You are not logged in.
Not sure what to do
# echo '/usr/sbin/rfkill unblock all' >> /etc/rc.localSee also https://www.freedesktop.org/software/sy … rvice.html
but wouldn't it be as fast a thinkpad x230?
I doubt it, those A53 cores aren't very fast at all and the Coretex-M4F is only 32-bit. It would be much better if they used a Snapdragon 8cx but I suppose that would hit their profit margin too much.
But at least it uses an open-source compatible graphics chip, unlike the PineBook Pro and it's shitty Mali GPU.
That is an interesting device but it's over-priced and under-powered (IMO). And they claim it runs Debian 11, which doesn't exist (yet). If they actually mean Debian bullseye then Devuan chimaera should also work on it.
No, the GRUB version in beowulf doesn't support f2fs.
I wonder if there is a workaround? I tried copying to the custom grub menu the entry that artix linux creates and it complains of not having /boot/grub/i386-pc/f2fs.mod loaded (and yes i do have f2fs-tools installed in devuan and artix), so i wonder if i could just move that module there into stable grub
I was thinking about suggesting that but a better solution is probably to just install GRUB in Artix instead of Devuan. The only problem with that is having to manually update the configuration every time the kernel version in Devuan changes.
Another alternative is to install Artix's GRUB version in Devuan by running these commands from Artix:
# mount /dev/sda1 /mnt
# pacman -S grub
# grub-install --target=i386-pc --boot-directory=/mnt /dev/sdaThen remove the grub-pc package in Devuan but keep the grub2-common package installed so that the configuration file is updated every time the kernel version changes:
# apt install grub-pc- grub2-commonAnd yet another alternative is to backport the grub-pc, grub-common & grub2-common packages from ceres.
I presume this is a non-UEFI system?
does that seem fair ?
Indubitably. Good work!
when i update-grub there but this is not the case as update-grub just finds unknown linux distro and it dont boot
Devuan's GRUB can't deal with the multiple initramfs images that are created in Arch-based distribustions if the CPU µcode package is installed.
If you post the actual /boot/grub/grub.cfg file then we could confirm if this is the problem here.
In /etc/grub.d/40_custom
menuentry "Artix" { set root=(hd0,3) linux /boot/vmlinuz-linux-lts initrd /boot/initramfs-linux-lts.img }
You need to pass the location of the root partition to the kernel and it's best to identify it with a UUID because the block device assignment might change from one boot to the next.
Try something like this (replace $uuid with the actual UUID for the Artix root partition):
menuentry 'Artix' {
search --fs-uuid --set=root $uuid
linux /boot/vmlinuz-linux-lts root=UUID=$uuid rw quiet # add any other boot parameters here
initrd /boot/intel-ucode.img /boot/initramfs-linux-lts.img
}^ That presumes the intel-ucode package is installed, check /boot in Artix for the actual initramfs images and adjust as required.
There is already a guide for Debian buster that should be applicable to beowulf:
Head_on_a_Stick wrote:If you're using VirtualBox (which you shouldn't because it's shit) then you need the "guest additions" for proper graphics support. Ask the interweb about it (or switch to QEMU/KVM, which is faster and generally better all-round).
Well, I followed your recommendation. Installed Beowulf on KVM (`virt-manager`) with `backports` and `firmaware-amd-graphics`.
Max. chooseable resolution is 1920*1200. Does it mean, due the fact that in KVM hasn't 'guest.iso', that if I install `Beowulf`on my NVMe I will have a max of ca. 1k instead of 4k?
You don't need a backported kernel or the AMD graphics firmware if you're running in a VM because the VM isn't exposed to the AMD card (as I already explained and as you already claimed to have understood).
I don't use virt-manager but I know that the default -vga std option for QEMU supports up to 2560x1600. The -vga qxl option will support up to 2560x1440 and higher resolutions are available if the VGA memory size for the QXL drivers is increased: https://wiki.archlinux.org/index.php/QE … or_support
I presume your guest has the xserver-xorg-video-qxl package installed?
Baremetal yes, 4k (3840*2160) not at all. EDID is OK, firmware is OK, xrandr is ok, resolution not, the thread will never close this way.
No, it will never end unless you start providing some logs that are relevant to the problem (in this case the X.Org log) and clearly state which version of the kernel and firmware packages are installed.
@rolfie says that the RX580 works for them, I don't have that card or a 4k monitor so I can't really comment directly.
vbeinfo is deprecated, videoinfo is actual command
Oops, sorry.
I meant the installation-representation, under *buntus is called `ubiquity`. There is much more to read and write. Maybe I can switch shortly to terminal command and force `os-installer-resolution` to higher value and be able to read the installer-text.
Well just use the non-graphical installer then. I was under the impression that the beowulf ISO didn't provide a graphical installer anyway.
No newbie explanation here?
Apparently not, no.
Dell Vostro A90 netbook running Devuan Buster with LXQt
So that's a 32-bit system with 1GiB of RAM then? Not really comparable to my (64-bit) ThinkPad E485 with 16GiB of RAM — Linux eats all the memory it can and with 16GiB available usage shoots up in comparison to a 1GiB system.
Perhaps its time to switch to Devuan
No way, glibc is a bloat-monster compared to musl ![]()
But seriously, I have compared live versions of some of the Openbox Devuan derivatives and Alpine is still significantly lighter on my hardware.
EDIT: and the Debian-derived abstractions always irritate me after a while, Alpine's simplicity is more to my liking.
All Lives Matter
Yes, that's the whole point of Black Lives Matter — at the moment black lives seem to matter less than non-black lives ![]()
And I am very pleased to offend such an ignorant fuckwit as yourself ![]()
Devuan ~ # mkdir tmp2 Devuan ~ # cd tmp2 Devuan tmp2 # wget https://pkgmaster.devuan.org/devuan/pool/main/e/eudev/eudev_3.2.7.orig.tar.gz --2020-08-29 07:19:32-- https://pkgmaster.devuan.org/devuan/pool/main/e/eudev/eudev_3.2.7.orig.tar.gz Resolving pkgmaster.devuan.org (pkgmaster.devuan.org)... 54.36.142.183, 2001:41d0:2:1f68::3624:8eb7 Connecting to pkgmaster.devuan.org (pkgmaster.devuan.org)|54.36.142.183|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 1945415 (1.9M) [text/plain] Saving to: 'eudev_3.2.7.orig.tar.gz' eudev_3.2.7.orig.tar.gz 100%[====================================================================================================================>] 1.85M 110KB/s in 15s 2020-08-29 07:19:49 (126 KB/s) - 'eudev_3.2.7.orig.tar.gz' saved [1945415/1945415] Devuan tmp2 # tar xf eudev_3.2.7.orig.tar.gz Devuan tmp2 # cd eudev-3.2.7 Devuan eudev-3.2.7 # wget https://pkgmaster.devuan.org/devuan/pool/main/e/eudev/eudev_3.2.7-6.debian.tar.xz --2020-08-29 07:21:08-- https://pkgmaster.devuan.org/devuan/pool/main/e/eudev/eudev_3.2.7-6.debian.tar.xz Resolving pkgmaster.devuan.org (pkgmaster.devuan.org)... 54.36.142.183, 2001:41d0:2:1f68::3624:8eb7 Connecting to pkgmaster.devuan.org (pkgmaster.devuan.org)|54.36.142.183|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 61868 (60K) [text/plain] Saving to: 'eudev_3.2.7-6.debian.tar.xz' eudev_3.2.7-6.debian.tar.xz 100%[====================================================================================================================>] 60.42K 133KB/s in 0.5s 2020-08-29 07:21:09 (133 KB/s) - 'eudev_3.2.7-6.debian.tar.xz' saved [61868/61868] Devuan eudev-3.2.7 # dpkg-buildpackage -us -uc -b dpkg-buildpackage: error: cannot open file debian/changelog: No such file or directory
WTF are you doing? Why didn't you unpack the tarball containing the debian directory? ![]()
Try this instead:
# apt install build-essential devscripts
# tee -a /etc/apt/sources.list <<<'deb-src http://deb.devuan.org/merged/ beowulf main'
# apt update
mkdir -p ~/builds/eudev && cd ~/builds/eudev
apt source eudev && cd eudev-3.2.7
# mk-build-deps --install --remove
debuild -us -uc
# aptitude purge eudev-build-deps^ That works for me.
Maybe exacually LOOK at the config ?
I did:
override_dh_auto_configure:
dh_auto_configure --builddirectory=build-deb \
-- $(CONFFLAGS) $(CONFFLAGS_deb)
dh_auto_configure --builddirectory=build-udeb \
-- $(CONFFLAGS) $(CONFFLAGS_udeb)Does anyone here know any thing about eudev ?
Not much but if you have an actual question then please post it so I can demonstrate how useless your search-fu is ![]()
Why can I not rebuild it with 'dpkg-buildpackage -us -uc -b' ?
Because you don't know what you're doing. Please post the actual errors so that we can help you.
FWIW I've just managed to rebuild it from the Devuan source in my Debian buster box with no problems whatsoever.
What kind of convoluted crap does your build system do ??
What ever happened to the KISS method ??
Debian is not and has never been in any way KISS, it is one of the more abstracted and complicated distributions; Devuan inherits this complexity, for better or worse.
If you want a systemd-free KISS distribution then perhaps try K1SS ![]()
Why does it get built with --disable-dependency-tracking ??
[...]
The prefix variable does NOT seemed to be passed properly either, lol.I mean it's configured with --disable-maintainer-mode, which is not even a valid option for eudev, lol.
Why do you think this?
The debian/rules file has no such options:
CONFFLAGS = \
--exec-prefix= \
--with-rootprefix= \
--with-rootlibdir=/lib/$(DEB_HOST_MULTIARCH) \
--disable-silent-rules
CONFFLAGS_deb = \
--enable-selinux \
--enable-manpages \
$(NULL)
CONFFLAGS_udeb = \
--disable-selinux \
--disable-manpagesWhat was the actual number of tabs you had open in each browser?
8
Were the websites in each tab identical for each browser
Yes.
Were some Web sites using JavaScript/HTML5/WebSockets or other memory/CPU intensive operations?
No, javascript was disabled in both browsers.
What command(s) did you use to get these number?
no one knows why
Well the reason for not allowing a regex symbol in the user name seems pretty clear to me.
EDIT: and if you mean the appendix then it's a vestigial organ that is used to digest non-starch polysaccharides in ruminants but we use it as a storage area for commensal micro-organisms.
For the time being I may just work around it by doing an "ln -snf <zonefile> /etc/localtime.
That's exactly what the postinst script does:
if [ "$1" = configure ]; then
# If the user prefers to manage the time zone by itself, let him doing that.
if ! [ -e /etc/timezone ] && [ -z "$DEBCONF_RECONFIGURE" ] ; then
db_stop
echo
echo "User defined time zone, leaving /etc/localtime unchanged."
else
# Get the values from debconf
AREA=Etc
ZONE=UTC
db_get tzdata/Areas && AREA="$RET"
db_get tzdata/Zones/$AREA && ZONE="$RET"
db_stop
# Update the time zone
echo $AREA/$ZONE > /etc/timezone
ln -nsf /usr/share/zoneinfo/$AREA/$ZONE /etc/localtime.dpkg-new && \
mv -f /etc/localtime.dpkg-new /etc/localtime
which restorecon >/dev/null 2>&1 && restorecon /etc/localtime
echo
echo "Current default time zone: '$AREA/$ZONE'"
fi
# Show the new setting to the user
TZBase=$(LC_ALL=C TZ=UTC0 date)
UTdate=$(LC_ALL=C TZ=UTC0 date -d "$TZBase")
TZdate=$(unset TZ ; LANG=C date -d "$TZBase")
echo "Local time is now: $TZdate."
echo "Universal Time is now: $UTdate."
if [ -z "$DEBCONF_RECONFIGURE" ] ; then
echo "Run 'dpkg-reconfigure tzdata' if you wish to change it."
fi
echo
fiEDIT: added the rest of the configure option.
Regarding the backports-repository, I could install `Header files for Linux 5.4.0-0.bpo.2-amd64` from `backports/main` but... there are also things like `libreoffice 1.7` and generally I made bad experience with backports at least under kubuntu.
Well that's because Kubuntu is a pile of crap ![]()
For Devuan the backports have a pin value of 100 by default and so will only be installed and upgraded from there if the --target option is used: https://backports.debian.org/Instructions/
Actually (see above) I run Kubuntu as host and Beowulf in the VM
Kubuntu include the firmware blobs in the kernel (bunch of degenerates) and a beowulf VM will not "see" your video card unless you pass through the PCI address.
Check the lspci output in the VM, it will almost certainly show a virtual video card.
on beowulf want not start at all on my main machine giving me only black-screen and cursor
Well the RX580 card is definitely supported, as confirmed by rolfie earlier in this thread.
Installed former Devuan3.iso in VM and cannot get a bigger resolution than 1024, no matter what I try
If you're using VirtualBox (which you shouldn't because it's shit) then you need the "guest additions" for proper graphics support. Ask the interweb about it (or switch to QEMU/KVM, which is faster and generally better all-round).
Installed former Devuan3.iso as main system and got black screen, can login, make apdates, install software but was not able to start GUI.
For a bare metal installation you will need the firmware-amd-graphics package, as I have mentioned several times now.
Does GRUB use Xorg?
No.
How can I increase the resolution during Installation? I cannot see anything!
Press c at the GRUB screen, use the vbeinfo command to list the supported resolutions then press escape and edit the installer GRUB menuentry to add the desired resolution to the end of the line that starts with linux; for example:
video=800x600@75How to create the `Xorg-configuration-file` properly and how to point the system to look at that during start?
Read man xorg.conf. The system will automatically read the file if it is placed in the correct location.
How to let read the system at EDID and configure Xorg properly?
If Kubuntu displays the correct resolution then the EDID block is not the problem.
I won't get in the details now but after installation I was able to create BTRFS raid, create my subvolumes, and run Devuan from them.
Is this to be a challenge thread then? ![]()
Can we at least see the contents of /boot/grub/grub.cfg?
BadWolf
Hmm, interesting. It's quite nice and a bit lighter than firefox-esr in my Alpine system:
53.6 MiB + 71.8 MiB = 125.3 MiB badwolf
115.5 MiB + 170.6 MiB = 286.1 MiB WebKitNetworkProcess (8)
371.3 MiB + 500.0 MiB = 871.3 MiB WebKitWebProcess (8)
947.4 MiB + 1.2 GiB = 2.1 GiB firefox (12)Same tabs open in both, a saving of ~800MiB is quite significant ![]()
But I do think FF has a memory leak somewhere...
Firefox 79 is compatible with Devuan, it was available last time I checked on Devuan repositories from Ceres release
Adding the ceres repositories to beowulf will break your system, that is an entirely expected result: https://wiki.debian.org/DontBreakDebian … nkenDebian
To use the non-ESR version of Firefox in beowulf download the tarball from Mozilla and then unpack and run it from your home directory.
Interesting paper analysing the information transmitted by various browsers, including Brave:
have noticed firefox 79 on linux requires the removal of libreoffice
I've just managed to install both FF v80 and LibreOffice in my Alpine Linux system (edge repositories) and there are no mentions of such a problem either on the Arch news page or in their forums so that doesn't appear to be the case.
Are you running one of the Devuan development branches? Package transitions can result in some things being un-installable for a while so perhaps you have encountered that issue.
(android users all over not happy)
LibreOffice is available for Android? ![]()
OpenBSD is always going to be slower than Linux because the developers prioritise security over features — for example their kernel is still mostly big-locked[0].
@OP: the post from @penguin is full of crap, that user doesn't know what they're doing and you shouldn't follow their advice.
How to install 'kernel 5.4.0-42-generic'?
Do not do that, that kernel is hopelessly outdated and only available from Ubuntu.
If you want to try a newer kernel than the 4.19 LTS branch supplied with beowulf then install the linux-image-amd64 metapackage from the beowulf-backports repository.
And to repeat: you *do not* need to install the AMDGPU-PRO drivers, they will offer worse performance than the stock open-source drivers supplied with the kernel (and written by AMD themselves).
Please post the output of
apt policy firmware-amd-graphics
# dmesg | grep firmware # needs to be run as root
uname -aIt's not (meanwhile) a question of `sources.list`, it's a question of to small resolution
Your desired resolution will only be supported if you have the firmware package from the non-free component of the repositories so they are entirely relevant here.
It is possible that your monitor isn't supplying a valid EDID block (or perhaps the cable or connector is sub-standard) but the most likely cause is a lack of firmware.
EDIT:
Another 'nice to have' could be 10 Bit deep color representation
Create an X.Org configuration file with a Screen section containing this option:
DefaultDepth 30See also https://linuxreviews.org/HOWTO_enable_1 … r_on_Linux
What vulnerabilities could be in gnu grep?
Probably none, there have only been two known vulnerabilities reported for it: https://www.cvedetails.com/product/2380 … ndor_id=72
But I was speaking in general terms — it is reasonable to assume that all code contains a number of bugs per thousand sloc so smaller code tends to be better (which is one of the reasons why I prefer OpenRC to systemd).
Be aware that BSD tools are different than GNU. For example, BSD grep is at least 10 times slower than GNU grep.
Be aware that OpenBSD has GNU's grep version in their ports tree: https://openports.se/sysutils/ggrep
OpenBSD avoids the GNU tools because of licensing incompatibilities but it's also worth noting that GNU's software tends to be rather bloated in respect of the codebase, which can expose potential vulnerabilities.
EDIT: for a rough example of this compare OpenBSD's grep.c (487 sloc) with GNU's grep.c (2177 sloc).
root@miyolinux:/# grub-install /dev/sda4
Install to the device rather than the partition.
Better method:
# dpkg-reconfigure grub-pc