You are not logged in.
boot.ini symlinks to
/media/computerbob/liveiso/isolinuxand
package_list symlinks to
/media/computerbob/liveiso/20210311_1353
This is the mountpoint for the usb:
/media/computerbob/
The volume label on the iso file is liveiso, so that's the usb stick, and the symlinks point to two directories on that usb.
liveiso/isolinux
liveiso/pkglist_snapshot-20210311_1353
So where are the symlinks? If they are on the usb, then
ls /media/computerbob/liveisoshould list
boot.ini isolinux live package_list pkglist_snapshot-20210311_1353I'm guessing this is the ventoy usb and not the isohybrid usb. boot.ini sounds like a windows thing.
I just made a snapshot today and it boots in qemu as well as on usb. Check the error log - /var/log/refractasnapshot.log Look for errors around 'initrd' and around the end of the log.
If you look at the mounted usb from another system, you should see real directories: live, isolinux, and maybe boot and efi. There should also be one with the package list. No symlinks. This is also what you would see if you mounted the iso.
The live directory should contain vmlinuz, initrd.img and filesystem.squashfs.
I'm curious - what symlinks are you seeing?
Here's a live-iso I made today. It's a proof-of-concept, experimental, unofficial, and it looks like it works just fine.
Beowulf, LXDE, it has both lxdm and lightdm with lxdm the active display manager. I forgot to remove lightdm. To change the default, run dpkg-reconfigure lightdm
There's not much more than what comes with lxde. Just a few apps. No gimp or libreoffice.
https://get.refracta.org/files/experime … 2_0818.iso
Login: Password
user:user
root:root
sha256sum:
6d4247852ea41865b0f1977ea88f8399f6ddab4b55c91da079ce1536c443b859 snapshot_beowulf_lxde_amd64-20210312_0818.iso
You don't want to copy the iso to a formatted partition. You need to copy the file directly to the device. If the usb is /dev/sdb, then one way to do it is cat snapshot-whatever.iso > /dev/sdb
This will overwrite anything that was on the usb, so using an empty one is good.
When I chose the second menu choice, I was shown several different installation choices (I don't remember what they were. I quit without choosing any of them.
Does that sound like what's supposed to happen -- no LIVE image, but the option to RE-INSTALL my current system, exactly the way that it was installed and configured when I used RefractaSnapshot to create that .iso?
I've never used ventoy, so I'm not sure how it arranges things. If it lets you boot live-isos that are on the usb, then I would guess that the first thing you'd see would be the refractasnapshot boot menu, which is just an isolinux boot menu. It lists several different boot options. It does not list any install options. So I'm not sure where you were.
It's an isohybrid image, so you could directly image a spare usb stick with it and boot the live system.
The iso you made with refractasnapshot should be an exact copy of the system that made it, with a few things excluded so it will boot properly and also to remove some private stuff. And you should be able to install it from a live-usb or live-cd/dvd.
And yes, it should work the first time without knowing what you're doing, as long as you don't ask it to copy your whole music and video collection. If you look under the hood, you can modify it to do things differently.
I just installed lxde in beowulf to see how it is. Starting with just a standard (no-X) system, apt refused to install the lxde metapackage or any individual parts until I tried intsalling lxde and policykit-1 together. Then it worked.
I did not test removable drives because I'm testing in a qemu VM. But the user could not shutdown or reboot from the desktop without entering the root password and not at all from the lightdm login screen. All the polkit and elogind stuff was there. Turned out that udisks was missing. When I added that, shutdown and reboot were active on desktop and login screen.
Question? Which package or procedure provides the magic where you get your special home folders generated? I seem to keep doing net installs and not getting a proper home directory
If you mean the subdirectories in your home for MUSIC PICTURES and whatever else, that would be xdg-user-dirs and xdg-utils. If those are installed, look in /etc/xdg/ if you want to edit the configs. There are a few xdg- commands, too. You might be interested in xdg-user-dirs-update.
If you mean /home/<username> does not exist, then you somehow failed to create a user during the installation. You would have had to say no when asked if you wanted to create a user. In an already-installed system, the adduser command does all the things you need to create a user. You can change those defaults in /etc/adduser.conf.
That single name folder is hiding in that whole file system. It's /usr/share/backgrounds/xfce.
Any time you change settings in xfce, they get stored in ~/.config/xfce4. One way to return to default settings is to remove the files that hold those settings. The same is true for program settings in ~/.config. If you delete all of that hidden dir, your desktop goes back to the default the same as if you had created a new user and logged in for the first time.
I added a few more services to runit using the three-step procedure I posted above.
- copy the sample files to /etc/sv/
- stop the service with /etc/init.d/<service> stop
- add the service with update-service --add /etc/sv/<service>
Added dbus, elogind, lightdm and all seems to be working.
# sv status /etc/sv/*
run: /etc/sv/acpid: (pid 1894) 1238s; run: log: (pid 1891) 1238s
down: /etc/sv/anacron: 1238s, normally up; run: log: (pid 1895) 1238s
run: /etc/sv/cron: (pid 1874) 1238s; run: log: (pid 1872) 1238s
run: /etc/sv/dbus: (pid 1885) 1238s; run: log: (pid 1884) 1238s
run: /etc/sv/elogind: (pid 1886) 1238s; run: log: (pid 1883) 1238s
run: /etc/sv/getty-tty1: (pid 1873) 1238s
run: /etc/sv/getty-tty2: (pid 1852) 1238s
run: /etc/sv/getty-tty3: (pid 1855) 1238s
run: /etc/sv/getty-tty4: (pid 1853) 1238s
run: /etc/sv/getty-tty5: (pid 1876) 1238s
run: /etc/sv/getty-tty6: (pid 1869) 1238s
warning: /etc/sv/getty-ttyS0: unable to open supervise/ok: file does not exist
run: /etc/sv/lightdm: (pid 1878) 1238s; run: log: (pid 1877) 1238s
run: /etc/sv/mdadm: (pid 1901) 1238s; run: log: (pid 1900) 1238s
run: /etc/sv/network-manager: (pid 1871) 1238s; run: log: (pid 1870) 1238s
run: /etc/sv/ssh: (pid 1899) 1238s; run: log: (pid 1898) 1238sEdit: fixed typo. Thanks.
df /boot Filesystem 1K-blocks Used Available Use% Mounted on /dev/sdb2 945144 100260 779656 12% /boot
What about space on /boot/efi? Don't you have a separate efi partition mounted there?
I'm not sure if I got the logging right. NetworkManager is working, but puts very little in the runit logs and puts more in syslog. It was spamming the syslog with everything it did until I added --log-level=ERR to the start command.
Also, mdadm gives an error (-1) when it stops. But it seems to be running. I don't have any raid array to test it right now.
pstree:
├─runsvdir─┬─6*[runsv───getty]
│ ├─runsv─┬─acpid
│ │ └─svlogd
│ ├─runsv─┬─NetworkManager───2*[{NetworkManager}]
│ │ └─svlogd
│ ├─runsv─┬─sshd
│ │ └─svlogd
│ ├─runsv─┬─cron
│ │ └─logger
│ ├─runsv─┬─mdadm
│ │ └─svlogd
│ ├─runsv─┬─haveged
│ │ └─svlogd
│ └─runsv───svlogdNetworkManager messages in syslog:
# on stopping network-manager:
Feb 28 02:25:08 chimaera-2 dbus-daemon[1506]: [system] Activating service name='org.freedesktop.nm_dispatcher' requested by ':1.35' (uid=0 pid=2919 comm="/usr/sbin/NetworkManager --no-daemon --log-level=E") (using servicehelper)
Feb 28 02:25:08 chimaera-2 dbus-daemon[1506]: [system] Successfully activated service 'org.freedesktop.nm_dispatcher'# on starting network-manager
Feb 28 02:26:50 chimaera-2 dbus-daemon[1506]: [system] Activating service name='org.freedesktop.nm_dispatcher' requested by ':1.38' (uid=0 pid=3006 comm="/usr/sbin/NetworkManager --no-daemon --log-level=E") (using servicehelper)
Feb 28 02:26:50 chimaera-2 dbus-daemon[1506]: [system] Successfully activated service 'org.freedesktop.nm_dispatcher'network-manager runit log on exit:
2021-02-28_02:25:09.02206 invoke-run: network-manager : run exit code is 0
2021-02-28_02:25:09.02209 invoke-run: network-manager stoppedmdadm runit log error on exit:
2021-02-28_01:26:40.87497 invoke-run: ERROR -1 in mdadm: runscript didn't exit normally
2021-02-28_01:26:40.87504 invoke-run: mdadm : run exit code is -1
2021-02-28_01:26:40.87505 invoke-run: mdadm stoppednetwork-manager/log/run
#!/bin/sh
set -e
NAME=network-manager
LOG="/var/log/runit/$NAME"
test -d "$LOG" || mkdir "$LOG" && chown -R _runit-log:adm "$LOG"
exec chpst -u _runit-log svlogd -tt "$LOG"mdadm/log/run
#!/bin/sh
NAME=mdadm
LOG="/var/log/runit/$NAME"
test -d "$LOG" || mkdir "$LOG" && chown -R _runit-log:adm "$LOG"
exec chpst -u _runit-log svlogd -tt "$LOG"Pulseaudio is installed and used in Devuan 3.x?
Only if you install it. If you don't want it, then either don't install it or remove it or use a derivative distro that doesn't include it.
Pulseaudio comes in if you install the full desktop with the task-* packages. It's easier to customize the desktop package selection if you install individual packages instead of metapackages.
Lorenzo, thanks for helping us with this and thanks for doing it.
Hi all,
Something like the following could work ( not tested)#!/bin/sh -e ## uncomment the following if you have dbus as runit service #sv start dbus && sv check dbus || exit 1 exec 2>&1 exec /usr/sbin/NetworkManager --no-daemonHope this helps,
Lorenzo
Tested. And it does work just as it is. I haven't tried dbus yet.
1. /etc/init.d/network-manager stop
2. copy files into /etc/sv/network-manager
3. update-service --add network-manager
4. (maybe) sv start network-manager Not sure if you need to do this. I started/stopped and check status a bunch of times until I was able to get the logging out of the syslog.
Here are the files I used:
network-manager/run
#!/bin/sh -e
## uncomment the following if you have dbus as runit service
#sv start dbus && sv check dbus || exit 1
exec 2>&1
exec /usr/sbin/NetworkManager --no-daemon
network-manager/log/run
#!/bin/sh
set -e
NAME=network-manager
LOG="/var/log/runit/$NAME"
test -d "$LOG" || mkdir "$LOG" && chown -R _runit-log:adm "$LOG"
exec chpst -u _runit-log svlogd -tt "$LOG"
network-manager/finish
#!/bin/sh
set -e
. /lib/runit/finish-default "$@"I also set up mdadm with runit. I used the same procedure as above with the sample runscripts but had to change the last line in mdadm/log/run from a hard-coded log file to "$LOG". That file looks like this now:
#!/bin/sh
NAME=mdadm
LOG="/var/log/runit/$NAME"
test -d "$LOG" || mkdir "$LOG" && chown -R _runit-log:adm "$LOG"
exec chpst -u _runit-log svlogd -tt "$LOG"@dice -
It looks like n-m is not being managed by runit. Look at what's under runsvdir - just fgetty.
Try sv status network-manager and compare to sv status fgetty
You put your NetworkManager directory in /etc/runit. My run dirs are in /etc/runit/runsvdir. (see tree below)
My run dirs are named the same as the init script. I don't know if that matters. So maybe you should have /etc/runit/runsvdir/network-manager.
I didn't examine the content of your run script or review the procedure I used, so I can't comment on either. I'm new at this, too. I can try it with n-m later today or this weekend.
Here's my tree:
# tree /etc/runit
/etc/runit
├── 1
├── 2
├── 3
├── ctrlaltdel
└── runsvdir
├── current -> /etc/runit/runsvdir/default
├── default
│ ├── acpid -> /etc/sv/acpid
│ ├── anacron -> /etc/sv/anacron
│ ├── cron -> /etc/sv/cron
│ ├── getty-tty1 -> ../../../sv/getty-tty1
│ ├── getty-tty2 -> /etc/sv/getty-tty2
│ ├── getty-tty3 -> /etc/sv/getty-tty3
│ ├── getty-tty4 -> /etc/sv/getty-tty4
│ ├── getty-tty5 -> /etc/sv/getty-tty5
│ ├── getty-tty6 -> /etc/sv/getty-tty6
│ ├── haveged -> /etc/sv/haveged
│ └── ssh -> /etc/sv/ssh
├── single
│ └── sulogin
│ └── run
├── solo
│ ├── getty-tty1 -> ../../../sv/getty-tty1
│ └── getty-ttyS0 -> ../../../sv/getty-ttyS0
└── svmanaged
20 directories, 5 filesHere's some pstree. My n-m is started by runit but not managed by it.
runit─┬─NetworkManager───2*[{NetworkManager}]
├─acpi_fakekeyd
├─at-spi-bus-laun─┬─dbus-daemon
│ └─2*[{at-spi-bus-laun}]
├─at-spi2-registr───2*[{at-spi2-registr}]
├─blueman-tray───2*[{blueman-tray}]
├─bluetoothd
├─2*[dbus-daemon]
├─dbus-launch
├─runsvdir─┬─6*[runsv───getty]
│ ├─runsv─┬─acpid
│ │ └─svlogd
│ ├─runsv─┬─sshd───sshd───sshd───bash───pstree
│ │ └─svlogd
│ ├─runsv─┬─cron
│ │ └─logger
│ ├─runsv─┬─haveged
│ │ └─svlogd
│ └─runsv───svlogdIf you want gpt with legacy bios boot, you need a special partition for grub that is at least 1MB size, has no filesystem on it and has flag bios_grub (in gparted) or type ef02 (in gdisk). Then you get the advantage of more than four primary partitions.
Edit: Fixed what I wrote in my sleep. Thanks, HoaS!
Look for those packages in /var/log/apt/history* to see when they were installed and what else was installed at the same time.
fsmithred wrote:Removing systemd service files is the equivalent of removing sysvinit scripts. Doing so breaks interoperability and we absolutely should not be doing that. It also smacks of hypocracy.
build-deps are another story. The current advice is to build in a chroot so you don't contaminate your system. Maybe someday we'll have enough devs to fork everything.
nonsense.. i removed complety those files in jessie.. was a hard job of course and i not are a fashion software stupid guy so my jessie will be here almost 9 years more, cos i not need updates if i do not run server and nobody break a double nat protected provider of internet
It's not nonsense at all. You miss the point. I think it's fine for you to modify your system any way you want. The ability to do that is one of the best features of linux. I think it would be very bad if debian made it official policy to remove all sysvinit scripts. Likewise, I think it would be very bad if devuan made it official policy to remove all systemd service files.
Removing systemd service files is the equivalent of removing sysvinit scripts. Doing so breaks interoperability and we absolutely should not be doing that. It also smacks of hypocracy.
build-deps are another story. The current advice is to build in a chroot so you don't contaminate your system. Maybe someday we'll have enough devs to fork everything.
Is this something new that a boot partition needs to have a boot flag, or are you talking about an efi partition?
Here are the current docs for packaging.
https://git.devuan.org/devuan/documenta … aintainers
Quick summary:
Clone from upstream git repo to preserve the history (salsa.debian.org if it's there)
Build it locally to make sure it works.
Push to your devuan git.
Looks like your projects are all personal projects. To get into the repo someone would need to transfer the projects to the devuan group, then build (usually for suites/unstable). I guess we'd need to figure out where the packages are going. Are they backports, or are you forking for chimaera or ceres?
Same here. 1.20 from devuan beowulf main repo
Sorry, I don't know the answer. I see the "not provided by any .service files" a lot, and I just ignore it. I assume it's there to remind us all that we're not running systemd.
Try this:
apt -s remove consolekitand see what it wants to take with it.
Thanks. They're working on it.
https://bugs.devuan.org/cgi/bugreport.cgi?bug=548
groucho@devuan:~$ aptitude why consolekit i slim Depends default-logind | logind | consolekit groucho@devuan:~$What does this mean in practical terms? eg: for slim which I am fond of.
It means slim needs something that provides default-logind or one of the others on that line. If a package depends on libpam-system | libpam-elogind, it'll complain about missing libpam-systemd if neither of those is there. Install libpam-elogind first and anything that wants libpam-systemd will be happy. (See the Provides line below)
$ apt show libpam-elogind
Package: libpam-elogind
Version: 241.4-2
Priority: optional
Source: elogind
Provides: default-logind, libpam-systemd, logind
Depends: libc6 (>= 2.28), libcap2 (>= 1:2.10), libpam0g (>= 0.99.7.1), elogind (= 241.4-2), libpam-runtime
Conflicts: libpam-ck-connector
Breaks: libpam-systemd
Replaces: libpam-systemdThanks. I never noticed that. Most of my installs are tests, so I use very short passwords. The yad dialog window only takes 18 characters, but it stores more if you type them. I'm not sure why it didn't use your long password.
I tried running the code in a terminal (as root), and it works. I'll have to try it in an install.
Here's the code for that window with the output of my 20-character password entry.
~# pass_entry=$(yad --form --title=$"Configure $pass_dialog password" --center --borders=10 --button=$"OK":0 \
--text=$"You should reset the $pass_dialog password.\nUse TAB to change fields." \
--field=$"Enter new $pass_dialog password::H" \
--field=$"Confirm new $pass_dialog password::H" \
--field=$"Use current password\? (not recommended)":CHK \
"$field_four")
12345678901234567890|12345678901234567890|FALSE|Trim the output, check it and feed it into the passwd command, and it works.
newpass=$(echo $pass_entry|awk -F "|" '{print $1}')
echo $newpass
12345678901234567890
/bin/bash -c "echo -e \"$newpass\n$newpass\n\" | passwd $newusername"
New password: Retype new password: passwd: password updated successfully