You are not logged in.
I am currently booted in chimaera, same Xfce 4.16 packages, and the "Default Applications" is shown under Menu > Settings > Default Applications.
I don't know if this is supposed to be an exclusive "announce" thread or not, but I do have a support request and I think this is the best place to put it. If it needs to be in a thread of its own, please move it.
I have a webscript that requres the package php-recode. When I try to install php-recode, I get this:
root@FC-srv:/home/nixer# apt install php-recode
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:
php-recode : Depends: php7.4-recode but it is not installable
E: Unable to correct problems, you have held broken packages.
root@FC-srv:/home/nixer# apt policy php7.4-recode
php7.4-recode:
Installed: (none)
Candidate: (none)
Version table:
root@FC-srv:/home/nixer#
The package php7.4 is missing. Can it be created or am I not finding it? I am using the tdrnetworks php builds.
The package php7.3 is present, but I am not running php version 7.3. I did a web search and I don't see any mention of "recode" being deprecated in version 7.4.
If I may, the command to download and add the keyring is slightly off on the linked page above. This worked:
wget http://hezeh.org/packages/antofox-keyrings && apt-key add antofox-keyrings
I have not seen or used gnome in over 5 years. I will give it a test. Thank you.
@nixer . . . I'd love to see that horse site. Care to share a link?
At the request of the site owner, it was taken down a few months ago after being online for about four years.
But here is the photo that I used to get the site colors from.
I don't even have a screenshot of the site before I took it offline, sorry. I even checked a few web archive sites to see if they had it, but can't find it, nor does google have it.
It's done folks and not going to change (much).
Why change, I like it! Really it is only a start, and will be changed very soon after installing - just like blackhole said.
Because colors are relative to the other colors that are proximate and only the human eye can see that.
So true! I once made a website for a horse, a very expensive horse. I had the hardest time choosing a color scheme to use, so I used the horse itself for my color selection. From a nice photo of the horse, I lifted a light and dark color for my color scheme. It worked quite well.
Don't be surprised if this thread gets deleted.
I for one appreciated the introduction to the new theme colors, and its inspiration. Thank you.
Nice to know and thanks again. My friends install was originally a jessie install from 2016, updated to ascii, and now to beowulf. His system is powered by an amd processor, hence, this is why I installed that firmware on it.
I use the refractasnapshot script on other amd and intel powered cpus with and without the amd64-firmware package, and I have never had this problem. I do not know why he has it, yet. Maybe something will show later on, but I don't know now.
Appreciate your help!
file -L /initrd.img
root@desktop:/home/user# file -L /initrd.img
/initrd.img: ASCII cpio archive (SVR4 with no CRC)
I removed the
amd64-microcode
package and the snapshot successfully created.
Removing the amd64-microcode did not cause any problem with the system as far as we can tell. It rebooted and displayed the xfce session fine.
Thanks for your help!
I was hoping it would be an easy fix, but not yet. xz-utils is already installed
root@desktop:/home/user# apt install xz-utils
Reading package lists... Done
Building dependency tree
Reading state information... Done
xz-utils is already the newest version (5.2.4-1).
After ~6 years of using refracta-tools, I finally ran into the first problem creating a snapshot. I do not know what to make of this:
The terminal output that displayed immediately after rsync copy completed:
sent 5,207,488,792 bytes received 3,294,422 bytes 77,196,788.36 bytes/sec
total size is 5,193,829,718 speedup is 1.00
/tmp/extracted /home/work
Archive is cpio archive...
Decompression error
root@desktop:/home/user#
The refractasnapshot_errors.log file noted:
ls: cannot access '/home/snapshot/*.iso': No such file or directory
62 blocksgzip: stdin: not in gzip format
cpio: premature end of archive
cpio: kernel/x86/microcode/AuthenticAMD.bin not created: newer or same age version exists
62 blocks
xzcat: (stdin): File format not recognized
cpio: premature end of archive
This occurred after a dist-upgrade from ascii to beowulf with the xfce4 desktop. Missing dependency perhaps? This upgrade was done remotely on a friends computer in a ssh terminal. A lot of packages were removed with "apt autoremove", so something could have been removed.
Thanks for any ideas.
I got it working. The major problem that I had was that the entry in both the sshd_config file on the remote system and the ssh_config file on the local system had the entry as the default:
XAuthLocation $HOME/.Xauthority
It would not work with this entry in either file.
So I changed this entry in both files to:
XAuthLocation /usr/bin/xauth
And it works.
So for anyone that runs across this thread, I will put the config entries in both the files that are working now.
On the remote sshd_config file, I have:
X11Forwarding yes
X11DisplayOffset 10
X11UseLocalhost yes
AllowTcpForwarding yes
XAuthLocation /usr/bin/xauth
On the local system and its ssh_config file, I have in the ~/.ssh/config file:
Host IP_ADDR
ForwardAgent yes
ForwardX11 yes
ForwardX11Trusted yes
XAuthLocation /usr/bin/xauth
To use, I open a terminal window and enter the Xephyr command to open a blank and black window:
Xephyr :2 -screen 1368x768 -resizeable &
Next, enter the command to ssh into the remote box and forward the xsession to display two on the local machine:
DISPLAY=:2 ssh -Y USER@IP_ADDRESS
Lastly, in the same terminal window which is now showing connected to the remote machine, I entered
startxfce4
And it works.
Thanks for your input fsr, I will keep tinkering with this some more. It may be something small on my system that is the problem, but I don't know what it could be and I can't think of what else to check. Thanks again, at least now I know that it does work on someone else's system, and I am motivated to keep looking and trying to find a resolution.
Thanks for responding rr. I have never needed xvfb before and up until now I have never seen it mentioned as being a dependency or needed for nested xsessions.
This is an example of what I have used in the past:
https://nek0.eu/posts/2014-10-25-Deskto … ephyr.html
or
https://www.dedoimedo.com/computers/xephyr.html
But this depends on Xforwarding successfully working, which I am getting errors on at the moment. Thanks again.
Sorry for such a long post, but I wanted to include all that I thought was relevant to my issue. I am trying to use nested X-sessions between a client and server using the Xephyr-xserver package. The Xephyr program appears to be working but I keep seeing these errors when I try to ssh into the server machine with the following command:
$ ssh -v -Y user@xxx.xxx.xxx.100
Warning: No xauth data; using fake authentication data for X11 forwarding.
...
debug1: Remote: No xauth program; cannot forward X11.
X11 forwarding request failed on channel 0
TL/DR-
- xauth is installed on both the client and server
- .Xauthority file is located on both client and server and it has the correct permissions
- sshd_config on server and ssh_config on client "appear" to have all needed options active.
These are the commands used with Xephyr,
user@host:~$ Xephyr :2 -screen 1368x768 -resizeable &
[1] 4819
This loads a blank black screen, which is normal.
user@host:~$ DISPLAY=:2 ssh -Y user@xxx.xxx.xxx.100
...
Warning: No xauth data; using fake authentication data for X11 forwarding.
...
Common configurations of client and server:
# xauth is installed on both the client and server
# apt policy xauth
xauth:
Installed: 1:1.0.10-1
Candidate: 1:1.0.10-1
Version table:
*** 1:1.0.10-1 500
500 http://deb.devuan.org/merged beowulf/main amd64 Packages
100 /var/lib/dpkg/status
# xorg is installed on both client and server
# apt policy xorg
xorg:
Installed: 1:7.7+19
Candidate: 1:7.7+19
Version table:
*** 1:7.7+19 500
500 http://deb.devuan.org/merged beowulf/main amd64 Packages
100 /var/lib/dpkg/status
# Xauthority has the proper permissions on both the client and server
$ ls -n .Xauthority
-rw------- 1 1000 1000 105 Sep 15 16:07 .Xauthority
I added this to both the server and client in the ~/.bashrc file just to try "something"
export XAUTHORITY=$HOME/.Xauthority
The username on the account is the same for both server and client.
Server setup
# sshd_config contains
AddressFamily any # ipv6 is not disabled
AllowAgentForwarding yes
X11Forwarding yes
XAuthLocation $HOME/.Xauthority
X11DisplayOffset 10
X11UseLocalhost yes
AllowTcpForwarding yes
# Client setup
# ssh_config contains in ~/.ssh/config
Host xxx.xxx.xxx.100
ForwardAgent yes
ForwardX11 yes
ForwardX11Trusted yes
XAuthLocation $HOME/.Xauthority
There is a lot written about this on the web and it seems rather straightforward to implement. However, I am overlooking something. I have tried several examples that I found on the web, but still I get these errors.
Warning: No xauth data; using fake authentication data for X11 forwarding.
...
debug1: Remote: No xauth program; cannot forward X11.
X11 forwarding request failed on channel 0
Additional info:
Even with the ssh x-forwarding errors, the ssh connection will succeed, so if I go ahead and issue the command to "startxfce4" in the terminal window connected to the server, I expect the xsession to start in the blank black Xephyr window. Instead, I get this output:
$ startxfce4
/usr/bin/startxfce4: Starting X server
/usr/lib/xorg/Xorg.wrap: Only console users are allowed to run the X server
xinit: giving up
xinit: unable to connect to X server: Connection refused
xinit: server error
But if I go into /etc/X11/Xwrapper-config and change
allowed_users=console
to
allowed_users=anybody
then the xsession will start, but not within the Xephyr nested x window. It starts on the monitor connected to the physical machine.
It is not forwarding the xsession to Xephyr, and I am out of ideas on what to check and I can not find anything new on the web to help me out, yet. So I am asking for help. Any ideas?
Just throwing an idea or two out, if you want it to refer to Devuan hacks
D-hacks
If you want it to refer to systemd hacks or workarounds,
d-hacks
IMHO, this does not deserve a capital "D"... (-;
Perhaps,
What ever happened to "KISS" and "if it works, don't fix it"
"KISS Hacking"
"Simple Hacks"
This is the way that I would do it:
- Open a text editor, leafpad or geany are my favorite.
- Type out a long password with a combination of uppercase letters, lowercase letters, numbers, and symbols. Highlight this new password with a click and drag motion, right click with a mouse, and select copy.
- Click here to change your password. Input the old password and then click on the field to input the new password - so that the cursor indicator is there - right-click on the mouse device and select "Paste". This will paste the newly created password into this field. Do this again for the confirmation field.
- Click "Submit".
- Click on "Logout".
- Delete the text editor file containing the new password. Do not save.
I am not familiar with this particular forum board, but some do not allow for accounts to be "deleted" to the point where your posts will disappear. It has to do with database issues, for lack of a more accurate explanation. But this method should suffice that you, nor anyone else, will be able to use the account.
Edit: Spelling correction.
FWIW - for what it's worth, I just migrated from Kali to Beowulf using the kali netinstall.iso.
Using the instructions on this page, I only slightly deviated from these instructions.
To keep it short and sweet, I -
- updated the kali install, added devuan's sources in sources.list, commented out the kali sources
apt install eudev
(will get errors, do not reboot)
apt -f install
This is the extra step not noted on the link above. sysvinit-core did not get pulled in
apt install sysvinit-core
reboot
apt purge systemd* libnss-systemd
apt dist-upgrade
- alter /etc/network/interfaces to rename en0ps (??) to eth0
service networking restart
apt install libelogind0
- check with command
dpkg -l | grep systemd
Returns nothing, so sysd here.
I downloaded the kali virtualbox file and will try this out on it now. Will report back.
Edit:
While the netinstall.iso worked, the full kali install did not. At least I got a dead end. Now I see above why the recommendation above was to start at ceres and see what broke.
A current kali to beowulf migration was not successful in large part because kali uses a newer version of dbus. An attempt to use ceres or chimaera may be more successful.
root@kali:/home/kali# apt policy dbus
dbus:
Installed: 1.12.20-1
Candidate: 1.12.20-1
Version table:
*** 1.12.20-1 100
100 /var/lib/dpkg/status
1.12.16-1+devuan2 500
500 http://deb.devuan.org/merged beowulf/main amd64 Packages
And trying to downgrade dbus to devuan's version resulted in 934mb of packages to be removed. Oh well, it was fun trying.
May be because Devuan itself is good enough for servers.
+1
Start with a minimal build, customize it, rebuild with refracta-tools, and you have your server OS ready to install wherever you want it. Servers usually don't use that many installed programs, so a few apt commands to install what you need is not much work. For me, it takes longer to configure what I install than to install what I need. I do save my configuration files though, naturally.
Virage 3.1 Beowulf
But beowulf is not mentioned in the sources.list file. The repository mentioned in the sources file is:
deb http://es.deb.devuan.org/merged ceres main contrib non-free
deb-src http://es.deb.devuan.org/merged ceres main contrib non-free
I tested in vbox and it installed fine. The first update pulled in 430+mb of updates. Ceres related maybe? Beowulf is commented out in the sources.list.d folder also.
I will hesitate to install on a physical partition as I am having a hard time isolating a system freeze on ceres at the moment. I like the packages installed and the theming is excellent.
tdrnetworks,
With jessie support scheduled to end June 30, 2020 (5 days from now), what is your plan for the jessie php builds? Will you continue to build them after this date?
Do you have any plans to build php for Chimaera (current testing) any time soon?
I, and others, appreciate you building and offering your php builds to devuan. Thank you.
just want to know if this would work on your devuan system
It works pretty good here. I noticed the editor line at the bottom was empty though. Is that because I do not have a default editor configured?
-------------------------------------
os: Devuan GNU/Linux 4 (chimaera/ceres)
kernel: 5.6.0-2-amd64
uptime: 4 hours, 15 minutes
pkgs: 2028
memory: 4.0Gi / 15Gi
wm: Xfwm4
shell: /bin/bash
editor:
I did have to install wmctl as it was not installed.
Try it with http:// instead of https://. I don't know this but I think there may be a problem with the ssl cert.
I saw this error yesterday and used the above workaround, and it worked.
Great work. Thank you !
+1
I'm using gocryptfs
+1
From reading about gocryptfs, it was started by a person who assisted with EncFS, who wanted to address the security vulnerabilities in it. Thus, gocryptfs was born. A couple of additional facts about it:
First, it pulls in no dependencies. I can pin a particular version and run it in Ascii, Beowulf, Chimaera, and Ceres and mount/use the same remote nextcloud instance in all versions.
Second, with the -config option, you can use gocryptfs to mount a remotely encrypted folder which is decrypted locally, then the files will be automatically stored encrypted in the mounted remote location. EncFS did this, and so does gocryptfs.
I have been using it a while now, and really like it.
FlibberTGibbet mentioned that the Buster to Beowulf migration did work in WSL. I had read about WSL but never tried it, so I gave it a spin. It works.
This process is divided into three main parts:
Activate WSL in Windows 10 > Install Debian from MS Store > Migrate to Beowulf
Why do this? > Why not? It is there, give it a try and see if it helps with your productivity in some way. My purpose was to run a native linux app "Virtual-Manager" (virt-manager) within windows for a friend of mine who wants to access KVM and its vhosts. Your purpose may be different.
Install/Activate WSL in Windows
You will need to install an xserver in windows if you wish to display native linux apps. After all, this is my reason for testing this WSL. Xming seems to be the most popular, but this page mentioned that vcxsrv was the most pure opensource xserver for windows available. It seems to work fine.
Boot up the "program" Debian in Windows 10. You should have sudo rights by default, but if you want a root user, set the root password with:
sudo passwd root
Migrate Debian Buster to Devuan Beowulf:
My migration steps, mostly taken from the link above.
This step may be optional, but I have done it on all Beowulf upgrades and migrations that I have done.
export the PATH in /root/.bashrc with:
export PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
Change apt sources:
nano /etc/apt/sources.list
Make your sources.list look like the one provided. Comment out all other lines.
deb http://deb.devuan.org/merged beowulf main
deb http://deb.devuan.org/merged beowulf-updates main
deb http://deb.devuan.org/merged beowulf-security main
#deb http://deb.devuan.org/merged beowulf-backports main
Now you have to update the package lists from the Beowulf repositories.
Recent changes in APT mean this is not allowed but we can override this behaviour.
apt update --allow-insecure-repositories
The Devuan keyring should now be installed so that packages can be authenticated.
apt install devuan-keyring --allow-unauthenticated
Update the package lists again so that packages are authenticated from here on in.
apt update
Upgrade your packages so that you have the latest versions. Note that this does not complete the migration.
apt upgrade
Once this is done eudev needs to be installed.
apt install eudev
The last command is known to cause package breaks but we will fix this as part of the migration process.
apt -f install
The command to install eudev pulled in sysvinit-core, but not sysvinit, so:
apt install sysvinit
Just to make sure all is updated:
apt dist-upgrade
Systemd* packages were removed during this process, but just to make sure:
apt purge systemd libnss-systemd
apt autoremove --purge
apt autoclean
Before you start installing any packages in WSL, I recommend that you make apt to not install recommended packages.
I suspect this will help keep the WSL smaller and leaner. Add the following to /etc/apt/apt.conf.d/00norecommends:
APT::Install-Recommends "no";
Reboot > But the shutdown/reboot commands do not work in WSL. The WSL process has to be stopped/started like any other service in Windows.
If you install openssh-server, you will notice that it does not start automatically with WSL, nor will most, or all, other linux services that you think should start. Installing sysv-rc-conf and rcconf programs do not help, unless I did something wrong. This may be a bug but I see no error message anywhere, yet. However, this little script seems to control the start/stop of both WSL and its system services at the windows boot. Beware, this script will start WSL with Windows, and it will start the (linux services) that you add to the command.txt file list. I don't know about the security of this, but it seems to work.
Like FlibberTGibbet said, there is a niche group of people who would want or need this. I feel this is true, but I am only tinkering.
I did not take exact notes during this testing, so if I made a mistake or overlooked something, please add your input. After finding some good links, this was pretty easy. And FlibberT was right, it is nice to have a (Devuan) linux terminal available in windows. If I am going to use windows, I do want some Devuan on it.
The option to use virtualization in windows still stands, but the ability to use a native linux program in windows without using a virtualization program is pretty neat too. It removes a layer of complexity for the end user. Instead of learning an entire program like virtualbox, the user only has to learn a couple of terminal commands to keep the WSL updated. I hope my friend appreciates this new capability, and others may as well.
The only problem that I have seen thus far, is that the virt-manager program does not read the keys from the preferred $HOME area of the WSL user. It reads the keys from the global /etc/pki/libvirt location within WSL. The program "virt-manager" does work. Will need to look into this more, later on.