The officially official Devuan Forum!

You are not logged in.

#1 Re: Hardware & System Configuration » "Partial ACPI Installation" on MacBook5,2 » Today 16:43:01

rebuild your kernel

Been bit by this before on an Intel (not mac) laptop, the kernel did not have the correct ACPI motherboard support.
Hit the MacBook forums...

#2 Re: Installation » actuallization from Ceres impossible » Today 16:33:01

who knows why they changed it

The current sources.list is not particularly friendly [with multiple bracketed differing options on one line] to being parsed and edited programmatically, hence the new format. Read into that what you want ~ SystemD control is where my money is.

#3 Re: Desktop and Multimedia » sayonara: undefined symbol in libsqlite.so » 2025-10-03 20:29:22

System's Excalibur and fresh, untainted

As excalibur is not released and still in testing...
when it breaks you get to keep all the pieces

I tried with devuan's package

And that is the *only* source of packages you should install from if you don't want system wide breakage (whats wrong with vlc ?)

#4 Re: Other Issues » [SOLVED] difficulty compiling openssl » 2025-10-03 20:16:01

Just done a quick google, as of Jan 2024 brotli is an external module and not part of nginx (so will unlikely be in daedelus sources).
You build brotli on its own, no need to build OpenSSL or nginx

I googled for a solution and hit this, there may be better options out there...
https://thelinuxcode.com/enable-brotli- … ion-nginx/

#5 Re: Other Issues » [SOLVED] difficulty compiling openssl » 2025-10-03 18:54:00

Ha, just tried it for fun and even that fails

#6 Re: Other Issues » [SOLVED] difficulty compiling openssl » 2025-10-03 17:54:30

This is absolutely, mindboggingly complicated for no reason

One reason it is absolutely mindboggingly complicated is that you are compiling outside of the packaging system.
I would not even attempt it this way, my approach would be to

apt-get source openssl
apt-get source nginx

*Build them both without modification*
Once that works, for each package modify the compilation options (most likely debian/rules) to achieve the desired result.

#7 Re: Other Issues » [SOLVED] difficulty compiling openssl » 2025-10-03 17:07:59

nginx is still trying to compile OpenSSL (which looks like it requires 'cmake' not 'make' anyway).
Where did these come from, they look suspicious to me ?

--with-cc-opt="-I /home/drm/openssl-3.5.1/include" \
  --with-ld-opt="-L /home/drm/openssl-3.5.1/lib"

Shoudn't the paths be

/home/drm/openssl-3.5.1/install/include
/home/drm/openssl-3.5.1/install/lib64

Your guide is

./configure --help

#8 Re: Other Issues » [SOLVED] difficulty compiling openssl » 2025-10-03 15:39:16

As it is unclear to me why you are compiling OpenSSL (my fault TL;DR;)

g4sra@localhost:# apt-file search libssl.a
libssl-dev: /usr/lib/x86_64-linux-gnu/libssl.a
g4sra@localhost:# apt-file search libcrypto.a
libssl-dev: /usr/lib/x86_64-linux-gnu/libcrypto.a

So installing the libssl-dev package would satisfy both those previously mentioned libraries (for me)

#9 Re: Other Issues » [SOLVED] difficulty compiling openssl » 2025-10-03 15:32:04

That error is caused by a missing system header file.
Ensure you have installed the package corresponding to your compiler version,

gcc -v 2>&1 | tail -1
gcc version 12.2.0 (Debian 12.2.0-14+deb12u1)

for me that is

apt install libgcc-12-dev

#10 Re: Installation » Also UEFI install problems » 2025-10-01 21:24:42

I have never had any different issues booting\installing between 5.0.0 and 5.0.1 installation media.
Just use 5.0.0 and dist-upgrade or full-upgrade to 5.0.1

Sorry, can't help further.

#11 Re: Installation » Also UEFI install problems » 2025-10-01 21:07:44

it supports tpm since it has a connector for the module

Do you have a module plugged in ?

The installation media will install itself in the same mode that it is booted in.
Turning off Legacy is one way to ensure it boots and install in UEFI.

#12 Re: Installation » Also UEFI install problems » 2025-10-01 20:51:22

In computing terms that motherboard is ancient and likely does not support TPM.
Check the motherboard manual and check your firmware is the latest.

#13 Re: Installation » Cant do upgrade Dadedalus to Ceres » 2025-09-22 20:49:54

@steve_v

the same ARTS

Oh we are, Trinity Desktop Environment ships with a (probably updated for compatibility) KDE 3 implementation.

the operative word there is "video"

Ok gotcha, thanks for the clarification.
I always think of video with audio as two separate streams ...
Been using ALSA with RTP for a while, but only on the local net.

#14 Re: Desktop and Multimedia » (alsa) changin soundcard easily via terminal » 2025-09-22 17:04:52

Do not use indexes, use the designated device names consistently and totally...

# PlayStation®VR USB Streaming used in 3D Mode Only (HDMI used in 2D Cine Mode)
pcm.psvr_hw {
  type hw
  card 'PlayStation®VR'
  device 0
  channels 2
}

This is an example from my config of a sound device which comes and goes without breaking anything.

#15 Re: Installation » Cant do upgrade Dadedalus to Ceres » 2025-09-22 17:00:11

@steve_v, nice to read that history of ALSA sound, I can relate to your comments.
I currently use arts, and have not had one issue with it in the last eight years.
There is nothing that I have wanted to do that ALSA has not been able to do, once I have learnt how.
However I am confused by your statement that ALSA is not capable of 'streaming', could you please elaborate on exactly what you mean ?
Being a dinosaur 'streaming' may have a different meaning for me than the context you are using it in.

#16 Re: Other Issues » [SOLVED] Growing file bloat in ~/.dbus/session-bus and /root/.dbus/session-bus » 2025-09-19 10:56:14

There seem st be a lot of confusion in this thread about what a 'machine id' is and what a 'dbus session id' is. They are not the same thing despite dbus using the machine id as a unique but determinable temporary filename.
The reason '~/.dbus/session-bus/' fills up is evidence that the machine id is changing. Under systemd the machine id is never supposed to be changed so the filename never changes and the contents are simply overwritten.

NB This is also why you should avoid changing the machine id once logged in, it can have side-effects on dbus.

root@localhost:~# cat ~/.bash_logout
rm -f ~/.dbus/session-bus/*

#17 Re: Other Issues » [SOLVED] Growing file bloat in ~/.dbus/session-bus and /root/.dbus/session-bus » 2025-09-18 13:50:11

root@localhost:~# ls -l /var/lib/dbus/machine-id
-rw-r--r-- 1 root root 33 Sep 18 09:09 /var/lib/dbus/machine-id

#18 Re: Installation » [SOLVED] Should the live desktop auto start RAID arrays? » 2025-08-28 10:57:51

@fsmithred, been playing.
Blacklisting md_mod is not sufficient although it does prevent passive loading.
When 'mdadm' is invoked with certain parameters it will then in turn trigger the kernel to load md_mod.
That is why 'chmod ugo-x /usr/share/initramfs-tools/hooks/mdadm' and thus knobbling 'mdadm' fixed it for me.

NB, and knobbling the UDEV rules as well...

#19 Re: Other Issues » [SOLVED] excalibur: grub boot menue entry for archlinux disappeared. » 2025-08-28 10:40:58

@delgado, do you have multiple /boot/grub/grub.cfg files, that is, multiple /boot partition mounts ?
Which EFI/<directory>/grub64.efi is being loaded by your PC UEFI, so which EFI/<directory>/grub.cfg is chain loading which /boot/grub/grub.cfg ?

Hit F1,F2,F11,<add whatever key here> during boot to access your PC UEFI boot priority to see which one it is going for.
Installing UEFI bootloaders can rewrite your PC UEFI variables without telling you it has done so.

N.B. UEFI has been forced on you to make your PC more secure and straight forward to configure because as the owner of the PC you are not capable of making those decisions for yourself.

#20 Re: Installation » [SOLVED] Should the live desktop auto start RAID arrays? » 2025-08-27 17:14:22

@fsmithred

Re script:/etc/modprobe.d/mdadm.conf mine comes from the installed package:

ii  mdadm          4.2-5        amd64        Tool to administer Linux MD arrays (software RAID)

With the execute mode removed from script:/usr/share/initramfs-tools/hooks/mdadm mdadm is not invoked (not installed) in the initramfs.
That in conjunction with sysvinit 'mdadm' service being disabled md_mod is not automatically loaded at boot.

NB: If dm_raid has loaded that would use md_mod.

#21 Re: Installation » [SOLVED] Should the live desktop auto start RAID arrays? » 2025-08-26 11:01:46

@Eeqmcsq, have you blocked md autodetction at boot ?
The results that you have posted suggest md raid was activated during boot, not after.

blacklist the md? modules or pass something like raid=noautodetect?.

nodmraid is an alternative to removing the UDEV, rule.

@Eeqmcsq, looking at my daedalus /lib/udev/rules.d/64-md-raid-assembly.rules as an alternative to removing the script, 'nodmraid' on the kernel command line will cause the script to jump to the end bypassing all of the 'ACTION' directives.

#22 Re: Installation » [SOLVED] Should the live desktop auto start RAID arrays? » 2025-08-25 14:26:56

There seems to be differing methods of activation depending whether md raid is compiled in or a module, plus that udev rule that would run.
I found this type of thing a bane when performing data recovery by attaching faulty\corrupt media to a workstation for repair\recovery.

Kconfig CONFIG_MD_AUTODETECT

file drivers/md/md-autodetect.c

#ifdef CONFIG_MD_AUTODETECT
static int __initdata raid_noautodetect;
#else
static int __initdata raid_noautodetect=1;
#endif
        if (raid_noautodetect)
                printk(KERN_INFO "md: Skipping autodetection of RAID arrays. (raid=autodetect will force)\n");
        printk(KERN_INFO "md: Waiting for all devices to be available before autodetect\n");
        printk(KERN_INFO "md: If you don't use raid, use raid=noautodetect\n");

#23 Re: Installation » [SOLVED] Should the live desktop auto start RAID arrays? » 2025-08-25 13:19:32

@Eeqmcsq, looking at my daedalus /lib/udev/rules.d/64-md-raid-assembly.rules as an alternative to removing the script, 'nodmraid' on the kernel command line will cause the script to jump to the end bypassing all of the 'ACTION' directives.

# "nodmraid" on kernel command line stops mdadm from handling
#  "isw" or "ddf".
IMPORT{cmdline}="noiswmd"
IMPORT{cmdline}="nodmraid"

ENV{nodmraid}=="?*", GOTO="md_inc_end"

What needs to be considered is whether a user is expecting to be able to install (onto md storage) from a Live image ?

#24 Re: Devuan » Old Stable Devuan 5 (daedalus) » 2025-08-24 23:56:04

Devuan follows Debian, so you need to know the Debian release timetable.

#25 Re: Installation » [SOLVED] Should the live desktop auto start RAID arrays? » 2025-08-24 23:48:54

FWIW: My experience has been mainly with hardware raid, but if i remember correctly the kernel md modules probed for raid at boot.
That means that the Live media kernel version which may be different from the on disk version may come into play.
If it is still done that way you might be able to blacklist the md? modules or pass something like raid=noautodetect?.
Looking...
Running 'modinfo md-mod' gives two parameters ' start_dirty_degraded:int' and 'create_on_open:bool' but grepping
drivers/md/md-autodetect.c does give a hit on 'noautodetect', suggest that warrants further investigation.

Board footer

Forum Software