* Hydragyrum upgraded to ceres when chimaera hit freeze | 01:39 | |
carlos | hi all what network manager should i install in chimaera? just uninstalled wicd | 02:47 |
---|---|---|
Hydragyrum | I use networkmanager mostly | 02:50 |
debdog | carlos: there are connman, cmst and the one Hydragyrum mentioned | 02:52 |
mystified | hey guys, what is the way to install testing , with either open rc or runit | 03:03 |
mystified | is there a link to specific iso or is it configured by the installer as a option | 03:04 |
rrq | nope. there nearest you can come is the daedalus preview and then upgrade to ceres | 03:05 |
mystified | thanks rrq: | 03:10 |
mystified | rrq: where is the link for daedalus preview | 03:13 |
rrq | pick a donwload mirror | 03:14 |
mystified | all good, thanks | 03:15 |
mystified | all good, thanks, i've not used devuan since version 1. | 03:16 |
rrq | we all suffer in our own ways ;) | 03:16 |
mystified | it's matured greatly since then, time to give it another go. | 03:16 |
rrq | welcome back. | 03:17 |
mystified | will the installer allow me to select the init | 03:17 |
mystified | thanks | 03:17 |
rrq | it will, but only in a limited set of three options. | 03:18 |
mystified | hey guys, thansysv, openrc & runitwhat is the way to install testing , with either open rc or runitI've not installed | 03:18 |
mystified | lol. christ, what garbage | 03:18 |
mystified | sys v, runit & openrc | 03:19 |
Hydragyrum | use whichever init you intend to use after updating | 03:29 |
Hydragyrum | any of them will work, and I've switched inits without issue on ceres, but it's best to pick the one you want to use | 03:30 |
carlos | hey all, what can i use instead of wicd in chiamera? | 05:06 |
Hydragyrum | connman, cmst, and networkmanager | 05:07 |
Hydragyrum | or if you'd prefer, you can just use ip + wpa_supplicant + iw{,config} | 05:07 |
carlos | thanks | 05:08 |
carlos | Hydragyrum i was looking for some gui network manager like wicd | 05:09 |
carlos | *i am looking for... | 05:09 |
Hydragyrum | networkmanager has gui | 05:09 |
Hydragyrum | (nm-tray package has a tray applet that's DE/WM independent as long as there's a systray) | 05:10 |
DPA | My pxe boot environment with a readonly pxe root started failing to boot today. It uses openrc. The root cause was the checkroot-bootclean init script failing, which was required by the mountall.sh init script, and because of that some file systems weren't mounted. | 10:19 |
DPA | It used to work, I'm not sure what changed. There is a message about mounting /tmp as tmpfs because it's readonly, but I don't know where that's done, and it seams to happen too late. | 10:19 |
DPA | For now, I just moved checkroot-bootclean from the "Required-Start:" in mountall.sh to "Should-Start:", to work around the problem. | 10:19 |
DPA | *nfs root | 10:19 |
HumanG33k | Tenkawa: back for report | 12:58 |
HumanG33k | my other server face to the same issue | 12:58 |
HumanG33k | libapache2-mod-php7.3 install as local libapache2-mod-php7.4 not install and system is upgrade to chimaera | 12:59 |
jk000 | hello, devuan-people! I am new in this init-freedom-ecosystem :) and want to ask a rookie question: how do I shutdown my devuan system from the commandline? ^^ | 13:58 |
ShorTie | same as always, shutdown now -h | 13:59 |
debdog | or just halt | 14:02 |
debdog | I totally forgot the command to change runlevel. | 14:06 |
debdog | init 0? | 14:07 |
tr-amarsh04 | init 0 works for me with sysvinit on Devuan | 14:16 |
ShorTie | defualt runlevel is 2 | 14:16 |
jk000 | when I type shut+tab-autocomlete I get no options o.O | 14:17 |
jk000 | same with halt | 14:17 |
sadsnork | sudo shut+tab? | 14:18 |
debdog | need to be root | 14:18 |
debdog | there's the possibility to "chmod +s /sbin/halt", on you own risk, of course | 14:19 |
ShorTie | wouldn't be doing that | 14:20 |
jk000 | yes, I was root # shut+tab | 14:20 |
jk000 | ...oh may be I habe a PATH-problem | 14:20 |
debdog | root via su? | 14:21 |
jk000 | sbin has halt and shutdown | 14:21 |
debdog | then "su -", IIRC | 14:21 |
jk000 | jup, no sbin in PATH for me :/ | 14:22 |
debdog | root via su? | 14:22 |
debdog | then "su -", IIRC | 14:22 |
jk000 | yes, root via su | 14:22 |
jk000 | I'll try su - | 14:23 |
debdog | the ASCII release notes have more on that topic | 14:23 |
tr-amarsh04 | there was a change in the behaviour of su in an update to util-linux | 14:23 |
tr-amarsh04 | I kept the old behaviour | 14:23 |
jk000 | oh! intereesting. thanx! | 14:24 |
jk000 | so, yes "$ su # echo $PATH" has no sbin and "$ su - # echo $PATH" does | 14:25 |
tr-amarsh04 | so in /etc/default/su I have: | 14:28 |
tr-amarsh04 | ALWAYS_SET_PATH=yes | 14:28 |
jk000 | ok. a new file, right? | 14:31 |
tr-amarsh04 | man su should answer your questions | 14:36 |
tr-amarsh04 | man su should answer your questions | 14:36 |
tr-amarsh04 | believe it or not I hit the issued in August 2018 | 14:36 |
ShorTie | shut+tab is a bash completion thing, most likely not setup for su/root | 14:54 |
ShorTie | does sudo shut+tab work ?? | 14:55 |
ShorTie | there are more option then just -h for shutdown too | 14:56 |
sharpie | Hey, I upgraded from beowulf to chimaera, thought everything went ok, but uname -a still says I'm on 4.19.0-9-amd64 | 15:12 |
* ShorTie wonders, and rebooted ?? | 15:13 | |
sharpie | hmm, no thought linux could do that in place, don't want to lose my 477 days uptime :D | 15:14 |
sharpie | but maybe i'll have to | 15:14 |
ShorTie | gotta reboot to change a kernel | 15:20 |
hyrcanus | wonder if some OS can persist processes between (compatible) kernels | 15:24 |
hyrcanus | as long as they don't require on complex state in hardware like a gpu | 15:24 |
hyrcanus | probably ibm's z-series could do it | 15:24 |
carlos | cant write anything in my unbound.conf file (its empty) after upgrading to chimaera because reports error and unbound service doesnt reload. | 15:43 |
carlos | also network manager completly ignores unbound | 15:45 |
carlos | also network manager completly ignores unbound with chimaera | 15:46 |
hyrcanus | idk either of those apps | 18:04 |
golinux | hyrcanus has much to learn | 18:13 |
hyrcanus | who doesn't | 18:13 |
hyrcanus | i know *of* them | 18:14 |
golinux | Even I got unbound working with a little help. | 18:17 |
hyrcanus | i'm not sure we have services in devuan | 18:18 |
hyrcanus | i have some daemons though | 18:18 |
rwp | Isn't "service" just a generic word describing any... what's the word, oh yes, service provided by a system usually by a system daemon? | 18:24 |
rwp | I still use bind most places but I'll tip my hat to unbound which handles DNSSEC more easily on my mobile laptop. | 18:26 |
rwp | But sadly the coming loss of wicd has slowed my upgrade on my laptop because I am not quite ready to make the change yet. I need to work through it more first. | 18:27 |
hyrcanus | what's wicd for. | 18:28 |
Hurgotron | no more wicd? What replaced it? | 18:28 |
Hurgotron | hyrcanus: desktop network management | 18:28 |
hyrcanus | i use /etc/network/interfaces | 18:30 |
morruth | hyrcanus: for wifi too? | 18:35 |
hyrcanus | nope | 18:35 |
Hurgotron | For wifi and moving between networks, wicd is quite nice. | 18:35 |
hyrcanus | 'ceni' is a nice little utility if i need to setup a wifi | 18:36 |
hyrcanus | but i don't now | 18:36 |
Minnesotags | I got a problem with libvirt now that I've upgraded from Beowulf to Chimeara.... It looks like it's a Devuan package issue? | 18:37 |
hyrcanus | what version you got there btw | 18:38 |
hyrcanus | apt-cache show libvirt0 |grep Version | 18:39 |
Minnesotags | Version: 7.0.0-3+devuan3 | 18:54 |
Minnesotags | apt-get install libvirt-daemon-system | 18:54 |
Minnesotags | gives | 18:54 |
Minnesotags | The following packages have unmet dependencies: | 18:55 |
Minnesotags | qemu-system-common : Breaks: libvirt-daemon (< 7.2.0-1) but 7.0.0-3+devuan3 is to be installed | 18:55 |
Minnesotags | E: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages. | 18:55 |
hyrcanus | i have 7.6.0-1 here so i don't have that issue | 18:59 |
Minnesotags | Hmm. How can I upgrade libvirt0 to that? | 19:10 |
hyrcanus | it's in devuan ceres | 19:11 |
Minnesotags | Ugh. | 19:12 |
Minnesotags | But seriously.... | 19:23 |
gnarface | honestly you don't need it, and i highly doubt the people who make it are serious about getting work done | 19:26 |
Minnesotags | Confused.... What do you mean? | 19:26 |
gnarface | contrary to the information fed to the public by virt-manager, you don't actually need virt-manager to use qemu, and in most cases due to quality control issues or glaringly missing features it is more trouble than it is worth unless your sole goal is to get a job maintaining virt-manager | 19:28 |
gnarface | furthermore, when a piece of software is still exhibiting the same integration problems en-masse today as it was almost a decade ago when i discarded it initially, i think that's sufficient cause to classify it as malware | 19:29 |
Minnesotags | Shots fired. | 19:29 |
gnarface | i'm an angry old man and that's just my opinion | 19:30 |
gnarface | but i assure you you can start qemu by command-line and if you're only running one guest, it's almost certainly easier | 19:30 |
Minnesotags | I'm running several guests on that server. | 19:31 |
gnarface | sure, you have to figure out alot of qemu command-line details but i guarantee you're still gonna have to figure those out to debug virt-manager because the people who made it clearly don't use it for work | 19:31 |
trmin | Hi, back to the freeze problem (2 days ago). | 19:31 |
Minnesotags | So what is an alternative method to console in to qemu virtual machine? | 19:32 |
Minnesotags | NOT ssh | 19:33 |
HumanG33k | gnarface: as an old man, can you explain me why merge /bin and /usr/bin is wrong ? Do you know why debian choose to merge it ? (real question) | 19:33 |
gnarface | trmin: i didn't read the german parts, statistically freezes are usually a driver issue, particularly video drivers... is that the type of freeze you're talking about? like a i/o lockup where nothing responds? | 19:33 |
gnarface | HumanG33k: dunno, looks like the type of change someone inexperienced would make if a manager put them in charge on day one and said "put your mark on this in as cheap and destructive way possible" | 19:34 |
gnarface | HumanG33k: just fits the pattern of the new guard just throwing out anything they never used themselves and are too inexperienced to understand the purpose of | 19:35 |
gnarface | HumanG33k: and it's primarily wrong just because it's a needless change that breaks things for little value other than the sophomoric elation of having technically a few less characters to remember out of literally billions | 19:36 |
HumanG33k | i read some stuff about index (to access), userland load | 19:37 |
trmin | I got to get used to not to use the Enter key ... the computer is still running while not responding to input an freezing displays. I checked it by recording from a tv card. Next I will find out by a screen clock, at and after what time I/O freezes. Meanwhile I updated nvidia driver with no effect and today I removed every screensaver. | 19:37 |
gnarface | trmin: which nvidia driver version? one from the repos? | 19:38 |
gnarface | HumanG33k: yea if you have a lot of stuff installed it's also gonna slow down access to /usr/bin a bit | 19:39 |
gnarface | HumanG33k: that's hardly the worst part about it though, the worst part about it is just that these aren't the types of changes they should be spending their time on | 19:39 |
trmin | gnarface: exactly, complete I/O-lockup except Magic_SysRQ. | 19:40 |
gnarface | trmin: did you try the one from backports or are you on unstab?E | 19:40 |
gnarface | *unstable | 19:40 |
HumanG33k | in your opinion what things need more time focus ? | 19:41 |
gnarface | HumanG33k: this shit trmin is having trouble with is a big problem for everyone and it also is direct evidence of illegal collusion, so you know... maybe something like that? | 19:42 |
trmin | Chimaera ist stable now. The former nvidia was from unstable (chimaera) and updated from current nvidia installation. | 19:42 |
gnarface | trmin: hmmm. any programs running or is it just idle? i've had a lot of problems with nvidia drivers in the past but with screensavers usually only the opengl ones; you should also consider the compositor (if your window manager uses one, but most popular ones do now these days) suspect as well as firefox, wine and mplayer | 19:43 |
gnarface | trmin: those are all the usual suspects for me. usually it was some combination of one or two of those being on while dpms engages as the screen goes idle | 19:44 |
gnarface | trmin: oh, so yea that's another thing to test; disable dpms (xset dpms 0 0 0) | 19:44 |
gnarface | trmin: but then without dpms you risk screen burn-in, so you're gonna want to find a safe old screensaver that just pixel blits from algorithm rather than anything fancy | 19:45 |
gnarface | trmin: (or just unplug/power off the display if that's possible instead) | 19:45 |
gnarface | trmin: additionally, having a ssh terminal connected from a remote machine might help figure things out because sometimes when this happens it's only the local console that locks up | 19:46 |
gnarface | trmin: (if you can ctrl+alt+f2 on the local console though quickly after it locks up, sometimes there is a short window of opportunity where you can kill Xorg and relaunch it to recover without reboot) | 19:47 |
gnarface | trmin: sorry none of that is a direct fix, really just more tests to try | 19:47 |
trmin | I was considering dpms, but do it as next step - otherwise I won't know what the problem was if it's gone. But once the freeze alos happend while I was working. | 19:49 |
gnarface | HumanG33k: definitely the things they should not be doing are the things they currently have been doing that constitute breaking expected behavior of the system and making sweeping alterations to what has basically become the de-facto industry FHS and tools | 19:49 |
gnarface | trmin: sometimes a minor shuffle of driver or library versions can avoid weird issues, the thing is i don't think NVidia even knows what's going on with this thing anymore | 19:50 |
hyrcanus | trmin: just to confirm, you are running Xorg? | 19:51 |
gnarface | trmin: but dpms has been an Xorg lockup issue for me on other hardware too, and many of the freezes i've had with NVidia hardware seemed to coincide with dpms engaging while two or more of either firefox, wine or mplayer were going at once | 19:51 |
gnarface | trmin: oh, yea i assumed this was Xorg in question... wayland is a whole other matter entirely and i haven't tested it much at all | 19:52 |
gnarface | trmin: (in mplayer the achilles heel seemed to be related to vdpau, so if you have some other program running that uses the nvidia driver's vdpau implementation that could be suspect as well) | 19:53 |
gnarface | but opengl stuff is the primary suspect since firefox and wine were always using that | 19:53 |
gnarface | and the opengl screensavers from xscreensaver can *reliably* lock up the nvidia driver on all hardware | 19:53 |
trmin | I'm running an ASCII on the same computer with a nvidia driver version between the former and present one in chimaera. No problems. Okay, that does'nt automatically mean that it's not the nvidia driver. | 19:54 |
gnarface | trmin: hmmm, no it doesn't but it also doesn't mean you might not be able to avoid the issue by changing the version of something else in the install, possibly Xorg | 19:55 |
gnarface | often the outcome of extensive debugging of this stuff was an inexplicable bug that Xorg, NVidia, and Wine staff all blamed the others for and nobody would budge on or admit any guilt themselves | 19:56 |
gnarface | and just as often it would evaporate for months at a time with some minor supposedly unrelated version increment of some other part of the system | 19:56 |
hyrcanus | basic troubleshooting; swap out components | 19:57 |
gnarface | trmin: oh, check the heat too, but since it is stable with the other install that's probably not the issue | 20:00 |
gnarface | trmin: but i did conspicuously catch a couple OEMs over the years building planned obsolescence in with thermal compound under the heatsincs that actually expires after 6 months and starts insulating | 20:01 |
gnarface | trmin: (if it's black in color be highly skeptical) | 20:01 |
trmin | gnarface: No one likes to be guilty. I'll see, what the present changes will lead to tonight. Enough possibilities to test next, if nothing changed. By the way: One result of the change to new nvidia driver was a .xsession-errors size of 13.2 GB within an hour, so I had to redirect the messages to /dev/nul. Many thanks for now. I'll be back. | 20:03 |
gnarface | trmin: oh wow that's a sign of something really badly wrong; i have to imagine someone else must have seen that and reported it somewhere... maybe just to the nvidia.com forums or maybe one of the distro bug trackers if you're lucky. | 20:05 |
Minnesotags | Ok. My main problem is this: | 20:05 |
Minnesotags | error: Failed to start domain 'cwmm' | 20:05 |
Minnesotags | error: internal error: cannot load AppArmor profile 'libvirt | 20:05 |
gnarface | Minnesotags: this is a recent chimaera install or upgrade? | 20:05 |
Minnesotags | Upgrade | 20:05 |
Minnesotags | Upgrade from Beowulf | 20:06 |
gnarface | Minnesotags: someone else ran into an issue earlier with missing a bunch of apparmor config files, but i think just reinstalling the packages they came from fixed it, i don't recall if it was determined where the install had went off the rails initially though | 20:06 |
Minnesotags | /usr/bin# virsh start cwmm | 20:06 |
Minnesotags | error: Failed to start domain 'cwmm' | 20:06 |
Minnesotags | error: internal error: cannot load AppArmor profile 'libvirt-3fe994f9-caff-4340-9cdf-4d2ca435358b' | 20:06 |
gnarface | yea i gotta assume it's just missing files from /etc/apparmor/ | 20:07 |
Minnesotags | Hmm | 20:07 |
Minnesotags | So uninstall and reinstall apparmor? | 20:07 |
gnarface | or /etc/apparmor.d/ | 20:07 |
gnarface | no i would say virsh | 20:07 |
gnarface | or libvirt | 20:07 |
gnarface | the apparmor profile it's complaining about is libvirt's | 20:07 |
Minnesotags | Aha copy that | 20:07 |
gnarface | dpkg -S /etc/apparmor.d | 20:08 |
gnarface | this should show you all the packages that put stuff in there, i think | 20:08 |
hyrcanus | oh wow thank you gnarface | 20:09 |
Minnesotags | ntp, telepathy-mission-control-5, libvirt-daemon-system, evince, man-db, apparmor, libreoffice-common: /etc/apparmor.d | 20:09 |
gnarface | or all the packages that were supposed to have anyway, the issue in question was that some of them apparently didn't and last i heard we still don't know why yet but reinstalling them should fix it | 20:09 |
Minnesotags | Got it! | 20:10 |
gnarface | apt-get --reinstall install [package name] | 20:10 |
Minnesotags | Hmm. Reinstalling isn't working to fix that.... | 20:59 |
_ds_ | Missing conffiles? dpkg --force-confmiss $PACKAGE.deb | 21:12 |
gnarface | Minnesotags: yea maybe you have to do that or uninstall/reinstall first... weird though, were any changed by hand? | 21:32 |
gnarface | they'd usually be replaced unless there's a collision, maybe there's another bug involved here... | 21:32 |
XJ8 | Congratulations on 4.0 release matching debian 11 | 22:15 |
DPA | I've now completed the upgrade of vm & container server. I had some trouble initially getting libvirt-qemu working again. I had to put "namespaces = []" in /etc/libvirt/qemu.conf . | 22:16 |
DPA | Without that, it created a mount namespace and mounted /dev as tmpfs, presumably to limit access to device files. But it didn't create most of the needed device files. Took me hours to figure all this out... | 22:16 |
DPA | * of my | 22:17 |
XJ8 | I'm also having some difficulties using the package view search online. It returns no results. | 22:17 |
gnarface | XJ8: pkginfo.devuan.org? | 22:18 |
gnarface | XJ8: put .* at the end of your search string | 22:18 |
XJ8 | Yeah. I figured out what was wrong. It's very picky and will take the results literally unless I add that to the string | 22:19 |
gnarface | DPA: just curious, were you running the guest without udev? | 22:19 |
DPA | No, I use eudev. /dev was mounted. I also verified it was fine in /proc/$(pidof libvirtd)/cwd/dev/ | 22:21 |
DPA | I replaced /usr/bin/kvm with a script to check out the environment it got from libvirt, and without "namespaces = []" or with "namespaces = ["mount"]" it was skrewed up like this. | 22:23 |
gnarface | good to know | 22:23 |
XJ8 | snapd and other snap libraries is in devuan right because I like the convenience of using it for select packages? | 22:29 |
gnarface | i don't think it is actually but i'm not sure | 22:30 |
gnarface | yea, sorry, banned: http://packages.devuan.org/bannedpackages.txt | 22:31 |
XJ8 | Oh well. | 22:31 |
DPA | It uses systemd's container / namespace abstractions, so it can't work on a system without systemd. | 22:32 |
XJ8 | That makes sense | 22:32 |
XJ8 | I'll give google the benefit of the doubt that this isn't malicious, but searching for devuan gives me nothing but debian results despite the very clearly different name. | 22:33 |
gnarface | yea well they have plausible deniability because nobody has proof that devuan is not a common misspelling of debian | 22:35 |
gnarface | they should at least provide you with a link to the right search results | 22:35 |
gnarface | but most of the relevant results are probably gonna be on the forum | 22:36 |
gnarface | so just searching it directly might be a shortcut | 22:36 |
gnarface | in the case of init stuff, old debian wheezy documentation should still mostly match up correctly though | 22:37 |
XJ8 | gnarface I was talking about using google search directly to find out information about devuan. | 22:41 |
hagbard | How do i report a bug to debian instead of devuan with devuans reportbug? | 22:54 |
gnarface | you probably shouldn't do that | 22:56 |
gnarface | i don't think they'll like that at all | 22:56 |
hagbard | Kexec wasn't upgradable to the version in daedalus, due to a bug in its postinst skript. | 22:57 |
hagbard | I reported the the bug to devuan at first. | 22:57 |
hagbard | Then I got a Mail back, that told me to report to debian instead. | 22:57 |
gnarface | hmmm | 22:58 |
hagbard | But i'm sure they won't like it, too. | 22:58 |
hyrcanus | if the package comes from the debian maintainer, that's the proper place to report | 22:58 |
gnarface | well then you should do it, but i just think that forging a reportbug submission is gonna be met with hostility | 22:58 |
gnarface | but if it's something that doesn't affect debian i expect them to still not care | 22:59 |
gnarface | actually i expect them to break it worse on purpose in spite, please try to be careful | 23:00 |
hagbard | Mark Hindley told me to to report to debian. But the bug is about how the postinst script invokes the /etc/init.d/ script. | 23:00 |
hagbard | And I not to sure either, that debian is overly keen to care about that. | 23:01 |
hagbard | So how do I get the fix in? | 23:01 |
gnarface | diplomacy? | 23:02 |
gnarface | file the bug report just don't use reportbug to do it | 23:02 |
gnarface | i could be wrong about this advice so do what you want but my assumption is that if you can get it to submit a bug report to debian it'll still expose you as a devuan user by http headers or something and they'll probably be mad if automatic filtering doesn't just skip it entirely | 23:03 |
hagbard | yeah, i'll just write a mail | 23:04 |
gnarface | and either way the first question someone is gonna have when they go to fix it is probably gonna be "wait, how did you run into this bug exactly?" | 23:04 |
gnarface | so yea, diplomacy | 23:04 |
gnarface | probably would help a lot if the fix is only one or two lines changed and you can actually provide the literal change needed so nobody has to think more than they want to about it | 23:05 |
hagbard | it's just a few autogenerated lines removed | 23:06 |
gnarface | all the better | 23:17 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!