Re: [DNG] ASCII installation static IP problem
On Thu, 14 Jun 2018 19:54:11 +0900 Olaf Meeuwissen wrote: > Hi Don, > > Don Wright writes: > > > [ ... ASCII using Expert (text) from > > devuan_ascii_2.0.0_amd64_dvd-1.iso ...] > > > > Upon successful boot into the system things looked good locally, > > until I tried to SSH to the box. Not there! > > While /etc/network/interfaces has the settings I expected, the GUI > > showed wicd had ignored them and called DHCP to create all new and > > mostly wrong settings. > > > > #apt remove wicd soon cleaned that up, but who the systemd thought > > it was a good idea to ignore! working! static! IP! settings! and > > install an unwelcome network mangler in the first place? Take a > > purgative, get your heads out of your ASCII, and stop your wicd > > ways from overriding traditional handcrafted, all-natural, > > artisanal, text-based config files. > > The output of `apt-cache rdepends wicd` using various combinations of > the --recurse and --no-* options indicate that just about any, if not > all, of the task-*-desktop packages recommend it, either directly or > indirectly. Some may even prefer network-manager ... putting you > between a rock and a hard place. > > > The guilty parties should lose an inch of *nix beard each in > > penance. > > The guilty parties would mostly be the task-*-desktop packagers ;-) > but if you are comfortable with the installer's Expert mode, why not > forego the installation of a desktop and run > > apt install task-desktop wicd- > > after the initial system install? > > > [ Semi-humorous howls of rage aside: Does the installed system > > ignore static IP by design? ] > > Not if you don't install a desktop ;-) > # You mentioned installing on a Lenove Think*Server*. I *never* put a > # desktop on my servers ... > > Hope this helps, If I understand this correctly, installing any desktop (does this include window managers like openbox?) brings in wicd in a mode that breaks hard coded IP addresses. I would sure find this behavior surprising. Is there a way Devuan can eliminate the "recommends" for wicd and networkmanager with "desktops"? Meanwhile, if whatever distro you're working with does weird stuff when all you want is a hard coded IPV4 address, consider the following distro-independent shellscript, which can be run at boot and also any time some foolish daemon thinks it knows more about your desired IP address than you do: #!/bin/sh hostname=`grep -v "^\s*#" /etc/hostname | head -n1` ip link set dev lo up ip link set dev eno1 down ip addr add 192.168.100.6/24 dev eno1 ip addr add 192.168.100.106/24 dev eno1 ip link set dev eno1 up ip route add default via 192.168.100.96 The preceding sets the box's IP at 192.168.100.6 with an alias at 106, and sets the default gateway to 192.168.100.96. The assumed device name here is eno1. Obviously, you need to modify it to fit your needs, but after that, you can override the inconveniences almost every distro throws at you if all you want is a fixed IP address. I've used this shellscript in several distros. SteveT Steve Litt June 2018 featured book: Twenty Eight Tales of Troubleshooting http://www.troubleshooters.com/28 ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Ascii LVM Encrypted Issue
Thanks, I’ve know been through the Graphic and Text based installs and did not see any such prompt. I’ve even done what was suggested and did the install from a USB key and still the issue persists - installing Devuan ASCII and using LVM Encryption results in the requirement for a PS2 input to unlock the encryption. Unless there is something I am missing here, I believe this is in fact a bug. Any suggestions or input would be very much appreciated! Thanks again, Mike -- Michael McConnell WINK Streaming; email: mich...@winkstreaming.com toll free: 877-GO-4-WINK x 7400 direct: +1 312 281-5434 cell: +506 8706-2389 skype: wink-michael web: http://winkstreaming.com > On Jun 15, 2018, at 2:37 AM, Didier Kryn wrote: > > Le 15/06/2018 à 02:05, Michael McConnell a écrit : >> Thanks for the reply, but at no point during the install do I see an option >> to install the “necessary” drivers? I just ran through the install once more >> and still don’t see any prompt or indication asking about supporting USB >> drivers. I did the install using a USB keyboard on both the VM instance and >> the dedicated hardware. >> > AFAIR, this choice has been present in Debian installer for at least a > decade and I confirm it's still present in Devuan-ASCII - I don't remember in > which step though. > > The choice is rather radical and the option to only install the drivers > for devices present at install time is mostly meant for industrial SBCs and > hand-held devices which don't accept any hot-pluggable device. It is also > interesting for security reasons. > > It would be nice to have an intermediate choice including all common > hot-pluggable devices, but, there isn't. It might help to install from a USB > stick, or just pug one in during install. > > Didier > > > ___ > Dng mailing list > Dng@lists.dyne.org > https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] More Sheiße on the way
On Fri, 15 Jun 2018 20:28:38 +0100, Simon wrote in message <13821b74-b783-429a-a769-53a92d172...@thehobsons.co.uk>: > > So it sounds like this one isn’t the big issue it initially looks > like. ..how about this?: https://www.theregister.co.uk/2018/06/14/microsoft_r_open_debian_dev/ ..they _could_ make their rm /bin/sh depend on the presence of systemd, and not... This time it looks like a sloppy accident, waaay too early, and waaay too obvious: https://www.preining.info/blog/2018/06/microsofts-failed-attempt-on-debian-packaging/ -- ..med vennlig hilsen = with Kind Regards from Arnt Karlsen ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] devuan from scratch?
On Fri, 15 Jun 2018 12:55:10 +0300, Eric wrote in message <20180615095510.GA12693@orphan.zombinet>: > On Thu, Jun 14, 2018 at 08:27:26AM +0200, Jaromil wrote: > > *SKIP* > > it's interesting because it provides an alternative "new" way to > > bootstrap a new apt based system in addition to the classical > > debootstrap method. I also noticed the instructions have no mention > > of systemd, wonder where that will sneak in later... > > I've made myself to believe that some alternative facts (I guess) must > be pointed out. > > Linux-from-scratch builds from source. Debian-from-scratch builds > from respository. Quite a difference, eh? ..aye, totally the wrong way, it _would_ be useful, if it could be set up to build .deb packages and package repositories from source, any source, and independently from any other distro, automagically. ..it has to be set up to happen automagically, or it will stall, and die. > As of svchost -- it *will* come with dependencies, don't you worry. > What makes me wonder, this, so to speak, manual is dated as 20180610. > Yet it doesn't mention anything for sources.list or versions of any > .deb. What seems sneaky to me, because... > > I dare to state my IMCO (as in 'competent') -- it's impossible to > build debian-stable. The only thing debian-* that can be built is > debian-unstable (not debian-testing). But to achieve this they > already need debian-unstable. What LFS is not. > > So, place apropriate *.deb s all over incomplete LFS, to make it > debian-unstable. Then build dpkg and apt. Why would anyone do that? > Just unwind apropriate .deb s. > > *CUT* > > p.s. Also. For those ignorant out there. The Linux-from-Scratch has > gone The Dark Side. Close that book. ..I slightly disagree, the LFS approach is different from the Debian, Slackware, RH etc ways of doing thing and may yield insight we might miss doing our things the current Devuan way. ..if automated, it would also be handy when trying to keep an eye on e.g. systemd "progress" and its impact on system integrity etc, etc. -- ..med vennlig hilsen = with Kind Regards from Arnt Karlsen ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Refracta no-dbus experiment
El 15/06/18 a las 21:34, fsmithred escribió: Refracta no-dbus build (experiment) The subject of running without dbus comes up from time to time in various places. I decided to try it and see how far I could get. I started with a debootstrap install of devuan ascii, pinned dbus to a priority of -1, and proceeded to make the same changes as I do to make Refracta live isos. Normally, the Refracta isos use xfce, but that's not possible without dbus. I was surprised to see how much did install without dbus. So I thought I'd share it. This build uses openbox, lxpanel, lxterminal and spacefm. Maybe someone will want to use it. Maybe it will inspire someone else to do something better. Feeback is welcome. http://distro.ibiblio.org/refracta/files/experimental/refracta9_nodbus_amd64-20180612_0156.iso Thanks fsr, i'll give it a try tomorrow. Cheers, Aitor. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] elogind and startx
On 06/15/2018 04:12 PM, Haines Brown wrote: > > xserver-xorg-video-dummy, xserver-xorg-input-void, ... > Result is that when xserver is run by user, it brings up a frozen > fluxbox window manager. No keyboard or mouse input. My keyboard is > dead. So had to do a hot shutdown. > Description: X.Org X server -- dummy display driver This package provides a 'dummy' display driver, which does not actually display anything. Description: X.Org X server -- void input driver This package provides the void input driver, which never actually registers any input events -- it is, as such, a 'fake' driver. I don't know the purpose of these packages, but I suspect they may be the cause of the problem. Maybe you need some real drivers, such as xserver-xorg-input-all xserver-xorg-video- >> I find udev and eudev to be both installed, but apparently udev is being >> used. 'dpkg -l | grep udev' should show that udev a devuan package. (devuan is in the version) and eudev will also be there. fsmithred ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] elogind and startx
On Fri, Jun 15, 2018 at 07:41:52PM +0100, Dave Turner wrote: > You must install the xserver files to get X working. > > More precisely, they MUST be installed. Whether elogind and libpam-elogind > drag them in as dependencies I do not know. I did as you suggest, but with an unhappy outcome. I installed these packages on a raw ascii 2.0.0 system: aptitude xserver-xorg-video-dummy, xserver-xorg-input-void, xserver-xorg-core xinit, x11-xserver-utils xfonts-100dpi, xfonts-75dpi, xfonts-scalable fluxbox, xfonts-terminus, feh Result is that when xserver is run by user, it brings up a frozen fluxbox window manager. No keyboard or mouse input. My keyboard is dead. So had to do a hot shutdown. After rebooting I take a look at ~/.local/share/xorg/Xorg.0.log and find ... mouseset 0: Damage tracking initialized config udev; adding input device device Power Button (/dev/input/event5) No input driver specified, ignoring this device This device may have been added with another device file. This stanza is repeated for many devices such as device Video Bus, Power Button, device Sleep Button, device logitek USB Receiver, and many others. The last is device BRLTTY Linux 5.4 Screen Driver Keyboard I find udev and eudev to be both installed, but apparently udev is being used. Haines ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
[DNG] Refracta no-dbus experiment
Refracta no-dbus build (experiment) The subject of running without dbus comes up from time to time in various places. I decided to try it and see how far I could get. I started with a debootstrap install of devuan ascii, pinned dbus to a priority of -1, and proceeded to make the same changes as I do to make Refracta live isos. Normally, the Refracta isos use xfce, but that's not possible without dbus. I was surprised to see how much did install without dbus. So I thought I'd share it. This build uses openbox, lxpanel, lxterminal and spacefm. Maybe someone will want to use it. Maybe it will inspire someone else to do something better. Feeback is welcome. http://distro.ibiblio.org/refracta/files/experimental/refracta9_nodbus_amd64-20180612_0156.iso sha256sum: 633634c3ac2beb06252b29bc78b3135f5f5ded473a72f42e5dc6c17d326d1f17 Login/Password: user/user root/root No display manager. Run 'startx' to get a desktop. # These can be installed without dbus and without libsystemd0 rsync bash-completion busybox kbd locales firmware-linux-free deborphan unzip lvm2 cryptsetup sshfs \ hwinfo alsa-utils moc pppoeconf pppconfig pppoe ntfs-3g dosfstools curl \ live-boot live-config live-boot-initramfs-tools live-config-sysvinit squashfs-tools xorriso pmount pv \ syslinux syslinux-common syslinux-utils isolinux xz-utils gdisk parted hexedit iftop smartmontools lm-sensors \ hdparm testdisk fdupes irssi iptraf ethtool scrot wipe mlocate wireless-tools wpasupplicant \ # get libdbus-1-3 here gddrescue screen feh hddtemp p7zip-full partimage pm-utils sysv-rc-conf tree wodim htop bzip2 whois \ lsb-release file setnet net-tools cifs-utils mdadm arp-scan \ dialog live-boot-doc live-config-doc refractainstaller-base refractasnapshot-base \ btrfs-tools btrfs-progs pciutils psmisc rename tcpd usbutils uuid-runtime dnsutils \ eject telnet usbutils util-linux-locales vrms mutt sudo # These were installed after allowing libsystemd0 xorg openbox spacefm lxterminal lxpanel obconf lxappearance lxappearance-obconf lxrandr \ linux-headers-4.9.0-6-amd64 build-essential xserver-xorg-legacy xserver-xephyr xterm aptitude \ icewm xarchiver leafpad links2 xpdf mpv yad ***grub-of-your-choice*** x11vnc xtightvncviewer grsync bleachbit meld asunder winff \ mplayer ffmpeg volumeicon-alsa tilda geeqie dkms transmission-gtk gftp \ xserver-xorg-video-intel xscreensaver xinput libnotify-bin hexchat \ abiword hardinfo gdmap gimp geany firejail firefox-esr deadbeef http://sourceforge.net/projects/deadbeef/files/debian/deadbeef-static_0.7.2-2_amd64.deb/download firemenu https://sourceforge.net/projects/refracta/files/Extras/firemenu-1.2.deb refracta2usb https://sourceforge.net/projects/refracta/files/tools/refracta2usb-2.3.6.deb These will NOT install. (and probably a lot more that I didn't try.) audacious xfburn wicd connman libpam-elogind synaptic gdebi fsmithred ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] More Sheiße on the way
Alessandro Selli wrote: > I read that, as this is a CPU hardware bug, it affects all OSes: > > https://www.bleepingcomputer.com/news/security/new-lazy-fp-state-restore-vulnerability-affects-all-intel-core-cpus/ > > According to Intel this new vulnerability affects all Intel Intel > Core-based microprocessors and is a bug in the actual CPU, so it does > not matter what operating system the user is running. It could be > Windows, Linux, BSD, or any other operating running an an Intel > Core-based CPU and using "Lazy FPU context switching". Yes, it affects any OS - but only (AIUI) if the do this “lazy FPU context switch”. From the report I read (on TheRegister https://www.theregister.co.uk/2018/06/13/intel_lazy_fpu_state_security_flaw/ ) recent Linux kernels don’t use that and so aren’t vulnerable. > Modern versions of Linux – from kernel version 4.9, released in 2016, and > later ... are not affected by this flaw ... The Linux kernel team is > back-porting mitigations to pre-4.9 kernels It then goes on to say that > The fix is to employ a mechanism called eager FPU state restore, which modern > Linux, Windows and other kernels use. These mitigations do not carry a > performance hit – in fact, eager state switching can increase performance. So it sounds like this one isn’t the big issue it initially looks like. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] elogind and startx
On 06/15/2018 02:24 PM, Haines Brown wrote: > I installed ascii 2.0.0 without desktop. Usually I have installed the > packages associated with the xserver: xserver-xorg-video-dummy, > xserver-xorg-input-void, xserver-xorg-core, xinit and x11-xserver-utils. > > But in ascii release notes it says, "In Devuan 2.0 ASCII it is > sufficient to install 'elogind' and 'libpam-elogind', and then use > either 'startx' or 'xinit' as usual from a regular user account." > Is this implying that I do not need to install the xserver files listed > above? Or must they be installed, and the elogind simply relocates the X > server log? > Adding elogind makes it so you can run startx without installing xserver-xorg-legacy and adding root rights to Xwrapper.config. You will be running all parts of X as user. You do still need to install those parts. fsmithred ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] More Sheiße on the way
Alessandro Selli [2018-06-15 16:30]: > "New Lazy FP State Restore Vulnerability Affects All Intel Core CPUs" According to the DragonFlyBSD list, this doesn't affect Linux. -- Hilsen Harald ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] devuan from scratch?
Eric Pozharski [2018-06-15 11:55]: > p.s. Also. For those ignorant out there. The Linux-from-Scratch has > gone The Dark Side. Close that book. Not really: "In coordination with this release, a new version of LFS using the systemd package is also being released. This package implements the newer systemd style of system initialization and control and is consistent with LFS in most packages." So, it's a choice, just like with Gentoo. -- Hilsen Harald ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
[DNG] elogind and startx
I installed ascii 2.0.0 without desktop. Usually I have installed the packages associated with the xserver: xserver-xorg-video-dummy, xserver-xorg-input-void, xserver-xorg-core, xinit and x11-xserver-utils. But in ascii release notes it says, "In Devuan 2.0 ASCII it is sufficient to install 'elogind' and 'libpam-elogind', and then use either 'startx' or 'xinit' as usual from a regular user account." Is this implying that I do not need to install the xserver files listed above? Or must they be installed, and the elogind simply relocates the X server log? Haines Brown ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] devuan from scratch?
On Thu, Jun 14, 2018 at 08:27:26AM +0200, Jaromil wrote: *SKIP* > it's interesting because it provides an alternative "new" way to > bootstrap a new apt based system in addition to the classical > debootstrap method. I also noticed the instructions have no mention of > systemd, wonder where that will sneak in later... I've made myself to believe that some alternative facts (I guess) must be pointed out. Linux-from-scratch builds from source. Debian-from-scratch builds from respository. Quite a difference, eh? As of svchost -- it *will* come with dependencies, don't you worry. What makes me wonder, this, so to speak, manual is dated as 20180610. Yet it doesn't mention anything for sources.list or versions of any .deb. What seems sneaky to me, because... I dare to state my IMCO (as in 'competent') -- it's impossible to build debian-stable. The only thing debian-* that can be built is debian-unstable (not debian-testing). But to achieve this they already need debian-unstable. What LFS is not. So, place apropriate *.deb s all over incomplete LFS, to make it debian-unstable. Then build dpkg and apt. Why would anyone do that? Just unwind apropriate .deb s. *CUT* p.s. Also. For those ignorant out there. The Linux-from-Scratch has gone The Dark Side. Close that book. -- Torvalds' goal for Linux is very simple: World Domination Stallman's goal for GNU is even simpler: Freedom ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
[DNG] More Sheiße on the way
https://www.bleepingcomputer.com/news/security/new-lazy-fp-state-restore-vulnerability-affects-all-intel-core-cpus/ "New Lazy FP State Restore Vulnerability Affects All Intel Core CPUs" Similar but different from Meltdown. https://slashdot.org/submission/8290070/intel-x86-processors-affected-by-lazy-fpu-saverestore-security-flaw "Reportedly, AMD processors are not affected with this vulnerability" Alessandro ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Ascii LVM Encrypted Issue
Am Freitag 15 Juni 2018 schrieb KatolaZ: > Hi Stefan, > > if you choose to install only the necessary drivers, you will be > guaranteed to have only the drivers needed for the hardware present in > the system at install time. That's the reason why you don't get USB > modules in the initrd (which is exactly what I would expect to happen > if I asked to include only the necessary drivers, TBH). > > I guess there is genuinely no bug here, just a feature ;) I am sure > you get the point. > > HND > > KatolaZ > Yeah, certainly, the installer did as advertised ;-) as I surprisingly found out. Regards, Stefan ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Ascii LVM Encrypted Issue
Le 15/06/2018 à 02:05, Michael McConnell a écrit : Thanks for the reply, but at no point during the install do I see an option to install the “necessary” drivers? I just ran through the install once more and still don’t see any prompt or indication asking about supporting USB drivers. I did the install using a USB keyboard on both the VM instance and the dedicated hardware. AFAIR, this choice has been present in Debian installer for at least a decade and I confirm it's still present in Devuan-ASCII - I don't remember in which step though. The choice is rather radical and the option to only install the drivers for devices present at install time is mostly meant for industrial SBCs and hand-held devices which don't accept any hot-pluggable device. It is also interesting for security reasons. It would be nice to have an intermediate choice including all common hot-pluggable devices, but, there isn't. It might help to install from a USB stick, or just pug one in during install. Didier ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng