You need to make the first partition big enough to hold the iso (packed or unpacked) plus kernel and initramfs. If you plan on adding more than one iso to the stick, then make it big enough to hold all the isos. ISO_1 or ISO_2 will work. One unpacks the iso file and the other one keeps it intact and uses the 'fromiso=' boot option. That option is useful if you think you want to have a copy of the intact iso to use it somewhere else or share it.

If you think you might want to use the stick to share files with windows computers, make the first partition big enough to hold those files, too.

Yes, you need to give the fat32 partition the boot flag. You can do that in gparted.

To boot to ram, you need to add the boot option: toram=/path-to/live/filesystem.squashfs  (where path-to is the name you give the directory that holds the OS.) I'm not sure how or if that will work with persistence. You would only boot the read-only live system without any of your persistent changes. There might be a way to do it using overlays. Let me know if you figure it out.

You could boot to ram and manually mount a data partition to hold selected files, but that's not the same thing as persistence.

Look in the refracta2usb help to see what the disk layout looks like on the usb stick. Diagrams are near the bottom.

If you run refractasnapshot in a live session, be sure to edit the config file to use some real storage space for the work directory and the snapshot directory.

If you make a live-usb with an extra partition for full persistence, you can install packages and they will still be there on the next reboot. If you run apt-upgrade and apt-update on such a system, it will install the updated packages into the persistent volume. You could keep updating such a system until you fill up your usb stick. Better to update the iso and re-image the read-only part of the stick.

I meant the setting for the gamma. For example, if you set it to 0.8 and then run the script again, the gamma settings window comes up and shows the value as 1.4 instead of 0.8.

If you need the desktop to come up at a decreased brightness, you could add the one-liner script that was posted above in your startup applications.

I didn't have to do anything special to get sddm to work. It is being handled by runit. Here are the relevant commands I used to install kde.

apt install xorg xinit xterm
apt install kde-standard  xserver-xorg-video-all xserver-xorg-input-all --no-install-recommends
reboot
apt install sddm --no-install-recommends
reboot

I don't have any good ideas for why sddm is running but you don't see it. Is there anything in the sddm log file? (/var/log/sddm.log)

I'd say check the sddm config file, but there doesn't seem to be a simple sddm.conf. Another option might be to replace sddm with lightdm or some other display manager that has a normal config file instead of unintelligible stuff in /etc/pam.d.

I don't know what's going on with excalibur size. I just made a refracta excalibur i386 No-X iso with the last 686-pae kernel and it comes out to 1.5G using xz compression instead of 700MB like the daedalus iso. Looks like the excalibur iso has fewer packages installed, too.

Every release gets bigger, but this is the biggest jump I've seen.

Perfect! Thanks, I forgot about snapshot.debian.org.

I got linux-image, linux-headers and linux-source for 6.10.9-1 696-pae.

testing=excalibur NOW. When debian trixie (current testing) becomes stable, then sid (unstable) will move down to testing and all our forked packages will still be for trixie until excalibur goes stable.

That's why we recommend using codenames in sources.list.

I did a minimal install of daedalus with runit in qemu and then added xorg and kde-standard without recommends. I can confirm that on reboot into the installed system I don't get a login prompt until I press ENTER. After installing sddm, it boots to a graphical login screen.

User's subdirectories get created on first login to the graphical desktop. I got them.

The version in ceres will probably be in excalibur in less than a week. You didn't do anything wrong. You would need to install more packages or do an upgrade with the ceres repo enabled to screw up the system.

After 'init 3' command, you don't need to use the service command. They will start (unless you screwed with all the symlinks.) I use sysv-rc-conf to turn services off in runlevel 2.

Not sure what you have in mind, but one way to get root would be to have yad start a terminal that runs su -c 'init 3' and then it'll ask for root pass in the terminal. If you don't want the user to see a terminal, I think you can use yad to take a password and feed it to the command. I'm not as clear on this. See /usr/lib/refracta2usb/functions/function.mkloop around lines 104 and 176 for an example.

Turn them off in runlevel 2, and then the command to turn them on is

init 3