You are not logged in.
Perhaps you would find turnstile of interest in this discussion as a system service for running user services. It's of course just the service management part and not any actual service.
A web search on linux-u-boot-orangepi3b didn't give me any exact match, so I can't verify that it makes the same assumptions about partitioning as you are. If u-boot doesn't find boot.scr it won't boot.
Perhaps it assumes an initial type FAT boot partition, which wouldn't be unusual. That would at least have boot.scr for booting, and usually that will also contain kernel and initrd. Or for example, if your u-boot supports the sysboot command you may use that for an extlinux boot from a differt, type ext partition, which then would contain the extlinux.conf configuration file as well as kernel, FDT and initrd.
(EDIT: added mentioning of FDT = Flat Device Tree, aka DTB = Device Tree Blob)
Ideally you get a serial cable for the box so you can see u-boot logging.
But my suggestion is that you first set up your box with an initial FAT partition for boot.scr, The command to use in boot.cmd would be like this
sysboot mmc 0:2 ext4 0x7ffff000 /boot/extlinux.confwhere mmc is the drive type (for sdcard), 0:2 identifies the boot sdcard, partition 2, ext4 indicates the partition type, and it's then followed by the pathname of the configuration file.
You'll then need to mkimage that into boot.scr.
EDIT: The argument 0x7ffff000 is the load address for the extlinux.conf file, and you must choose a RAM address within your actual RAM. High up in your RAM is good.
The second partition can be a single partition for everything, i.e. boot and root filesystem. Note that I've used ext4 rather than ext3, but that's only because I'm incredibly modern
Though you might have better luck telling sysboot it's an ext2 even if it's an ext4, or perhaps use ext3... whichever.
EDIT: actually, you want rootfs to be encrypted, so then you'll need 3 partitions...
didn't you move off github yet?
don't use github = microsoft !!
Maybe you could use dates in "international" format, or like "7 Apr 2026". USA format is rather local to the USA.
My quick package search suggests version 10 is too old to be found, but there is the package libsnmp40 that contains a similarily named dynamic library. So perhaps it works by installing version 40 and implementing a lie by setting a link... thusly:
# apt-get install libsnmp40
# ln -s libnetsnmp.so.40 /usr/lib/x86_64-linux-gnu/libnetsnmp.so.10hth
Ralph.
Mabye you could check which package the files belong to (dpkg -S), which apparently has built in some assumption about context.
Good advice @steve_v. I hope the OP finds it useful/.
Do you have a reference on the "most admins" statistic?
Isn't LVM just an abstraction over the device mapper?
And why mention LVM as if it was an alternative to raid? Aren't they two independent concerns?
Usually only the last 16 hex digits are used.
Amprolla synchronizes repo versions bi-minutely, i.e. every 2 minutes. It also runs "collections synchronization" daily, and a major cleanup weekly.
Note that "Devuan repositories" comprise *meta files* only. It is those that "define" the Devuan repositories. They have the "pointers" (URLs) to the deb files that one install, and most debs are directly sourced from Debian's package pools.
The Devuan package pools have a comparatively small number of packages that replace and add to the Debian package pools. Those that are forks of same-named Debian packages are of course regularly behind their origins, since it takes developer effort to roll in origin updates.
Actually the man page advice is to use two iface snippets in /etc/network/interfaces
iface wlan0 inet manual
wpa-roam /etc/wpa_supplicant/wpa_supplicant.conf
iface default inet dhcpThe first snippet is for declaring the conf file to be used, and the second is the one wpa_action will use when setting up an access point link. (using ifup wlan0=default)
In that way you can also define additional, different handlings of different setups with different identities.
I have no idea.
I don't use it;
I just use plain ifupdown configurations, plus wpagui on some hosts.
I think of that as easy and straight-forward.
In particular, it's well documented.
A note for anyone holding their breath:
turnstile is now available on ceres
(debian package version 0.1.11-2)
what is PFS now? Is language too hard to be written in full?
You must OK the "load components" dialog (with or without selecting any non-default components) in order to get all filesystem type choices in the partitioner. Or, don't choose "expert install".
If you want to boot the ISO in a VM, then obviously it is the VMs (emulated) hardware that is interesting, and not at all the host hardware.
So what does the "failsafe" boot use that is different from other boot options?
So those where reports of the graphics setup of the virtual machines?
In which world are those two logs "much the same"??? Maybe in a world where you ignore differences...
Perhaps you need to install sane-airscan.
@Atalnte: you should have installed the package printer-driver-escpr and not confused yourself and your system with dowwnloaded drivers. You will still need to go through the printer setup after installing the package.
Everyone makes their own choice about what is important.
I jumped off github when I learnt they sold out to Microsoft.
github = Microsoft. Stay away!
Beware! That's still on github
You may check devuan repositories using pkginfo.devuan.org
For example, type https://pkginfo.devuan.org/qps to your browser.
Then, typically, a version code without "devuan" in it indicates that the package is directly a Debian package, whereas a package with "devuan" in it is a Devuan package; i.e., forked typically from Debian packaging sources, modified and built in the Devuan build pipeline and published in the Devuan specific repository overlay.