[00:01] TJ-, yes, the file is just copied "grub-install : information : copying `/usr/lib/shim/shimx64.efi.signed' -> `/boot/efi/EFI/@dingo/shimx64.efi'. [00:01] " [00:02] when I "md5sum /boot/efi/EFI/@dingo/*" or "md5sum /boot/efi/EFI/@cuttlefish/*" the md5sum are the same expect for grub.cfg [00:02] but it's always "/boot/efi/EFI/ubuntu/grub.cfg" that is taken anyway [00:02] so it means it's hardcoded [00:03] TJ-, I know there something about signed images, but I don't see how to not use that [00:03] don't know how [00:05] mirak: are you using Secure Boot? [00:05] mirak: "apt list --installed grub*signed* " [00:08] TJ-, grub-efi-amd64-signed/disco,now 1.115+2.02+dfsg1-12ubuntu2 amd64 [installé, automatique] [00:08] it's installed [00:09] mirak: there's the cause then, when you do grub-install all it does is COPY the binary from that package which obviously has the same path [00:10] mirak: just watch "grub-install -v ..." [00:10] TJ-, yes, obviously [00:11] mirak: to work around it you'd have to add your own Machine Owners Key (MOK) to the system, build your own grub core.image files and sign them [00:11] TJ-, however what doesn't make sense is that you can't really prevent ubuntu from changing things that don't break your setup [00:12] TJ-, if it was automated it would be be ok, but I guess it's not [00:12] mirak: yes you can; only all grub-install to run in ONE ubuntu install and then let GRUB's own menu pick up the other OSs using os-prober functionality [00:12] s/only all/only allow/ [00:13] TJ-, the other solution, would be that ubuntu and grub, doesn't change EFI/ubuntu/grub.cfg . This way it could point to another grub.cfg [00:14] that grub.cfg would boot core.img from other partitions [00:14] that's what I do on bios systems [00:14] mirak: that's what I said if you build your own core.image [00:15] TJ-, but each time the system update-grub, it will mess [00:15] I am not sure what is your idea here [00:16] mirak: but to do that you have to either 1) disable secure boot and remove the grub-signed package 2) enroll your own MOK, remove the grub-signed package, and sign your own core.image or 3) only allow ONE GRUB to do grub-install/update-grub for all Ubuntu installs, and rely on os-prober in that install to build a menu that includes all versions [00:17] TJ-, as far I remember, grub-install can only be run manually, so it's ok [00:17] mirak: my idea is, if you DO NOT use the grub-signed package, then you get to build your own custom core.image every time you run grub-install, so it'll take and honour the options you give it, like --bootloader-id=XXXX [00:17] however update-grub is run at each kernel upgrade [00:18] and it seems update-grub is calling efibootmgr [00:18] mirak: "sudo dpkg-divert --divert /usr/sbin/update-grub.divert --rename /usr/sbin/update-grub; sudo rm /usr/sbin/update-grub; sudo ln -s /bin/true /usr/sbin/update-grub" [00:19] mirak: do that ^^^ in all but 1 of the Ubuntu installs [00:19] TJ-, it seems it doesn't touch efibootmgr, sorry [00:20] mirak: grub-install calls efibootmgr [00:20] only grub-install does [00:20] you'd want to dpkg-divert /usr/sbin/grub-install too as a precaution [00:20] I see [00:21] that way any scripts/tools that do call those 2 get a /bin/true reply and assume it worked [00:21] and any package upgrades only change the diverted+renamed file [00:21] in fact there is no way to avoid grub-install to modify EFI/ubuntu/grub.cfg [00:22] that would do the trick, if I put the multiboot info straigth into EFI/ubuntu/grub.cfg [00:22] mirak: there isn't for the *signed* core.image, but there is for the regular unsigned custom-built-linked version [00:22] and if it doesn't change efibootmgt, because as you say, you still need to run grub-install to build the core.img [00:23] TJ-, what is the trick ? sorry, i missed it [00:23] I kind of get it [00:24] without the grub-signed package, setting GRUB_DISTROBUTOR= should be sufficent to handle both update-grub and grub-install using a different path in /EFI/${GRUB_DISTRIBUTOR_/ [00:24] My typing is going AWOL; too tired [00:26] but what I did on my bios install was a grub .cfg with multiple entries like "multiboot /@dingo/boot/grub/core.img" . When I want to add a new system, I just create another line. And create the core.img with grub-install when chrooting the new system [00:26] TJ-, ok thanks [00:26] TJ-, but will my windows 10 still boot ? [00:27] mirak: BUT it doesn't (at least on 18.04) - must be a bug there somewhere [00:27] TJ-, ha [00:27] mirak: this won't affect Windows, assuming it also boots using UEFI (unless Windows insists on Secure Boot and you've gone the disable-secure-boot route) [00:31] ok [00:31] good night [00:31] thanks [00:37] mirak: OK, I was wrong, GRUB_DISTRIBUTOR="xxxx" does work BUT xxxx cannot contain anything but the ASCII chars A-Z (I had a space in it!) [00:41] and they get converted to lower-case for use [00:46] I'm trying to intall the `hstr` package, following the instructions here: https://github.com/dvorka/hstr/blob/master/INSTALLATION.md#ubuntu I'm getting an error: https://pastebin.com/XiT3qWYd Not sure if it's a good idea to force-download it, with `trusted=true` or something...anyone know how to fix this? :) [00:46] install* [00:49] i read unicorn ... Ubuntu 14.10 Utopic Unicorn ? [00:49] man, what a mess with ppa's and foreign debian stuff [00:55] remove that apt.typesafe.com, and https://launchpad.net/~linaro-maintainers/+archive/ubuntu/toolchain does not have bionic packages [00:56] :-) [01:01] i need help creating a grub iso to boot a physical partition. i have a linux dual boot with windows using grub-efi. windows, linux, and efi partitions are all on the same drive. i also want to boot the linux partition as a guest under vmware in windows. vmware can't open the efi partition (which is in use by windows), so i need to make a new grub iso and put it in a virtual optical drive. how do i do ... [01:01] ... this? tutorials i've seen tell me to use files from under /usr/lib/grub/i386-pc/, but my system doesn't have that, as it doesn't use BIOS grub, it just uses efi grub. i have /usr/lib/grub/x86_64-efi and /usr/lib/grub/x86_64-efi-signed. here's one such tutorial: https://lifehacker.com/how-to-dual-boot-and-virtualize-the-same-partition-on-y-493223329 [01:02] are there screensaver settings in ubuntu? === zbenjamin is now known as Guest18813 === zbenjamin_ is now known as zbenjamin === westor{off} is now known as westor [01:05] I'm having problems with tracker package. Every time I start downloading with qBittorrent, tracker starts up and eats A LOT of CPU. [01:05] When I stop download, tracker also stops. [01:07] The process name is tracker-extract and tracker-miner-fs [01:08] And I cant even delete the tracker package bc ubuntu desktop depends on it [01:08] Very frustrating, please tell me what to do [01:09] tracker is a search indexer. bit torrent does a lot of writes all the time. [01:09] you can probably set up tracker to exclude specific directories. [01:10] the maintainers of the qbittorrent package should have added that exclusion for where ever qbittorrent downloads to. [01:10] you should contact them and complain. [01:10] note: complain the maintainers of the ubuntu package, not of the qbittorrent software source code [01:11] I honestly dont think tracker should reindex the whole file every time a single byte is written. [01:12] neither do i [01:13] cheater: you could create a second EFI-SP and copy the GRUB boot files into it, and attach that to your VMware guest [01:13] TJ-: i'm not sure how to do that. i wouldn't like to fuck with EFI since it's fragile as it is. i'd much rather set up a bios grub that i can boot from. [01:13] cheater: GRUB's core image (grubx64.efi) rarely needs to change so it would be pretty reliable [01:14] TJ-: i'd still like to use a BIOS grub, just to be happy [01:14] Ubuntu bug tracker is a mess, I wish someone would direct me to where do I post my complaint. [01:14] And I'm much more interested in disabling this thing. [01:15] Disabling usage history in setting did not help. [01:16] cheater: create an additional (small) partition (by small I mean between 2 and 64MB) then format it "mkfs.vfat -F 16 /dev/sdX9", mount it: "sudo mkdir /mnt/efi; sudo mount /dev/sdX9 /mnt/efi; sudo mkdir -p /mnt/efi/EFI/ubuntu; sudo cp /boot/efi/EFI/ubuntu/* /mnt/efi/EFI/ubuntu/; sudo umount /mnt/efi " [01:16] fastfresh: i don't think it's about usage history [01:17] TJ-: thanks. i'll copy that to my notes and if i can't figure out how to get a bios grub happening, i'll use your method. [01:17] cheater: if you want to do BIOS boot you could create a 2MB partition (rarely needs more than 1MB) in any spare space on the disk (type EF02) and "grub-install /dev/sdZ9" it from "grub-pc" package [01:18] type EF02 (in gdisk tooling speak) is a BIOS_BOOT GPT type [01:19] cheater: that's allow a VMware BIOS to boot GRUB [01:19] cheater: but as I said, you'd need to use 'grub-pc' package to do the grub-install correclty for BIOS mode boot [01:20] TJ-: could i do the following? 1. boot vmware using ubuntu iso placed in a virtual cdrom, the physical linux partition on my ssd, and a vmdk for grub. i'm booting ubuntu in live mode. 2. somehow install grub in live mode to the small additional partition, and make its menu boot the physical ubuntu? [01:21] cheater: e.g. to create the partition "sudo sgdisk --new=0:0:0 --typecode=0:EF02 /dev/sdZ" will find the largest remaining free space on sdZ and use it for the next partition number that is available [01:21] cheater: the Ubuntu ISOs have hybrid boot capability; however, in BIOS mode they use Syslinux NOT GRUB. GRUB on those is only used for UEFI mode boots [01:22] right, but even if they use syslinux, can i still install grub-pc on the live system to use it as you described? [01:22] OerHeks was that meant for me? [01:22] cheater: correct [01:23] ok, great! let me try that and see what happens :) thank you [01:23] cheater: your plan sounds feasible [01:23] cheater: try to do a walk-through without writing any files to disk initially though! [01:23] thanks. how do you do that? [01:24] cheater: um! tell vmware everything is read-only maybe? Or just don't take any steps that'll write into the SSD, only into your vmdk file [01:24] oh right, yeah [01:24] well, i've checked writing to the ssd, and that works [01:25] i wrote a directory [01:25] (there was an issue where virtualbox couldn't do it) [01:25] cheater: yes, but I'm referring to writing a partition :) [01:25] cheater: if you made a mistake that could mess the partitioning up seriously [01:25] funnily enough virtualbox can access the efi partition, but not write to the linux partition. vmware can write to the linux partition, but not access the efi partition. i like this more [01:26] cheater: I'd highly recommend you save a backup of the current partition tables before doing anything [01:26] oh, no, i'm not repartitioning my disk drive at all [01:26] i'll not even be using sgdisk. i'll use cfdisk instead. i prefer it much more. [01:26] cheater: if you're working with GPT I'd recommend using the *gdisk versions (g = GPT ) [01:27] cheater: so cgdisk [01:27] cfdisk does support gpt though [01:27] cheater: not as well! [01:29] OK thanks for the warning [01:29] let me reboot and try this now :) [01:33] Hey. I have an wacom intuos s (small) wired version (non-bt) and the tablet works in 18.04 for the most part, but doesn't show in the device settings applet. Wouldn't be a thing but I have multiple monitors so without that I'm either doing xsetwacom shenanigans to map to one monitor [01:34] Any tricks for getting the control panel to recognize that it's a wacom? [01:36] Hello. I want to use VeraCrypt to encrypt a Samsung EVO SSD (enclosed in a USB3.0 case). Need to know how do I (or do I need to), run FSTRIM on it? Thanks ... [01:36] Ubuntu 18.0.2 is my Operating System [01:40] I used to be able to double-click an icon on the tray at the top to bring the application back, now it doesn't do that anymore. [01:40] How can I get that back? [01:40] Mr_Cyclops: I answered this for you yesterday 20 seconds after you left! [01:40] 20:36 FYI: Mr_Cyclops enquiry. for LUKS, in crypttab, supprting FSTRIM with the "discard" option. For cross-platform cryptsetup can support veracrypt [01:41] Why would GNOME remove a feature like that? [01:41] gnome exists to remove user features [01:42] I really hate gnome and wish there was something better. [01:42] * styledash laughs in KDE Plasma [01:42] Can you double click icons in the tray to bring them back to view in KDE plasma? [01:43] Right now that's all I want. [01:44] So, anyone have ideas how to stop tracker without crashing pc with a hammer? I'm frighteningly close to that point. [01:45] You might be interested in Tails OS [01:46] No im not. I wonder if everyone else is having this issue without realizing it. [01:47] TJ-, aha! :) Thanks buddy ... I am gonna try it out on an SSD with dummy data tomorrow [01:48] TJ-: are you sure the syntax is grub-install /dev/sdZ9? [01:48] TJ-, Sorry I missed the info yesterday, I realized I rebooted my computer, which took away the session I guess [01:49] leonardus, may I suggest Ubuntu Budgie. Ever since I put that on an year ago, never looked back at anything else. You could try it out on one of the virtualization softwares, like Virtual Box (which is free) [01:50] TJ-: i ask because this seems to use a different syntax and i'm not sure why. https://askubuntu.com/a/757130/854479 [01:50] Mr_Cyclops: So in ubuntu budgie you can double click icons in the task tray to bring the applications back? [01:50] tracker, this centos post says something about tracker-preferences .. https://unix.stackexchange.com/questions/481206/tracker-extract-high-cpu-usage [01:50] fastfresh ^ ^ [01:50] yes @leonardus .. single click actually [01:50] Looks like I'll be using Ubuntu Budgie. [01:50] cheater: yes, but of course you change sdZ9 for your actual device name [01:50] Ubuntu budgie takes its desktop environment from popular Solus OS [01:51] I'm on "a start job is running for "Raise network Interface"... [01:51] I remember a hotkey to get past this wait... [01:51] I can send you a screenshot if you like to get you a glimpse of how it looks [01:51] but can't seem to figure it out. Any clue? [01:51] cheater: if you're executing grub-install from somewhere other than the installed OS, then you will need additional arguments since grub cannot figure out many settings [01:52] TJ-: yes, i am executing it from the live dvd, not from the installed os. i can't boot the installed os without grub, which is why i'm making grub. [01:52] TJ-: so i use a syntax like there, where /mnt is my installed os on my ssd? [01:52] cheater: OK, time to explain a bit more then ... [01:52] go on! [01:52] thanks [01:53] OerHeks: thanks, will try [01:53] btw, the live dvd comes with grub-install, but it's from the grub2 package, not grub-pc. is that fine? [01:54] cheater: grub2 is grub-pc [01:54] its fine [01:54] ty [01:58] TJ-: in case you're typing something, i'll brb (need to go to the kitchen) [01:59] cheater: ... assuming your OS is in /dev/sdZ4 and it includes the /boot/ directory and content (as in, you're NOT using a separate partition/file-system just for /boot/) and your separate 'device' (the vmdk?) for GRUB is /dev/sdY ... so you do "sudo mkdir /target /mnt/boot; sudo mount /dev/sdz4 /target; sudo mount /dev/sdY /mnt/boot; sudo grub-install --verbose --root-directory=/target [01:59] --boot-directory=/mnt/boot/ --recheck /dev/sdZ " ... and I'm not 100% sure I have all the options on grub-install you may need there [02:00] how do i mount /dev/sdY? how do i make it mountable? [02:00] thank you btw [02:01] cheater: although I think I missed steps there since /dev/sdY may need to be partitioned, and certainly needs a file-system format with "mkfs.ext4 ..." [02:01] cheater: I'm assuming you're inside the VMware guest and you've added a GRUB vmdk file as an additional device? [02:01] cheater: so you replace sdY with the actual device name that device has [02:02] yes, it's a vmdk [02:02] can you mount unformatted drives without a partition table, or do i need to give it a partition table and format it somehow? [02:03] i know how to mount a formatted partition, but not a whole drive [02:03] cheater: if it needs partitioning you can do "sudo sgdisk --new=:0:0:0 --typecode=0:EF02 /dev/sdY" which will assign all its space to a single partition [02:04] cheater: I'm struggling to figure out what actually you're expecting to put on the vmdk 'sdY' - is it grub's bootstrap + core.image, or does it also want the grub modules (in which case it needs a 2nd partition formatted with mkfs.ext4 ) [02:04] cheater: if the vmdk is supposed to be for 'all' of grub then we need to revise things a bit [02:04] TJ-: i'm not even at the stage where i'm trying to understand that yet [02:05] TJ-: the way i have it right now is there's one partition on it. the vmdk is /dev/sda, and the partition is /dev/sda1. but why did you tell me to mount the whole disk? is that possible? [02:05] cheater: "sgdisk --new=1:0+2MB --typecode=1:EF02 --new=2:0:0 --typecode=2:8300 /dev/sdY" then "mkfs.ext4 /dev/sdY2" [02:05] eeek, typo! [02:06] cheater: I missed a colon out. "sgdisk --new=1:0:+2MB --typecode=1:EF02 --new=2:0:0 --typecode=2:8300 /dev/sdY" then "mkfs.ext4 /dev/sdY2" [02:06] can you describe what this command does? i'd like to create it with cgdisk, which i trust more than using a cli [02:07] cheater: then you could do "mount /dev/sdY2 /mnt/boot; grub-install --root-directory=/target --boot-directory=/mnt/boot /dev/sdY" [02:07] cheater: creates 2 partitions, first is bios_boot of 2MB, second is linux using remaining space [02:08] how much remaining space do i need? [02:08] cheater: for the 2nd partition that'll contain grub modules? no more than 15MB :) [02:08] thanks! why does the second partition contain linux? [02:09] cheater, "linux" is the typecode name for 8300 [02:09] cheater: do "sgdisk -L" to see the typecodes and their friendly names [02:09] ah, i thought you meant "bios_boot" was a label. it's the type, right? [02:09] correct [02:10] bios_boot is EF02 [02:10] great! [02:11] grub-install will read the files it needs from the existing OS mounted on /target and write the files to /mnt/boot as well as writing its core.image into /dev/sdY1 and its bootstrap into sector 0 of /dev/sdY [02:13] re - I ended up finding a solution that works for the intuos S (aka CTL-4100) https://github.com/linuxwacom/wacom-hid-descriptors/issues/41 [02:13] cheater: what you then need to take care of is generating a correct /mnt/boot/grub/grub.cfg for which you'll likely need to use a chroot of /target with /mnt/boot/ over-mounted on /target/boot :P [02:13] huh? [02:13] seems that ubuntu doesn't have the updated input db packaged. [02:14] cheater: grub-install only writes the executable parts of GRUB; it doesn't handle the config script /boot/grub/grub.cfg [02:14] cheater: and that is generated by update-grub running in the OS [02:15] oh... so... what do i do then? [02:15] sorry about being so clueless about this [02:16] cheater: so you'd need to do something like "mount --bind /mnt/boot /target/boot; for n in proc sys dev etc/resolv.conf; do mount --rbind /$n /target/$n; done" then "chroot /target" and "update-grub" to write it, then "exit" to leave the chroot, then "for n in etc/resolv.conf dev sys proc; do umount /target/$n; done; umount /target/boot" [02:16] I'm a little confused about how I should be using this to install a wifi receiver. https://github.com/mathew19/rtl8814AU [02:17] I believe I've used it in the past and that, when my kernel was updated, I lost whatever clever thing I did last time. [02:17] But I can't seem to reproduce. [02:17] mustmodify: did you use DKMS to install it? [02:17] let me try that, thank you [02:17] Hm... [02:17] #linux [02:18] mustmodify: it might be there but failed to build? "dkms status" might help [02:18] Well dkms isn't in my path, but let me dig a little more. [02:18] mustmodify: OK, if 'dkms' tool isn't installed then presumably you haven't used it [02:19] Well there was a catastrophic failure at some point in the past, and that may have been after I was using this. [02:19] mustmodify: do you want to install it again now? [02:19] Drat. So now I have no internet connection. [02:19] and I need to install something. [02:19] Ugh. [02:19] ok I'll have to work that out and I'll get back with more questions. Thanks. [02:19] if i write a windows iso to usb in ubuntu using the format dd if=windows.iso of=/dev/sdb bs=1M the resulting usb drive doesnt boot windows. is there something i need to do to make it bootable [02:20] devslash: that looks correct to me [02:20] devslash: did you 'sync' after doing the 'dd' to ensure the data was flushed from cache to the device? [02:20] no [02:21] devslash: also, you'll find 'dd' a lot faster with a larger blocksize; For USB2 I'd recommend bs=50M [02:21] & for usb3? [02:21] devslash: sky's the limit :) [02:22] devslash: "dd if=windows.iso of=/dev/sdb bs=100M status=progress conv=fdatasync,fsync" will write it fast, tell you what its doing, and sync data and metadata to the device [02:28] devslash: i dont think windows iso is bootable, if you want to use it with uefi you could create a fat32 partition and put the contents of the iso there [02:28] otherwise you'll need some sort of program that makes it bootable [02:29] maye rufus or something like that could help? [02:29] maybe [02:29] not bootable? really, did that change? [02:29] I recall helping a friend with Windows 7 I think it was, and that definitely bootable [02:29] i mean, why else is there win-usb [02:29] dunno [02:30] how would Windows install if the ISO/DVD wasn't bootable ? or are we referring to something else with "windows.iso" ? [02:30] I'm assuming we're talking about the windows *installer* ISO of course [02:32] win7 iso isnt efi and the computer has efi enabled [02:34] win7 x64 did [02:35] TJ-: i dont think its dd ready dunno about win7 but win10 for sure [02:35] (it doesnt have a bootloader in the mbr) [02:35] this program used to be a thing but its not maintained anymore [02:35] !winusb [02:35] Ubuntu does not have a recommended tool for preparing bootable Windows installation USBs. If you need one, consider using Microsoft's media creation tool from a Windows computer (see ##windows for support). If this is not an option, there are third-party unsupported tools like WoeUSB ( https://github.com/slacka/WoeUSB , sudo snap install --edge woe-usb ) that might work. [02:39] EriC^^: that'd only affect BIOS mode if it has no bootstrap code, but I'm pretty sure (from having slipstreamed Windows installers) it does have the bootstrap code. [02:44] EriC^^: ahhh, the reason is, the Windows 10 ISO is a pure ISO9660-El-Torito boot mode, it isn't hybrid [02:44] anyone using multiple monitors and have night light enabled? [02:45] Win10_1903_V1_English_x64.iso.part: ISO 9660 CD-ROM filesystem data 'CCCOMA_X64FRE_EN-US_DV9' (bootable) [02:45] kinghat: wha'ts night-light, the red-tinge adjustment? [02:45] ya [02:46] kinghat: I used to have something like that on 1 of my PCs, but don't recall its name now. I found it made things worse :) [02:49] I just killed my server with a stupid mv command and need help fixing [02:50] I was reading on how to mv all the subdirectories in a directory, and didn't read it close enough and ran "sudo mv /*/ ../../folder" [02:50] I realized after a few error messaged the mistake and ctrl^c, but it was too late [02:53] I need to mv /dev/ and /bin/ and /lib/ and /etc/ and /boot/ and /home/ back, but the mv command no longer works [02:54] any ideas? [03:07] Anyone? [03:08] !patience | qwebirc15353 [03:08] qwebirc15353: Don't feel ignored and repeat your question quickly; if nobody knows your answer, nobody will answer you. While you wait, try searching https://help.ubuntu.com or https://ubuntuforums.org or https://askubuntu.com/ [03:11] qwebirc15353: When you say server - leaves a lot out ... raid ? LVM ? EXT4 file system ? depending on what we know is how we can help. [03:13] i swear there was a usb writer that came with ubuntu(usb iso) but now that i have it installed i cant find one. [03:13] kinghat: the ubuntu usb creator? [03:14] ya [03:14] kinghat: its the icon, with a harddisk and ubuntu logo and green arrow [03:14] "startup disk creator" [03:14] ya i guess its not installed for me.. [03:14] kinghat: your ubuntu release? [03:15] fresh install of 18.04.2 [03:15] kinghat: ubuntu-desktop? [03:15] just used the store to install it now [03:15] ya [03:15] then it has to be there [03:17] i literally just had to install it [03:18] Bashing-om:Yes, LVM and Ext4. No raid. [03:18] I can run most commands by using the full path to the newly placed bin directory, but mv fails now due to permission [03:19] that's becasue etc was moved and I now get "sudo: unknown uid 1000: who are you?" [03:20] qwebirc15353: I can not assist with LVM ,, a lack on my part :( encryption ? But I would think one could mount the individual partitions from an extermal source and copy the data back - somehow. [03:20] so I can't move etc back withoud sudo/ID working, and I can't get sudo/ID working without moving etc back [03:21] Yeah, my searching is suggesting booting to a live USB and moving back that way [03:21] but it's a headless server, so I'm not sure what to do about that... except unplug it and take it somewhere with a monitor and keyboard [03:21] qwebirc15353: Yeah .. why you work from say a liveUSB. [03:22] I could configure the liveUSB from my laptop, but still not sure if it would boot once I plug it into my server... I guess I'll just need to try that before getting a monitor and keyboard hooked up [03:23] I was just hoping there was some way to could switch to root or something. I still have an SSH session open, but it isn't letting me establish new sessions [03:24] su root gets me "su: Cannot determine your user name." [03:29] is there a lightweight live distro you can recommend for live booting to do this task? [03:30] OK, I've installed dkms. [03:30] Now, how do I get the rest of this working? [03:30] https://github.com/mathew19/rtl8814AU [03:30] I tried `sudo make install` and that seemed to work, but then I restarted and no network connection. [03:34] mustmodify: Something like ` sudo dkms install rtl8814AU ' . Assumming that is the correct driver for your card. [03:37] ok, working on that and I got something surprising. "Error! Your kernel headers for kernel 4.15.0-45-generic cannot be found. Please install ... or use --kernelsourcedir [03:37] So I'm assuming I must have it installed ... if that's what I'm using, right? [03:37] So I just likely need to point it to where the kernel lives... [03:40] mustmodify: Don't know much, but, ' dpkg -l | grep linux- 'show the needed package as installed ? === [b0b] is now known as b0b [03:56] I can't get past the network connections page of the LiveUSB "Network configuration timed out". I've tried DHCP and static [03:56] how long did you wait? [03:56] until the timeout message came up === mundus2018 is now known as mundus [04:10] what is LTS after 18? [04:10] kinghat: 20.04 [04:11] so 18.10, 19.04, 19.10, 20? [04:11] kinghat: .04 april .10 october [04:11] Frisky Feline [04:12] 20.04 april next year? [04:12] kinghat: yes [04:12] wasnt sure if you were talking about the 19 series.. [04:13] Eoan Ermine is such a bad name tho... [04:13] !lts | kinghat [04:13] kinghat: LTS means Long Term Support. LTS versions are supported for 5 years on the desktop and server. The latest LTS version of Ubuntu is !Bionic (Bionic Beaver 18.04). Ubuntu !flavors may have different support durations, check their release notes for information. [04:14] kinghat: 19.04/19.10 are just regular releases [04:25] can you make an app open on a specific desktop? [04:27] kinghat: check out devilspie [04:28] (and gdevilspie) [04:34] i need help creating a grub to boot a physical partition. i have a linux dual boot with windows using grub-efi. windows, linux, and efi partitions are all on the same drive (drive 0, /dev/sda; linux is on /dev/sda5). i also want to boot the linux partition as a guest under vmware in windows. vmware can't open the efi partition (which is in use by windows), so i need to make a new grub and put it on a ... [04:34] ... virtual drive (drive 1, /dev/sdb). I want to use a vmdk attached as a separate virtual scsi drive. i'd like to boot into a ubuntu live dvd, and use that to put a bios version of grub on /dev/sdb that will let me boot /dev/sda5. there is /boot on /dev/sda5, including the vmlinuz image. how do i do this? I spoke to TJ- earlier but I wasn't able to follow his advice, as several points were not clear ... [04:34] ... to me and some of the steps yielded errors. [04:49] ducasse: https://hastebin.com/cawugetuna.rb [04:54] kinghat: If ' dpkg -l gdevilspie ' shows installed ,, 1) I expect the g is for Graphic such that there is an icon ,, and 2) I would expect also that trminal command `gdevilspie` will also start the tool . [04:56] no packages found [04:57] !info gdevilspie bionic | kinghat [04:57] kinghat: Package gdevilspie does not exist in bionic [04:58] !info devilspie bionic | kinghat [04:58] kinghat: devilspie (source: devilspie): find windows and perform actions on them. In component universe, is optional. Version 0.23-2build1 (bionic), package size 45 kB, installed size 227 kB [05:04] cheater, i'm having trouble following your requirements there. but it sounds like you need to chroot and run grub-install ? [05:12] !ping [05:12] pong! === onohiroko is now known as sarashina_ [06:52] Hi all. I am trying to see which versions of gcc and cmake ubuntu has by browsing https://packages.ubuntu.com/ [06:52] Strangely enough, when I select any branch: xenial, bionic, cosmic - all are marked as dev. Which of them is stable branch? I am a bit lost. [06:55] !info | bs0d [06:55] 'bs0d' is not a valid distribution: bionic, bionic-backports, bionic-proposed, cosmic, cosmic-backports, cosmic-proposed, disco, disco-backports, disco-proposed, eoan, eoan-backports, eoan-proposed, kubuntu-backports, kubuntu-experimental, kubuntu-updates, partner, precise, precise-backports, precise-proposed, stable, testing, trusty, trusty-backports, trusty-proposed, unstable, xenial, xenial-backports, xenial-proposed [06:55] !info gcc | bs0d [06:55] bs0d: gcc (source: gcc-defaults (1.176ubuntu2.2)): GNU C compiler. In component main, is optional. Version 4:7.4.0-1ubuntu2.2 (bionic), package size 5 kB, installed size 64 kB [06:56] Version 4:7.4.0 means gcc 7.4.0? [06:56] !info cmake [06:56] cmake (source: cmake): cross-platform, open-source make system. In component main, is optional. Version 3.10.2-1ubuntu2 (bionic), package size 3362 kB, installed size 18363 kB [06:57] looks good to me. I am looking for an OS for my container to run a recent C++ toolchain. If I may ask a stupid question - which distro is stable and which is development? [06:58] bs0d= LTS is the most "stable" and longest supported. non-LTS has the latest/greatest shiny stuff. not knowing the details of your use case, can't advise. personally: LTS only. [06:59] bs0d: the 4: is a 'epoch' purely for packaging purposes, so yes, the upstream version follows the : [07:00] bs0d: non-lts means you need to upgrade every 6-9 months, lts you can go for 5 years [07:01] thanks a lot. Now the list of distros makes sense :) xenial and bionic are LTS. cosmic, disco and eoan are non-LTS [07:02] exactomundo! [07:18] hi everybody, is there any security tools in ubuntu like netstat, tracert and so on, obviously i haven't got the tools as i tried in the shell [07:19] tuxi: try 'apt install inetutils-traceroute' - all these tools can be installed [07:20] ducasse thank you very much [07:20] hm does not work in the terminal as root, though :-( [07:22] tuxi: there is also 'apt install traceroute' - unsure what the difference between the two are, but look at the package descriptions [07:23] ah sorry it worked i made an error [07:33] Hi. I am using Ubuntu 18.04 and in Software & Updates I have checked "Automatically check for updates:" = "Never" and "Notify me of a new Ubuntu version" = "Never". Yet each time I restart it checks for updates. [07:33] I would like to make it check manually only. [07:34] I mean when I restart it checks and then opens a new dialog box "there are new updates". I don't want that, which is why I checked those settings in the dialogs. [07:34] hey TJ- [07:34] oh, i was in scrollback, TJ isn't here [07:34] cheater: TJ- is afk for now [07:34] yup [07:35] nblade42: disable the startup service or uninstall update-manager [07:45] lotuspsychje: How? I tried the command systemctl list-unit-files --type=service | grep -i update [07:45] But I didn't see anything interesting. [07:46] nblade42: type this in terminal: sudo sed -i "s/NoDisplay=true/NoDisplay=false/g" /etc/xdg/autostart/*.desktop [07:46] nblade42: then check your startup services icon === rkta_ is now known as rkta [08:08] When i tried to boot kdeneon from usb on new laptop [08:08] i had to add 'acpi=off' to the grub on boot [08:08] what's acpi? [08:09] !acpi | Jonopoly [08:09] Jonopoly: to debug ACPI issues on ubuntu make sure your bios is up to date and follow the procedure here: https://wiki.ubuntu.com/DebuggingACPI [08:09] ah a bios thing! okay [08:12] i've just got an asus (gaming) laptop the other day and i wanna put linux on it.. for development i want windows too because i need windows for gaming/VR [08:13] just for safety would it be advised to back up my entire windows? as it contains some ASUS files 'controlling fan' etc.. before proceeding to dual boot? [08:14] Jonopoly: its always adviced to make a backup of your data [08:14] lotuspsychje: would that backup include windows etc..? 'bootloader' [08:14] just worried i'd get stuck with grub if i had to revert [08:15] Jonopoly: to backup windows you can sk in ##windows [08:15] ask [08:16] !uefi | Jonopoly [08:16] Jonopoly: UEFI is a specification that defines a software interface between an operating system and platform firmware. It is meant as a replacement for the BIOS. For information on how to set up and install Ubuntu and its derivatives on UEFI machines please read https://help.ubuntu.com/community/UEFI [08:16] !dualboot | Jonopoly [08:16] Jonopoly: Dual boot instructions: x86/AMD64: https://help.ubuntu.com/community/DualBoot/Windows - Macs: https://help.ubuntu.com/community/MacBookPro https://help.ubuntu.com/community/YabootConfigurationForMacintoshPowerPCsDualBoot [08:16] !neon | Jonopoly [08:16] Jonopoly: KDE Neon ( http://neon.kde.org/ ) is a KDE project to package the latest stable and development versions of KDE software on top of an Ubuntu base. As it is not an official Ubuntu or Kubuntu project, please use #kde-neon for discussion and support. [08:16] wow thanks [08:19] hello. how to "dmesg -T" but also showing last 1000 lines? it shows not enough by default. or how to check the log with human readable timestamps otherwise etc [08:21] ansyeb, dmesg -T|tail -n1000 [08:28] lotuspsychje: What do you mean by the startup service icon? [08:28] nblade42: are you on ubuntu-desktop with gnome? [08:33] Yes, it is standard Ubuntu 18.04. With the standard GNOME desktop. [08:33] nblade42: ok, then you have a startup applications icon (purple with greeg arrow to the left) [08:33] green [08:35] Yes I checked that list and it does not mention anything about updates. [08:36] nblade42: did you enter that command i gave you? [08:38] nblade42: i see the service here, check for updates [08:58] lotuspsychje: I see in /etc/xdg/autostart/update-notifier.desktop it is set with property NoDisplay=true. I changed that to NoDisplay=false as you suggested (but not for all .desktop files). And now I can see it in startup applications list and can uncheck it. [08:58] +1 nblade42 [09:03] Thanks it is better now. === EriC^^ is now known as ME === ME is now known as EriC^^ [09:11] anyone know where I can get a very low res tux penguin icon? [09:11] eg one for a brand on a directory? === BrianG61UK_ is now known as BrianG61UK [10:36] Hi! I want to install ubuntu desktop on mdadm raid0, i want to raid 3 ssds together into one. [10:36] I managed to install ubuntu on it, but cannot get grub to work. My setup uses EFI. [10:37] halvors: do you have a fat32 partition with grub on it? [10:37] Yeah that is what i am trying to do. [10:39] But then metadata needs to be on the end right? [10:40] Or should the ESP just not be raided? [10:40] Either way i could not get ubuntu to boot. [10:40] eric: Installer was unable to install grub altogether. [10:43] Using this: https://gist.github.com/umpirsky/6ee1f870e759815333c8 i was able to install Grub on sda1, and grub starts but promts to shell as it was unable to load ubuntu (which was on raid0) [10:45] is raid0 even supported by grub [10:47] halvors: yeah i think the esp shouldnt be part of the raid [10:48] hey folks [10:48] can you load a live usb to troubleshoot? [10:48] Yeah. [10:48] But just have esp on sda then? [10:48] Like this, make a raid0 consisting of sda2 (sda1 is esp), sdb and sdc? [10:53] If i want 2 linux distro installs should they share the raid array or and paritioning that? [10:53] i dunno much about raid sorry [10:53] i guess it's fine? [10:54] given they are on separate partitions and stuff i dont see why not [10:55] ok thanks :) [11:00] halvors: firmware (UEFI) doesn't understand RAID so it won't work *unless* you use mdadm metadata level 0.9 or 1.0 (so the emetadata isn't put where the partition table/boot loader need to be) [11:24] Hey everybody, I am trying to configure my bionic as multiarch with arm64 packages being available. I have already set the sources.list to reference the ports.ubuntu.com Now the apt install git:arm64 attempts to install but fails because liberror-perl:arm64 is not installable. In fact it is all package. Is there a way to resolve? [11:43] Anybody know why /etc/sudoers would be ignored? I have an ubuntu container and my users can simply use sudo without supplying any authentication, not even on login. [11:53] adrian_1908: are they all running as UID 0 inside the container? [11:53] adrian_1908: what kind of container? LXD? [11:57] TJ-: Yes, LXD. My ubuntu user does run as itself from the looks of it. If I use sudo, the command runs as root. Sorry, I don't know enough about this. [11:58] TJ-: I use `$root: su - ubuntu` to switch to that user. Is that the wrong approach maybe? [12:03] sudo -u user ;) [12:08] adrian_1908: It's not clear from your description whether you're talking about the experience inside the container, or outside it. alos, which distro/version is installed in the container? what is in its /etc/sudoers and /etc/sudoer.d/* [12:13] adrian_1908: /etc/sudoers.d/90-cloud-init-users [12:14] I am working to restore my Ubuntu 18.04 server after I accidentially moved a few system directories, but now my login screen is stuck in a loop [12:15] meaning, when I enter my username and password, I briefly see the welcome header flash, then I go back to the login prompt [12:15] I believe my problem may be that my home user directory ownership is a UID not a username [12:15] so /home/username is showing 1000:1000 as ownership [12:16] leftyfb: ffs, I think that's it! "ubuntu ALL=(ALL) NOPASSWD:ALL". My /etc/sudoers has the `includedir` line commented out, so I didn't investigate further. === jstein_ is now known as jstein [12:17] I'll check my lxd config for that container and report back, I presume cloud-init is fed from there. [12:17] I tried to chown in archboot liveUSB, but it says it is an invalid user [12:19] my passwd file has username:x:1000:1000:username,,,:/home/username:/bin/bash [12:38] hey folks, I tried to install a HA cluster in AWS (proof of concept) but couldn't find fence-agents and pcs packages, is it normal ? [12:41] Would the bash alias sudo='/usr/bin/sudo -S' be an security issue? I'm trying to get around an LXD issue regarding sudo and tty. I reckon if sudoers is enforced, this cannot be abused somehow, right? [12:46] adrian_1908: I wouldn't recommend that approach. If you want to alias it, give it something unique, like ssudo or something similar. This will prevent other scripts and applications from breaking [12:50] adrian_1908: the #includedir is NOT a comment, # is the prefix. It is NOT shell script [12:50] pragmaticenigma: fair enough, but plain `sudo` is broken anyway. [12:51] TJ-: Oh, I see! [12:51] adrian_1908: right! I know because that caught me out badly some years ago :D [12:52] adrian_1908: I don't understand what you mean by "plain `sudo` is broken anyway." The command sudo has worked just fine for me [12:53] pragmaticenigma: I think adrian_1908 was caught out by #includedir /etc/sudoers.d/ :) [12:54] pragmaticenigma: An alias cannot break a script. [12:54] pragmaticenigma: the original issue, in LXD container, was "all users can use sudo without a password" which, as leftyfb pointed out, was due to /etc/sudoers.d/90-cloud-init-users [12:54] pragmaticenigma: Maybe it wasn't clear, but my comment was regarding an LXD issue: https://github.com/lxc/lxd/issues/3218 that seems to only be solvable with hacks. A developer suggested running `script /dev/null` which somehow creates the tty needed, but I have no idea if that's a better idea than that sudo alias. [12:54] For example: alias ls='ls --color=always' [12:55] For scripts they will not get the color parameter. [12:55] TJ-: two separate issues actually, now that I got sudoers working (the leftyfb hint), I ran into the thing where I want to alias. Sorry for the confusion. [12:56] adrian_1908: you want to force "sudo -S", yes? [12:56] I would like help installing my wifi driver. I think I have most of the pieces in place... I have a repo I believe I have used in the past for this. I know it needs to be installed with dkms. Just having a few errors I can't get past. [12:56] TJ-: right, since that seems to fix the issue. [12:58] adrian_1908: I'm looking/wondering if there's way via sudo.conf [13:00] adrian_1908: the obvious way is to replace /usr/bin/sudo with a shell wrapper script [13:02] How can I get rid of these failed units: https://dpaste.de/skL2 ? They're from a DVD that I had to force-eject because it's scratch and was unreadable. [13:03] s/scratch/scratched [13:04] TJ-: right, I might just do that then. Bit odd, but I doubt I'll use sudo a ton inside the container, aside from the occasional configuration change. [13:04] cxl: A reboot should clear that up [13:08] pragmaticenigma: no problem in keeping it like this and not rebooting for a while? [13:09] cxl: I don't see any issues, the only thing might be applications like file managers that keep looking their for information [13:14] adrian_1908: use dpkg-divert to prevent your wrapper being replaced. "sudo -Si" then " dpkg-divert --divert /usr/bin/sudo.real --rename /usr/bin/sudo; echo -e '#!/bin/sh\nexec /usr/bin/sudo.real -S $@' > /usr/bin/sudo ; chmod +x /usr/bin/sudo " [13:15] TJ-: Awesome, was thinking about that issue. Thanks so much for your help! [13:16] adrian_1908: and make sure the new wrapper works before leaving 'sudo -i' mode! [13:20] what's the latestCan you guys please clarify what is the best way to mount files in windows 10 from ubuntu [13:20] What is obsolete and what is the latest [13:20] samba, smb, cifs? [13:20] I'm getting conflicting information. [13:21] I mean mount shares, not files. [13:21] smb == samba ~= cifs [13:22] and that (samba) is more or less the only recommended thing you can use from Ubuntu, to mount Windows shares over the network. [13:22] so I must get rid of cifs [13:22] ? [13:23] cifs is just a variant of smb, which is protocol. samba is the software suite for smb/cifs protocols [13:23] Hi all just a quick question is there a quick comand to refrest the menu [13:23] if you use smb:// URLs, then that's all you need to do [13:24] HiDeHo= ??? explain [13:25] cfhowlett, add a new .desktop file how to refresh the menu so its added [13:25] why else would the menu need to be refreshed. [13:25] cfhowlett, i am talking about the menu on the panel that shows all the apps [13:31] HiDeHo: where is the .desktop file you're added/changed? I'm not sure which, if any, DE shells rescan or inotify watch for changes but if they do it should be detected and refreshed in the background [13:37] this is from mount.cifs manpage "The CIFS client can get and set POSIX ACLs (getfacl, setfacl) to Samba servers version 3.0.10 and later." [13:38] hi [13:38] Client: HexChat 2.14.1 • OS: Ubuntu "bionic" 18.04 • CPU: Intel(R) Xeon(R) CPU E5640 @ 2.67GHz (1.60GHz) • Memory: Physical: 38.4 GiB Total (33.5 GiB Free) Swap: 20.1 GiB Total (20.1 GiB Free) • Storage: 26.2 GB / 209.1 GB (182.9 GB Free) • VGA: Matrox Electronics Systems Ltd. MGA G200eW WPCM450 @ Intel Corporation Xeon 5600 Series QuickPath Architecture System Addres [13:38] s Decoder • Uptime: 33m 6s [13:38] mount.cifs also supports the vers= option which supports 3.0 - The SMBv3.0 protocol === jstein_ is now known as jstein [13:56] can anyone help me with dns issues on my VM Owncloud? [13:56] !ask | vpeters [13:56] vpeters: Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) See also !patience [14:01] #mode #ubuntu [14:09] Greetings! When I ssh into my fresh install, and I leave a screen session, if I exit, the sceen session is killed too! How do I stop this? I'm on ubuntu 18.04 lts [14:09] Draggor: You can use a application like "screen" which allows you to restore a previous session. [14:09] pragmaticenigma, he is [14:10] Draggor, how are you exiting? [14:10] sorry, misinterpreted the screen session [14:10] I detach from screen, then exit the ssh session. SystemD seems to kill the agent and screen [14:10] Draggor: When you detach, what command are you giving? [14:10] ctrl a + d [14:10] Draggor, can you re-attach before exiting? [14:11] yes [14:11] Draggor: what sort of hardware is this install on? [14:11] A 4 year old thinkpad x260, and a modern thinkpad x1 extreme [14:12] When the ssh session exits, every process gets killed with it [14:12] even screen which has a valid detached named session [14:12] If I ssh in again while still already being in, i can attach on that other ssh session [14:12] but if they all exit, it kills everything [14:12] Draggor: have you tested any command? Something like top? [14:14] ssh in; screen -S test; top; ctrl a + d; exit; ssh in again; screen -list produces nothing; [14:15] after I detach, screen -list shows the properly named session. It will be gone when I exit ssh. [14:15] There's a bunch of articles complaining about this from years ago when systemd got introduced and KillUserProcesses=yes as a default, but changing that to no doesn't seem to do anything. [14:16] Draggor, Try enabling logging in screen with C-a H [14:16] Perhaps that will log a signal [14:17] It does get a signal [14:17] because systemd is sending it [14:17] I would like systemd to not kill it [14:17] Is this not an issue other default ubuntu installs have? [14:17] I am on Ubuntu 18.04 and I notice after installing some applications from the Ubuntu Software Center I see a wrong app icon when starting certain KDE applications. Some are OK (e.g. Amarok and Kdenlive), but others (KmPlot and Kalzium) just show a strange "blank screen with a red 'circle-backslash' symbol in it" icon on the dock. [14:17] Draggor: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=825394 [14:17] Debian bug 825394 in systemd "systemd kill background processes after user logs out" [Normal,Fixed] [14:18] I would like help installing my wifi driver. I think I have most of the pieces in place... I have a repo I believe I have used in the past for this. I know it needs to be installed with dkms. Just having a few errors I can't get past. [14:18] Inside the apps menu (Super+A) the applications have correct app icons, so I think the icons are there but just somehow not getting picked up by the Dock and Alt+TAB. [14:18] leftyfb: Yes, and I have KillUserProcesses=no, and yet I still see this issue. [14:20] Draggor: did you run daemon-reload after making the change and restart systemd-logind? You might even have to log out completely. [14:22] nblade42: Try seeing if there is a *-icon package for your application. Check your logs, you might see what path the application is trying to load it's icon from and then you can place the desired one there. If neither of those are options for you, you can file a bug report [14:26] leftyfb: I restarted the machine in full [14:26] Draggor: I think you still need daemon-reload regardless of rebooting [14:26] oooh, okay [14:28] NopeNope === im0nde_ is now known as im0nde [15:12] pragmaticenigma: If I run the malfunctioning programs in a terminal I see lines like: Invalid Context= "stock" line for icon theme: "/usr/share/icons/ubuntu-mono-dark/stock/16/" [15:14] Draggor: did you figure the screen problem out? [15:15] akik: nope [15:16] looking into sshd service options? [15:17] Draggor: i thought it would've been fixed in systemd. try if this command lets you keep running screen: systemd-run --scope --user screen [15:17] it's documented here: https://www.freedesktop.org/software/systemd/man/systemd-run.html [15:18] there's another command there: loginctl enable-linger which might help [15:45] nblade42: does that path exist? [15:47] Yes. It is installed by the ubuntu-mono package, which is installed. Also I've got a kile.svg icon in /usr/share/icons/breeze/apps/48/kile.svg but somehow it doesn't get picked up by the application. [15:48] So maybe it is a bug in the program itself (Kile) or possibly the way it was prepared for Ubuntu. [15:48] akik: failed to create a bus connection, that's new [15:50] nblade42: No it's an issue with poorly implemented attempts at theming everything. If you were to copy or symlink that image file like this "ln -s /usr/share/icons/breeze/apps/48/kile.svg /usr/share/icons/ubuntu-mono-dark/stock/16/kile.svg" it would probably get picked up by the application [16:04] hello, i am getting "System program problem detected" every time i log in [16:04] how do i view the error report to see what crashed [16:12] solidfox: press more details or similar [16:12] solidfox: they're in /var/crash [16:13] solidfox: often what happens is if you don't submit the error, it hangs around and nags you next time [16:13] EriC^: there is not "more details" [16:13] just "report" or cancel [16:13] but i already clicked send report [16:13] solidfox: if you absolutely don't want to send it, you can just "sudo rm /var/crash/*" [16:13] if you look in that folder you can see which program and date/time [16:14] popey_: strangely after clicking send error, it still didnt create a ".uploaded" file like the other errors [16:14] i'll remove them i guess [16:15] thanks [16:16] np [16:19] I have 18.04.2 LTS and I want to install openmodellica [16:19] But apt says no such package exists [16:19] And I have seen other people on forums install it with the same version as me [16:20] s/version/Ubuntu version/ [16:21] Any suggestions on how to solve this issue? [16:21] whereistejas, indeed, that package is not in our repos [16:22] follow their tutorial : https://openmodelica.org/download/download-linux [16:22] good luck [16:36] OerHeks: Is there any way for me to get in touch with the package manager of that package on other versions of Ubuntu? I'm a novice at this stuff. I would really appreciate some help. [16:38] whereistejas: The page that was linked to is the only support we can provide. Applications not provided/found in the Ubuntu software repositories are something this channel cannot provide support for. The publisher/developer of that software will have to help you as they have the documentation to do so [16:38] whereistejas: we -- Ubuntu -- do NOT package openmetalica. They themselves do it. So you have to get in contact with them === Richard is now known as Guest82343 [16:46] pragmaticenigma: hggdh: Okay. Thanks. I will try to get in touch with them. [17:04] does hastebin work on ff for you guys? [17:04] the buttons dont seem to do anything [17:12] how do i install oracle 8 jdk? [17:12] jre* [17:14] echelon: If it is available, from the archive in the usual way. But Java 8 is out of support now (I think) - so it'll depend on which ubuntu release you're working with [17:14] TJ-: bionic [17:14] I'm having this problem on Ubuntu Trusty, does anyone know of a way on Ubuntu to call GCC with -lto, but override the unique Ubuntu default of _FORTIFY_SOURCE=1 : https://stackoverflow.com/questions/34907503/gcc-fread-chk-warn-warning [17:15] I tried passing -D_FORTIFY_SOURCE=0 and I get error: :0:0: error: "_FORTIFY_SOURCE" redefined [-Werror] [17:15] !info openjdk-8-jre [17:15] openjdk-8-jre (source: openjdk-8): OpenJDK Java runtime, using Hotspot JIT. In component universe, is optional. Version 8u212-b03-0ubuntu1.18.04.1 (bionic), package size 68 kB, installed size 250 kB [17:15] kinghat: Saving doesn't work for me, I see a 503 in network logs when I try [17:16] TJ-: yeah, openjdk won't work unfortunately.. will need oracle jre 8 :/ [17:16] same [17:16] echelon: I thought it was Oracle who ended Java 8 support? [17:16] yes [17:17] echelon: are you telling us the current openjdk-8-* packages are broken? [17:17] TJ-: no, i'm saying i need oracle 8 jre [17:18] echelon: they're here: [17:18] https://www.oracle.com/technetwork/java/javase/downloads/java-archive-javase8-2177648.html [17:18] akik: thanks :) [17:27] akik: So, my solution was a combination of the kernel parameter, KillUserProcesses=no, enable-linger, and getting systemd-run wrapper working for screen. [17:54] Draggor: which kernel parameter? that should be a userland thing [17:57] how do i prevent ubuntu from installing a different jre/jdk than the one that's already installed? [17:57] i'm trying to install maven, but maven has openjdk 11 as a dependency [17:58] but i already have a java runtime that i installed manually just now [17:58] i'm having huge problems with my ubuntu installation.. it can freeze pretty much all the time [17:58] echelon, why did you manually install a java runtime? [17:58] the only resolution is to reboot the host [17:58] physically [17:58] zamba, what Ubuntu version and flavor? [17:58] lordcirth: i needed a java version that was out of support [17:58] lordcirth: 18.04.2 LTS [17:59] zamba, Desktop or Server? [17:59] lordcirth: stuff like doing an google image search could overthrow the system [17:59] if i don't react immediately by ctrl + w-ing the tab when i notice the mouse pointer starting to jerk, i have lost my system [18:00] lordcirth: desktopå [18:00] zamba, ok. What CPU do you have, how much RAM, what graphics card, and what graphics driver? [18:01] i5-3470, 4 GB RAM, some integrated graphics card.. [18:01] and i'm running gnome session fallback [18:01] 00:02.0 VGA compatible controller: Intel Corporation Xeon E3-1200 v2/3rd Gen Core processor Graphics Controller (rev 09) (prog-if 00 [VGA controller]) [18:01] i915 kernel driver [18:01] zamba, 4GB is low, but shouldn't freeze. Do you have swap enabled? 18.04 has it by default [18:02] zamba, .. to reboot the host.. what OS is the host? [18:02] 1GB swap enabled [18:02] A colleague is not in the sudo group and can't install applications. They say they used to be the root user so they don't have a root user. How can you add a user to the sudo group? [18:02] OerHeks: eh? we're in #ubuntu now [18:02] OerHeks: so yes, this is ubuntu :) [18:02] OerHeks, I believe he means "host" as in the machine, not as in VM host [18:02] yes, this is not a virtual machine [18:02] oke, that is not how i read it [18:03] lordcirth: but.. it seems somehow related to thunderbird running [18:03] Siecje, sudo usermod -aG sudo username [18:03] lordcirth: Can't use sudo though. Will it work. [18:03] Siecje, you have no user on the system that can use sudo or su? [18:04] lordcirth: Correct. [18:04] lordcirth: i don't think i have experienced the issue without thunderbird running.. and also with thunderbird running i can lock the screen and come back some hours later being unable to log in.. this time around i was away for 4 days and i paid attention to NOT leave thunderbird running when i left the office and i could then successfully log back in [18:04] Siecje, reboot, choose other options, recovery, root shell. Then run that. [18:04] zamba, can you please pastebin the output of "free -m" when the system is stable? [18:05] lordcirth: coming right up [18:05] Then 'watch free -m" and trigger the freeze [18:05] https://pastebin.com/qywEBE4N [18:05] unfortunately i'm not able to trigger the freeze [18:05] zamba, when it freezes, you can still see the screen, right? [18:06] lordcirth: yes [18:06] but i'm not sure if it actually updates [18:06] Actually, either way: "watch 'free -m' | tee ~/free.log'" [18:07] Just leave that running, and the next time it freezes and you need to reboot it, just look at the end of the log file [18:07] lordcirth: yup [18:07] but this will just show the last entry.. would it be more helpful to show all entries? [18:07] zamba: have you reviewed the system logs after it has failed for clues as to what was happening? [18:08] zamba, sorry, tee -a [18:10] lordcirth: ah, thanks :) [18:12] TJ-: wel, maybe this could be related: [18:14] zamba: e.g. "journalctl -b -p warning --since=yesterday" [18:14] https://pastebin.com/4FJzDrR7 [18:15] TJ-: that command only shows the boot up [18:18] zamba: indeed; I am wondering if there's something systemic happening early on that sets the scene for the later failure [18:19] zamba: your log shows some unusual/unexpected causes of "Permission denied" and the frequent dhclient DHCP renewal requests is weird too, but nothing exactly obvious there === ErrantEg_ is now known as ErrantEgo` [18:21] TJ-: i feel quite confident that my system will run just fine as long as i don't have thunderbird running [18:21] TJ-: i'll try that for some days and see how that works [18:48] hey is there a mailing list that I can subscribe to which will notify me of any security issues with the os? [18:48] zamba: something else you could do, is call thunderbird from within a shell script that logs TB's stdout/stderr to a file, in case it leaves clues behind as it runs. [18:49] schmeets: https://lists.ubuntu.com/mailman/listinfo/ubuntu-security-announce [18:49] sarnold: I totally checked, but couldn't find it :P [18:49] thanks :D [18:49] zamba: I use TB with lots of IMAP accounts with one receiving ~5000 emails a week without problems, so I wonder if it might be related to the types of account you've got it connected to [18:51] schmeets: there's a *lot* of ubuntu lists :) [18:52] TJ-: i'm experiencing issues on another system i'm running thunderbird on [18:52] TJ-: so i strongly believe there is something there [18:54] zamba: which version of TB? [18:54] zamba: i have: "thunderbird/bionic-updates,bionic-security,now 1:60.7.0+build1-0ubuntu0.18.04.1 amd64 [installed] " [18:54] TJ-: 60.7.0 [18:55] ping replies are being dropped by the iptables rule "-A INPUT m conntrack --ctstate INVALID -j DROP", even though `conntrack -L` shows an entry for the icmp packet created [18:55] zamba: it could also be related to TB <> Gnome shell - I don't use Gnome [18:55] TJ-: what are you using? [18:55] zamba: Xubuntu/XFCE [18:56] TJ-: ok [18:56] zamba: there were some famous memory leaks in Gnome in the run-up to 18.04 release and I still see some reports that suggests there may still be some remaining, which is why I mention it [18:56] with just 4gb ram, you might be happier on xfce or lubuntu or similar [18:57] I know I'm getting responses, because tcpdump shows the replies before they get dropped [18:57] zamba: is TB using IMAP accounts only? [18:57] and the count on the iptables rule ticks up every time I send a ping [18:58] TJ-: yup [18:59] Forty-3: do you mean the count is increasing on *reception* of the ICMP reply ? [18:59] yes [18:59] e.g. packet gets sent, gets replied to, shows up in tcpdump, gets dropped by the rules [18:59] zamba: I know if the IMAP account is large, TB locally can build a large cache, and can have struggles pruning it from time to time [18:59] Forty-3: and presumably if you dop the --cstate INVALID it works? [19:00] no, because I also have -P INPUT DROP [19:00] TJ-: but this seems to be related to something happening in the browser.. it's usually a medium to heavy web page that causes this [19:02] Forty-3: right, but if the policy wasn't DROP :P [19:02] zamba: oh? I thought your earlier comments referred to this only happening when TB is running? [19:03] zamba: is it possible it's something to do with TB rendering HTML emails? [19:04] Forty-3: you referred to seeing the ping in the conntrack table; did you mean you see the outgoing icmp flow in there (e.g. src= dst= ? [19:04] after removing all other reject rules, it goes through [19:04] Where should I put .appimage files so that I can execute them from the terminal yet I won't break anything? [19:04] but it should be caught by RELATED,ESTABLISHED [19:04] which I have right before\ [19:05] Forty-3: also, me being pedantic but, did you miss out a hyphen before the "m" when you wrote this? "-A INPUT m conntrack --ctstate INVALID -j DROP" [19:05] for the conntrack table, there is a line for the outgoing icmp flo [19:05] I did, but it's in the original [19:05] Forty-3: OK, just checking :) [19:05] Forty-3: thought it may be a copy/paste [19:05] unfortunately not [19:06] Forty-3: hmmm, I'll try to replicate it here [19:06] TJ-: while TB is running [19:06] TJ-: if i load heavy web pages as long as TB is not running, it seems to be running just fine [19:06] .. at least so far [19:07] Forty-3: it's not you; it breaks my pings here too [19:07] uh [19:07] *huh [19:13] anyone using the extension to make dark input fields readable in ff when using a dark theme? [19:14] dark reader ? [19:15] https://addons.mozilla.org/en-US/firefox/addon/text-contrast-for-dark-themes/ [19:15] nah because text input fields get jacked in ff. [19:18] does ubuntu 18.04 have "^ dir-up" icon in directory explorer ? like winXP used to have ? [19:19] (sorry for windows teminoloy) [19:22] It shows you the path in the bar, you just click on the directory you want to go [19:23] Forty-3: would adding a "[!] --ctproto l4proto" discriminator help? ICMP is layer 3 /but/ it appears in the conntrack table with its correct protocol number 1 (see /etc/protocols) so you might be able to add a "! --ctproto icmp" to your rule [19:26] TJ-: I ended up just accepting icmp before the conntrack rule [19:26] Forty-3: easy way out :D [19:27] Forty-3: interesting issue though [19:27] Forty-3: it makes the definition of INVALID very fragile [19:27] yeah, now I just have to fix my other (maybe-not-iptables-related) ping problems [19:32] Ool: https://bugzilla.mozilla.org/show_bug.cgi?id=1283086 [19:32] Mozilla bug 1283086 in Widget: Gtk "Some form fields are black on Gtk3 dark themes" [Normal,New] [19:54] I installed 19.04 and now when I play AisleRiot Solitaire it randomly freezes, hangs, and acts all jittery. What's up with that? === im0nde- is now known as im0nde [19:57] xamithan, is there any tweak to add such an icon ? That list is not very handy [20:00] There is a few unofficial patches floating out there. If you really want an icon like that just switch to a file manager that has it, Like Nemo [20:01] jonvonb, did you install video drivers? [20:03] gallomimia: No, this is an Intel NUC w/ integrated graphics. I haven't noticed these issues in any other app except AisleRiot [20:06] 18.10 had finally gotten pretty good. I probably should have stayed there, or just installed 18.04. [20:08] anyone doing preseeds of bionic? i'm doing it and trying to get the hwe kernel installed, but not having much luck [20:09] swills, not sure you can during install [20:09] i had 'd-i base-installer/kernel/altmeta string hwe-16.04' in preseed.cfg for xenial, changing that 16 to 18 in the preseed.cfg for bionic didn't work [20:11] (it worked that way on 16.04 of course) [20:12] https://help.ubuntu.com/lts/installation-guide/amd64/apbs04.html#preseed-base-installer makes me think that this will work, but it doesn't: d-i base-installer/kernel/image string linux-generic-hwe-18.04 [20:14] that setting that worked came from https://bugs.launchpad.net/ubuntu/+source/debian-installer/+bug/1663337 fwiw [20:14] Launchpad bug 1663337 in debian-installer (Ubuntu) "base-installer did not pick hwe kernel during automatic preseed" [Low,Opinion] [20:14] https://askubuntu.com/questions/953430/using-preseed-how-do-i-select-the-hwe-kernel also mentions a similar but slightly different setting [20:15] d-i base-installer/kernel/override-image string linux-generic-hwe-18.04 # is what i find ?? [20:15] i tried that as well, without success [20:15] but i would double check [20:15] swills, and 'd-i keyboard-configuration/xkb-keymap' is given as clue https://askubuntu.com/questions/953430/using-preseed-how-do-i-select-the-hwe-kernel [20:16] why, no clue .. [20:18] swills, don't you automatically get HWE if you use a .2 ISO? [20:19] Maybe that's not true for PXE [20:22] lordcirth, hmm, good find [20:22] with .3 you would get 5.x [20:27] I didn't realize 18.04.3 existed [20:27] there's one thing i don't understand about the release schedule [20:27] it's like there's two development trees.... why? [20:28] gallomimia, how so? [20:28] Oh, .3 isn't yet out. [20:28] well... 18.04 comes out [20:28] then 18.10 and 18.04.2 at the same time? [20:29] or .1 or whatever [20:29] gallomimia, the point releases are generally pretty minor. Bugfixes plus HWE. [20:29] * gallomimia looks up what HWE is [20:30] new drivers? updated kernels? [20:30] gallomimia, Hardware Enablement Stack. Basically newer kernels - and thus associated driver support - backported to LTS [20:31] So you can run 18.04 on 2019 hardware, etc. [20:31] !hwe [20:31] The Ubuntu LTS enablement stacks provide newer kernel and X support for existing LTS releases, see https://wiki.ubuntu.com/Kernel/LTSEnablementStack [20:31] so, the differences are.... backend versus frontend? [20:31] for example, 19.04 has all these snaps and a new GUI? [20:31] i skipped a few versions so i don't quite know what i'm talking about. trying to learn [20:32] gallomimia, 19.04 is a whole new release, with lots of software updated, etc. 18.04.2 is 18.04.2 with bug fixes and support for newer hardware. [20:32] is 18.04* [20:32] I'm trying to set up a router, and I'm missing a route http://sprunge.us/AS2o2J [20:33] Forty-3: what version of ubuntu? [20:33] 18.04 [20:33] Ubuntu 18.04.3 LTS, August 2019 .. sorry [20:33] in any case, the version shouldn't matter [20:33] these particular commands have been working for ages :P [20:34] okay thanks. i do recall using 16.04.1 and 2 and being confused about it [20:34] Forty-3, is that your netplan.yaml? [20:34] no [20:34] Forty-3, what if you just do "ip r get 10.2.0.9"? [20:34] it's a terminal session [20:35] that works [20:35] Forty-3, ok, so what's the problem? [20:35] oh, netplan is preferrably to setup such task as router [20:35] it needs to route stuff from 192.168.0.0/16 [20:35] OerHeks, he's not changing settings there, only debugging [20:35] as it happens I am also using systemctl's thing since I need dummy devices [20:36] Forty-3, you are writing networkd files? [20:36] yeah [20:36] but that's not relevent atm [20:36] i'm having this 100% repeatable bug where my boot-up hangs until i press ctrl-alt-f2, ctrl-alt-f1 [20:36] (probably) [20:36] gallomimia, as in, ssh doesn't come up until you do that, or you just can't get a login locally? [20:37] i don't get a login [20:37] it honestly hangs until i come press the buttons. it's a really minor issue, but it seems vastly unreported [20:38] gallomimia, so, you cannot ssh until you do that, and then you can? [20:41] gallomimia: could it just be the GUI has started on a different TTY to the one that is being used for console? [20:41] gallomimia: it may be related to the vt_handoff kernel command-line parameter set by GRUB scripts [20:44] no, i have to change to another console and back to the same one [20:44] i have not tried to ssh into the machine under this condition [20:45] i have always always always had odd boot issues with this motherboard. i hate it [20:47] gallomimia: "systemd-analyze critical-chain" and "systemd-analyze blame" [20:47] blame!? [20:48] is this a root-required set of commands? [20:48] gallomimia: no, user level [21:12] what the hell is this cloud-init nonsense? === ben_r_ is now known as ben_r [21:25] how can I enable the TRACE target for iptables [21:26] packets show up with -j LOG but not -j TRACe [21:26] in syslog [21:27] I have already run `modprobe ipt_LOG` [21:47] I find myself "double navigating" a lot of the time. cd'ing around in a terminal window, getting things done, then when it comes to attaching something to an email, I have to navigate to the same place in the File Open dialog. Is there a better way? [21:49] gbellinoz: I wonder.. what happens if you cd somewhere in your terminal and then type 'xdg-open .' [21:49] gbellinoz: will that open a file browser to the current working directory? or will it try to open a web browser? [21:51] sarnold: I've set that up in years past... lessee, currently it opens Nautilus. [21:51] gbellinoz: would that then let you drag the file icon to your mail client? [21:51] good idea... let's see. [21:52] xdg-open that's the one? [21:52] yeah, absolutely. good idea. sort of a bridge between the CLI and the GUI world. [21:52] there's probably also a Thunderbird event or something I could trigger. [21:53] huh [21:53] Thanks for that. It's sort of a more general problem - finding myself "double navigating" a lot. Maybe I should see how many of those situations could be solved by xdg-open plus click+drag. [21:53] xdg-open . is the command [21:53] the GUI navigator already has a command for "Open In Terminal" this is the secret i needed [21:56] Another example would be setting up a project via the terminal, vim'ing a notes file, etc., then needing to save a file from Gimp to that same project. So, I end up navigating to the project dir from Gimp. [21:56] "Recently Used" doesn't have dirs, otherwise I might be able to find a way to populate that from the CLI. === jje_ is now known as jje === linewalk is now known as list === list is now known as linewalk [23:09] Hello! [23:10] Could someone help me telling the ubuntu server channel anme please? [23:10] name* [23:10] #ubuntu-server ? [23:11] thank you xamithan ! [23:14] I'm running Software Updaterand getting "Failed to download repository information\n Check your internet connection". [23:17] puff: Show us :) Pastebin ` sudo apt update ; sudo apt upgrade ` . [23:18] https://paste.ubuntu.com/p/sbqVccgWZN/ [23:20] Looks like you need to switch to a working repository [23:20] That one is dead [23:30] xamithan: Hm, it died? [23:31] Looks like it, no release file, no public key [23:31] Should be able to switch to another mirror under software and updates [23:31] aha [23:33] I don't see anything about mirrors in Software & Updates. [23:33] It should be on the first tab as "Download from" [23:35] Thanks. [23:35] That *appears* to be working, fingers crossed and bated breath. === jje_ is now known as jje [23:41] That fixed most of the errors, still getting errors on the oddball repos like dl.bintray.com, openprinting.org, packages.ros.org, [23:41] Guess I'll just disable them for now. [23:45] So I guess now's the time to upgrade to Bionic Beaver. [23:45] puff: If you trust Open Source Lab - you can get their key. [23:45] No wait, xenial. [23:45] You're already on xenial [23:45] so ancient [23:46] * jack is running cosmic [23:46] Ah, weird... I googled on "latest ubuntu LTS" and it lists 18.04 LTS Bionic as July 26, 2018 and 16.04 LTS Xenial as February 28, 2019. [23:46] Okay, so Bionic? [23:46] I prefer to stick with the LTS releases. [23:46] The latest point release for 16.04 probably released on that date [23:46] Also, I'm actually running xubuntu. [23:46] and eoan...but i like my cosmic install better [23:47] wait, actually it's disco [23:47] cosmic is dead I thought [23:48] just getting "old" [23:48] disco rules :) [23:48] Hm, how do I upgrade to bionic via the GUI? I was told a while back that the GUI is safer for dist-upgrades than the shell, these days. [23:48] !eol 19.04 [23:48] Gr I forgot that command [23:49] !upgrade | puff [23:50] puff: For upgrading, see the instructions at https://help.ubuntu.com/community/UpgradeNotes - see also http://www.ubuntu.com/desktop/get-ubuntu/upgrade [23:50] puff: what version of ubuntu are you running? [23:50] puff: Ubuntu 16.04 is supposed until 2021 [23:50] supposed/supported [23:50] leftyfb: 16.04 LTS, yeah. [23:51] leftyfb: I'm not necessarily in a hurry to upgrade, today just seemed like a good day to get it done. [23:57] what was between xenial and bionic? [23:58] yakkety, zesty, artful [23:59] uh, thx [23:59] .. nothing supported, between xenial and bionic