[00:00] which shells were you able to reproduce the problem in? [00:00] sh and bash [00:00] oerheks: the image is definitely mounted as read-write [00:02] five623468263: What are you using to create the chroots? [00:03] What do you mean? I mount the image with kpartx, then I mount /proc/, /dev/ and /sys/ inside, then I simply run 'chroot /mnt/chroot' [00:04] And again, this exact same process (which is a script) worked flawlessly in all previous versions of Ubuntu [00:04] The only thing that has changed is my host Ubuntu installation, which is now 21.10 === Chex_ is now known as Chex [00:07] and /dev/pts [00:07] Yeah, I do that too, just forgot to mention, sorry [00:07] oh oke [00:11] I just tried the *exact* same workflow with Ubuntu 21.04 as a host, and it works flawlessly [00:11] So the issue is definitely in the host's update... but I have no idea why [00:14] five623468263: how about zsh or ksh or csh? [00:14] five623468263: have you been able to reproduce this with those shells? [00:15] I will try right away. Should I do zsh on host or chroot or both? [00:16] I think within the chroot [00:18] five623468263: try also installing the procenv package, and run procenv -A both via chroot /whatever procenv -A and chroot /whatever -- and then running procenv -A from within that shell [00:22] I can't seem to be able to use apt from within chroot, even with the "apt apt update" workaround [00:22] I get: Unknown error executing apt-key [00:22] ouch [00:23] I guess that answers that question :) [00:23] Seems like the whole environment is very broken, but why :/ [00:23] five623468263: https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1947860 [00:23] Launchpad bug 1947860 in qemu (Ubuntu) "qemu-arm-static swallows the program name, and shifts all arguments left" [Critical, Fix Committed] [00:24] five623468263: man I'm glad someone else filed this bug report already :) [00:24] Ahhh wow, so it's a proper bug [00:25] bummer, the packages are still building :( [00:26] Since I haven't ever been this close to bleeding edge fixes, how long will it take until this fix will appear in my host's apt update? [00:28] five623468263: excellent question, I can't recall if there's a "seven days in -proposed" thing for stable releases or if that's only for the development release [00:29] I see. Well, I am in no rush to use Ubuntu 21.10, so I will patiently wait :) [00:29] Thanks so much for all the help to everybody here, especially for managing to dig out the existing report sarnold and ubottu [00:29] I thought I was going insane... but I guess I just hit a very neat bug === five623468263 is now known as five6 [00:30] five6: yeah, it's *crazy* and I really dreaded trying to find the guilty component. :) [00:31] Tbh I suspected it would be qemu, there was just nothing else left, but I know nothing about how it works so I couldn't even attempt to troubleshoot it [00:32] yeah, qemu was my first choice of suspect, but linux wouldn't have surprised me either. I just wanted to get the easy cases of shells covered first :) [00:32] Makes sense. Well, I shall be going off now. Thanks for all the help again! [00:59] Hello i'm trying to setup auto fs [01:01] and what autofs howto are you following? === Starmina_ is now known as Starmina [01:13] hello i'm trying to setup autofs [01:13] C0nundrum: "20:01 < oerheks> and what autofs howto are you following? [01:13] " [01:15] and what ubuntu version and so on [01:17] * oerheks pings timeout and walks outside for a smoke break [01:36] I added it as "+/mnt/auto /etc/auto.master.d/auto.nfs" [01:37] Is that the correct format ? === genii is now known as genii-core [01:41] again, what howto are you following, for what ubuntu version? [01:43] let me guess; not ubuntu :-D [02:05] "Ubuntu 16.04.7 LTS" [02:06] !16.04 | C0nundrum [02:06] C0nundrum: Ubuntu 16.04 LTS (Xenial Xerus) was the 24th release of Ubuntu. !End-of-life was April 30th, 2021. Paid support (ESM) is available. See also !esm, !eol, !eolupgrade === bloodninja123 is now known as bloodninja [02:11] People have certain versions for certain reasons i'd asume [02:12] what does that have to do with the autofs package syntax... [02:12] your getting lost in the details [02:12] wait, you are the one not sharing details.. [02:12] I don't understand lol [02:12] you asked for my os and i gave it [02:13] and yes, we people have reasons not to support EOL versions [02:13] idk what else you asked for ? [02:13] my hardware [02:13] can only run [02:13] 16.04 [02:13] I'm not trowing it out because software went outdated.... === M4he is now known as mahe [02:20] is it possible to remove or change the the transition color between gdm and login [02:20] ? [02:20] its currently purple [02:24] yukiup, yes, it is a kind of a hack, editting .css https://www.reddit.com/r/Ubuntu/comments/clgj93/is_there_a_way_to_change_default_purple_bringle/ [02:25] not the login background... [02:25] the color inbetween login and desktop [02:25] it shows for like 3 seconds [02:27] hmm, no idea there [02:27] yah, its annoying and dosen't need to be there [02:34] looking at the css file, it seems quite likely that adjusting the #2c001e value in #lockDialogGroup would do just that [02:35] hi everyone [02:35] got. some. urhent, issue [02:35] ! [02:35] tomreyn, that was my 1st thought too, but he mentions the transition between [02:36] im, VIA [02:36] yverx, ask wait and see [02:37] all. of a sudden. the. taskbae [02:37] taskbar** is. gone [02:38] managed to do. some. old. 'restore [02:39] but. im, lackimg. everything. i ned. [02:40] what taskbar on what ubuntu version? [02:41] i. have no. internet connectivitu.... i. disabled. it. before. logoff.. [02:41] im. om 20.04.xfce [02:42] allso.sruy bout my writing [02:43] had. a. phome. reset. n. it. see [02:44] this answer might be your help too https://itectec.com/ubuntu/ubuntu-resetting-xfce-panels-to-default-settings/ [02:44] xfce4-panel --quit ; pkill xfconfd ; rm -rf ~/.config/xfce4/panel ~/.config/xfce4/xfconf/xfce-perchannel-xml/xfce4-panel.xml ; xfce4-panel; [02:45] meedto. restore. all. snnoying. crap. lile. aicomplete//automess [02:45] xfce4-panel -r # might work too [02:48] yverx: Later releases ' mv ~/.config/xfce4/panel/whiskermenu-1.rc ~/Desktop ' ; ' xfce4-panel --restart ' . [02:49] think i ahould follow, tje commans u lisye oearheks? [02:50] yverx: ' ls -al .config/xfce4/panel/ ' shows >> whiskermenu-1.rc ? [02:54] i get only wierd errors [02:55] i assume ineed sudo? [02:55] no need for sudo [02:56] and how about a simple reboot? [02:56] yverx: No sudo ! try as ' ls -al ~/.config/xfce4/panel/whiskermenu-1.rc ' . [02:58] no idea what. u mean bashing-om [03:01] Anyone now if there is a problem with launchpad.net login? It currently fails for me with a bad request page. "Bad bot, go away! Request aborted." [03:02] tomreyn , changing the color didn't work [03:02] bluap: similar was previously reported here (which is the wrong place), i pointed people to #caonical-sysadmin (which might still be the wrong place, but they should know) [03:02] yukiup: that's a pity. i just guessed. [03:02] OK, thanks. [03:02] bluap, you are the 2nd one tonight, no clue there, try again in a minute? [03:03] Yep, try again tomorrow, its very late here. Thanks. [03:03] u speak in riddles to me, bashing-om [03:05] k shur boot .. just afraid its gonnA fck up more [03:06] yverx: Sorry - we want to know if a particular file is present. in terminal execute command ' ls -al ~/.config/xfce4/panel/whiskermenu-1.rc ' . [03:07] riddles [03:07] explain to a. noob or not at all.. [03:09] yverx: there's an application called "terminal", are you able to start this? [03:09] yverx: I am sorry - I do not know how to simplfy further the instruction to execute. Can you help me help you with where I may be failing ? [03:09] when i do the first kfc kill panel command, thers still 2odd panels [03:10] ahwait ok now only 1 odd on top [03:11] no bashom. stay smart [03:12] ofc tomreyn, where else would i input the commands [03:14] yverx: well, you said you did not understand the instructions. those mentioned "terminal" and provided a command you were supposed to enter there. === Sven_vB_ is now known as Sven_vB [03:20] bashom: it gibes [03:20] *** gives me an output with A DATE [03:21] so. i assume, that file is present? [03:23] nov6 0321 so almoexactly 1 hr old/modifiedd,... whatebeee [03:23] OOo [03:23] yverx: Then a later release where the whisker menu is in play. If you remove that file and ' xfce4-panel --restart ' then a new user config will be created with system defaults when the panel is restarted. [03:23] FML my. typimg. is. pire. shit. since. phome reset... apologies.. [03:24] how do i do that? [03:26] the "rm" command takes one argument, a path to a file to remove/delete. the file you may want to remove / delete is ~/.config/xfce4/panel/whiskermenu-1.rc [03:27] yverx: execute 2 commands: ' mv ~/.config/xfce4/panel/whiskermenu-1.rc ~/Desktop ' and ' fce4-panel --restart ' . [03:28] Bashing-om: i think that's missing an x there in "fce4-panel" [03:29] yverx: the 'mv' moves a files to your Desktop directory. either 'mv' or 'rm' will work - generally mv is the sager option in the event of ooopps, [03:30] tomreyn: Ouch ! yeah ! [03:30] yverx: ^^ ' xfce4-panel --restart ' . [03:32] sry dc [03:33] pls repost last cmd [03:33] th files. been copied to. desltop [03:33] yverx: Should be ' xfce4-panel --restart ' after running mv ~/.config/xfce4/panel/whiskermenu-1.rc ~/Desktop. [03:35] back to 0 now [03:37] yverx: "back to 0 now" is that a good thing ? [03:38] 1 useless. panel on top,. 2nd dummy panel on bottom [03:39] no its back to uzseless [03:41] where can i check networkability? [03:42] also,. how to purge *blumantray'? [03:44] BMT ans "redshift* mostly only eber caused issies [03:46] considering a hardkill, is thT wise? === gschanuel2 is now known as gschanuel [03:55] fck [03:55] yverx: A reboot should not hurt and may reset the panels. [03:56] i meant hardkill not boot [03:57] boot oftn doesn work to negin witj [03:58] yyeyee] e [03:58] yverx: A hard kill (power button) is not a good thing in the least! as could leave the system in an inconsistent state. [03:59] look like i got. som. fimctiomaliyunback leys see [04:00] uk3ee(ekekkek [04:01] you will never know, without trying reboot. [04:03] cant. boot [04:03] gets stuck in loop [04:04] ola amigos [04:04] Bashing-om: !!!! [04:04] no help but i learned smthn + patience <3 [04:05] yverx: As you say "boot oftn doesn work to negin witj" - do suggest a file system check from a live USB, at this point. [04:05] and in the end always use brute force [04:05] :D [04:06] VIA: :D any time I add to my store of knowledge is a good day. [04:09] <3 [04:09] thx everyone fortheir time [04:10] this is my first linux distro after 1.5 centuries of win [04:10] i manage to f*ck everything to sh!ts [04:11] "sont use sudo blah blah blah" [04:11] VIA: Hang in and ubuntu will grow on you too :P - We are here to help you over the rough spots. [04:12] oh yea no doubt debian is already a winner [04:13] with win 8.1 etc ++ im so lost i dont even know where to start loosing time to unf*ck the whole system [04:13] also appreciate nobody sraight suggested a kayer 8 problem [04:13] l* [04:14] i really know what im doing : }}}} [04:16] or so i think :] [04:16] after hardkill i had an option to restore panel settings from 2 month back [04:17] where is it decided when to make such backups [04:18] i need to remove/purge some soft and reinstall proprietary drivers i think then ill run smooth [04:18] debootstrap stable bullseye ... where can I find a kernel? wget and unzip a .deb is fine [04:18] er debootstrap ubuntu :p [04:18] I'm doing both [04:18] but its usually running 247 so little problems [04:19] CarlFK: which ubuntu release? === starz_ is now known as starz [04:20] debootstrap focal focal http://archive.ubuntu.com/ubuntu/ [04:20] you might want to do so through https [04:21] VIA: "reinstall proprietary drivers" -- no, unless the graphic's are Nvidia or some exotic card. [04:21] for the kernel image you can choose between vanilla and HWE. i guess the latter makes most sense [04:21] where do I get either? [04:22] CarlFK: hwe would be package linux-generic-hwe-20.04 [04:23] for vanilla / general availability (GA) kernel, it's just linux-generic [04:23] you probably also want grub [04:24] apt install linux-generic ... The following NEW packages will be installed: (lots of packages...) [04:25] Is it easy and more stable to have multiseat on ubuntu with virtualbox or with qemu/kvm? I guess latter is free too? multiseat = multiple people user one system at the same time with their own set of keyboard,mice,monitor,audio etc [04:25] I just want initrd.img-5.4.0-89-generic vmlinuz-5.4.0-89-generic [04:26] the drivers rnt working [04:26] no other way than clean/ourge n try again [04:27] i tried like 20 distros in 1 week this one is most to my liking [04:28] VIA: ' sudo lshw -C display ' to get a start on finding ourt what the problem might be. [04:28] how do you tell, driver not working? [04:29] and it took me me 30~+ installls to figure out what packs to install in what order to make it atually WORK and not endup with a broken sys again [04:29] ayght lets see what it do [04:30] vidal72[m]: If you have to try hard in ubuntu ......... you are doing something wrong :D === mkv is now known as m4v === erlon_ is now known as erlon [05:20] What is better for having windows for gaming? Virtualbox or qemu/kvm or anything else? Which will give best performance? === mauved_ is now known as mauved [06:34] Sircle: dual boot would be the best, without emulation [06:59] Hi, i just installed ubuntu 20.4 to surface pro 6 [06:59] unable to find battery indicator [06:59] anyone else been there? fixed it? [07:00] Hello [07:00] Anyone setup bind9 in 20.04? I could use some guide [07:01] i have getting error when i try to update [07:01] it says " GPG error: http://deb.debian.org/debian jessie Release: The following │ ac5tin [07:01] │ | signatures couldn't be verified because the public key is not available: NO_PUBKEY │ acerbic [07:01] │ | 7638D0442B90D010 NO_PUBKEY CBF8D6FD518E17E1" this [07:02] it says " GPG error: http://deb.debian.org/debian jessie Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY 7638D0442B90D010 NO_PUBKEY CBF8D6FD518E17E1" this [07:02] killadi: this channel is about ubuntu, your error message is about debian. [07:02] oh sorry === killadi is now known as mekilladi [07:17] when aptitude why shows Depends and it lists them separated by |  , what does that mean? Does it mean "or"? [07:19] i think so, but it's been a long time since i used aptitude, so I might remember incorrectly [07:19] !aptitude [07:19] Like apt and apt-get, aptitude is a terminal frontend for Ubuntu's package management system. Unlike the others, it is not recommended in Ubuntu because its behavior differs significantly from other Ubuntu package management tools and can cause issues. [07:30] I use aptitude to search [08:16] hello there! [08:16] morning Rockerul [08:18] how is everybody? [08:19] !chat | Rockerul [08:19] Rockerul: #ubuntu is the Ubuntu support channel, for all Ubuntu-related support questions. Please register with NickServ (see /msg ubottu !register) and use #ubuntu-offtopic for other topics (though our !guidelines apply there too). Thanks! [08:23] !guidelines [08:23] The guidelines for using the Ubuntu channels can be found at https://wiki.ubuntu.com/IRC/Guidelines [08:23] I stick with apt and dpkg. [08:35] how can I register? :( [08:36] I mean...I know the commands, but I have a problem on picking the email.. [08:36] Rockerul: for help wit IRC related stuff please ask in #libera [08:37] ok, thx [08:37] this channel is for support with ubuntu and it's flavors only [08:37] I am new to this :D [08:37] sure [08:37] I just got back to Ubuntu :) [08:37] but wanted to join some IRC for good old times [08:37] still offtopic chatter... [08:38] anyway, welcome back then [08:38] ty, so here I can only ask for Ubuntu related things [08:41] Rockerul: yep, as said this channel is for support with Ubuntu and it's official flavors [08:41] good, need a bit of help then :) [08:42] if you just want to chit chat then there's #ubuntu-discuss and #ubuntu-offtopic [08:42] got an SSD where I installed Ubuntu, but still have an HDD that I want to use for Datas (files) [08:42] my question is: how to format the HDD, and the mount on startup [08:43] since i never have done that i'll leave that to someone else [08:43] you can use gparted to format it and then edit /etc/fstab to mount it automatically [08:44] make sure you backup anything you wanted to keep before formatting [08:44] well...I think I did it in a way, but don't know if it's the correct way, or if there is an "correct" way [08:44] !fstab | Rockerul [08:44] Rockerul: The /etc/fstab file indicates how drive partitions are to be used or otherwise integrated into the file system. See https://help.ubuntu.com/community/Fstab and http://www.pclosmag.com/html/Issues/200709/page07.html and !Partitions [08:45] also...the thing I never understood: SWAP, do I need it? Have 8Gb RAM...on Ubuntu install I followed basic instalation...let Ubuntu pick partition settings [08:45] is it ok? [08:46] Rockerul: should be fine, nowadays ubuntu tends to use a swapfile rather than a separte partition [08:46] separate [08:47] If the RAM is 8GB you'll get a swapfile of 2GB [08:48] so that should be ok than? I want to use it to learn programming on this laptop [08:48] should be fine [08:49] how much swap space you have has no relation to programming [08:50] how often to search for updates? I use "sudo apt update && sudo apt upgrade" like once / week, is it ok? [08:50] sorry if I ask stupid questions...but I prefer to be stupid for my questions for 5 minutes than remain stupid for how long I use Ubuntu :)) [08:51] There's a Software Update which pops up once there are updates available, so no need to search manually [08:51] I don't think upgrade would get you kernel updates so you'll want full-upgrade [08:52] Update/Updater [08:52] just sudo apt full-upgrade? [08:52] after sudo apt update yes [08:52] sudo apt update && sudo apt full-upgrade [08:52] aaa I see [08:53] and what is the difference if I use sudo apt-get update? I got used to put that -get after apt....don't know why but... [08:54] update refreshes the repositories to the latest and upgrade fetches the newer versions to install and full-upgrade fetches newer version that also installs new dependancies (such as kernels) [08:55] apt-get is just the older one, it is recommended to use apt (meant to be cleaner) [08:56] https://askubuntu.com/questions/445384/what-is-the-difference-between-apt-and-apt-get [08:56] As I understand it, aptitude and apt-get should be considered deprecated. I'd get used to using apt from now on. [08:57] i will use only apt from now on than. Thanks guys :) [08:57] yw [08:59] will go to #libra later and figure it out with registration, and when I have questions about my system I will come here :) I will go now and thest if my HDD it's auto-mounting now. Again, thank you for your help guys and catch you later. Take care, bye :) [08:59] Rockerul: there is a package called something like unattended-updates [08:59] it will do the job for you [08:59] what that doese? [09:00] Rockerul: the name is pretty much self-explanatory :-) [09:00] I cannot remember off the top of my head how often it checks for updates [09:00] !info unattended-updates [09:00] Package unattended-updates does not exist in impish [09:01] perhaps it is unattended-upgrades ? [09:01] the "unanttended" part was there for sure [09:01] !info unattended-upgrades [09:01] unattended-upgrades (2.8, impish): automatic installation of security upgrades. In component main, is optional. Built by unattended-upgrades. Size 47 kB / 436 kB [09:01] https://www.cyberciti.biz/faq/set-up-automatic-unattended-updates-for-ubuntu-20-04/ [09:02] there you go [09:25] Hello, does anyone know how netplan is called at boot time? man netplan-generate says that it is, but how exactly? I could not find a systemd unit and there did not seem to be a related file in /etc/init.d. Is it somehow called by systemd-networkd? [09:27] Just out of curiosity, how is ifupdown2 called? I assume it would be the same place? === Daniel is now known as Guest6708 [09:58] Can someone explain why Ubuntu installs Python 2 with npm? [09:58] It's not listed as a dependency, it just does. [09:59] I just don't get why anything in Ubuntu installs Python 2 when Python 2 hasn't had a security update for almost two years. [10:09] qskwood, why would ubuntu install python 2? [10:09] That would be my exact question. [10:09] you surely should look at the package you install with npm [10:09] I'm not using npm. [10:10] I'm using `apt install npm` [10:10] then it is a dependencie of nopm, likely [10:10] Yes [10:10] Yes it is [10:10] And my question is "Why" [10:11] apt-cache depends npm # shows why [10:12] First off, if you actually read what I wrote you'd see it's not a dependency [10:12] Second, that wouldn't say why it's a dependency, just that it is one (which it is not) [10:13] apt show npm [10:13] Again [10:14] that doesn't provide any information. [10:14] If you don't know the answer to a question it's perfectly fine to disengage [10:14] Wasting people's time is rude [10:14] installing npm pulls in 210 packages, good luck [10:19] https://packages.ubuntu.com/focal/gyp [10:19] This was "fixed" in groovy [10:20] It will be fixed in the next LTS by extension, so I'll be able to use an LTS environment that doesn't rely on Python 2 [10:20] good thing groovy is EOL [11:16] Sorry I upgraded my server and my IRC client was offline. Was there a response to my question? [11:34] Hi :) Question: I have managed to somehow make my LAN networkcard dissapear, how do i install it again? Its a Realtek card [11:35] gnask: is it built-in or USB? [11:35] Built-IN [11:35] gnask: lets see if we can identify the hardware: "lspci -nnk -d ::0200" [11:36] (this lists PCI devices in the ethernet network class (0200) [11:36] 4:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [10ec:8168] (rev 16) Subsystem: Lenovo RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [17aa:36d9] [11:36] gnask: good, does it also report a kernel module in use? [11:36] That is all the text it outputs [11:37] did you miss the "k" in the options? (that reports kernel modules either in use or available) [11:37] Nope -nnk are the flags i used [11:37] gnask: OK, so we now know what the issue is - you've 'lost' the kernel driver [11:37] Sounds bad :> [11:37] gnask: so next we figure out which one it should be and where it went! [11:38] Allright! [11:38] gnask: as this a a Realtek device its possible the driver is not included in the mainline kernel and you've got an out-of-tree kernel module installed to drive it [11:38] Ok [11:38] gnask: see if a driver that looks to match that device is listed with "dkms status" [11:39] rtl8821ce, 5.5.2.1, 5.13.0-20-generic, x86_64: installed [11:40] gnask: that looks like it. it shows 2 kernel versions installed. Check what the current running kernel version is with "uname -r" [11:40] 5.13.0-20-generic [11:40] gnask: OK, lets see if we can find the module [11:41] gnask: "find /lib/modules -name 'rtl8821ce*' " --- do you get any matches? [11:42] Yes [11:42] /lib/modules/5.13.0-20-generic/updates/dkms/rtl8821ce.ko [11:43] gnask: all looks good so far so now we need to find out why that module doesn't load [11:43] gnask: let's first try manually inserting the module and seeing if the kernel has a problem [11:43] Ok [11:43] gnask: "sudo modprobe rtl8821ce" ... if that works you won't get any messages, just back to the shell prompt [11:44] gnask: then check kernel messages with "dmesg | tail -n 20" [11:45] modprobe worked, no messages [11:45] i got alot of kernel msgs though from dmesg [11:45] Can someone please help to reslove the udev issue? https://askubuntu.com/q/1373624/21005 [11:45] ^resolve [11:45] gnask: that system has an internet connection so we can share them in a paste? [11:46] Yep, im on it now via WIFI [11:46] gnask: if so " dmesg | tail -n 50 | nc termbin.com 9999 " [11:46] https://termbin.com/jwoy [11:48] gnask: something wrong here, those are just UFW firewall messages, no reports of the module loading or the device being initialised as we'd expect [11:48] Oh ok [11:49] gnask: let's see if the module did load. "lsmod | grep rtl" - do you see the rtl8821ce ? [11:50] https://termbin.com/z4wu [11:51] gnask: this is getting weirder [11:51] Welcome to my life lol [11:52] gnask: when did you install that rtl821ce-dkms package? Looking at it, it is for a wifi device NOT wired :D [11:52] gnask: which would explain this [11:52] gnask: let me go back to basics and figure out which driver your hardware should be using [11:53] TJ-: I dont know really, was on a spree yesterday and tried to update stuff [11:53] Ok! [11:55] gnask: show me " grep . /etc/modprobe.d/* | nc termbin.com 9999 " I want to see if there's been a blacklist for the expected kernel driver [11:56] https://termbin.com/4p7z [11:56] gnask: it looks like that device ought to be using either the kernels' r8169 driver or the one from the r8168-dkms package [11:57] Yes ok [11:58] gnask: nothing obvious there. You said you did some house-keeping and updates. Is that when this problems tarted? [11:58] Yep it was [11:59] aha! so, lets look at the apt log! " cat /var/log/apt/history.log | nc termbin.com 9999 " [11:59] https://termbin.com/j8s2 [12:01] Hello everyone. I am doing "do-release-upgrade" in non-interactive mode. How do I prevent/exclude a single service from being autorestarted during the upgrade process? [12:02] gnask: so I see yesterday you removed and reinstalled dkms and the rtl8821ce packages at 09:28 - was the device still working at that point, or was that an attempt to fix it? [12:03] If i recall correctly, i belive it was my attempt to update it to the "latest" driver lol [12:03] gnask: OK, we'll discount all that and just install the one that looks to be correct [12:04] Yes [12:06] gnask: at least at my end the kernel's r8169 driver claims to support your device. Your device is 10ec:8168 and if you do "modinfo -F alias r8169 | grep 8168" you should see a match for "pci:v000010ECd00008168sv*sd*bc*sc*i*" [12:06] modinfo: ERROR: Module r8169 not found. [12:07] gnask: but apparently, according to the notes in "apt-cache show r8168-dkms" suggest the kernel's r8169 doesn't fully support all devices correctly [12:07] gnask: huh!?! [12:07] gnask@gnaskBuntu:~$ modinfo -F alias r8169 | grep 8168 [12:07] modinfo: ERROR: Module r8169 not found. [12:07] gnask: OK, let me go check if your kernel version should have that module. This could be the issue [12:07] Oh ok ok! Much thanks for your time. [12:09] gnask: is this package installed? "apt list linux-modules-extra-5.13.0-20-generic" [12:09] Yes it is [12:09] you should see "installed" there [12:09] [installed,automatic] [12:10] OK, so that should install the file /lib/modules/5.13.0-20-generic/kernel/drivers/net/ethernet/realtek/r8169.ko [12:10] and modinfo should have reported from that [12:10] check once again "modinfo r8169" [12:10] linux-modules-extra-5.13.0-20-generic/impish-updates,now 5.13.0-20.20 amd64 [installerat,automatiskt] [12:10] modinfo: ERROR: Module r8169 not found. [12:10] if that fails do "sudo depmod" [12:11] then do the "modinfo r8169" again [12:11] gnask@gnaskBuntu:~$ sudo depmod [12:11] gnask@gnaskBuntu:~$ modinfo r8169 [12:11] modinfo: ERROR: Module r8169 not found. [12:11] hmmm, something major is wrong here [12:11] Yeaaaaah :p [12:12] gnask: "find /lib/modules/5.13.0-20-generic/kernel -name '*r8169*' -ls " [12:12] 2370834 180 -rw-r--r-- 1 root root 184273 okt 15 15:53 /lib/modules/5.13.0-20-generic/kernel/drivers/net/ethernet/realtek/r8169.bak [12:13] "bak" ...!!!! [12:13] gnask: what have you been up to!?! [12:13] haha, what is a bak? a backup? [12:13] yes, it should be .ko [12:13] oooo [12:13] gnask: who did that?! [12:13] no idea, i did not rename it manually thats for sure [12:14] the date suggests it was done on October 15 [12:14] Uhm ok, it worked yesterday [12:14] until i started tinkering [12:14] very strange, lets change it anyhow [12:14] Ok [12:15] are you copy/pasting my commands? [12:15] Yep [12:15] good you'll need it for this next one :) [12:15] gnask: "sudo mv /lib/modules/5.13.0-20-generic/kernel/drivers/net/ethernet/realtek/r8169.{bak,ko} " [12:16] gnask: then "sudo depmod" and then "modinfo r8169" [12:16] now modinfo should report [12:16] It didnt like mv , "filoperand saknas" which translates to fileoperation missing [12:17] hmmm, what shell are you using? "echo $SHELL" [12:17] bash [12:17] that should work then, make sure the {bak.ko} at the end of the line aren't corrupted [12:18] oh I know, become root first with "sudo -i" then do: [12:18] "mv /lib/modules/5.13.0-20-generic/kernel/drivers/net/ethernet/realtek/r8169.{bak,ko} " [12:18] Ok now it worked [12:19] gnask: then "sudo depmod" and then "modinfo r8169" [12:19] if modinfo reports then we can get the device going [12:19] Yes it reports back :) [12:19] I'll assume you're still root (prompt is a # not a $) [12:19] "modprobe r8169" [12:20] done [12:20] then "dmseg | tail -n 20" and you should see the device come alive [12:20] oops, typo [12:20] dmesg not dmseg [12:20] woooo! its working :) [12:21] Will it still work after a reboot ? [12:21] finally! how the heck did that get renamed to .bak ... there must have been a reason [12:21] yes, it will work. [12:22] Superthanks for this. Not everyone is so kind <3 [12:22] now, before you think this is over, one reason that may have been renamed is to allow another module to load and operate that device. [12:22] Ok, like what module? [12:22] if you find the device is flakely or doesn't perform as expected, then you may need to install a different package "apt install r8168-dkms" instead [12:23] Ok, will save that advice [12:23] gnask: read the info on that just in case you do need it later "apt-cache show r8168-dkms" [12:24] gnask: but I think you'll be OK [12:24] Will do. Thanks once again, your a champ [12:28] Howdy all [13:04] lt]>bJ]?$J{HA;Y< [13:04] omg cringe, pasting password into irc [13:05] greatest one BinarySavior! [13:05] but it's not binary! [13:07] thanks for all your crypto, BinarySavior! [13:07] :D [13:07] luckily this host is not outward facing [13:07] i'm still gonna change the password anyway [13:07] +1 good practice [13:08] stay alert when you're pasting stuff! [13:09] alt+tab , shift+insert failed me [13:09] i dont know how hexchat ended up in-between my ssh session and keepassxd but it did [13:23] hi, I'm trying the ubuntu universe and multiverse repos, but I get an authorisation error "Authorization required, but no authorization protocol specified" [13:24] fmlatghor: for what Ubuntu release? [13:25] uname says $100-Ubuntu SMP + [13:25] TJ-: uname says $100-Ubuntu SMP ? [13:26] fmlatghor: what!? [13:26] #100 [13:26] fmlatghor: "cat /etc/issue" [13:27] uh yes [13:27] let me go to the appropriate chan [13:35] is there a way to enable Ubuntu 20.04.3 LTS to use the apt 2.2.x version, instead of the 2.0.x version? [13:35] Firefishe: not easily, no. What functionality is missing you are seeking? [13:37] I'm on a m$ surface book 3 lap/tablet. I'm using the linux-surface ppa for the system as it's required for it to be functional. apt 2.0.x is currently having a problem with certain special characters in github url requests when upgrading. 2.2.x is supposed to work. [13:37] it's available in the 21.04 and later repos. you can edit the sources.list file but its not recommended [13:39] KBar: I don't know what effect a system upgrade would have on the current system. I depend on linux-surface for full functionality, so I need those things to be steady, even during an upgrade. It's essentially just the kernel and kernel headers, and some files specific to this laptop that are affected. [13:40] KBar: But apt is being difficult. The problem is being worked on, but I'm looking for quicker and simpler solutions. [13:41] KBar: What about compiling apt 2.2.3 from source, as an example, and replacing the current apt with it? [13:42] Firefishe: wich kernel are you currently on? [13:43] lotuspsychje: 5.14.15-surface [13:44] ok, i was reading https://www.phoronix.com/scan.php?page=news_item&px=Better-Surface-Linux-5.12 [13:44] lotuspsychje: There's an upgrade to 5.14.16-surface, but the upgrade is borked for the time being. [13:44] lotuspsychje: I have a surface book 3 [13:45] Firefishe: can you show us an example of the URL and these special characters? [13:49] TJ-: Here's the github link to the issue report: https://github.com/linux-surface/linux-surface/issues/625#issuecomment-962036723 -- It takes you to the middle of the report, more or less.. [13:49] Issue 625 in linux-surface/linux-surface "Ubuntu: Apt update fails with Error 401 Unauthorized" [Open] [13:53] Firefishe: does the surface suffer functionalitys on default kernel and HWE? [13:55] Firefishe: crikey; those URLs are terrible. no wonder it broke! [13:55] Firefishe: looks like the github devs think everything is a full blown web browser! [13:56] lol [13:56] TJ-: octet/stream just about made me drop my jaw [13:57] * TJ- fumes about 'dev-ops' and fetches the shotgun :) [13:57] TJ-: They are a little large for a git hub url. For as much as I know about it, which is only basic. [13:57] Firefishe: I'm figuring out if there's a sneaky way to help apt out here [13:57] TJ-: What is that ubuntu app that does system upgrades? [13:58] TJ-: If there is, posting a workaround in that thread would help! :-) [13:58] Firefishe: depends on what you mean with "upgrades" ... packages, entire release, etc [13:58] TJ-: Newer release with a newer kernel. I may still need the ppa though. [13:58] is there an ubuntu equivalent to something like Fedora/RHELs "dnf --installroot= --releasever=34 -y group install 'Fedora Server Edition'" to set up a full-fledged root ubuntu server filesystem under a directory? [13:58] Firefishe: what i would try in your case, is running latest ubuntu releases with higher kernels, see if you suffer still bugs [13:59] i'm guessing some invocation of debootstrap might do it, but i can't figure out what it would look like [14:01] TJ-: This is where I'm getting my special use kernel: https://pkg.surfacelinux.com/debian release main [14:02] currybullen: "debootstrap --include ubuntu-server focal /target/ " [14:03] currybullen: 'focal' is the SUITE so replace that as appropriate, e.g. it could be "impish" for the latest [14:03] currybullen: may need --include=ubuntu-server (note the = symbol) and no spaces [14:06] Firefishe: one workaround might be to deploy squid-deb-proxy and on the client squid-deb-proxy-client (proxy would need an allow ACL for the github URLs adding) but then squid would do the redirects for you [14:06] Firefishe: a local alternative on the host might be to install apt-cacher-ng [14:06] !info apt-cacher-ng [14:06] apt-cacher-ng (3.6.4-1, impish): caching proxy server for software repositories. In component universe, is optional. Built by apt-cacher-ng. Size 482 kB / 1,384 kB [14:09] TJ- apt-cacher-ng on the github server hosting the code, you mean? [14:11] Firefishe: eh? no, install it on the Ubuntu host that has the problem [14:11] Firefishe: I'm thinking a proxy engine will better handle the URIs and redirects [14:12] TJ-: ah nice, thanks! [14:28] TJ-: Here's my output of "sudo apt upgrade" -- https://pastebin.com/9HYDkyRG [14:30] Firefishe: is that with apt-cacher-ng installed? [14:31] Yes. Although apt-cacher-ng may not be configured correctly. This is the first time I used it, but I did say "YES" at the cli/curses window when it configured. [14:31] TJ-: hmm, the debootstrap command results in dependency errors. for impish there's "python3-cffi-backend-api-min is not installed" among others. shouldn't the dependencies be resolved automatically? [14:31] Firefishe: OK, so now we need to point apt at apt-cacher-ng [14:32] Firefishe, to see if the upstream server works, try: curl --show-error --head https://pkg.surfacelinux.com/debian/dists/release/Release [14:32] currybullen: depends on what else you've put on the command line! you may be asking for a package that is in universe or multiverse, in which case you need to add those components [14:32] Firefishe, then try with acng as proxy for curl [14:34] Firefishe: info for configuring apt itself to use apt-cacher-ng https://help.ubuntu.com/community/Apt-Cacher%20NG#Client_setup [14:34] TJ-: hmm, i only put "debootstrap --include=ubuntu-server focal /target/" [14:34] Sven_vB: https://pastebin.com/vt67h7nc [14:34] Firefishe: in your case I suspect you just need to use localhost [14:34] currybullen: really, let me test here [14:35] Firefishe: so " echo 'Acquire::http { Proxy "http://localhost:3142"; }' | sudo tee -a /etc/apt/apt.conf.d/proxy " [14:36] Firefishe: ensure apt-cacher-ng has started too, of course, with " systemctl status apt-cacher-ng" - you may to to edit /etc/default/apt-cacher-ng [14:39] TJ-: I ran that code then updated. Same errors. [14:40] Firefishe: hmmm, is apt saying it using apt-cacher-ng? [14:40] TJ-: Actually, there isn't anything in /etc/default [14:40] TJ-: How would I query apt to find that out? [14:40] Firefishe: the apt-cacher-ng package has a file /etc/default/apt-cacher-ng according to "apt-file list apt-cacher-ng" [14:41] Firefishe: If I recall correctly apt reports it is using the proxy server (but I may be wrong!) [14:42] Firefishe: does "systemctl status apt-cacher-ng" show it active and running? [14:43] TJ-: Yes [14:43] Firefishe, in the sudo…tee command, the -a seems odd. [14:44] Firefishe: for a different github resource for me it breaks too [14:44] Sven_vB: Eliminate the -a flag? [14:44] Sven_vB: no, that is append, to avoid over-writing and existing file [14:44] Firefishe: no, the -a is fine [14:44] ok [14:45] TJ-, yeah. I assumed that there should only ever be one proxy configured in that file. [14:45] Firefishe: apt-cacher-ng didn't help for me, I see "Invalid response from proxy: HTTP/1.0 403 CONNECT denied (ask the admin to allow HTTPS tunnels) [IP: ::1 3142] " [14:46] Sven_vB: always better not to let an existing extry be deleted, append will mean the last entry takes effect [14:46] Firefishe: I'll try allowing the tunnels now [14:47] Here's my /etc/default/apt-cacher-ng file: https://pastebin.ubuntu.com/p/CgtDFcz9C9/ [14:47] TJ-: thanks! [14:47] Firefishe: it worked! with github [14:48] Firefishe: I had to enable tunnels [14:48] TJ-: How do I do that? [14:48] sometimes you can also use the non-encrypted version by changing the URL to http:// . GPG signatures should be safe enough for content validation; the only attack I know of would be for a MitM to make you believe an update is not available yet. [14:49] Firefishe: "sudo dpkg-reconfigure -p high apt-cacher-ng" [14:49] (with HTTPS, the attacker could still block access to the update but it would result in an apt error.) [14:49] Firefishe: then restart it "sudo systemctl restart apt-cacher-ng" [14:50] Sven_vB: that isn't the issue we're resolving here [14:51] Firefishe: you'll see apt reporting "[Connecting to proxy ..." [14:52] https://pastebin.ubuntu.com/p/BBY2JTmFb5/ === Anton is now known as anton [14:52] Firefishe: I see "Get:1 https://github.com/retorquere/zotero-deb/releases/download/apt-get ./ zotero 5.0.96.3 [64.1 MB]" ... "Fetched 64.1 MB in 31s (2,090 kB/s)" [14:53] Firefishe: did you see any of those "[Connecting to proxy..." messages flash up and then disappear when connection is established? [14:53] TJ-: No. [14:55] and the service is running [14:57] Firefishe: let's collect more info. What is the complete apt command you're using ? [14:58] sudo apt upgrade [14:58] Firefishe: use this "sudo apt -o=Debug::Acquire::https=true upgrade |& tee /tmp/apt.log " [14:58] Firefishe: then share the log with "pastebinit /tmp/apt.log" [14:59] that last command here? [15:00] pastebinit is a command and it'll send the /tmp/apt.log to paste.ubuntu.com [15:00] then you can share the URL it gives [15:01] TJ- sorry to interrupt. Does it require a login? [15:01] it does not [15:01] TJ-: https://paste.ubuntu.com/p/Y5szTQkfGB/ [15:02] KBar: no, the problem is nasty characters in the github redirect that trip 'apt' up, so we're trying to avoid that by putting apt-cacher-ng in front of apt [15:02] TJ-: Taking a quick break. Back in five [15:03] Gallomimia, woah! much more convenient than going to paste.ubuntu.com, logging in and doing all that type of stuff. good to know! will come in handy. thank you [15:03] TJ- I see. Thanks. You are all awesome! [15:03] Firefishe: ok, looks like your proxy isn't bein used [15:06] TJ-: Back [15:06] TJ-: So it's not being used, eh? How to make it work? [15:07] TJ-: I appreciate your time spent. Thank you for the assistance today. [15:10] Firefishe: with the same Debug:: option here in the log I see things like "0% [Connecting to HTTP proxy (http://localhost:3142)]HTTP/1.0 200 Connection established " [15:10] I see [15:10] Firefishe: as you do not see those in your log that makes me think the apt.conf.d/proxy isn't being used [15:11] Firefishe: what does this show? " cat /etc/apt/apt.conf.d/proxy " [15:11] Firefishe: I created that file here which has "Acquire::http { Proxy "http://localhost:3142"; }" [15:13] TJ-: https://paste.ubuntu.com/p/yZQshhXRzD/ [15:14] Firefishe: so we have the same. [15:17] TJ-: I'm trying to wrap my head around what apt-cacher-ng is doing. [15:18] TJ-: I get what we're trying to get accomplished. [15:19] Firefishe: apt-cacher-ng is a HTTP proxy so it has the code to follow Location: redirects itself, so it does those not apt, which should work-around the bug [15:19] Unless those port numbers are borked or something [15:20] TJ-: Maybe a reboot? [15:20] Firefishe: no, this isn't windows [15:20] TJ-: I get it. [15:20] 3142 is the standard port :) [15:21] TJ-: Oh, I meant the port number after the github url. Never mind. I'm comlicating it and thinking in the wrong areas. [15:21] complicating [15:23] lol. not windows :) [15:23] i love comedy when i first wake up. makes my day [15:25] TJ-: I think I saw those popups you were talking about. Getting headers in brown/gold fonts? [15:26] TJ-: I need to skedaddle. I've been up far too long from my night audit hotel job in the commons area and I'm getting loopy. Gotta drive 20 miles home, too. Thanks for your help today. I'll keep working at it. firefishe@gmail.com is my address if you want to add to this. Thanks TJ-, Sven_vB, and everyone. I appreciate it. Be Well! [15:30] is https://pkg.surfacelinux.com/debian the repo that caused the github URL problem in apt? [15:35] Sven_vB: not directly; it redirects to a github URL which in turn generates a really long querystring'ed URL with that breaks apt on 20.04 [15:35] Sven_vB: seee https://github.com/linux-surface/linux-surface/issues/625#issuecomment-962046373 [15:35] Issue 625 in linux-surface/linux-surface "Ubuntu: Apt update fails with Error 401 Unauthorized" [Open] [15:37] TJ-, thanks [15:54] unfortunately apt-cacher-ng also fails. It looks like there's a space in the URL [15:54] (fails on 20.04, but works on 21.10) [15:55] https://paste.ubuntu.com/p/ry6K8KtWb4/ [16:51] Q: What'd be the quickest VPN solution I can setup in Ubuntu? My use-case is for remote desktop access (windows), clients (mac, linux, windows). Launched a OpenVPN on DigitalOcean but seems to be a "required" license version, plus would like to have a local network users can ping each other out of the box. Any tips? [16:51] My biggest issue is time, as I'm trying to help my friends but currently lacking time [16:56] wireguard === Ricardus_ is now known as Ricardus [18:30] hi [18:33] hi [18:35] I am trying to install several ubuntu distros side by side on encrypted partitions. unfortunately I am not "allowed" for securita reasons to mount one distos disk from another distro -I want them to be fully isolated. How can I have all distros listed and bootable from grub ? [18:36] is there any trick that I am missing ? [18:36] grub can't boot off of encrypted zfs [18:36] tmp1: there's only one ubuntu distribution, with multiple flavors. there are also ubuntu derivative distributions. [18:36] for now I can only make one of the distros bootable [18:36] normally there's an unencrypted /boot, and an encrypted / [18:36] what's the commands for list the channels here? [18:37] s0mbr4C3123K: /msg alis help [18:37] or list [18:37] that's what I an doing. /boot unecrypted and different / encrypted. but grub only see one of the installed distros at a time [18:38] /msg alis LIST *freetopic* -mode -t -show mt [18:38] tomreyn: I am even installing the exact same flavor but multiple times to isolate software from each other, [18:38] tmp1: can those installations share the same boot loader? [18:39] tomreyn: yes [18:39] but I am not able to make it multiboot. because at the moment of installation grub only see one of the partitions. [18:40] tmp1: if you really want the installations to be fully isolated from one another, i.e. actually making it impossible that they access one another, you'd need to use virtualization and place them in different networks. [18:40] why different networks? [18:40] tmp1: The hardware will always access one physical device for boot. It may offer a menu, but you can't have more than one active boot partition. [18:40] but why cant' I simply use different partitions ? [18:40] tmp1: You should be able to mount the others and chroot to them, but you're creating your own problems. [18:40] they can't run at the same time, so they can't communicate even if they have the same network [18:40] you can alternatively install them side by side, encrypted or not, but that will not prevent one installation writing to (overwriting) the other installations storage [18:41] maybe zfsbootmenu can do something [18:41] but it'd be far from automatic [18:42] and of course, while the systems can't read each other's data, they can destroy it still. [18:42] what I need is 2 installs of ubuntu 20.04 that have two encrypted and separate partitions with different passphrase and when I boot I wanna choose what distro to boot and what / to uncrypt [18:42] cbreak: i think you were the first one to bring up zfs in this context. [18:42] hmm... what else is there? [18:42] I am not using zfs [18:42] hmm... ok [18:42] dmcrypt-luks would be the common approach still. [18:43] does that require a separate /boot too? [18:43] just a separate unencrypted /boot that may or may not be the same for the two distos, but separated and encrypted / [18:44] /boot may be encrypted with luks1, grub can boot this. [18:44] the problem is that I am not able to install grub in a way it allows to boot both distros [18:45] tmp1: is this uefi or bios booting? [18:45] uefi [18:46] so if you don't actually need full isolation then you could either place different copied of grub on the ESP, configured to boot one ubuntu variant... [18:46] I currently use refind to select which boot loader to use. You could maybe do something similar. [18:46] ...or you could have one of them, ideally the one with the newer grub version, manage the grub installation for both [18:47] os-prober would not be able to identify the other installation if that was encrypted, so you'd need to manually manage the other systesm' grub configuration then. [18:48] but since partitions are encrypted , and that I do not mount other's distro partition from the second distro... it only see and configure one single ubuntu in grub [18:48] tomreyn: that seems to be the problem precisely. [18:48] but how to do it manually ? [18:48] if each ubuntu has its own copy of grub, you can chose between the grubs from refind [18:49] (or the uefi firmware boot chooser, which usually sucks) [18:50] tmp1: copy, or note down, what the grub other installations' grub configuration would be, and set that up on the original installations' grub configuration, i'd say. [18:50] tmp1: though with uefi, having multiple copies of grub installed, and their configuration managed by each of the OS installations individually, may actually be easier. [18:51] it's probably easiest if you have separate efi partitions, but I don't think that's needed, you can have different subdirs too probably [18:51] you'd just need to ensure that their grub shims on the ESP will not overwrite one another, i.e. will be installed to different locations on the ESP [18:51] Full isolation: Fast swap drive rack. Other drives aren't physically present when not needed. [18:52] yeah. Or usb sticks :P [18:52] hmm, yes, physical swapping of storages would also work [18:52] cbreak: That's a pretty cruel form of toture. [18:52] it worked for me on freenas... [18:52] it's a notebook. I have one drive only , and would like to be able to select partition to boot at boot. ;) does not look to difficult ? [18:53] but somehow is apparently .. [18:53] :/ [18:53] Esternal drive rack. [18:53] tmp1: if you can figure out how to configure your grub to end up somewhere else than in EFI/ubuntu, then it should be easy [18:53] tmp1: i don't think it is this difficult. maybe you're just overloaded with options, because there are so many ways to set this up? [18:54] because then you can have separate grubs [18:54] (and I'd again recommend refind to actually chose between the grubs, but there are of course other options here as well :) [18:54] I'm pretty sure that doesn't work at the partition level. You need one common grub level with the boot menu from there. [18:54] I just need one that works ;) === rexbinary_ is now known as rexbinary [18:55] cbreak: I think completely seperate grubs requires the BIOS/EFI boot menu to choose, and that usually means choosing among physical devices. [18:55] refind produce some "security violation" on my laptop (secure boot enabled) [18:55] jhutchins: efi can chose refind [18:56] jhutchins: refind then offers all the grubs and so on to boot [18:56] the entries for boot options in UEFI are not partition based, they point at a specific file in a specific partition [18:57] you can list what you've registered with `efibootmgr -v` [18:57] for me, among others, it lists Boot0001* rEFInd Boot Manager HD(1,GPT,101abacd...,0x800,0x100000)/File(\EFI\REFIND\REFIND_X64.EFI) [18:57] that's the one I boot normally [18:58] native ubuntu has entry HD(1,GPT,101abacd-...,0x800,0x100000)/File(\EFI\UBUNTU\SHIMX64.EFI) [18:58] but you could naturally have an other grub in XUBUNTU, LUBUNTU, UBUNTWO or what ever [18:59] I will try with refind. but for now it's not booting [18:59] is it secureboot compatible ? [18:59] I think it can be made, but I don't use secure boot, so I'm not sure how. I think by default it is not [19:00] adding another bootloader in front just to have a menu to select other boot loaders from doesn't seem to make much sense to me [19:01] refind isn't a boot loader [19:01] the uefi already provides that [19:01] it's a boot choser [19:01] one that's much better and much more reliable than the one in UEFI [19:01] which problem does it solve there? [19:01] refind auto-discovers grubs, clovers, opencores, windows, and a bunch of others [19:02] in this concrete scenario [19:02] so you don't have to be sure that all the different loaders are correctly registered with efi via efibootmgr [19:02] and if you add / change something, you don't have to make sure the entries are kept up to date all the time [19:03] I use refind for convenience and reliability [19:03] but it isn't neccessary [19:03] if you know how, you can register the different grubs via efibootmgr (and I think ubuntu will do that on its own too normally) [19:04] seems refit is my way [19:05] trying. [19:05] i see. maybe it's just that i lack experience with refind then. but adding it into the loop seems to increase complexity rather than decreasing it, especially with secureboot. [19:05] refit is a different thing [19:05] refind is in apt [19:05] refind, sorry [19:05] (and I think it's newer) [19:06] well. No comment on secure boot. [19:06] I've never seen the point in that, and never tried it [19:07] tomreyn: I've just had bad experiences with efi entries getting lost / overwritten / ... [19:07] not particularly with linux, but also with OpenCore / Clover on hackintosh [19:07] cbreak: that doesn't seem to be part of tmp1's scenario [19:08] well, it is multi-boot related === coconut_ is now known as coconut [19:15] tmp1: I don't see any reason why you cannot have several EFI-SP partitions, one per distro install, when using UEFI boot (as long as you ensure each distro does use a different EFI-SP - the installer can sometimes ignore your instructions so you'd need to do that part manually) The UEFI firmware boot entries include the partition index to use to read the boot loader [19:16] let me try this [19:18] tmp1: "grub-install" won't like multiple EFI-SPs; you have to trick/fake it [19:18] how ? [19:20] TJ-: is that easier than just using different subdirs in one EFI? [19:22] cbreak: well, tmp1 said they want to keep them all isolated. If you allow all the ubuntu installs to use the same EFI-SP they'll all think they own /EFI/ubuntu/ and overwrite files there, which for LUKS protected installs will include the cryptdisk device for that installation [19:23] cbreak: if the GRUB_DISTRIBUTOR variable is over-ridden then yes, separate directories in a single EFI-SP would work [19:24] I would try to make ubuntu use a different subdir name [19:24] tmp1: ^^^ the path in EFI-SP is formed as /EFI/${GRUB_DISTRIBUTOR}/ [19:25] trying [19:25] tmp1: so you could do for each distro " echo 'GRUB_DISTRIBUTOR="distroX"' > /etc/default/grub.d/local.cfg " [19:26] tmp1: changing distroX for each of course [19:26] but grub-install from one distro do not see the others.. [19:28] tmp1: that's what you want if you want them isolated. You then use the UEFI boot menu stored in the firmware NVRAM to choose the distro to boot [19:29] tmp1: because each distro is inside a different LUKS container, each needs its own GRUB that knows which LUKS device to unlock [19:29] is there any online ressource documenting it ? [19:29] yep, that's my "problem" [19:29] if you have distinct, separate grubs, that should work I think... [19:30] they're basically as separate as windows from linux would be, dualbooting [19:30] p [20:33] anyone know why my apt download is going so slow? It's going at only 40kbps over mobile broadband, meanwhile my speedtest is showing 8MBps [20:34] Get:2 http://us.archive.ubuntu.com/ubuntu focal-updates/main amd64 firefox amd64 94.0+build3-0ubuntu0.20.04.1 [56.0 MB] [20:34] 21% [2 firefox 46.2 MB/56.0 MB 82%] 70.5 kB/s 33min 38s^ [20:35] nshirelaptop, idk, probably your isp transparent cache ? [20:35] I am on a connection that is ipv6 btw [20:38] apparently the ubuntu servers just son't work with ipv6 [20:40] the workarounds all say to disable ipv6 and force it to use ipv4... which is a problem when the connection is ipv6-only [21:14] aren't there ipv6 mirrors? [21:16] nshiirelaptop needs to deploy DNS64/NAT64 on the gateway :) [21:17] isn't that the provider's job? === guntbert_ is now known as guntbert [21:23] I'm trying to configure vsftp but when i'm testing with filezilla i got this error: GnuTLS error -15 in gnutls_record_recv: An unexpected TLS packet was received. [21:30] why does it try tls? [21:30] does vsftp use tls? [21:30] (normal sftp is tunneled via ssh I think) [21:35] yes [21:36] do you want to see the vsftpd.conf? [21:58] who was it had the apt issue earlier!? I've built apt for focal with the URL patches for testing [22:02] got it [22:03] lol that username [22:03] slightly inappropriate but also finny [22:03] hi. Trying to build a rust project for ubuntu 18.04, it throws an error that I don't have `ncurses.h`, which package provides that file? [22:04] heloo [22:05] ah, looks like libncurses5-dev [22:05] thanks! [22:05] FreeBDSM, wouldn't that be a header file you have to install yourself [22:05] oh [22:05] you're welcome [22:07] anyone know of a workaround to get apt working better on ipv6? I've had this issue for several years actually. Previously I chalked it up to a bad data connection but in reality it's something to do with apt repos [22:08] basically my mobile hotspot ONLY does ipv6, no ipv4. all the workarounds say to disable ipv6 connectivity.. which only works for people who have dual ipv4 and ipv6 connectivity [22:09] nshirelaptop: Do not recall how old - but look at https://www.vultr.com/docs/force-apt-get-to-ipv4-or-ipv6-on-ubuntu-or-debian . [22:27] hi all! i posted a bug on https://bugs.launchpad.net/ one month ago. the bug is not assigned yet. its my first bug. can somebody explain how this process usally handled? [22:40] okaaaaa: it depends on a package maintainer seeing the bug and being interested in fixing it. If the package is directly imported from Debian for example, then it needs reporting to Debian directly, since Ubuntu imports and builds the Debian package but doesn't maintain patches [22:40] okaaaaa: give us the bug URL [22:52] tommy` / cbreak: vsftpd is an FTP server, which also supports FTP+TLS=FTPS. It does not support SFTP, a file transport protocol based on SSH, which in use (available commands), not in implementation (different code), bears some similarities to the FTP protocol. SFTP (think OpenSSH) provides much better security than FTPS (think vsftpd) does, should be preferred. I think FileZilla supports both protocols. [22:52] tomreyn: i just solved by fixing user and group problem [23:11] tommy`: very well. keep in mind that FTP(S) is annoying (and somewhat resource intensive) to use with firewalling and NAT, though. https://en.wikipedia.org/wiki/FTPS#Firewall_incompatibilities [23:11] tomreyn: now I'm struggling to understand how give access to the user other folders [23:11] with vsftpd it's not so immediate [23:12] i want give him access to a folder that is from another user [23:13] tommy`: i do not remember how to configure vsftpd, but i think they have good documentation. i haven't used it in years, and would recommend replacing it by an openssh server. [23:14] openssh allow transfer big files? [23:14] sure [23:14] I am using Ubuntu 21.04 (about to upgrade) and using KDE. I set my colour scheme to Haptic (HapticDark) and now System Settings won't open. [23:14] Help? === Guest6346 is now known as swahili [23:42] looks like the chromium snap that ships with ubuntu 21.10 does not have the Ozone platform patches... any way I can work around this? [23:45] is there a non-snap? [23:50] diniwed: https://github.com/ungoogled-software/ungoogled-chromium-debian should probably work [23:57] TJ- see pm