You are not logged in.
To make --no-install-recommends the default for your system,
echo 'APT::Install-Recommends "0";' > /etc/apt/apt.conf.d/00norecommends
I can't reproduce the error. I'm running daedalus, installed mariadb-server, then installed the version from ceres. So I upgraded from 1:10.6.8-1 to 1:10.6.9-1 without error. The service got stopped and started successfully.
I don't know what else to do. Maybe someone running ceres would get a different result.
Well, maybe dpkg --configure -a or dpkg-reconfigure mariadb-server-10.6
Not ours. (There's no "devuan" in the version.) So you won't find it in our bug system. Check debian bug reports.
The package still contains an init script, so it can use systemd or sysvinit.
What exactly happened? Did you lose the package upgrading to ceres from daedalus, or on a regular upgrade in ceres, or the package just refuses to install. Did you get any error messages?
apt-cache search mariadb-server
mariadb-server - MariaDB database server (metapackage depending on the latest version)
mariadb-server-10.6 - MariaDB database server binaries
mariadb-server-core-10.6 - MariaDB database core server files
apt policy mariadb-server
mariadb-server:
Installed: (none)
Candidate: 1:10.6.8-1
Version table:
1:10.6.9-1 10
10 http://deb.devuan.org/merged ceres/main amd64 Packages
1:10.6.8-1 500
500 http://deb.devuan.org/merged daedalus/main amd64 Packages
$ apt policy mariadb-server-10.6
mariadb-server-10.6:
Installed: (none)
Candidate: 1:10.6.8-1
Version table:
1:10.6.9-1 10
10 http://deb.devuan.org/merged ceres/main amd64 Packages
1:10.6.8-1 500
500 http://deb.devuan.org/merged daedalus/main amd64 Packages
$ apt-file list mariadb-server-10.6
mariadb-server-10.6: /etc/init.d/mariadb
<snip>
There's an old version of setnet in ascii and an older version in experimental. I'm pretty sure it's still in the minimal-live isos.
The default network manager in jessie, ascii and beowulf was wicd. I don't know why that's missing for you. In Beowulf, you could use network-manager or connman instead if you want.
Install either wicd or network-manager-gnome. I think that either one of those will get put into your desktop startup apps automatically.
Make sure that policykit-1-gnome is installed.
See the section on Session Management and policykit backends in the beowulf release notes:
https://files.devuan.org/devuan_beowulf … _notes.txt
You might want to switch from consolekit to elogind.
My mistake. The version of devuan-baseconf with the fix is only in ceres and daedalus. If the package just gets removed on upgrade, then I guess nothing else depends on it anymore. That's good.
If someone wants to try the newer version, it's here: http://deb.devuan.org/merged/pool/DEVUA … .2_all.deb
Removing the file, removing the package or doing nothing are all equally valid courses.
If you have not already deleted that file, please wait. Below is the last changelog entry for the package that provided that file in the past. Latest version of the package should remove all the old files, including avoid-systemd. Please see if devuan-baseconf is installed. Let me know if it is.
If it's not, please install it and see if it removes avoid-systemd. Let me know that, too. Thanks.
If you already deleted it, then don't worry about it. It's no longer needed.
devuan-baseconf (0.6.4+devuan3.2) unstable; urgency=medium
* Disable debian/install and debian/postinst.
* Keep postrm to remove old files.
* Keep package until nothing depends on it.
-- fsmithred <fsmithred@gmail.com> Sun, 05 Jan 2020 17:12:00 -0500
tomasz, what network issue did you run into on migrating from bullseye to chimaera? If you already explained it in a previous post, then yes, it was forgotten. I've been away for a few days. Help me out here and provide a link to the previous post or explain it here. Thanks.
I've used refractainstaller to install to lvm. You have to use the cli version of the installer, not the gui. And you have to do a lot of the work manually, which you seem to be doing anyway. Maybe something in my Howto will help you.
https://dev1galaxy.org/viewtopic.php?id=2323
One important change. The howto says that your logical volume names should end in a digit. I fixed the installer so that is no longer true. You can use all alpha characters in your /dev/mapper/names.
I can get the same error with a crash in chimaera or daedalus if I press shift-pgup/down a few times. It doesn't matter if I have the Glass theme there or not. It also crashes if I do have the glass theme and select it in the gui.
Maybe some library is missing?
Here's the themerc for openbox with the sapphire hexes. The one in the current package is unchanged from the original clearlooks-phenix theme. Replace the one in the sapphire theme package with this one. If you can figure out why it doesn't work, please let us know.
https://paste.debian.net/1248088/
paste is good for 90 days.
Ditto the warnings above. Make sure you have a good backup before you try it.
The latest release of Devuan is Chimaera (=Bullseye).
Be the first to write a review. (Let us know if it works.)
First mistake was excusable - OP believed the screen that asked him to load firmware from external media. Almost everyone who sees this screen should ignore it and continue. The firmware usually gets installed automatically.
We should probably put a screenshot of that in the install guide. I'm not sure how to get that. The screen doesn't seem to come up when booting a VM. Maybe a camera shot of a hardware boot?
Refracta has a package installed to disable CSD. To undo that:
apt remove gtk3-nocsd
Also, there's a problem with the installer failing to undo a change that live-boot makes to anacron. You probably need to run the following commands.
# First make sure /usr/sbin/anacron.orig.anacron exists.
ls -l /usr/sbin/anacron*
# If you don't see anacron.orig.anacron, you don't need to run the following commands.
rm /usr/sbin/anacron
mv /usr/sbin/anacron.orig.anacron /usr/sbin/anacron
Welcome to Devuan. In Debian and Devuan, openrc uses the sysvinit scripts. That's about all I know about it. I think there are some on-screen instructions when you install it that talk about fixing links that might be relevant if you're replacing sysvinit but get done automatically if you choose openrc when installing the system. That's based on my memory of doing a couple of test installs in the past few years, worth what you paid for it.
Removed packages sometimes leave config files behind. Purging the package instead will remove the configs. To see all leftover configs and remove them:
aptitude search ~c
aptitude purge ~c
# or be specific
aptitude purge <package package...>
This might be of interest. (more package manager commands)
https://dev1galaxy.org/viewtopic.php?id=511
There are a few people here using openrc. You'll get better answers from them. Also, search the forum for past discussions on openrc. There have been a few.
how can I boot, can YOU boot it?
As mentioned previously, there are currently no bootable installer isos.
https://dev1galaxy.org/viewtopic.php?pid=36517#p36517
This bug needs to get fixed:
https://bugs.debian.org/cgi-bin/bugrepo … ug=1013916
I see a new version of cron-daemon-common in sid, but there's no mention of the relevant bug. I guess we'll see if the next set of isos work when that moves down to bookworm/daedalus.
https://metadata.ftp-master.debian.org/ … _changelog
Fixed version of refractainstaller is now in ceres and will migrate down to daedalus. I'll have to pull some strings to get it into chimaera-proposed-updates. If you're in a hurry to get it, you can download it from the repo.
refractainstaller-base
https://pkgmaster.devuan.org/devuan/poo … .2_all.deb
refractainstaller-gui
https://pkgmaster.devuan.org/devuan/poo … .2_all.deb
This one will correctly remove the diversion of anacron and rename it so it runs. It also contains a postinst script, so that when you upgrade the installer to this version, it will check your system for the leftover cruft from the diversion and correct it. It does not restart anacron, so logs won't rotate until after a reboot or a manual restart of anacron.
lzma is an edge case that's come up a couple times in the past. I think maybe ubuntu uses it, too. That section of the script needs an overhaul, and I keep putting it off. Changes in the way amd and intel microcode are applied broke my code. I have some replacement code for that, but I don't look forward to all the testing needed to make sure it's right. But I do need to get to it soon.
In addition to lzma, I see in daedalus that zstd is the new default in /etc/initramfs-tools/initramfs.conf. When I run update-initramfs there, I get
W: No zstd in /usr/bin:/sbin:/bin, using gzip
No, we're not archiving the testing isos. To install daedalus or ceres now, you can either install a chimaera base system and upgrade or else do a debootstrap install. All the live isos have debootstrap installed.
There are currently no daedalus (testing) installer isos with an installer. This week's isos were not installable due to a new package with dependency problems. We pulled them to save everyone the time of getting part way through the install only to fail. The isos that are there are the ones that contain extra packages. They are not bootable.
The offending package is cron-daemon-common. Here's the bug report:
https://bugs.debian.org/cgi-bin/bugrepo … ug=1013916
Type menu at the boot prompt and press ENTER.
I think the easiest way to do it is to make a fourth partition to use as a separate /boot for the second linux. Then decide which linux will rule grub and only let that one do grub-install (or do it last from that one.)
Whichever one rules grub, you will have to create a boot entry manually for the other one, because os-prober won't find encrypted root filesystems. Put your entries in /etc/grub.d/40_custom and run update-grub to generate the boot menu.
I think it's possible to let the two systems share a boot partition, but there may be some tricks involved to avoid problems if/when grub packages are updated. It's also possible to have the /boot directory inside the encrypted root partition if /etc/default/grub has a line with GRUB_ENABLE_CRYPTODISK=y.
Edit: Here are a couple of old discussions that might be helpful:
https://forums.debian.net/viewtopic.php?p=352368
https://forums.debian.net/viewtopic.php?f=16&t=115362
The second pair of commands I posted are what the installer should do. They could also be run in a live session immediately before installing the system.
Currently, the installer removes the diversion but does not rename the diverted file. To fix an installed system, you remove the symlink and rename the file. I think restarting anacron is sufficient, but a reboot would certainly work.
rm /usr/sbin/anacron
mv /usr/sbin/anacron.orig.anacron /usr/sbin/anacron
/etc/init.d/anacron restart
anacron and cron are started from init scripts. anacron checks /etc/anacrontab for what jobs it should do. (It runs the cron scripts.)
Edit: Here's the bug report https://bugs.devuan.org/cgi/bugreport.cgi?bug=682
IMO, asking the user to OK the fix is asking for trouble since some users will likely be confused by this.
I agree, but I worry about messing up some edge case that I didn't think of. I will get a new version out soon.
For the record, the correct fix is
rm /usr/sbin/anacron
dpkg-divert --rename --remove /usr/sbin/anacron