Elias's questions

On my laptop with Ubuntu 20.04.1 LTS there is sometimes a delay during shutdown, with the message "A stop job is running for Thermal Daemon Service". It looks like this:

Then it just hangs like that until the time 1min 30s has passed, then it finally shuts down. This happens only sometimes, most shutdowns are quick without this happening, but sometimes it comes, seems random.

Why is this happening? How could I get rid of it, or at least get more info about what happens? Does it indicate a bug somewhere?

When some program crashes and an automatically generated error report is sent as discussed here What happens when I click "Send" after Ubuntu experienced an internal error? then the error report is added to the info shown at https://errors.ubuntu.com/ where the most frequent errors are shown at the top of the list.

The rightmost column in the list at https://errors.ubuntu.com/ is "Bug report" and for most of the common errors there is a bug report number there, and by clicking on that we can see the status and discussion regarding that bug report, which is great. But for some of the most common errors, the "Bug report" column is empty.

To me it seems like there should be a corresponding bug report for each error, especially for those errors that have hundreds or thousands of occurrences per day. Should I create a bug report in case the "Bug report" column is empty?

If I do that, how can I make sure my bug report is properly linked to the error reports on the https://errors.ubuntu.com/ page?

Edit: as an example, consider the error "seahorse (seahorse) glib-watch.c → 195 → timeout_update → Assertion `!t->dead'' failed." that is currently listed at https://errors.ubuntu.com/ with Rank 9, with 225 occurrences during the past day. This particular error happens on my computer every few days, I usually choose to send an error report so I expect that my reports are then seen as additional occurrences of the error in that list. But there is no linked bug report, which makes me worry that developers of the seahorse package are still not aware of it. The same error has been occurring for more than a year, which also makes me think the developers are unaware of it, otherwise they would have done something about it by now. The question is then, what should be done to make sure there is a bug report what the developers become aware of? Note however that my question is mostly about how this is meant to work in general, this particular seahorse error is just an example, I'm interested in that case but even more interested in understanding how this works in general.

In the book "TCP/IP Illustrated, Volume 1" by Kevin Fall and Richard Stevens, many examples are using a utility program called "sock" that can do various TCP/IP related things, for example it can act as a UDP server or it can send UDP datagrams to such a server. I would like to test some basic things in the same way they describe in the book, for example to start a UDP server on one computer and send a UDP datagram to it from another computer.

To give an idea of the kind of utility I'm looking for, according to the book the sock program can be used like this to start an IPv4 UDP server on port 7777:

sock -u -s 7777

and like this to send some UDP datagrams with a given destination IP address:

sock -v -u -i 10.0.0.3 discard

The source code for that "sock" program is available at https://www.icir.org/christian/sock.html which is great, but I am still wondering: are there some Ubuntu packages that I could install (or that are already available by default) that would provide similar simple utilities for UDP and TCP?

Edit: in addition to the nc (netcat) tool suggested in the answer below, I just learned that sending a UDP datagram can be done very easily using bash, like this:

echo "hello" > /dev/udp/example.com/8000

See for example https://unix.stackexchange.com/questions/191655/using-bashs-dev-udp-how-would-i-know-if-the-port-was-open

I recently learned how easy it is to get the source code for any given package using apt-get source so that I can get the source code, make changes and install my own modified version of any package. This is great!

Until today I was assuming that each package would have its own source code, and that different packages would have different source code.

However, now I just discovered that different packages can have identical source code. Here follows an example of that:

The following 4 packages seem to have identical source code:

gir1.2-mutter-4
libmutter-4-0
mutter
mutter-common

All four of them are installed on my Ubuntu 19.04 computer. Doing apt-get source gir1.2-mutter-4 gives exactly the same result as apt-get source libmutter-4-0, and also for the mutter and mutter-common packages.

Here is how I checked it:

mkdir a
cd a
apt-get source gir1.2-mutter-4
cd ..
mkdir b
cd b
apt-get source libmutter-4-0
cd ..
diff -r a b

The recursive diff on the last line above gives no output, showing that the directories have identical contents.

Now to my question: How can different packages have identical source code?

Assuming that this is intended and not some kind of error, what is the difference between the packages and how can I see that difference?

Could it be that the packages are different in the way the source code is configured and compiled, e.g. different parts of the code are included in the different packages? If so, where can I find information about how to configure each package?

Edit: forgot to add that if you want to test this, to make apt-get source work properly you may first need to enable it using software-properties-gtk as described here: https://askubuntu.com/a/857433/874649

Edit 2: thanks for the excellent answers! I also found this helpful https://askubuntu.com/a/246721/874649 -- about the apt-get build-dep and dpkg-buildpackage commands that are very useful. After modifying source code for a source package, dpkg-buildpackage -us -uc can be used to build new .deb file(s) that can be used to install the modified program(s).

I have two laptops, both with Ubuntu 19.04 installed. One of them is a couple of years older and has had the Ubuntu version upgraded a few times. The other one is newer and got 19.04 installed from the beginning, not upgraded from any previous Ubuntu version.

On my older laptop I sometimes get error messages looking like this:

which is nice because it allows me to click the "Show details" button to see lots of details about what happened.

On the newer laptop there are sometimes messages looking like this:

Unfortunately, clicking the "Report problem..." button above just makes the window disappear, no further details are shown. The "..." on the button makes it look like the intention was that something more should be shown when clicking that button, but nothing comes up, it just closes the window.

I suspect that it may be the same kind of problem but for some reason handled differently by Ubuntu despite the fact that both are running Ubuntu 19.04.

Do these messages mean the same thing?

If not, what is the difference between "application ... closed unexpectedly" and "system program problem detected" and how can I get more details in the latter case? (the "system program problem detected" leaves no file in /var/crash/ unfortunately)

One theory I have about this is that the older laptop, having originally run older Ubuntu versions and then been upgraded, has some settings in a config file somewhere left from a previous Ubuntu version telling it to use the older, more detailed (and in my opinion much better) way of showing errors. Could this theory be correct? If so, which config file would I need to modify in order to get the behavior I want, allowing me to show details about the error?

Edit to explain why this is not a duplicate of how to read and use crash reports? -- that issue is about how to make use of files located in /var/crash/ but in my case, when the "system program problem detected" window pops up, there is no file generated in /var/crash/. So, that question and the answers there unfirtunately do not help much for my question.

I think it would be really useful (and cool) to be able to get all the source code for my current Ubuntu installation. I mean everything, every last package from the kernel all the way up to user interface details.

(One situation where this would be useful is when there is some kind of error message that I don't understand what it means and don't know exactly where it came from, like the infamous "system program problem detected" popup. I don't know which package is responsible for presenting that. If I had all the source code I could simply search for the error message string to find the relevant part of the code. That could help me understand what happened and possibly modify the code slightly to add more details in the error message, or whatever, and rebuild that package.)

Assuming that I installed everything using apt, I can get the source code for any individual package x using apt-get source x so I guess it should be possible to write a script that first gets a list of the names of all currently installed packages, and then use a loop to get the source code using apt-get source for each package one by one. My question is: is there a more convenient way of getting all source code for all software I currently have installed, perhaps using a single command?

As I understand it, Ubuntu uses lots of free (FOSS/FLOSS) software but often also some closed-source proprietary software, which is why Ubuntu is not included in the FSF list of Free GNU/Linux distributions.

While I'm not a fanatic about this, I do want to support the free software movement and use as much free software as possible. So I wonder: is there some way to get a list or overview of all non-free software and firmware that is used/installed on my Ubuntu system?

When running something from the command-line, the which command can be used to find out which executable is actually called. That is great. I would like to be able to find out that information also for programs that are started using the graphical launcher.

For example, if I press the super key (the one with a windows logo on it) and start typing "sound", the program "Sound Recorder" pops up and I can click on it to start it. Then it starts, fine, but how do I know which executable it was that I just started? The name is not necessarily the same as the name "Sound Recorder" that I clicked on.

For this example I can figure it out doing some guesswork, like this:

ps -A | grep sound
 1848 tty2     00:00:00 gsd-sound
13645 tty2     00:00:00 gnome-sound-rec

So it is probably gnome-sound-recorder and then I can do

which gnome-sound-recorder 

to find out it is actually /usr/bin/gnome-sound-recorder.

But that was assuming the executable same was something with "sound", what if that was not the case? Is there a better way to find out which executable is launched?

SConnect is a browser extension (made by a company called Gemalto) that makes it possible for web pages to use security devices (e.g. a card reader) connected to the computer.

According to this page https://aginsurance.mobileconnect.gemalto.com/sconnect/faq/#compatibility SConnect is supposed to work for Firefox in Linux. But for me, it is not working.

The installation is supposed to be fairly automatic, when you go to a web page that uses SConnect it detects that SConnect is not installed and asks if you want to install it. I follow the installation steps but when trying the web page again it still acts as if SConnect was not installed. Firefox shows the SConnect extension as installed but "inactive".

In my case I need SConnect in order to use an electronic identification system that the Swedish tax agency is using (you insert smartcard in card reader to login), but I think SConnect is used in other contexts as well, also in other countries. SConnect works in Windows 10 but I need it to work in Linux, preferably in Ubuntu.

Steps to reproduce the problem: in Firefox, go to https://skatteverket.se/ and click the "Logga in" button in the top-right corner, then click "AB Svenska Pass". At that point, the SConnect extension is supposed to be activated if it is installed, or you are asked if you want to install it. That works properly in Windows 10, but in Ubuntu it always asks to install the extension again, even if it was already installed.

This is in Ubuntu 19.04 using Firefox 67.0.4 and SConnect version 2.10.0.0.

I have tried using the Chrome browser instead (Chrome should also work according to the compatibility info link above) but that behaves in the same way.

I suppose that ideally this problem should be reported to the company (Gemalto) and they should fix it, I have tried but so far I have not been able to get any meaningful answer from them. So I decided to try asking a question here hoping that someone else has faced this problem and perhaps solved it.

Can the SConnect Firefox extension work in Ubuntu, and if so, are some special tricks needed to make it work?

I am using Ubuntu 19.04. When I need to install some new package I usually do it by running sudo apt install from the command line. To keep things updated I use "Software Updater" (update-manager), sometimes it pops up to remind me and sometimes I run it myself to make sure everything is up to date.

However, recently I have installed some packages using snap install instead of apt, and now I am not sure if I should run something more than "Software Manager" to make sure everything is updated.

Does "Software Updater" only update software installed using apt, or also software installed using snap?

Do I need to run something else in addition to "Software Updater", and if so, what?

Also: is it a bad idea to mix the "apt" and "snap" ways of installing things?

I am using Ubuntu 19.04 (Disco Dingo). My question is the same as this old one: How to restart X Window Server from command line?

However, the much upvoted answer there does not work for me. When I run sudo systemctl restart display-manager as suggested there, the window system is restarted but then after that I cannot login, it seems like the window system is restarted again at each login attempt. So then I anyway need to reboot to make things work again. So that does not achieve what I want. What I would like is to be able to restart the window system without doing a full system reboot.

The other suggestion in the old answer was to check which display manager is used by doing cat /etc/X11/default-display-manager (in my case this gives /usr/sbin/gdm3) and then run sudo restart gdm but that gives me an error message: sudo: restart: command not found.

Is there a way to restart the X window system in Ubuntu 19.04 without doing a full reboot, and be able to login again afterwards?

Edit: I also tried systemctl restart gdm but that gives the same problem as sudo systemctl restart display-manager -- the window system is restarted but after that I cannot login, need to reboot to make things work again.

On my computer I have Ubuntu 19.04 intalled along with Windows 10 and Fedora 29.

In Ubuntu, running sudo update-grub gives the following:

Sourcing file `/etc/default/grub'
Sourcing file `/etc/default/grub.d/init-select.cfg'
Generating grub configuration file ...
Found linux image: /boot/vmlinuz-5.0.0-13-generic
Found initrd image: /boot/initrd.img-5.0.0-13-generic
Found linux image: /boot/vmlinuz-4.18.0-18-generic
Found initrd image: /boot/initrd.img-4.18.0-18-generic
Found Windows Boot Manager on /dev/nvme0n1p1@/EFI/Microsoft/Boot/bootmgfw.efi
Found Fedora 29 (Workstation Edition) on /dev/mapper/fedora-root
Adding boot menu entry for EFI firmware configuration
done

which I think looks fine, it detects both Windows Boot Manager and Fedora 29 and it sets up a grub menu letting me choose between the different operating systems at boot time.

My problem is that Fedora needs some special configuration in order to workaround some graphics driver problems. Specifically, the options rd.driver.blacklist=nouveau modprobe.blacklist=nouveau need to be added for that grub menu entry.

I am able to fix this by manually editing the file /boot/grub/grub.cfg but that change is overwritten next time update-grub runs, so it is not a good solution. Also, the file /boot/grub/grub.cfg that I am then editing starts with a comment saying "DO NOT EDIT THIS FILE", also indicating that I should perhaps not edit it.

My question is: how can I add some extra boot options for the Fedora grub entry that is generated when I run sudo update-grub in Ubuntu?

There are some options in the /etc/default/grub file that is used by update-grub, for example I can choose GRUB_TIMEOUT there which works fine, but as far as I know that file does not allow me to specify boot options for other operating systems, only for Ubuntu itself.

To further clarify what I mean, the auto-generated /boot/grub/grub.cfg file (created by update-grub) contains a section looking like this, specifying how Fedora is booted if that grub menu option is chosen:

menuentry 'Fedora 29 (Workstation Edition) (on /dev/mapper/fedora-root)' --class fedora --class gnu-linux --class gnu --class os $menuentry_id_option 'osprober-gnulinux-simple-6398dc3c-28e3-4947-ad82-d9defe5d9a28' {
    insmod part_gpt
    insmod ext2
    if [ x$feature_platform_search_hint = xy ]; then
      search --no-floppy --fs-uuid --set=root  b00e70c6-820f-4615-b95d-c933eaea0d3f
    else
      search --no-floppy --fs-uuid --set=root b00e70c6-820f-4615-b95d-c933eaea0d3f
    fi
    linux /vmlinuz-5.0.9-200.fc29.x86_64 root=/dev/dm-2
    initrd /initramfs-5.0.9-200.fc29.x86_64.img
}

and I would like the above to look slightly different, adding the options rd.driver.blacklist=nouveau modprobe.blacklist=nouveau in the end of the line starting with linux. Is there some way to tell update-grub to do that?

I did see the suggestion here about editing the /etc/grub.d/40_custom file but that solution means writing the whole menuentry manually so it will not adapt when the kernel version in Fedora changes. The auto-generated menuentry created by update-grub has the advantage that it gives the latest Fedora kernel, I would like to use that and just add something to the "linux" line inside the auto-generated menuentry. Can that be done, and if so, how?

I am using Ubuntu 18.10. Sometimes, when some program has just crashed, a window pops up saying the following: "Sorry, Ubuntu 18.10 has experienced an internal error. Send problem report to the developers?" Then there are three buttons: "Show details", "Don't send" and "Send".

If I click "Show details" then it does show details nicely, for example that "gjs-console crashed with SIGABRT" and lots of other info. It seems like a good idea to send this to the developers, so I click "Send". Then the window is closed. I do not get any confirmation that anything was actually sent. I would have liked to get some confirmation like "OK, problem report sent successfully" or something but there is no such confirmation so I am a little worried that something went wrong when I clicked the "Send" button.

My questions:

• Is anything actually sent to the developers?
• Is there any way for me to verify that something was actually sent?
• Can I even see the details of the resulting automatically generated problem report in a public bug tracking system somewhere?

EDIT: according to an answer here: How can I find my Bug-report on launchpad that I submitted from the Desktop? it should be possible to see the resulting bug report by going to https://bugs.launchpad.net/~/+reportedbugs but that does not work for me -- that page just says "There are currently no open bugs." So I am still not sure if any problem report was really sent and if so, where did it end up.

EDIT 2: this seems related also: How can I track a bug that caused a crash and was reported via apport / whoopsie? but after reading that I still don't know how to verify if a problem report was really sent.

EDIT 3: in /var/log/syslog there are a few lines mentioning "whoopsie" saying for example "Uploading /var/crash/_usr_bin_gjs-console.1000.crash." and "Sent; server replied with: No error" which sounds good. Also, the file /var/crash/_usr_bin_gjs-console.1000.crash does seem to contain all the collected error report info.

I have a Lenovo ideapad 720S-14IKB laptop with Ubuntu 18.04 and the built-in microphone does not seem to work.

In Windows 10 (also installed on the same laptop, dual boot) the microphone works fine.

The way I have tested this so far in Ubuntu is to use "Sound recorder" to do a recording, then the resulting recording is just silent no matter how loud I scream at it.

Since the microphone works in Windows I guess it should be possible to make it work also in Ubuntu -- what can I do to troubleshoot?

EDIT: Now, a few months later, the microphone has started working. I still don't know why it did not work earlier or what changed so that it started working. I did not knowingly change anything other than installing the normal updates using Software Updater. Anyway, the problem seems to have fixed itself.

EDIT 2: Now (8 Feb 2019) the microphone has stopped working again. It was working a week ago but now for some reason it is no longer working. As before, it works fine in Windows 10 but in Ubuntu it does not work. Any suggestions for how to troubleshoot would be much appreciated!