network share as a normal user (but under a folder that the user own), you will have the proper credentials to write file

Yes, I do want to "mount the network share as a normal user", and this might seem like an option - however, my username on Ubuntu is different from the Samba username, so:

An alternative is to specify the user and group ID that the mounted network share should used, this would allow that particular user and potentially group to write to the share. Add the following options to your mount: ...

$ sudo umount /tmp/myshare
$ 

$ ls /tmp/myshare
file_from_share01.txt  file_from_share02.txt ...

$ sudo umount -a -t cifs -l

Here is the tricky part - if I want to mount as normal user, and yet specify SMB username, I'd have to write the same command as above, but without sudo:

$ mount -t cifs -o user=MYUSER //192.0.2.1/myshare /tmp/myshare
mount: only root can use "--options" option

So, maybe I could have otherwise ran this as non-root, if I didn't have to use the user option - but I do. So I cannot really use this command.

One suggestion for going around this is to specify options in fstab - via mount share cifs folder without sudo :

To be able to mount that folder as a normal user, include the mount options for the cifs folder in the configuration file /etc/fstab, and add the options noauto,user,

However, I do not want to do this, because I do not want to keep records of the SMB credentials on my machine - and if I have to edit fstab, then the onus is on me to delete the entry from fstab once that is done, and usually at that time I'm tired, I could forget, blah blah.

I found that this is likely some security feature:

https://unix.stackexchange.com/questions/365308/use-mount-o-with-a-non-root-user

The -o limitation (only root can specify it) is in place to protect the system: the administrator can set up whatever options are necessary, either directly using -o or using /etc/fstab; users can only cause user-controllable file systems to be mounted or unmounted, without specifying options, because file system options allow a number of hostile scenarios. As a result, some file systems support other ways of setting certain options, such as the USER environment variable used above with CIFS.

Eh yeah, but for me its more secure not to enter any SMB paths in /etc/fstab at all.

So, I really would like to mount just by using a command line without changing any configuration files (including /etc/fstab) anywhere.

I hoped that there is another command, and I found gvfs-mount - but I have the same problem as Mount SMB share with gvfs from command line :

$ dbus-launch bash
$  gvfs-mount smb://192.0.2.1/myshare
This tool has been deprecated, use 'gio mount' instead.
See 'gio help mount' for more info.

Password required for share myshare on 192.0.2.1
User [locuser]: MYUSER
Domain [WORKGROUP]:
Password:
$

At this point, it appears that it mounted, but it in fact didn't:

$ ls /run/user/1000/gvfs/
$

... and via gvfs-mount -l, we can see there is no corresponding mount point:

$ gvfs-mount -l
...
Volume(3): test_nfs
  Type: GProxyVolume (GProxyVolumeMonitorUDisks2)
  Mount(0): test_nfs -> file:///tmp/test_nfs
    Type: GProxyMount (GProxyVolumeMonitorUDisks2)
Mount(0): myshare on 192.0.2.1 -> smb://192.0.2.1/myshare/
  Type: GDaemonMount

... and so we cannot access any files with this "mount" anyways.

$ ls ~/.gvfs
ls: cannot access '~/.gvfs': No such file or directory

$ gvfs-mount -u smb://192.0.2.1/myshare
This tool has been deprecated, use 'gio mount' instead.
See 'gio help mount' for more info.

And to address the mention of gio mount - first note, that without launching a dbus-session, it fails with:

$ gio mount smb://192.0.2.1
gio: smb://192.0.2.1: volume doesn’t implement mount

... then, within a shell started by dbus-launch bash, it behaves exactly the same as gvfs-mount:

$ gio mount smb://192.0.2.1/myshare
Password required for share myshare on 192.0.2.1
User [locuser]: MYUSER
Domain [WORKGROUP]:
Password:
$

Looks like it mounted, but:

$ gio mount -l
...
Volume(3): test_nfs
  Type: GProxyVolume (GProxyVolumeMonitorUDisks2)
  Mount(0): test_nfs -> file:///tmp/test_nfs
    Type: GProxyMount (GProxyVolumeMonitorUDisks2)
Mount(0): myshare on 192.0.2.1 -> smb://192.0.2.1/myshare/
  Type: GDaemonMount

... again there is no local directory mount point.

The above being said: is it possible to mount SMB share, from the command line, without changing any local files (such as, and especially /etc/fstab), without sudo, and with r/w permissions (which I've confirmed in Windows, the SMB user MYUSER has) - and if it is, how?

I work with a "virtual private server" in the cloud, which has Ubuntu 20.04 installed.

I've noticed that - occasionally - when I reboot the server, I cannot ping anything:

$ ping yahoo.com
ping: yahoo.com: Temporary failure in name resolution
$ resolvectl query yahoo.com
yahoo.com: resolve call failed: No appropriate name servers or networks for name found

As far as I understand, since this is Ubuntu 20.04, /etc/resolv.conf is not really used anymore, and systemd overtakes handling of DNS.

But, in any case, I would not want to manually specify DNS servers (as the old trick of putting Google's 8.8.8.8 in /etc/resolv.conf) - apparently, most of the time when DNS works: when the server boots, it somehow "obtains" a DNS server. So, I would like to repeat this process, except I have no idea how.

(By analogy, if I wanted to refresh/obtain new IP address for a machine, I'm aware that the machine should ask the DHCP server for it - so eventually, it is easy to lookup, that in order to do that, I should do sudo dhclient -r eth0 for instance)

Note also, that when DNS name resolution does not work, I do not have DNS Servers listed in this output:

$ systemd-resolve --status
Global
       LLMNR setting: no
MulticastDNS setting: no
  DNSOverTLS setting: no
      DNSSEC setting: no
    DNSSEC supported: no
          DNSSEC NTA: 10.in-addr.arpa
                      16.172.in-addr.arpa
                      168.192.in-addr.arpa
                      17.172.in-addr.arpa
                      18.172.in-addr.arpa
                      19.172.in-addr.arpa
                      20.172.in-addr.arpa
                      21.172.in-addr.arpa
                      22.172.in-addr.arpa
                      23.172.in-addr.arpa
                      24.172.in-addr.arpa
                      25.172.in-addr.arpa
                      26.172.in-addr.arpa
                      27.172.in-addr.arpa
                      28.172.in-addr.arpa
                      29.172.in-addr.arpa
                      30.172.in-addr.arpa
                      31.172.in-addr.arpa
                      corp
                      d.f.ip6.arpa
                      home
                      internal
                      intranet
                      lan
                      local
                      private
                      test

Link 2 (eth0)
      Current Scopes: none
DefaultRoute setting: no
       LLMNR setting: yes
MulticastDNS setting: no
  DNSOverTLS setting: no
      DNSSEC setting: no
    DNSSEC supported: no

I've seen that to clear the DNS cache, they (https://unihost.com/help/how-to-clear-the-dns-cache/) recommend:

sudo /etc/init.d/networking restart

... however, I'd rather not run this command, since I'm connected to the server over ssh, so I imagine that the restart of whole network, would also kill my ssh terminal session, and possibly have other unintended consequences.

So, I am wondering: is there a command I could run in an ssh session, that would refresh or obtain new DNS servers for this machine - using a similar mechanism that is used upon boot, when DNS actually works on this machine?

Let's say I want to build a package like network-manager; I can do, say, on 14.04 trusty:

apt-get source network-manager
cd network-manager-0.9.8.8
debuild -us -uc

... this, however, gives me a bunch of .deb files when it is done, such as:

gir1.2-networkmanager-1.0_0.9.8.8-0ubuntu7.4_amd64.deb
libnm-glib4_0.9.8.8-0ubuntu7.4_amd64.deb
libnm-glib-dev_0.9.8.8-0ubuntu7.4_amd64.deb
libnm-glib-vpn1_0.9.8.8-0ubuntu7.4_amd64.deb
libnm-glib-vpn-dev_0.9.8.8-0ubuntu7.4_amd64.deb
libnm-util2_0.9.8.8-0ubuntu7.4_amd64.deb
libnm-util-dev_0.9.8.8-0ubuntu7.4_amd64.deb
network-manager_0.9.8.8-0ubuntu7.4_amd64.deb
network-manager-dbg_0.9.8.8-0ubuntu7.4_amd64.deb
network-manager-dev_0.9.8.8-0ubuntu7.4_amd64.deb

Say I want to build only network-manager_0.9.8.8-0ubuntu7.4_amd64.deb - is there a debuild (or dpkg-buildpackage) switch that would allow me to do that?

This has been asked before at https://unix.stackexchange.com/questions/114364/nmcli-no-longer-allow-settings-modification - but never answered. So I'll try asking here:

You can see from diverse posts on the net, that the command-line Network Manager used to support something like (from https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/7/html/Networking_Guide/sec-Using_the_NetworkManager_Command_Line_Tool_nmcli.html):

$ nmcli connection show id 'MyCafe'
$ nmcli connection modify id 'MyCafe' 802-11-wireless.mtu 1350

Right now, I have on Ubuntu 14.04:

$ nmcli --version
nmcli tool, version 0.9.8.8

... and I get:

$ nmcli con modify
Usage: nmcli connection { COMMAND | help }
  COMMAND := { list | status | up | down | delete }

  list [id <id> | uuid <id>]
  status [id <id> | uuid <id> | path <path>]
  up id <id> | uuid <id> [iface <iface>] [ap <BSSID>] [--nowait] [--timeout <timeout>]
  down id <id> | uuid <id>
  delete id <id> | uuid <id>

Error: 'con' command 'modify' is not valid.

$ nmcli con show
Usage: nmcli connection { COMMAND | help }
  COMMAND := { list | status | up | down | delete }

  list [id <id> | uuid <id>]
  status [id <id> | uuid <id> | path <path>]
  up id <id> | uuid <id> [iface <iface>] [ap <BSSID>] [--nowait] [--timeout <timeout>]
  down id <id> | uuid <id>
  delete id <id> | uuid <id>

Error: 'con' command 'show' is not valid.

So, what is the equivalent for nmcli con show and nmcli con modify syntax, now that these keywords don't exist anymore?

On my laptop, I have Ubuntu 14.04. This laptop has both an RJ45 connector for wired Ethernet, and wireless/WiFi. I have noticed, however, that on this laptop, right after reboot, I have only this:

$ ifconfig
lo        Link encap:Local Loopback  
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
          RX packets:608 errors:0 dropped:0 overruns:0 frame:0
          TX packets:608 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0 
          RX bytes:76086 (76.0 KB)  TX bytes:76086 (76.0 KB)

wlan0     Link encap:Ethernet  HWaddr XX:71:XX:XX:b1:XX  
          inet addr:YY.0.XX.XX  Bcast:YY.0.XX.XX  Mask:255.255.0.0
          inet6 addr: fe80::XXXX:XXXX:fe3f:b1db/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:2589 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2044 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:1458453 (1.4 MB)  TX bytes:292837 (292.8 KB)

In fact, if I turn off wireless / uncheck "Enable Wi-Fi" from the Network Manager applet, I have only this:

$ ifconfig
lo        Link encap:Local Loopback  
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
          RX packets:1794 errors:0 dropped:0 overruns:0 frame:0
          TX packets:1794 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0 
          RX bytes:178068 (178.0 KB)  TX bytes:178068 (178.0 KB)

As you can see from the above ifconfig dumps, eth0 corresponding to the wired ethernet network is not listed at all! And in fact, it won't appear, until I connect the laptop with a router/switch, and the laptop gets an IP address!? In fact, while there is nothing connected to the RJ45 connector, the laptop reports just this via lshw:

$ sudo lshw -c network
  *-network               
       description: Wireless interface
       product: AR9485 Wireless Network Adapter
       vendor: Qualcomm Atheros
       physical id: 0
       bus info: pci@0000:01:00.0
       logical name: wlan0
       version: 01
       serial: XX:71:XX:XX:b1:XX
       width: 64 bits
       clock: 33MHz
       capabilities: pm msi pciexpress bus_master cap_list rom ethernet physical wireless
       configuration: broadcast=yes driver=ath9k driverversion=4.2.0-42-generic firmware=N/A ip=YY.0.XX.XX latency=0 link=yes multicast=yes wireless=IEEE 802.11bgn
       resources: irq:16 memory:f7c00000-f7c7ffff memory:f7c80000-f7c8ffff

Notice, the wired/eth0 adapter is not listed anywhere ?!

Otherwise, when I've connected the laptop's RJ45 connector with a network cable to a router or Internet cable, I get an ifconfig entry like:

eth0      Link encap:Ethernet  HWaddr XX:XX:2b:XX:XX:XX
          inet6 addr: fe80::YYYY:YYYY:fe0a:YYYY/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:6 errors:0 dropped:0 overruns:0 frame:0
          TX packets:61 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:1980 (1.9 KB)  TX bytes:12108 (12.1 KB)

In fact, in the above snippet, the eth0 didn't even get an IP4 address assigned; so apparently all it matters is that it is connected somehow to an Ethernet network.

Note, on this laptop, /etc/network/interfaces was containing just:

# interfaces(5) file used by ifup(8) and ifdown(8)
auto lo
iface lo inet loopback

... however, I added auto eth0 as a last line, saved and rebooted - and there was no change in behavior whatsoever (i.e. eth0 was still hidden, until connected).

Now, on the other hand, I manage a desktop machine as an Ubuntu server (also 14.04), and on it I've always seen eth0 show up in ifconfig - even if I have pulled the network cable from the desktop's RJ45 connector?!

So my questions are:

• Why does Ubuntu sometimes "hide" eth0 (or other network interfaces) until they are connected - and why is it not consistent (i.e. on desktop it is always shown, on laptop it is not shown until connected)?
• How can I list all network hardware, on an Ubuntu that "hides" network interfaces like this - as in that case, as shown above, not even sudo lshw works?

In Dia, I can add a flowchart box, then add text inside it, and then format the text using properties:

I can align the text horizontally (i.e. left, center, right) inside the box - but i don't see an option to align the text vertically (i.e. top, center, bottom) inside the box. This is on Dia 0.97.2.

Is there a way to align text vertically in a flow/text box in Dia?

EDIT: It can somewhat be achieved by adding carriage return/line feed/Enters either before or after the text (and also controlling the text padding) - but I was wandering if there was a built in function...

Firefox on Ubuntu 14.04 just updated to v 47, and it broke some of the stuff I was working on. That is however no biggie, I thought, because I can just download the previous version, and work with it, I thought. So, I do something like this:

wget https://ftp.mozilla.org/pub/firefox/releases/46.0.1/linux-i686/en-US/firefox-46.0.1.tar.bz2
tar xjvf firefox-46.0.1.tar.bz2
mv firefox firefox-46.0.1
./firefox-46.0.1/firefox # to run

So far so good, but unfortunately, as soon as I click Alt / Help / About Firefox, the damn thing starts updating:

... and it doesn't even ask me, which I find extremely disrespectful.

The problem here is that at startup, even this version uses the "Default User" profile that my normal system Firefox uses. Now I could set up a separate profile folder just for this Firefox, but then I have to manage more things - which I don't find worth the effort, since all I want to do anyways, is just to stop the updates.

So, I looked up a bit:

• Stop automatic updates. | Firefox Support Forum | Mozilla Support (2014)

You can set these prefs in about:config to disable automatic updating:
app.update.auto - false
app.update.enabled - false
app.update.silent - false
....
You can create a new file in a text editor ... called user.js ... Place this file in your Profile Folder.
// turn off application updates:
user_pref("app.update.enabled", false);

This would be OK, if I wanted to use a separate Profile directory, but as I said, I don't (and I don't want to mess with my "Default User" profile either). Then I found this:

• http://www.linuxforums.org/forum/applications/145630-firefox-where-does-user-js-go.html (2009)

<firefox_installation_dir>/defaults/profile

Ok, so this is something outside of a Profile folder, and in the application directory, this I could use, but does it still work? Currently I have this:

$ find firefox-46.0.1/ -name '*.js'
firefox-46.0.1/defaults/pref/channel-prefs.js
$ cat firefox-46.0.1/defaults/pref/channel-prefs.js
/* This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */

pref("app.update.channel", "release");

Well, I thought, if this file is read, then I can just add the preferences to stop updates here?

$ cat >> firefox-46.0.1/defaults/pref/channel-prefs.js << 'EOF'
> pref("app.update.auto", false);
> pref("app.update.enabled", false);
> pref("app.update.silent", false);
> EOF

Re-run Firefox, as soon as I see Help/About, it starts updating again :(

Tried also adding these to a separate user.js file in the same directory:

$ cat >> firefox-46.0.1/defaults/pref/user.js << 'EOF'
> pref("app.update.auto", false);
> pref("app.update.enabled", false);
> pref("app.update.silent", false);
> EOF

... but still no dice - again Firefox starts updating immediately once I see Help/About; and even about:config indicates that these preferences have not changed from their defaults.

So, is there a way to stop the updates of this "extern" version of Firefox on "application" level (i.e. in its application folder), such that is still uses the "Default User" profile otherwise user by the main Firefox version?

I've seen Apt wants to remove ubuntu-mate-desktop, but this is different.

I use MATE desktop, and my Ubuntu is:

$ uname -a
Linux MYPC 4.2.0-36-generic #42~14.04.1-Ubuntu SMP Fri May 13 17:27:22 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
$ cat /etc/issue
Ubuntu 14.04.4 LTS \n \l
$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description:    Ubuntu 14.04.4 LTS
Release:    14.04
Codename:   trusty

Just today, I was thinking of doing an upgrade:

$ sudo apt-get update
...
$ sudo apt-get upgrade 
Reading package lists... Done
Building dependency tree       
Reading state information... Done
Calculating upgrade... Done
The following packages have been kept back:
  compiz compiz-core compiz-gnome compiz-plugins compiz-plugins-default
  libcompizconfig0 libdecoration0 overlay-scrollbar-gtk2:i386
The following packages will be upgraded:
  apport apport-gtk libc-bin libc-dev-bin libc6 libc6:i386 libc6-dbg libc6-dev
  libc6-i386 libpam-smbpass libsmbclient libwbclient0 multiarch-support
  python-samba python3-apport python3-problem-report samba samba-common
  samba-common-bin samba-dsdb-modules samba-libs samba-vfs-modules smbclient
23 upgraded, 0 newly installed, 0 to remove and 8 not upgraded.
Need to get 26.4 MB of archives.
After this operation, 33.8 kB of additional disk space will be used.
Do you want to continue? [Y/n] ^C

Oh, I thought, some packages are held back, why not update them with dist-upgrade?

$ sudo apt-get dist-upgrade 
Reading package lists... Done
Building dependency tree       
Reading state information... Done
Calculating upgrade... Done
The following packages will be REMOVED:
  compiz-mate ubuntu-mate-desktop
The following packages will be upgraded:
  apport apport-gtk compiz compiz-core compiz-gnome compiz-plugins
  compiz-plugins-default libc-bin libc-dev-bin libc6 libc6:i386 libc6-dbg
  libc6-dev libc6-i386 libcompizconfig0 libdecoration0 libpam-smbpass
  libsmbclient libwbclient0 multiarch-support python-samba python3-apport
  python3-problem-report samba samba-common samba-common-bin
  samba-dsdb-modules samba-libs samba-vfs-modules smbclient
30 upgraded, 0 newly installed, 2 to remove and 0 not upgraded.
Need to get 29.7 MB of archives.
After this operation, 3,128 kB disk space will be freed.
Do you want to continue? [Y/n] ^C

B... but wait, I do not want to remove ubuntu-mate-desktop, I am using it! So what is going on, why does the update want to remove ubuntu-mate-desktop? How should I do the update, so I keep ubuntu-mate-desktop?

EDIT: I let the dist-upgrade go so ubuntu-mate-desktop is removed; thankfully it is just a metapackage, so I still have MATE running after reboot, but now I get:

$ sudo apt-get install ubuntu-mate-desktop 
Reading package lists... Done
Building dependency tree       
Reading state information... Done
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
 ubuntu-mate-desktop : Depends: compiz-mate but it is not going to be installed
E: Unable to correct problems, you have held broken packages.

Why, oh why, ... can't compiz-mate and compiz-gnome just coexist? Why is it necessary to break my system? ...

I managed to install TWRP Recovery image twrp-3.0.2-0-vegetahd.img on an Aquaris E5 phone with stock Android; thereby having a rooted device.

Out of the box, the stock Android allowed USB connection and MTP mounting of the device without a problem. However, after rooting, I simply cannot get MTP mount to work again - when I plug in the device on a Ubuntu 14.04, I get:

Unable to mount Aquaris
No MTP devices found

.... and /var/log/syslog says:

kernel: [ 4319.011008] usb 1-1.3: new high-speed USB device number 12 using xhci_hcd
kernel: [ 4319.100242] usb 1-1.3: New USB device found, idVendor=2a47, idProduct=201d
kernel: [ 4319.100254] usb 1-1.3: New USB device strings: Mfr=2, Product=3, SerialNumber=4
kernel: [ 4319.100261] usb 1-1.3: Product: Aquaris
kernel: [ 4319.100265] usb 1-1.3: Manufacturer: BQ
kernel: [ 4319.100269] usb 1-1.3: SerialNumber: UA0XXXXX
mtp-probe: checking bus 1, device 12: "/sys/devices/pci0000:00/0000:00:14.0/usb1/1-1/1-1.3"
mtp-probe: bus: 1, device: 12 was not an MTP device

... and also:

$ adb kill-server
$ adb devices -l
List of devices attached 

Btw, this device is:

$ lsusb
Bus 001 Device 012: ID 2a47:201d  

... and I have:

$ cat ~/.android/adb_usb.ini 
# ANDROID 3RD PARTY USB VENDOR ID LIST -- DO NOT EDIT.
# USE 'android update adb' TO GENERATE.
# 1 USB VENDOR ID PER LINE.
0x2a47

$ cat /etc/udev/rules.d/99-adb.rules
SUBSYSTEM=="usb", ATTR{idVendor}=="2a47", MODE="0666", GROUP="plugdev"
ATTR{idVendor}=="2a47", ATTR{idProduct}=="201d", SYMLINK+="libmtp-%k", MODE="660", GROUP="audio", ENV{ID_MTP_DEVICE}="1", ENV{ID_MEDIA_PLAYER}="1", TAG+="uaccess"

Is there anything I can do to have the device connected again - most importantly, so I can control it via adb?

I have an Aquaris E5 Ubuntu edition, and also an Android phone. As an experiment, I would like to make both of these dual boot both Ubuntu Touch and Android, and I've already found this:

• https://wiki.ubuntu.com/Touch/DualBootInstallation
• http://www.mibqyyo.com/en-articles/2015/09/16/ubuntu-android-installation-process-for-bq-aquaris-e4-5-and-e5

However, I'd like to have the second respective OS installed on an added microSD card on the phone; so:

• The Ubuntu phone should have Android on sdcard
• The Android phone should have Ubuntu touch on sdcard

I'm afraid, however, I'm not that familiar with partition layouts on Android, or in general, smartphones - I'm just trying to think in terms of partitions as used on a PC.

So my question is: is this kind of dual boot even possible - and if so, what sort of precautions do I need to take, or in other words, what should I be aware of in terms of partitioning before undertaking something like this?

For the Terminal app on Ubuntu Touch:

• There is a button to show keyboard, but it gets covered when the keyboard is shown. What is the proper way to hide the keyboard? So far, I've discovered that if you slide the app a bit so the list of background apps is shown, the keyboard gets autohidden, but this looks too hackish to me - there must be a better way?
• How can I press Ctrl-Shift-V to paste in Terminal? There is no Ctrl or Shift on the virtual keyboard, there are shortcuts defined for Ctrl-C though.. Can I edit these shorcuts somehow, so I can add Ctrl-Shift-V?

I had ubuntu-sdk installed, /var/log/apt/history.log has to say this about it:

Commandline: apt --no-install-recommends install ubuntu-sdk
Install: python3-simplejson:i386 (3.3.1-1ubuntu6, automatic), click-doc:i386 (0.4.43+16.04.20160203-0ubuntu3~0trusty1, automatic), libsybdb5:i386 (0.91-5, automatic), libjs-underscore:i386 (1.4.4-2ubuntu1, automatic), libxcb-xinerama0:i386 (1.10-2ubuntu1, automatic), libboost-program-options1.54.0:i386 (1.54.0-4ubuntu3.1, automatic), schroot-common:i386 (1.6.8-1ubuntu1.1, automatic), freetds-common:i386 (0.91-5, automatic), schroot:i386 (1.6.8-1ubuntu1.1, automatic), xserver-xephyr:i386 (1.15.1-0ubuntu2.7, automatic), ubuntu-sdk:i386 (1.126.2~0ubuntu1~trusty2), intltool:i386 (0.50.2-2, automatic), devscripts:i386 (2.14.1ubuntu0.1, automatic), dh-make:i386 (0.63, automatic), libpq5:i386 (9.3.12-0ubuntu0.14.04, automatic), mysql-common:i386 (5.5.49-0ubuntu0.14.04.1, automatic), gdb-multiarch:i386 (7.7.1-0ubuntu5~14.04.2, automatic), libmysqlclient18:i386 (5.5.49-0ubuntu0.14.04.1, automatic), click-reviewers-tools:i386 (0.35-0~550~ubuntu14.04.1, automatic), fcitx-libs:i386 (4.2.8.3-3, automatic), python3-magic:i386 (5.14-2ubuntu3.3, automatic), python3-yaml:i386 (3.10-4ubuntu0.1, automatic), libbotan-1.10-0:i386 (1.10.5-1ubuntu1, automatic), libjs-sphinxdoc:i386 (1.2.2+dfsg-1ubuntu1.1, automatic), click-dev:i386 (0.4.43+16.04.20160203-0ubuntu3~0trusty1, automatic), libxcb-xf86dri0:i386 (1.10-2ubuntu1, automatic), ubuntu-sdk-ide:i386 (3.5.1~133+201604191700~ubuntu14.04.1)

So, clearly, there were dependencies installed. So I first remove the ubuntu-sdk package:

$ sudo apt-get remove --purge ubuntu-sdk
Reading package lists... Done
Building dependency tree       
Reading state information... Done
The following packages will be REMOVED:
  ubuntu-sdk*
0 upgraded, 0 newly installed, 1 to remove and 0 not upgraded.
After this operation, 32.8 kB disk space will be freed.
Do you want to continue? [Y/n] y
(Reading database ... 303883 files and directories currently installed.)
Removing ubuntu-sdk (1.126.2~0ubuntu1~trusty2) ...

Ok, now I try to remove the dependencies using autoremove:

$ sudo apt-get autoremove --purge
Reading package lists... Done
Building dependency tree       
Reading state information... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.

Hmm.. what? Apt-get autoremove should by here have recognized the dependent packages for removal, but it doesn't?

Why does this happen, and how can I get autoremove to remove dependent packages of ubuntu-sdk once its removed?

I have already seen some answers related to this, but they didn't bring me to a definitive answer, so I'm going to try again.

Basically, my question is: how can I run Ubuntu Touch applications from the command line - specifically, so that I could debug them with gdb?

More details:

First, what that means is that I'd like to run them from a remote shell connection to a (phone/tablet) device on a PC, either from adb shell or phablet-shell (see Hacking Ubuntu Touch, Part 5: adb shell vs. phablet-shell). I already got the Terminal app, but - given that one cannot establish a proper (VNC) remote desktop connection to an Ubuntu Touch system, - it is of little use to me, as I cannot really work with it (haven't yet tried a Bluetooth keyboard attached to a device, though).

Second, I'd like, if possible, to run applications in the foreground; then when called from the terminal command line, I'd expect the applications to dump their stdout printouts to the terminal. Thus, the recommendation in Ubuntu Touch: How do I run an ubuntu touch app from the terminal? I cannot use much, since (this is from a BQ E5 (Ubuntu 15.04, OTA-10.1)):

phablet@ubuntu-phablet:~$ ubuntu-app-list
ubuntu-system-settings
com.ubuntu.calculator_calculator_2.0.233
phablet@ubuntu-phablet:~$ ubuntu-app-launch com.ubuntu.calculator_calculator_2.0.233
phablet@ubuntu-phablet:~$ 

... it simply launches the app in the background, and immediately exits. And AFAICS, webapp-container simply opens an URL in the browser, doesn't run arbitrary applications.

Some of the preinstalled applications are apparently run by a program qmlscene, as How do I simply run an ubuntu touch app from the terminal? notes; but even the accepted answer there isn't applicable to Ubuntu 15.04 Touch as-is:

$ /usr/bin/qmlscene \
> --desktop_file_hint=/usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/share/applications/ubuntu-clock-app.desktop \
> /usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/share/qml/ubuntu-clock-app.qml
Loading module: 'libubuntu_application_api_touch_mirclient.so.3.0.0'
file:///usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/share/qml/ubuntu-clock-app.qml:22 module "Alarm" is not installed
file:///usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/share/qml/ubuntu-clock-app.qml:20 module "WorldClock" is not installed

... and the application shuts down on device barely after being started. And below is the only command line I've found to work on a BQ E5 Ubuntu 15.04:

APP_ID=com.ubuntu.clock_clock_3.7.456 \
QML2_IMPORT_PATH=/usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/lib/arm-linux-gnueabihf \
QML_IMPORT_TRACE=1 \
/usr/bin/qmlscene \
--desktop_file_hint=/usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/share/applications/ubuntu-clock-app.desktop \
/usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/share/qml/ubuntu-clock-app.qml

The application then starts on phone, while on shell terminal messages are being dumped, like:

Loading module: 'libubuntu_application_api_touch_mirclient.so.3.0.0'
QQmlImportDatabase::addImportPath: "/usr/lib/arm-linux-gnueabihf/qt5/qml"
QQmlImportDatabase::addImportPath: "/usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/lib/arm-linux-gnueabihf"
QQmlImportDatabase::addImportPath: "/usr/lib/arm-linux-gnueabihf/qt5/bin"
...

... and after you press Ctrl-C in terminal, the application exits.

Why I like this, is for one, I know what is the application printout and how to redirect/save it - and for another, I know how to build a gdb command line in order to debug. I managed to set the device in writable mode, and from the shell install build-essential (see Recommended way to install regular(cli) .deb packages on Ubuntu Phone?), which pulls in gdb, so now it is available in shell also in read-only mode. Then I can do:

APP_ID=com.ubuntu.clock_clock_3.7.456 \
QML2_IMPORT_PATH=/usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/lib/arm-linux-gnueabihf \
QML_IMPORT_TRACE=1 \
gdb --args \
/usr/bin/qmlscene \
--desktop_file_hint=/usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/share/applications/ubuntu-clock-app.desktop \
/usr/share/click/preinstalled/com.ubuntu.clock/3.7.456/share/qml/ubuntu-clock-app.qml

... which will properly open gdb:

GNU gdb (Ubuntu 7.9-1ubuntu1) 7.9
...
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/bin/qmlscene...(no debugging symbols found)...done.

(but won't be of that much use in this example due to missing debugging symbols).

However, besides these apps ran by qmlscene, there are seemingly also "proper" executables on Ubuntu Touch 15.03, such as the Browser app:

phablet@ubuntu-phablet:~$ which webbrowser-app
/usr/bin/webbrowser-app
phablet@ubuntu-phablet:~$ file /usr/bin/webbrowser-app
/usr/bin/webbrowser-app: ELF 32-bit LSB shared object, ARM, EABI5 version 1 (GNU/Linux), dynamically linked (uses shared libs), for GNU/Linux 2.6.32, BuildID[sha1]=c173774045b98a97e831eda92dd5008d08b9a251, stripped

... but this kind I simply cannot get to run from the command line. Closest I got to (by analogy with above, and copying some environment variables that env ran on the Terminal app on the phone reported) is this:

APP_ID=webbrowser-app \
UNITY_MIR_SOCKET=/run/mir_socket \
MIR_SOCKET=/var/run/mir_socket \
MIR_SERVER_PROMPT_FILE=1 \
MIR_SERVER_NAME=session-0 \
/usr/bin/webbrowser-app \
--desktop_file_hint=/usr/share/applications/webbrowser-app.desktop

... which however fails immediately in terminal with:

Loading module: 'libubuntu_application_api_touch_mirclient.so.3.0.0'
UbuntuClientIntegration: connection to Mir server failed. Check that a Mir server is
running, and the correct socket is being used and is accessible. The shell may have
rejected the incoming connection, so check its log file
Aborted

What's more, all this business with --desktop_file_hint seems something required by Unity8 (i.e. Unity8 might inspect the app command line and look for it, even if the app itself doesn't use it) - but may be (or is) going away:

ubuntu-bugs - [Bug 1578319] [NEW] Some ubuntu-ui-toolkit AP tests failing to launch - rejected by unity8 as is using desktop_file_hint (2016-05)

...is not using upstart-app-launch - which would explain the fail. desktop_file_hint is not a supported means of launching an app (and can go away at any time), so UITK should change to use UAL in its AP tests.

This indeed used to work until recently, but a recent refactoring broke the desktop_file_hint behaviour. Since it's not a supported behaviour, it should be relied on, so it would be better to fix UITK than restore the old behaviour.

Ok, but then, how do I run executable files, possibly from non installed applications, with upstart-app-launch (i.e. ubuntu-app-launch) - and how would I run them in foreground in the terminal? I'm aware that, - just like Android has ndk-gdb, - the Ubuntu SDK has something similar via Qt Creator which apparently uses gdbserver on the device. However, I'd prefer not firing up a GUI for something like that - and in fact, would love to run gdb in the old-fashioned way, as described above, with an executable as argument on the command line.

I guess a lot of this has to do with me not quite understanding init/Upstart, X11/Mir, Gnome/Unity etc differences that are applicable on a device vs. a (classic) desktop - and I cannot really tell what runs what. I'd love to have enough of it explained, so I can understand how running (GUI) Ubuntu Touch apps from a terminal command line would work.

On Android devices, there is an option Settings > Developer Options > Stay awake when charging which keeps the device ON (prevents sleep) even if the device is not touched, when the device is connected via USB and is charging.

Is there something similar for Ubuntu Touch OS devices?

I'm trying to compile a program on a Vivid (15.04) Ubuntu Touch device; and the only problem I have is:

fatal error: mir/raii.h: No such file or directory

The code does:

#include "mir/raii.h"

... and apparently, that file should have been in the libmircommon-dev, which I both have:

$ apt-show-versions -r mircommon
libmircommon-dev:armhf/vivid 0.20.3+15.04.20160322-0ubuntu1 uptodate
libmircommon5:armhf/vivid 0.20.3+15.04.20160322-0ubuntu1 uptodate

It's definitely not on the system, because I'm doing:

sudo find / -xdev -name raii.h

... and it returns nothing.

Now, in Ubuntu – Package Contents Search (trusty) -- raii.h, there is a /usr/include/mircommon/mir/raii.h in mircommon-dev.

However, already in the next version, Ubuntu – Package Contents Search (wily) -- raii.h, it is not anymore in mircommon-dev - there's just a file /usr/include/wibble/raii.h in libwibble-dev, which I think is unrelated (and same for all the next versions)

However, I can also see code like ~kdub/mir/key-repeat-gmock-fixes : files for revision 3419, with commits in 2016, which still uses #include "mir/raii.h"?

So my question is:

• Is mir/raii.h fully removed from Mir, or does it exist in a different package? If so which package should I install?
• If it has been removed completely, what header could I use instead to have the program compile?

EDIT: just looked at the apparently latest mir/development-branch/tarball r.3489 (via http://bazaar.launchpad.net/~mir-team/mir/development-branch/files), and there is still a

./mir/development-branch/src/include/common/mir/raii.h

there... the question is then, why isn't it in any of the Ubuntu packages?

I was wondering, is there a tool that I could run on Ubuntu Desktop, and be connected to an Ubuntu Touch device through a remote desktop (or remote display) application, such that mouse clicks are also propagated as taps on the device (and also, copy/paste of text)?

Basically, if Ubuntu Touch used X11, then I could have used vinaigre or xvnc4viewer on Ubuntu Desktop as VNC clients, and Vino as VNC server on the device - however, Ubuntu Touch uses Mir, and I cannot really find any information about remote desktop/display via Mir.

So, is there anything like this out there?

How can I edit the print margins in Firefox in Ubuntu 14.04? I would like to print a page in A3, with custom page margins, to a PDF...

There are suggestions out there like "setting print margins is done in the printer settings" - but, I absolutely cannot find anything related to print page margins... Here's what I'm trying to print:

In the Print dialog/General, there's nothing related to margins:

In the Print dialog/Page Setup, there's nothing related to margins:

EDIT: As per comment, the contents of "Paper size" ("US Letter"):

In the Print dialog/Options, there's nothing related to margins:

I even installed the Print Edit extension, also that has nothing obvious related to page margins:

So, how can I control page margins in Firefox in Ubuntu?

EDIT: I found Where is the Print Setup setting in Firefox so top and bottom margins print on the page (now too high and too low)? | Firefox Support Forum | Mozilla Support (2015):

This is hidden and a little obscure... Firefox stores properties for each of your printers, including the "unwriteable" margin distance and the "edge" distance. ...
(1) In a new tab, type or paste about:config in the address bar and press Enter. ...
(3) Double-click the print.printer_BrandX.print_edge_top preference and enter your desired distance in hundreds of an inch. ...

So to update the question - is there a way to set print page margins in Firefox from a GUI (without changing about:config settings directly)?

I have just gotten my hands on an E5 Ubuntu Touch device, however, I don't want to register anywhere (including the Ubuntu app store) until I get the hang of it.

Now, I'm aware that on Android, the app store packages have an extension .apk, and they can be installed either through the Google Play store with a proper login, or through a USB connection using adb install *.apk.

I've learned that on Ubuntu, the format defined for Ubuntu Touch is called "Click," using the .click file extension, and it can also be installed either through the Ubuntu app store, or through a USB connection using pkcon install-local --allow-untrusted *.click.

However, I'm also aware that on Android, you can start a download of an .apk file, and then install it by doubleclicking (or its equivalent in tapping) once its downloaded; can Ubuntu Touch do the same?

If it can, is there an example online of a very simple .click application, that I could download and install the same way - possibly by allowing some developer options first?

I've noticed via Touch/Collection - Ubuntu Wiki, say, the xkcd-viewer app - but all it offers for download are .deb files, which apparently don't install on Ubuntu Touch, at least not through the GUI.

I am using Ubuntu 14.04 with MATE, and my problem is that when I use a local Ethernet cable to access local network via a router (which is not connected to Internet), and at the same time Wireless (or Mobile Broadband) to access the internet - I cannot access the Internet, because apparently Ubuntu tries to first do Internet communication via the wired Ethernet cable, and it locks in there (as the wired Ethernet is, as I mentioned, not connected to Internet).

So, I've seen Set specific interface for internet access , where the answer is:

These routes are automatically created by NetworkManager. Try checking Edit ▸ IPv4 Settings ▸ Routes ▸ Use this connection only for resources on its network on the two LAN connections to prevent them from becoming the default gateway.

Ok, so I go to Network Connections / Edit, Edit the "Auto Ethernet connection", and enable "Use this connection only..." and Save. Then ifconfig reports to me (say) 192.168.100.10 for eth0, and so I test first pinging other local PCs:

$ ping 192.168.100.11
PING 192.168.100.11 (192.168.100.11) 56(84) bytes of data.
64 bytes from 192.168.100.11: icmp_seq=1 ttl=64 time=0.554 ms
64 bytes from 192.168.100.11: icmp_seq=2 ttl=64 time=0.638 ms
...

Well, so far so good; so I enable my Broadband - and at this point, I have this route:

$ ip route
default via 10.64.64.64 dev ppp0  proto static 
10.64.64.64 dev ppp0  proto kernel  scope link  src 10.206.104.70 
192.168.100.0/24 dev eth0  proto kernel  scope link  src 192.168.100.10  metric 1 

... and let's see if pinging something online works:

$ ping google.com
PING google.com (168.95.1.1) 56(84) bytes of data.
64 bytes from 168.95.1.1: icmp_seq=1 ttl=224 time=1155 ms
64 bytes from 168.95.1.1: icmp_seq=2 ttl=224 time=436 ms

Oh nice, it works... or does it? Let's try retrieving something online:

$ wget http://google.com -O -
--2016-04-27 10:50:09--  http://google.com/
Resolving google.com (google.com)... 168.95.1.1
Connecting to google.com (google.com)|168.95.1.1|:80... ^C

Basically it hangs there for more than a minute, and I have to hit Ctrl-C to break/interrupt it.

And now, from Network Manager, I disconnect "Auto Ethernet", and when I try downloading the same resource again:

$ wget http://google.com -O -
--2016-04-27 10:51:12--  http://google.com/
Resolving google.com (google.com)... 95.209.200.29, 95.209.200.45, 95.209.200.38, ...
Connecting to google.com (google.com)|95.209.200.29|:80... connected.
HTTP request sent, awaiting response... 302 Found
Location: http://www.google.dk/?gws_rd=cr&ei=gX0gV4rVAYe6sQHrzbnICg [following]
--2016-04-27 10:51:13--  http://www.google.dk/?gws_rd=cr&ei=gX0gV4rVAYe6sQHrzbnICg
Resolving www.google.dk (www.google.dk)... 64.233.165.94, 2a00:1450:4010:c08::5e
Connecting to www.google.dk (www.google.dk)|64.233.165.94|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: unspecified [text/html]
Saving to: ‘STDOUT’

    [<=>                                    ] 0           --.-K/s              <!doctype html><html itemscope="" itemtype="http://schema.org/WebPage" ...

Well... now it works - but of course, NOT with the both connections active, but only with the Ethernet disconnected.

So let me repeat the question: how can I have two network connections active at the same time in Network Manager, such that the wired Ethernet connection is used to communicate only to the local network of the router it is connected to - and the wireless/broadband connection is used to access the Internet?