=== Guest8607 is now known as westor [00:13] Ubuntu 20.04 on a VPS. Where is an appropriate place to put self-compiled binaries that act as daemons (recieve connections from the internet). These programs don't require root, so I'd prefer to run them as their own user, but I'd like them to be able to be automatically started if the server reboots or goes down. What's the best way to go about accomplishing this for a complete noob? [00:18] /usr/local/bin is the traditional place for user complied stuff not managed by package distributions [00:42] tsapien perfect thank you. What's the best way to re-run them on server restart? Can I just 'sudo adduser --disabled-password --home /home/username username' then 'sudo su username' then 'crontab -e' and add the process I'd like to run as that user with @restart and it will start as that user after a reboot? [00:42] QuickBASIC_: create a systemd unit [00:42] ( .service file) [00:45] leftyfb Ah, thank you. I was struggling with search terms, so you've likely pointed me in the right direction. Can a systemd service start a .sh so I can pass parameters or can I include them in the 'ExecStart=' line? [00:45] yes [00:46] you pass the parameters in the ExecStart [00:46] Ah makes so much sense. Are systemd services running as root though? [00:47] QuickBASIC_: they can be, or you can run them as a user [00:49] Cool. Thank you so much. Now that I have some search terms I think I can figure out what I'm doing. === keypushe- is now known as keypusher [01:07] hey all [01:35] hi fella [01:41] When installing ubuntu on a 2 monitor setup, is there a way to tell ubuntu to prefer a monitor and to rotate it? I have the same issues with grub details on boot when all the userland settings are correct === [_] is now known as [itchyjunk] [01:51] @noarb have you messed with the Display settings already? [01:57] I haven't done anything before editing userland display settings after the machine has booted. Do I need to pass a boot param or something to get grub2 boot info on the right monitor and rotated correctly? I have a 2 monitor setup, 1 horizontal 1 vertical. Right now grub and installation are on the wrong monitor and rotated incorrectly [02:05] @noarb: go in `etc/default/grub`, check your resolution settings. comment out and/or edit what you need and then run `sudo update-grub` when you're done [03:31] tomreyn: things are back to being screwy [03:31] the resolv.conf got overwritten [03:32] maybe there was another systemd thing I had to disable [03:32] well anyhow it turned back into a symlink [03:32] * fmlatghor sighs [03:32] maybe I should just learn how to use systemd-resolved [03:33] What is cloud init and should I just disable it? [04:27] I am trying to build my own .deb packages for plymouth. So I downloaded the source (plymouth_0.9.4git20200323.orig.tar.xz and plymouth_0.9.4git20200323-0ubuntu6.debian.tar.xz) from packages.ubuntu.com. Then extracted plymouth_0.9.4git20200323-0ubuntu6.debian to ~/pbuilder and copied the other to ~/pbuilder. After that I ran `debuild -S -d -us -uc` [04:27] and `pbuilder-dist focal build ../plymouth_0.9.4git20200323-0ubuntu6.dsc`. But it is not generating any deb file. The first command generated lots of `ignored for deletion` warning. and the second one generated `pbuilder dependency error`. I am where I went wrong. [04:28] *correction I am unsure where I went wrong [04:31] Also If this is not the correct place to place this question, please refer me the correct channel. [04:50] Hey Guys! [05:31] webchat60: are you trying to contribute to ubuntu, or whats your endgoal? [05:47] webchat60: can't you just use checkinstall? [06:45] hi. i am the budweiser official monkey man of nigeria [06:45] i selling beers [06:45] pertho: welcome to FartLand [06:46] i sell bears [06:46] your mother has fungus in her cooch [06:47] much diseased [06:47] very smelly vagina [06:48] lotuspsychje Sorry for the delayed response. I am trying to make a customized ubuntu for persoanl use [06:49] Croran Can you please tell me the command [06:49] ? [06:50] I got an old Ivy Bridge laptop with a discrete Nvidia GeForce GT 650M card. In the old days, I used bumblebee/optirun (primus?) to use it. However it seems now you have to use prime-switch and a new nvidia driver. Only that it doesn't work on my old card. Any ideas? Is bumblebee still available? [06:51] !ops | sniffable bad news [06:51] sniffable bad news: Help! Channel emergency! (ONLY use this trigger in emergencies) - Pici, Myrtti, jrib, Amaranth, tonyyarusso, Nalioth, lamont, CarlK, elky, mneptok, Tm_T, jpds, ikonia, Flannel, wgrant, stdin, h00k, IdleOne, Jordan_U, popey, Corey, ocean, cprofitt, djones, Madpilot, gnomefreak, lhavelund, phunyguy, chu [06:52] The Holocaust is hoax [06:52] tnx [06:53] pertho: these these its nvidia-prime taking care of optimus cards, ubuntu-drivers list to see the drivers for your card [06:54] thanks lotuspsychje. Where's that list to be found? [06:54] pertho: in your terminal; ubuntu-drivers list [06:54] it's an older card and prime-switch nvidia makes the screen just go completely black when it boots.. I have to use the built-in intel chipset but the nvidia just for off rendering? [06:55] prime-select on-demand is useless.. nvidia-smi just says "not supported" [06:55] pertho: can you pastebin; sudo lshw -C video please? [06:56] sure hang on [06:57] lotuspsychje: https://pertho.net/pb/fDC5lfmo [06:58] 404 [06:59] sorry .. one sec [07:00] https://pertho.net/pb/avjxTDZ3 [07:01] pertho: yeah unclaimed means driver failed to load, lets have a look in your dmesg please? [07:02] when I 'modprobe nvidia' I get: modprobe: ERROR: ../libkmod/libkmod-module.c:838 kmod_module_insert_module() could not find module by name='off' [07:03] ah.. prime-select query is set to 'intel' [07:03] if I set it to 'nvidia' I get a black screen when I boot.. only way to fix is to ssh in and set it back to intel as root [07:03] on-demand does work though.. I think [07:05] when it creates the new initram disk, I see lots of these messages: W: Possible missing firmware /lib/firmware/i915/skl_guc_49.0.1.bin for module i915 [07:05] I'm guessing that's a linux-firmware packaging problem or something [07:05] Is there a GUI application that I can use to burn a windows ISO to USB? Googling lead me to WoeUSB, but I can't get it to run. [07:05] pertho: your intel graphics do get loaded, its only your nvidia module failing [07:06] pertho: more clues in your dmesg [07:06] I used to use an app called 'toaster', I think. But it isn't in the repos anymore. Maybe it doesn't exist anymore. [07:06] lotuspsychje: if I use prime-select on-demand and reboot, the nvidia module does load then [07:06] gparted is another option. But it's just overkill and too complicated [07:06] bobdobbs: woeusb [07:07] lotuspsychje: it won't run on my setup. I've tried to install it, but can't find the binary. If it installed, it didn't install in my PATH. [07:07] bobdobbs: try the snap woeusb [07:08] ah, ok [07:08] lotuspsychje: cool! snap search found it. I'll install it from there. [07:08] lotuspsychje: lshw output with 'prime-select on-demand' (after rebooting) https://pertho.net/pb/3sxQ3bfm and here's the dmesg of prime-select ondemand: https://pertho.net/pb/04KVCJqm [07:09] lotuspsychje: installing it from it's own repo is annoying. I couldn't find the command for it, and couldn't find the command to start it by googling. So frustrating! [07:09] pertho: ubuntu-drivers list please too? [07:11] lotuspsychje: https://pertho.net/pb/QYBf7i5b [07:12] nvidia-390 is what's installed [07:13] webchat60: well do everything to the normal build process except don't run make install [07:13] webchat60: run `sudo checkinstall` instead [07:13] webchat60: of course you need to install checkinstall itself first. it's in the ubuntu repos. [07:18] lotuspsychje: here's output of nvidia-smi: https://pertho.net/pb/BDRqiysN [07:19] Croran I already ran `dpkg-checkbuilddeps` and it returned `NULL`. [07:19] Seriously how do i get rid of these stupid pop-ups in Ubuntu. "! Software Updates Available" "Important OS and application updates are ready to be installed" "[Not Now] [View] [Cancel] [OK]" [07:20] webchat60: i don't know what that's supposed to do, and that's not part of the process as far as I'm aware. [07:21] hmm I wonder if I need to toggle on persistence mode [07:21] pertho: do you still have a previous non 5.11 kernel in your booting list? [07:22] lotuspsychje: I have a 5.8.0-43-generic [07:22] lets try that pertho [07:23] ok I'll try booting that up.. what should I look for afterwards? [07:23] ok, now that I'm running woe-usb, I'm not liking it. What alternatives are there? [07:23] pertho: that your 390 module is loaded correctly, sudo lshw -C video [07:24] bobdobbs: rufus [07:24] webchat60: do the recommended build process for your software. look at the README included with the tarfile. The last step described there is probably to do a 'sudo make install'. Instead of that do a 'sudo checkinstall'. You should end up with a .deb file. [07:24] Croran When I ran `dpkg-buildpackage -b -nc -tc` i am getting the following outputs https://termbin.com/mnkj [07:24] EriC^^: thanks! I'll check that out, [07:24] webchat60: I told you to use checkinstall, not dpkg-buildpackage. [07:25] bobdobbs: no problem [07:26] EriC^^: oh, it looks like rufus is a windows application. I'm looking for an app that runs on ubuntu. [07:26] lotuspsychje: https://pertho.net/pb/zxh6pHvn (5.8 kernel lshw -C video) [07:26] yeah seems like its working there pertho [07:26] EriC^^: like, I've downloaded a windows iso to my ubuntu machine, and I want to turn it into a bootable USB [07:27] lotuspsychje: in 5.11, nvidia-smi produced output, in 5.8, it gives an error: Failed to initialize NVML: Driver/library version mismatch [07:27] bobdobbs: i see, are you going to use it for UEFI or legacy MBR booting? [07:28] The ubuntu tarfile does not come with a readme. [07:28] uhm, I actually don't know. I guess it's been a while since I've done this. [07:28] NVRM: API mismatch: the client has the version 390.144, but this kernel module has the version 390.141. (on 5.8 kernel).. 5.11 does not produce this error [07:28] bobdobbs: will you be using it on your current pc? [07:29] pertho: could be due, your latest 390 mismatches and older kernel [07:29] EriC^^: ah, I see what you mean, I think. I want to make the USB so I can install windows on another machine [07:29] lotuspsychje: I'll boot back into 5.11.. the nvidia driver did load then [07:29] pertho: what i assume is your issue, we notice a lot of 5.11 bugs recently, might be a fault there [07:30] is there a way to fix the driver so it matches properly on 5.8? I'm just not sure how to tell whether it's actually using the nvidia discrete GPU [07:30] bobdobbs: how old is the hardware roughly? like 2010+ ? [07:30] (brb 10-15 min) [07:31] EriC^^: that I'm installing to? maybe about 2015. Relatively recent. [07:32] pertho: booting an older kernel was just for testing purposes, not adviced to daily driver anymore [07:33] bobdobbs: ah great, it should have UEFI support, what you want to do is create a fresh GPT partition table on the usb, and create a single FAT32 partition, mount it, then double click on the iso so the contents show, and copy the files to the FAT32 partition, it'll be bootable on UEFI, you'll have to enter the BIOS and switch the booting to UEFI enabled, or CSM legacy disabled [07:33] EriC^^: kk. From memory I can do that partitioning with gparted? [07:33] correct [07:33] :D [07:34] Croran With `sudo checkinstall` i got this [07:34] ========================= Installation results =========================== [07:34] make: *** No rule to make target 'install'. Stop. [07:34] **** Installation failed. Aborting package creation. [07:36] webchat60: what happens if you run 'sudo make install' instead? [07:42] webchat60: read the 'INSTALL' file that comes with plymouth. excerpt: "Plymouth isn't really designed to be built from source and get installed by end users. For it to work correctly, it needs integration with the distribution. Because it starts so early, it needs to be packed into the distribution's initial ram disk, and the distribution needs to poke plymouth to tell it how boot is progressing." [07:42] so probably no point in making a .deb file regardless. [07:45] I have read this. But my question is how Ubuntu creates the deb packages for plymouth and related? [07:47] lotuspsychje: will nvidia-prime work with my card or will I have to install bumblebee to get it working? [07:47] it is an older card/intel chipset (circa 2012).. laptop got a new lease on life when I installed a new SSD on it :) [07:47] EriC^^: oh crap. I think I did something to my main HDD [07:48] pertho: bumblebee is a depraced method, its nvidia-prime these days [07:48] bobdobbs: can you pastebin the output of "sudo parted -ls" ? [07:48] I closed gparted, but just before I closed it, I noticed that I was operating on a drive that isn't the USB [07:48] http://pastie.org/p/49RlfxDLvqIt4GPsVVxiMx [07:49] lotuspsychje: ok well, prime-select on-demand is the config that seems to enable both cards (prime-select intel just enables the intel chipset only, and prime-select nvidia just gives me a black screen) [07:49] but I have no way to know whether if I run, say, glxgears whether it's actually using the nvidia gpu === cacti is now known as ice [07:50] bobdobbs: the 2000gb one looks good, the 500gb one is empty, did you have stuff on it? [07:50] with bumblebee, at least I could do: optirun glxgears and it would run on the nvidia discreete graphics card [07:50] pertho: you need nvidia-settings to pick performance mode or powersaving mode [07:51] EriC^^: Yes, I did. I think I just accidently nuked it. However, that's ok. That was just a media bucket [07:51] EriC^^: serves me right for not taking a breath and being careful [07:51] EriC^^: I think I got off really lucky [07:51] I'd forgotten how risky this stuff was [07:52] I used to use an app called "toaster" it's main purpose was burning to CD and DVD's. I guess it doesn't exist anymore, cos nobody uses that old tech. But it was so easy to make bootable usb's with it. [07:53] you could probably track it down somewhere [07:54] bobdobbs: hmm so you have backups of the stuff, no problem? [07:54] EriC^^: yeah, it's no problem. Nothing really important there. [07:55] bobdobbs: actually it doesn't look like anything was formatted, you could probably get it right back to where it was at as if nothing happened easily [07:55] bobdobbs: do you happen to have an older pastebin of parted's output or fdisk/gdisk? [07:55] lotuspsychje: nvidia-settings can't save my settings.. if I select one, it prompts for authentication.. I put in my password and it fails [07:56] what is prime-offload? [07:57] EriC^^: I'm probably being a bit irrational, but I don't care abot that nuked data. I get fixated on stuff. Right now I'm fixated on making the bootable/installable USB. Probably because I remember that this used to be easy for me, and now I feel like I'm unfamiliar territory, so I feel like a clueless idiot. [07:57] No, unfortunately I didn't save a copy of the formatting information. [07:59] bobdobbs: ok, if you want let's get the USB ready, then restore the partitions on the 500gb [08:00] :) [08:00] bobdobbs: first, type "sudo gdisk /dev/sdc" [08:01] you have to recreate the partition table as GPT since it's msdos right now, and windows has a problem booting uefi if it's not GPT [08:01] http://pastie.org/p/6MN2VWLLsaOT2mqd52420X [08:01] bobdobbs: ok, type "o" and press enter [08:01] lotuspsychje: is /sbin/prime-offload meant to run when X starts? I see a script in /usr/lib/ubiquity/dm-scripts/oem/00-prime-offload and also /usr/share/lightdm/lightdm.conf.d/90-nvidia.conf .. so it seems to try to run prime-offload, but when I run it by hand I get: https://pertho.net/pb/rMFG9kd2 [08:01] bobdobbs: that should create a fresh gpt partition table [08:02] EriC^^: ok, 'p' now produces this: http://pastie.org/p/0qUQBzMAcp8oiec1eHpbB0 [08:03] bobdobbs: great, type "n" to make a new partition [08:03] "partition number" I guess I accept the default? 1? [08:03] yup [08:04] same for first sector and last so it uses all the space [08:04] EriC^^: ah, I was just about to ask that [08:04] bobdobbs: when it asks for the hexcode, put "ef00" [08:04] what does that mean? Is that an address? [08:05] bobdobbs: it's the partition type, in this case ef00 is for efi partitions [08:05] ah, I see. [08:05] this is the output of 'p': http://pastie.org/p/1CtPSfFA6ToTZ991lMPVmy [08:05] bobdobbs: great, type "w" to write it [08:06] cool. that's done [08:06] bobdobbs: ok exit the program with "q" iirc [08:07] then type "sudo partprobe /dev/sdc" so that the kernel reloads the new partition table [08:07] It dumped me out automatically after 'w'. The exit message: http://pastie.org/p/5Fh5nZYZZz5PXOqlUirDXR [08:07] looks good [08:08] so that's interesting: the kernal will still remember the previous data describing this drive? [08:09] yeah i guess it loads the partition table earlier, and after writing the new one to disk it has to re-read it [08:09] google tells me that partprobe informs the kernel of updates. Should I run partprobe? [08:09] yeah [08:10] http://pastie.org/p/6LGntkw3h4RP4raTqACkFS [08:10] that's the output from partprobe [08:10] I wonder if it would help if I just ejected the drive [08:10] bobdobbs: ah, it might be mounted, try "sudo umount /dev/sdc1" [08:11] from 'mount': "/dev/sdc1 on /media/mantis/7022-B1F2 type vfat" [08:11] so yeah, unmounting... [08:11] ok, try partprobe again [08:11] I also pulled the drive and and stuck it back in [08:12] k, partprobe returns with no message [08:13] bobdobbs: ok, type 'sudo mkfs.fat /dev/sdc1' [08:13] After re-inserting, my os makes a noise. But 'mount | grep sd ' doesn't show sdc* [08:13] and I can't see the drive in the GUI file manager [08:14] oh wait... but fdisk can see it [08:14] bobdobbs: what's the output of "lsblk"? [08:14] that mkfs command returned without error. [08:15] lsblk shows sdc/sdc1 [08:15] great, mount it and double click on the windows iso, and copy all the files to it [08:16] there should be a dir called 'efi' there so you know it supports uefi booting [08:16] sorry, silly question... where should I mount it to/ [08:16] anywhere? [08:16] bobdobbs: any place will do, i'd just click on the filemanager [08:16] kewl kewl [08:16] really appreciate this, btw [08:16] no problem [08:18] gaaaaah! [08:18] ? [08:19] EriC^^: As root, I mounted it at '/media/mantis'. Mantis is my username when I'm not root. Then I did 'chown -R /media/mantis'... [08:19] I did this as root. But the return from chown is "chown: changing ownership of 'mantis/': Operation not permitted" [08:20] bobdobbs: you don't want to change the permission of /media/user cause it has extra acl's there and stuff [08:21] oh. acl's. I've heard of those [08:21] access control lists? [08:21] bobdobbs: you dont need to chown anything, fat32 won't save the permissions anyways [08:21] yeah [08:21] ah, ok [08:22] lol. now I have to remember where I downloaded Windows to [08:22] ah, found it :) [08:23] :) [08:23] EriC^^: ok, as root I'm now copying windows to the device at the mount point. [08:23] now... inhaaaaaaaaaaaaling [08:24] breath bob. breath. [08:24] k, now what were you saying about enabling certain boot options on the target lappy? [08:25] yeah, enter the bios, set it to use UEFI and/or disable CSM legacy [08:26] bobdobbs: what kind of laptop is it, which brand? [08:29] It's an Asus. I'm hunting around for the model, but the light is kinda crappy in the zone it's in. [08:29] bobdobbs: my electricity just went out, running on the UPS currently so if i stop responding it's cause i disconnected, my nick won't though cause i'm using a bouncer, if it's an hp when the pc boots hit "f9" to get the boot options and choose UEFI usb, lenovo is f12 [08:29] bobdobbs: use f8 for asus [08:29] you really know your stuff [08:30] nah i googled asus :D [08:30] oh no! USB not big enough! That's weird: the file is 5.4G and I'm sure the USB is over 7G [08:30] doh [08:31] odd [08:31] I do have another unopened USB though. And I think I've got enough of a handle on the process to step though it myself [08:31] what's the output of 'df -h /dev/sdc1' ? [08:32] http://pastie.org/p/7CFYLM4GTXGaD9Ys1ihTYS [08:32] ls -alh says that the windows file is 5.5G [08:32] looks good it's just half full [08:33] hmm, that's odd [08:33] the cp operation errored out, saying "File too large". That is odd [08:34] bobdobbs: oh i see, you need to double click the windows iso to show its contents and then copy the individual files [08:34] the error is cause fat32 only supports 4gb max individual files [08:34] df -h returns http://pastie.org/p/758RPl0dzPDRvcrhmozYOt [08:35] hey, isn't there a block copy operation I can use? [08:35] is this a job for dd ? [08:35] bobdobbs: i'd erase the contents of /media/mantis first then copy the iso contents [08:35] ok [08:35] bobdobbs: no [08:35] ok [08:36] k, I've cleared out that USB [08:36] copy the contents of the iso? [08:36] you might need to open the filemanager with root permission to be able to copy/paste the contents [08:36] yeah [08:36] sudo -H should do the job [08:37] ok. lets give this a shot... [08:38] as for the 500gb lost partitions, it can be recovered pretty easily using testdisk, hopefully the electricity comes back in a few and we could do that if you want [08:38] that could be good :) [08:38] just for the learning in it [08:38] yup [08:41] EriC^^: ok, coping that iso file to /media/mantis now [08:41] sounds good [08:41] EriC^^: where are you at anyway? why is your power failing? [08:43] gah. The file manager is giving me the same error as cp did: "file too large". [08:44] I think nautilus must be using cp under the hood anyway [08:45] bobdobbs: electricity is back, im from lebanon [08:46] bobdobbs: did you double click the iso and show its individual files? [08:46] ah, no [08:46] doh! [08:47] ok, I'm gonna rm the usb contents and copy the iso contents [08:47] geez, I'm bumbling cos I'm so impatient tonight [08:47] too much coffee mayb [08:47] xD [08:49] uhm, I think I might have found the gui that makes this easy... [08:50] I right-clicked on the iso and the list of apps included something called 'brasero' [08:50] from memory, brasero is a tool that might be able to make bootable usb's [08:50] but I'm going ahead just copying the iso files anyway [08:55] EriC^^: I'm having trouble opening the iso. If I extract the file, only two files are extracted: a readme.txt and an empty folder [08:55] bobdobbs: i think it's for burning iso's to cds and stuff [08:56] I mean - I'm trying to extract the files using the file manager [08:57] bobdobbs: i think double clicking should mount it for you and you can open it as if it were the usb [08:57] should appear at the left of the file manager [08:57] double-clicking just opens up a new nautilus window showing the readme and the empty folder [08:58] ok, right-click gives me the option of opening the iso in the "disk image mounter". lets see what that does... [08:59] absolutely nothing [09:00] bobdobbs: ok, open a terminal and type "sudo losetup -f /path/to/windows.iso" [09:01] that command returns without errors or messages [09:01] bobdobbs: type "lsblk" you should see a loop at the bottom that has the same size as the iso [09:02] huh. I see three [09:02] loops 32, 33, 34 [09:03] bobdobbs: ok type 'sudo mount /dev/loop34 /cdrom' [09:03] done. returned without errors [09:04] cool! [09:04] try 'ls /cdrom' does it look like the windows iso? [09:04] I can see the files! [09:04] yep! [09:04] efi.. etc? [09:04] yes [09:04] ok great, type 'sudo rsync -av /cdrom /media/mantis/' [09:05] kewl [09:06] ok! copying in progress! [09:06] #feelsgoodman [09:06] :D [09:07] what a journey [09:09] still copying... [09:11] bobdobbs: after it's done, type 'sync' so it writes everything in cache to the usb then unmount and unplug the usb if you want [09:12] kewl. will do [09:12] hi [09:12] What is this? [09:12] I'm new to the whole linux thing [09:12] irc chat [09:12] ubuntu support channel on irc [09:13] Awesome [09:14] loop34 o_O [09:14] lots of snaps i guess :D [09:15] What are tar.gz files? [09:17] sapphire: compressed file archives. you're welcome to ask anywuestions here which a simple web search doesn't answer. [09:18] EriC^^: lol. rsync errored out with "file too large". lolololol [09:18] EriC^^: I've got a fresh usb in my drawer that is 16G. I'll just use that [09:19] bobdobbs: something's amiss [09:19] bobdobbs: type "ls -l /cdrom /media/mantis" [09:19] yeah? [09:19] the first fear that comes to mind: eroding hardware [09:20] nah earlier it was due to copying a single file, the error sounds like it, nothing about out of space [09:21] http://pastie.org/p/5v59M805kyUfZ7h2OnL999 [09:21] lotuspsychje: NVIDIA (Performance Mode) gives me a black screen when I boot up, Intel (Power Saving Mode) disables the nvidia driver.. the only time both GPUs are enabled is NVIDIA On-Demand [09:22] bobdobbs: try 'sudo cp -a /cdrom/!(sources) /media/mantis' [09:22] EriC^^: I've already put the new USB in [09:23] wait.. [09:23] ok [09:23] If I put the original one it, it will probably mount automatically [09:24] yeah [09:24] clicking in the filemanager and then lsblk should show the new mountpoint [09:28] lotuspsychje: something got kinda messed up.. glxgears now gives: Error: couldn't get an RGB, Double-buffered visual [09:29] lsblk shows those loops again [09:29] but the file manager isn't showing the USB. looks like it didnt mount automatically [09:30] I'll mount it manually [09:30] ok [09:30] oh no! sdc isn's showing under /dev [09:40] darnit, my os can't see either of the USB drives now [09:42] and gparted can only see the drive that we nuked earlier. It can't see anything else. [09:45] Hey all [09:46] Does anyone know to backup a directory from the command line in a way that does a snapshot? So like, I start backing up a directory that contains a database, and the database is being to written to while I am backing it up, but the backup I receive is the snapshot from when I started doing the backup. [09:52] Kolusion: you'll need a special file system (zfs, btrfs, others) or block device layer (LVM) to do this. [09:53] How is it done from there on? [09:53] LVM rings a bell [09:54] I read it is done with rsync [09:54] My understanding is rsync does not protect the files from being written to during backup. [09:55] I need something like shadow copy [09:58] Kolusion: even with a block-level snapshot, there is no guarantee the file-system is in a consistent state if there is active writing to open files at the same time [09:58] TJ: That's not true, such technology exists, on Windows at least. [10:01] Basically I just need to make a snapshot copy of a few directories while they are being written to during the backup process. [10:02] I checked out LVM, it doesn't seem to be the solution. From what I have so far read, it's for backing up entire volumes. I only need to backup a few directories. [10:03] Kolusion: Microsoft's VSS requires all applications to use the VSS APIs to achieve that; it isn't 'just' a block device level tool [10:03] ZFS or BTRFS are the prime candidates for file-level snapshots [10:04] Damn [10:04] That means needing to reinstall Ubuntu [10:04] That is not an option. [10:04] I take it then it's not possible on an default Ubuntu install? [10:07] you can create a BTRFS file-system inside a file via loopdev if it really is necessary. For databases though we usually run master/slave replicas (we use Postgresql with streaming WAL) to the secondary [10:18] TJ: That is too much effort for me. Yes, I am doing the same backup for my PostgreSQL as well. WAL allows for this. But I need a solution for my web and mail server files. [10:18] I want to snap all three at the same time. [10:18] Well no, with the database I am simply going to copy over the database directory. Not sure if this is a great idea. [10:18] stop the service, or make it stop writes, snapshot, restart the service / make it write again. [10:18] Exactly. [10:18] https://www.postgresql.org/docs/12/backup-file.html [10:18] i was speaking generally, not postgresql [10:18] Was hoping for a live way using some snapshot technology that isn't too hard for me to use. [10:18] yes; flush and freeze writes [10:18] Yes, generally as well. [10:18] TJ: What is flush freeze writes? [10:18] Kolusion: the actions required to safely snapshot the underlying files/block device [10:18] TJ: Yeah I get you. Kind of like that 'sync' tool yeah? [10:18] Kolusion: Flush all pending writes and finalise metadata, freeze any further writes until the snapshot is taken [10:18] Yes! [10:18] but in DB cases the DBMS needs to be told to do that [10:18] What is a DBMS? [10:18] DataBase Management System [10:18] https://www.postgresql.org/docs/12/server-shutdown.html [10:18] as in Postgresql [10:18] What is your opinion of the pg_dump tool? [10:18] that's the preferred method for single databases [10:18] Yes [10:18] Thing is, I want the time exact for the web and mail server as well. [10:18] mail shouldn't be a big issue, if the SMTP process creates 1 file per email [10:18] Yes, I will get big though one day. I want to be ready for it now. :) [10:18] and web-servers generally are read-only, leaving data storage to a database via whatever web application is in use [10:18] Yep === Exagone313 is now known as Exa === Nebraskka_ is now known as Nebraskka [10:18] we use things like DRBD to have real-time cross-network RAID mirrors for vital file-systems, Ceph is another choice there [10:18] It all sounds great, thing is it's too complicated for me and I don't have the time to learn it. [10:19] I was hoping rsync could do it but it can't, [10:19] rsync is a piece of cake [10:21] for mysql, there's innobackupex which you can use to reate hot backups using binary logs, then do PIT recovery from those (to the same mysqld version only) [10:21] i assume something similar will exist for postgresql [10:22] PostgreSQL pg_dump can do this, but the problem is I want to backup file directories of web and mails at the same exact time. [10:22] So I need something that executes outside of it so it can snap all of them together. [10:22] but for anything serious you'd want a cluster, of course, and then you can have a read replica which you can take the backup from. [10:22] I don't have the infrastructure for that. [10:23] The server will be a desktop [10:23] laptop even [10:23] then why do you have this need for same-timed backups? [10:24] It's just better. [10:24] and why is stopping services a problem then? [10:24] Everything on the same level. [10:24] I will be running a website. It's going to look so shit if I have to stop everything every 24 hours lol [10:25] well either it's an important website, then you'll have proper infr, or it isnt, then it doesn't matter at all [10:25] That is a flawed outlook. [10:26] Also, having everything backed up at the same time makes backing up cleaner. [10:26] I mean restoring [10:27] There has got to be something that we don't know about. [10:27] I will go looking [10:27] i don't think you'll be able to combine "serious project, serious requirements" and "server runs on a (single) laptop". [10:28] A serious project can run on a laptop. [10:28] The laptop has nothing to do with this issue. [10:28] This is operating system level [10:29] good luck! [10:29] Maybe I could create two virtual machines and have them replicated, then when I need to do the backup, take one offline and do the backup there. [10:31] What do you think? === Andrew is now known as CoHaxPenguino === CoHaxPenguino is now known as HaxPenguino [10:45] Anyone ever used lsyncd? [10:53] Ok, what do others reckon about this live server backup method. I am running a web server, mail server and database, each in their own virtual machine. To back them up live, I will create a fourth virtual machine and run the same web, mail and database in it, and use lsyncd to syncronise the files. Then when I want to back the files up, I shut down [10:53] the web, mail and database services... [10:53] in the second virtual machine, and do a backup on them. [10:53] selamlar [10:53] turk varmi [11:02] Loving with community support the FOSS community rages on about... === HaxPenguino is now known as CoCoHaxPenguin [11:28] pertho: thats a weird behaviour indeed [11:40] lotuspsychje: if I run prime-select intel, can do 3D stuff just fine.. but can't use the nvidia gpu discrete [11:42] pertho: you dont use like cuda or something right? [11:44] no [11:44] my problem is similar to this person's: https://askubuntu.com/questions/1188458/how-to-use-prime-select-on-demand-in-latest-kubuntu-19-10 [11:45] (but I'm on 20.04) [11:45] yeah im just seeing bug #1824282 [11:45] Bug 1824282 in nvidia-prime (Ubuntu) "Add a mode that load nvidia.ko only so that we display on intel and do cuda deep learning in NV GPU" [Undecided, Fix Released] https://launchpad.net/bugs/1824282 [11:46] https://download.nvidia.com/XFree86/Linux-x86_64/390.144/README/randr14.html talks about using xrandr.. but if I add those bits to /etc/X11/xorg.conf, X won't start [11:48] pertho: in a normal case scenario, optimus intel/nvidia would just load their modules and you be able to switch between them with performance mode/powersaving mode [11:48] "there are changes in ubuntu-drivers-common and in nvidia-prime that will make on-demand mode work for you. This work will land soon." .. on-demand mode does load both GPUs (because the nvidia driver loads as well), but 3D stuff seems disabled [11:50] lotuspsychje: the way my optimus is designed.. the nvidia GPU cannot be set as the 'main GPU'. It still has to use the Intel GPU for displaying things, but the idea was using the Nvidia GPU for rendering offscreen [11:50] pertho: that sounds like the way intel/AMD hybrid cards would behave [11:52] what is prime-offload? [11:52] these utils seem to be very poorly documented (nvidia-prime package) [11:53] pertho: PRIME render offload is the ability to have an X screen rendered by one GPU, but choose certain applications within that X screen to be rendered on a different GPU. This is particularly useful in combination with dynamic power management to leave an NVIDIA GPU powered off, except when it is needed to render select performance-sensitive applications. [11:53] right.. and you'd expect it to work in on-demand mode, right? except that it doesn't' [11:54] pertho: wich option is enabled right now performance mode or powersaving? [11:54] I don't know what to do with this. Would be nice to have nvidia GPU rendering things... it'd take a load off the intel chipset/cpu [11:55] right now.. powersaving (intel) [11:55] mesa/libGL seems to only work on powersaving (intel) [11:55] if I switch to performance mode, X won't load (black screen, no screen output anywhere) [11:55] thats pretty weird [11:56] (I have to ssh into the machine and run 'prime-select intel' or 'prime-select on-demand' and reboot to get any screen display) [11:56] i do recall there were passed bugs about that on-demand thing [11:56] were those bugs fixed in 20.04 or later? [11:56] !20.04 [11:56] Ubuntu 20.04 (Focal Fossa) is the 32nd release of Ubuntu and is the current !LTS release. Download at https://ubuntu.com/download - Release notes at https://wiki.ubuntu.com/FocalFossa/ReleaseNotes [11:56] lemme check rlsnotes [11:57] would have thought fixes would be backported to LTS? [11:57] pertho: long time since I messed with prime, but as you said the PC is circa-2012 I wonder if it is one of the early dual GPU systems that actually have a hardware multiplexer, rather than it done in software, in which case the current tools may indeed have a problem [11:57] TJ-: yeah it's circa 2012.. Ivy Bridge (Intel Graphics 4000) + Nvidia GeForce GT 650M [11:58] there is no BIOS switch for optimus.. at all [11:58] I ran bumblebee on it years ago and running apps with 'optirun' worked fine [11:58] the hardware multiplexer as I recall can only switch which GPU writes to output, whereas with the so-called 'muxless' (no multiplexer) one GPU is responsible for the output but the other can do off-screen rendering and pass it through to the other GPU [11:58] like, optirun glxinfo -B would tell me Nvidia was handling things [11:59] TJ-: his dmesg from earlier; https://pertho.net/pb/04KVCJqm [11:59] http://download.nvidia.com/XFree86/Linux-x86_64/460.84/README/optimus.html [11:59] there's a summary there about mux vs muxless [12:00] v460.84? I'm on 390.144 [12:01] I think 418 is the highest driver version from nvidia that supports my nvidia chipset [12:01] pertho: yeah but your ubuntu-drivers showed only 390 for your card [12:02] instructions say: # NVIDIA-Linux-x86_64-390.144.run --no-opengl-files [12:02] but I'm betting the ubuntu package doesn't do that [12:13] i need a copy of debian that will work without me having to download drivers for my hp [12:14] aresdeath666: this is ubuntu support, try #debian [12:17] TJ-: prime-select intel == 3D works but only uses Intel chipset, prime-select ondemand == both cards enabled in 'lshw -C video' (and nvidia module is loaded).. however I get: Error: couldn't get an RGB, Double-buffered visual .. prime-select nvidia and a reboot means I just get a black screen (nothing will display anywhere) and I have to ssh in and run 'prime-select intel' and reboot to get my display back [12:18] (intel = powersaving, nvidia = performance in this case) [12:18] I think 'nvidia-prime' package assumes you have a 'newer' or muxless optimus? [12:18] pertho: right; I dumped NVidia a long time ago due to their making it all messy and the drivers out of mainline [12:19] yeah my newer laptop I bought I avoided Nvidia.. but my older laptop only needed a new hard disk so I slapped a 256 GB SSD in and it all works.. except optimus :) [12:19] thing is, bumblebee worked great! [12:19] pertho: the ubuntu graphics ppa also provides 390.144 so we cant try that neither [12:20] lotuspsychje: I'm on 390.144 [12:20] pertho: is what i meant yeah [12:20] !nvidia [12:20] For nvidia and matrox graphics cards, see https://help.ubuntu.com/community/VideoDriverHowto . For AMD/ATI graphics cards, see « /msg ubottu ati » and « /msg ubottu fglrxmissing » For the latest Nvidia drivers see https://launchpad.net/~graphics-drivers/+archive/ubuntu/ppa [12:20] and the highest version nvidia driver that supports my card is 418 [12:20] pertho: another shot worth might be checking latest firmware for your clevo [12:20] https://my.hidrive.com/share/yze8mg-wf8#$/BIOS%20and%20EC%20Firmware/CLEVO/W_Series/W1_Series/W1x0ERx/W110ER/W110ER-Bios-R1.00.04-EC-R1.00.03 [12:23] pertho: i wonder, did you have previous ubuntu release(s) where your model did work properly? [12:24] Hey. I was trying to get skins on vlc to work. pointed it to the skins2.so file, and now vlc will not load. even when uninstall and run from the terminal I keep getting this message... [12:24] Theme parsing error: main-dark.css:2470:0: Expected a valid selector [12:24] [000055b05cabf6f0] skins2 interface error: unable to attach stream_directory, treat as XML! [12:24] [000055b05cabf6f0] skins2 interface error: no XML found in theme /usr/lib/x86_64-linux-gnu/vlc/plugins/gui/libskins2_plugin.so [12:24] [00007f3488c53690] xml xml reader error: XML parser error (line 1) : Document is empty [12:24] [000055b05cabf6f0] skins2 interface error: no skins found : exiting [12:25] !paste | CanisElSaviore [12:25] CanisElSaviore: For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [12:25] apt purging doesn't clear it. Logging out, then in again doesn't clear it. I may be doing steps out of order. [12:25] Sorry [12:25] ioria: just on time for a weird issue! [12:26] lotuspsychje, really ? :þ [12:26] lotuspsychje: my W110ER is on 1.02.09 [12:27] ioria: pertho here has a 20.04 on 5.11 with an optimus card on a clevo where the nvidia doesnt like to work https://pertho.net/pb/04KVCJqm [12:27] lotuspsychje: yeah I think it worked on 14.04 [12:27] https://paste.ubuntu.com/p/MgXYYhYmNX/ [12:27] I had 14.04 on this laptop back in the day [12:27] pertho: thats long time ago heh [12:28] ioria: yeah, worked back in 14.04 with bumblebee, optirun, etc.. but with these new-fangled nvidia drivers.. doesn't work [12:28] ioria: powersaving mode works, but the switch to performance mode gives black screen [12:28] CanisElSaviore: i'm not sure what exactly you did - this might be good to know. the messages seem to tell us that the vlc skin2 code tried to interpret the (shared library) file /usr/lib/x86_64-linux-gnu/vlc/plugins/gui/libskins2_plugin.so as (text based format) XML. [12:28] lotuspsychje, i see [12:28] Hi folks [12:32] CanisElSaviore: that could only fails since shared library files are compiled bytecode [12:32] I pointed it there thinking there were skins in that file. [12:33] CanisElSaviore: you probably followed some guide / how-to, and if so, should tell us which [12:33] pertho, what you have installed ? 390 ? [12:33] now, whatever is in my system making it go there will not purge. [12:33] I just did what i do in windows...I thought. [12:34] CanisElSaviore: thinking is good. :) but you must have made some changes as well, which now makes your vlc try to load skins from /usr/lib/x86_64-linux-gnu/vlc/plugins/gui/libskins2_plugin.so - probably something in your home directory [12:34] vlc is not currently installed, and I still get the same message. [12:34] yes i did [12:35] ioria: yeah he's on 390.144 [12:35] ioria: 390.144 yes [12:35] I used preferences instead of the vlc skin search, and install from the plugins menu that I din't realize was there untill a few minutes ago. [12:35] CanisElSaviore: so what did you change in your home directory? and what is it that you do that now triggers this message, with vlc uninstalled? [12:36] pertho, a bit old... what does 'ubuntu-drivers list' show ? [12:36] ioria: https://pertho.net/pb/QYBf7i5b [12:37] I'm pretty sure it's that I pointed the "use custom interface" option in the preferences menu to try getting a skin, and pointed it to that skins2.so file, and I shouldn't have. [12:37] pertho, dpkg -l | grep nvidia | nc termbin.com 9999 [12:37] That's the only actual change I made [12:38] CanisElSaviore: vlc configuration file for your user are likely stored in ~/.config/vlc/ [12:38] if you'll remove them or edit them to remove the reference to the sharede library, this will probably solve the problem [12:39] is there a way I can get a more verbose output when vlc is run from the terminal so I can find which file is being referenced, and I can clear that particular path from it? [12:39] hmmm ok [12:39] ioria: https://termbin.com/l8x1 [12:40] pertho, lspci -nnk | nc termbin.com 9999 [12:40] CanisElSaviore: there's cvlc, a command line interface variant of vlc which may provide more output on a temrinal by default. there may be ways of making vlc (and cvlc) log more verbosely, and there is strace which you can use to check which files a software is accessing. [12:41] ioria: https://termbin.com/j6ea [12:41] but i think just fixing/removing/moving/renaming the configuration files will likely be sufficient === CoCoHaxPenguin is now known as \[ [12:41] ioria: right now prime-select is in on-demand mode [12:41] which is why the nvidia and i915 are both loaded === \[ is now known as \} === \} is now known as a_ === a_ is now known as Andrew [12:42] pertho, dpkg -l | grep linux-generic | nc termbin.com 9999 [12:43] Andrew: please limit your nickname changes while on this channel [12:43] ioria: that's just a one-liner: ii linux-generic-hwe-20.04 5.11.0.27.29~20.04.11 amd64 Complete Generic Linux kernel and headers [12:44] I [12:44] m using whereis to delete all vlc folders first [12:44] pertho, sorry, dpkg -l | grep linux-iamge | nc termbin.com 9999 [12:44] pertho, sorry, dpkg -l | grep linux-image | nc termbin.com 9999 [12:45] ioria: https://termbin.com/67lyq [12:46] (BTW I checked booting into 5.8.. mismatched nvidia driver or something) [12:46] i was asking you to reboot with that [12:47] ioria: 08:27 | pertho > lotuspsychje: in 5.11, nvidia-smi produced output, in 5.8, it gives an error: Failed to initialize NVML: Driver/library version mismatch [12:47] CanisElSaviore: you should not delete files which belong to packages that are still installed. apt list --names-only *vlc* would tell you what potentially vlc-related packages apt thinks are still installed. and there could be snaps as well: snap list *vlc* [12:47] Does anyone know if Ubuntu 20.04 LTS has a way to freeze a directory for copying, copy the directory, which then gets released after copying is complete? [12:48] pertho, have you tried to purge nvidia , reboot and reinstall it ? [12:48] reboot into 5.8 and reinstall? [12:48] purge nvidia* ? [12:48] pertho, nope; purge nvidia , reboot and reinstall it with the 5.11 kernel [12:48] Kolusion: you could place that directory on separate file system and then remount that file-system in read-only mode. [12:49] OK will do [12:49] pertho, you know how ? [12:49] Kolusion: or you could recursively mark files as immutable, but that's probably not what you want. [12:49] what is it, just apt install nvidia-390 ? [12:49] Kolusion: why are you asking this? [12:49] tomreyn: I can't because the files will still need to be used during that process. [12:50] tomreyn: I need to backup a live database. [12:50] pertho, sudo apt purge nvidia-driver-390 [12:50] Kolusion: mysql? postgresql? [12:50] pertho: PostgreSQL [12:50] Kolusion: pg_dumpall will do what you want I believe [12:50] Kolusion: oh this is still about the 'snapshot backup', i told you what i think i know there already. [12:51] or if you use zfs you can stop postgresql, take a zfs snapshot, then start it back up [12:51] pertho: Yes, but I need the snapshot of the database at the same time of my web server files and mail server files. [12:51] pertho: Server can't be taken offline. [12:51] then you'll have to do pg_dumpall, and rsync the web/mail server files somewhere [12:52] ioria: OK gonna reboot brb :) [12:52] pertho, wait [12:52] pertho, sudo apt autopurge [12:52] hmm. I'm installing the newer vlc v4 I worked too long today to get my huion tablet just the way I want it on a multo monitor setup to let stupid vlc video window glitches stop me. [12:52] pertho: rsync doesn't freeze the source. [12:53] zfs snapshot would [12:53] pertho: rsync would also not be used at the same atomic time with pg_dumpall. [12:53] pertho: I don't have time to learn ZFS. [12:54] Kolusion: does the website change that frequently? [12:54] pertho: Not now it doesn't, but it may in future. [12:54] time to learn zfs then :) it's really handy [12:54] hahahahahaha [12:54] I'm sure it is. [12:54] if you want snapshots of the filesystem in time, zfs is the answer [12:54] I have a solution already, but was hoping for something easier than doing it. [12:55] I believe LVM does snapshotting too [12:55] but you need free space left over in the PE for it [12:55] pertho: Yeah, but I know a file volume method, but I only need a few directories and not the entire volume. [12:55] pertho: Yes, LVM is what I was talking about. [12:56] pertho: I will tell you the solution I already have. One moment while I find the tool name I need to explain it... [12:56] ioria: hmm ok I reinstalled nvidia-driver-390 [12:56] pertho, dpkg -l | grep nvidia | nc termbin.com 9999 [12:57] right now it's set to 'nvidia' for prime-select query .. I dunno what'll happen if I reboot (may lose display) [12:57] ioria: https://termbin.com/cmi8 [12:57] pertho: Fuckin hell, I can't find the name of the app! [12:58] pertho, ok [12:58] ioria: be back in a bit need to eat .. thanks for your help so far and thanks to lotuspsychje too :) [12:59] ok, me too [12:59] good luck! [13:00] Fuckin hell!!1 [13:00] Does anyone know the name of that tool like rsync that can copy files and keep them synced on the fly? [13:01] It's a replication tool. [13:01] please keep the language family friendly in support Kolusion [13:01] lotuspsychje: There are no families in here. [13:02] ALRIGHT! I FOUND IT! [13:02] It's called lsyncd! [13:04] What I plan on doing is making a virtual machine with a database running, using lsyncd to keep the database files synced in real time, then I can shut down the database server in the virtual machine and backup the database there while the live server stays running. What do others think? Clunky? Effective? === westor is now known as Guest8655 [13:08] tomreyn: I see you have some dungeons and dragons pic on your github account. You sound like one of those womanless loser Linux users that has been to a party like 3 times in your life. Please confirm. [13:08] LMFAO [13:08] Bye bye 👋 [13:08] Oh no!! [13:12] Way to never get helped again.. O_o [13:12] He never wanted help. Just attention [13:14] Well, I didn't read a whole lot of the relevant scrollback.. [13:24] making a vm and use lsyncd.. sounds overengineered anyway [13:24] sometimes the simplest solution is the most effective, but oh well [13:25] ioria: I am going to reboot and see if I get a black screen (prime-select query shows 'nvidia') [14:21] ioria: holy smokes.. it works! https://pertho.net/pb/q9Ok4UDX [14:21] I guess something got out of kilter with the nvidia driver.. purging and reinstalling seems to have sorted it [14:31] there's only one problem though.. was testing using steam/portal 2.. and got: traps: portal2_linux[23375] general protection fault ip:f2255a49 sp:ffa3c39c error:0 in engine.so[f2019000+8d4000] so I'm going to swap back over to the intel gpu and see if those errors stop [15:22] OK all works now [15:22] amazing to breathe new life into the old laptop.. thanks lotuspsychje and ioria for all your help! [15:22] pertho, no problem [15:25] hello [15:45] ioria pertho hehe i should not have skipped the good old purge/reinstall trick lol [15:45] nice find :p [16:00] I'm from Windows, and trying to get rid of old ubuntu's GPT record. How to do it without losing Windows? [16:00] I intend on installing another version of Ubuntu and don't want the old one to mess around [16:45] should i expect msbc to work with my bt headset when using the latest version coming from the ubuntu 21.04 repo? [16:46] i swapped to pipewire and enabled it as described in https://www.redpill-linpro.com/techblog/2021/05/31/better-bluetooth-headset-audio-with-msbc.html [16:47] i cannot select a codec for my input [16:47] there seem to have been some recent changes as suggested in https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/1006 but im not quite sure what is to be expected with the version shipped with current ubuntu, hence the kinda specific question in here [16:47] Issue 1006 in pipewire/pipewire "Regression in mSBC support?" [Closed] [16:47] gurki: this may depend on kernel version [16:48] 5.11.0-31-generic #33-Ubuntu SMP Wed Aug 11 13:19:04 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux [16:48] feel free to ask for anything else that might be relevant, im a bit puzzled at this point [16:51] ... possibly the headset, its a sony wh-1000xm4 [16:52] they dropped apt(x) going from xm3 to xm4, might that stop me from using msbc? [16:55] gurki: no, xm4 was reported to work at least once [16:57] gurki: msbc should work with linux-modules 5.11.0-11 or later [16:58] I am very surprised [16:58] gurki: but not sure about 21.04 pipewire; with impish at least pulseaudio should work with msbc and maybe pipewire as well [16:59] My nVidia GeForce GTX 1650 Ti is correctly recognized at system information, but when I do lspci | grep VGA, I get VGA compatible controller: NVIDIA Corporation Device 1f95 (rev a1) [16:59] why is that? [17:00] is there a reasonable way to determine where this goes wrong? === lordrishav_ is now known as LordRishav [17:02] oh my pw-cli output seems broken [17:02] but this might be a mere linebreak [17:05] hello [17:05] https://pastebin.com/ajfSbqPq [17:06] jumping to a daily-builds thing isnt really an option on this machine [17:06] i have an issue in update command [17:07] gurki: mind sharing 'pactl list cards' ? [17:07] https://pastebin.com/TjpaBmC4 [17:10] khushant: can you post this? grep -hEv '^([ ]*#.*)?$' /etc/apt/sources.list{,.d/*.list} 2>&1 | nc termbin.com 9999 [17:11] need to type that command in terminal ? [17:11] https://pastebin.com/ipmd4ajL [17:11] i-garrison: [17:12] khushant: you could copy and paste it there, if you dare to. otherwise we can do it manually [17:12] okay [17:12] gurki: did you tried changing profile to headset-head-unit [17:12] khushant: the idea there is to get the content of your apt repository configurations [17:12] https://termbin.com/jpqr [17:13] yes, that will leave me with the cvsd codec and no option to change it to another one [17:14] (to be more precise: i only get to select /headset audio (hsp/hfp)/ , there are no other options) [17:15] Nah, um [17:15] lspci | grep vga doesn't show me 1650 Ti, instead I still see NVIDIA Corporation Device 1f95 [17:15] in 20.10, it's displaying fine [17:16] khushant: remove the "universe" in the line which currently says "deb http://archive.canonical.com/ubuntu hirsute partner universe" [17:16] but idk how to make 20.04 to display it [17:16] how to do that ? [17:17] khushant: use the activities menu to look for the "Software & Updates" application [17:17] click on "other software" [17:18] maybe the .in mirror is on holliday? [17:18] happened before, no updates [17:18] oerheks_: in this case it's because canonicval partner repository has no "universe" section, just "partner" [17:18] thank you so much [17:18] now its fine [17:19] oh good spot [17:19] khushant: you're welcome. maybe you can tell whoever you copied this from (if you remember) that it's an incorrect configuration [17:20] gurki: you have 0.3.24, and setting 'bluez5.msbc-support = true' should enable msbc; you probably want to ask on #pipewire at oftc for more details [17:21] gurki: I can only debug pulseaudio and you need 15.0 (e.g. in impish) to test mSBC [17:21] can i make my files app semi- transparent ? [17:23] guess ill do that, the # on this network doesnt seem to be very active [17:23] *the pipewire # [17:34] can i use android apps in linux [17:37] khushant: there are emulators and virtualizations. the easiest way to set it up may be anbox and google's android studio [17:38] than 1st i go for anbox, i try to do it my self if i need help, will come back :) [17:39] make sure you read their faq first so you'll know what to expect. i don't think it comes with most applications, such as the play store, by default. [17:41] [ 2021-09-05 17:40:56] [session_manager.cpp:146@operator()] Failed to start as either binder or ashmem kernel drivers are not loaded === oerheks_ is now known as oerheks [17:55] Salve a tutti [17:55] Ho bisogno di aggiornare 3 pacchetti che non vogliono proprio aggiornarsi [17:55] vi allego lo screen così che possiate capire [17:55] Uploaded file: https://uploads.kiwiirc.com/files/6d1d3c70265795415d39f41eec2558d5/Schermata%20del%202021-09-05%2018-34-13.png [17:55] Non mi permette in alcun modo di aggiornarli [17:58] privato89: chances are higher someone answers if asking in english I guess [17:59] privato89, or check #ubuntu-it [18:00] Sorry, I tought it was the italian channel [18:00] I can't upgrade the packages on my system [18:01] I tried to help him in #ubuntu-it, but was unable to solve his problem, so I sent him here [18:02] it didn't occurred to me to specify it is an english speaking channel [18:02] gurki: their official channel is at oftc [18:04] I need to upgrade 3 packages, but I cant cause there are issues that I cant solve. [18:04] apt upgrade: https://bpa.st/5ISA apt list --upgradable: https://bpa.st/2HZQ cat /etc/apt/sources.list https://bpa.st/KJFQ [18:18] Hi, I'm having a very slow connection from a wireless usb-dongle with Ubuntu 21.04 (I'm lucky if it can reach 2 Mb) other devices in the same area can reach over 100 Mbit, and the dongle itself is rated as 54 Mbps [18:22] zteam: can you mention the chipset of the dongle and the kernel version too please, so volunteers can think along with you [18:25] lotuspsychje, how do I find out the chipset of the dongle?, kernel version is 5.11.0-31-generic #33-Ubuntu SMP Wed Aug 11 13:19:04 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux [18:26] zteam: sudo lshw -C network === Roedy is now known as Guest4857 === Guest4857 is now known as Roedy [18:39] sorry for the delay, this wifi is lika a rollercoaster [18:39] :P [18:40] lotuspsychje, https://paste.ubuntu.com/p/TW7vjK7sr3/ [18:42] Seriously, they require login to view paste.ubuntu.com nowadays? *sigh* [18:46] lotuspsychje, new output pasted here instead, https://pastebin.com/wtF1wvYG [18:47] zteam: says disabled, you might wanna take a look into dmesg what happened [18:48] Huh what's the reason they require you to log in [18:49] probably spam protection in some way [18:52] lotuspsychje, you are right, that's a silly mistake from me, I use my cellphone now to connect to IRC and since, my wifi connection was unstable I did choose "turn off" instead of disconnecting from the wifi network [18:53] lotuspsychje, this one is corrected https://pastebin.com/qQ5mj3Mm [18:55] ravage, I could buy this, but they clearly says you need to login to view the paste to, that kinda defeats the whole purpose of it [18:56] zteam: does iwconfig in terminal show power management enabled? [18:58] jeremy31, yes, and dmesg does show lots of complains about that too, but disabling it doesn't result in any increase in connection speed [19:02] zteam: did you change the conf file in /etc/NetworkManager? [19:03] output from dmesg | tail https://pastebin.com/VPE5ahX5 [19:04] jeremy31, no I only tried to disable it temporarily with iwconfig [19:08] zteam: This command will disconnect you for a few seconds but it might help> sudo sed -i 's/wifi.powersave = 3/wifi.powersave = 2/' /etc/NetworkManager/conf.d/default-wifi-powersave-on.conf && systemctl restart network-manager.service [19:12] jeremy31, it replied: "Failed to restart network-manager.service: Unit network-manager.service not found." [19:12] zteam: try a reboot then [19:12] jeremy31, will do [19:17] jeremy31, I'm back, unfortunetly it didn't improve the connection [19:18] Zteam run the 2 commands at https://gist.github.com/jeremyb31/a2bee9856d8c13f42f1835bc31bf9480 in terminal and post the termbin URL here [19:23] jeremy31, here we go https://termbin.com/ijcq [19:29] Zteam, you have a strange error but I don't know what the fix might be, it is in the last line of the script results [19:35] jeremy31, yes I have seen it too === erica is now known as birb === bugsniper_ is now known as bug_sniper [20:32] /join #grapheneos and meet the most secure mobile OS ever existed. Its developer strcat aka Daniel Micay is the best hacker ever existed challenges Linus Torvalds himself for linux kernel security debate. [20:33] hello [20:34] is anyone here? [20:34] many are here [20:34] good === birb is now known as erica [20:35] what's the purpose of the channel? [20:35] Ubuntu Support chat [20:35] this is a support channel for getting help with using ubuntu [20:35] always read the topic [20:36] yes thanks to everyone. I am new, i will read it === beaver is now known as pong [21:32] Hi [21:32] I cant upgrade 2 packages on Focal [21:37] privato89: that's a bummer. are you looking for help with it? then we'd need more info. [21:39] you could post a screenshot to imgur.com and link it here, or, better yet, use apt and share the command you ran and its output on a pastebin. [21:45] Hi [21:45] I post my issue [21:46] https://bpa.st/4RUA [21:46] can't upgrade these 2 packages: gnome-shell-extension-prefs libgcc1 [21:48] privato89: are you familiar with both "apt-get upgrade" and "apt-get dist-upgrade", and the difference between them? [21:49] also note that you could add -V to learn which versions those would be upgraded to (with dist-upgrade) [21:50] libgcc1 getting upgraded this way could be a problem, if it's not just a bug/security fix. [21:51] (i'm wondering whether you may have a third party package repository configured which provides a version of libgcc1 that's newer than what's in Ubuntu 20.04 LTS) [21:53] Yes I do [21:54] I can't get something [21:55] If I give the command: $ sudo apt-get dist-upgrade [21:55] it give me this output [21:55] https://bpa.st/HXGA [21:59] privato89: if you understand these commands and it seems safe to run them, please do: sudo /bin/true && cat &>/tmp/aptlog < <(sudo grep -hEv '^([ ]*#.*)?$' /etc/apt/sources.list{,.d/*.list} 2>&1; sudo apt-get -y update 2>&1; apt-cache policy 2>&1; sudo apt-get -syV full-upgrade 2>&1;); nc termbin.com 9999 this gathers more info on your apt configuration without actually changing anything. [22:00] if it doesn't feel safe, then please start by sharing the output of your apt sources configurations, e.g. grep -hEv '^([ ]*#.*)?$' /etc/apt/sources.list{,.d/*.list} [22:00] https://termbin.com/okno [22:02] privato89: side note: security updates should be taken directly from security.ubuntu.com to prevent delays due to mirroring. [22:03] oh thanks for the tip [22:03] privato89: are you cross compiling to arm, or why would you have that configured? [22:04] I use my own security mirror too. I make sure it's up to date. I don't see any general disadvantage here [22:05] ig uess if you make sure it's up to date that's fine [22:05] And this looks like the mirror his server Provider chose maybe [22:06] tomreyn no I don't [22:06] privato89: what about apt policy gnome-shell-extension-prefs libgcc1 2>&1 | nc termbin.com 9999 [22:06] https://termbin.com/2taf [22:09] privato89: hmm, so it's weird that apt-get upgrade shows that these packages will not be held back (that alone would be fine) AND apt-get full-ugrade does not even consider upgrading them. [22:10] newer versions are available for either package, so they should get upgraded unless some dependency could not be fullfilled that way [22:11] Yeah, that is so weird [22:11] you can get more verbose info on the decisions the dependency solver takes if you like [22:11] sure, why not [22:11] we could try [22:11] but i bet you would solve this issue by just moving to security.ubuntu.com mirrors [22:12] what will happen if I try to remove them and to re-install them again? [22:13] there are "-o 'Debug::pkgProblemResolver = 1'" and "-o 'Debug::pkgProblemResolver::ShowScores = 1'" according to the bottom of apt.conf(5) [22:14] i don't know what will happen if you remove and reinstall them. but you probably just need to wait until dependencies can be resolved for these new updates [22:15] So you think it doesn't depends about me? [22:15] one of the cogent mirrors is probably not up to date and that's why you see this [22:16] i do not understand this question, sorry, can you rephrase it? So you think it doesn't depends about me? [22:18] Sorry, I meant if I should wait that someone solve that dependencies for that new update or can I do something to solve it? [22:19] Cause I need to upgrade dist and with those packages I can't upgrade it [22:20] sorry I meant that I need to upgrade do-release and those packages doesn't permise me to do it [22:21] privato89: try replacing the security mirror, see whether this solves it. come back if it doesn't. [22:23] apt list --installed | grep ',local\]$' | nc termbin.com 9999 is another output that could be interesting in this context. but i don't think that's the cause. you should look through those packages before release upgrading, though. and remove any ppa's and packages + package versions obtained from those PPAs, though. [22:24] tomreyn: a little question: How do I replace the security mirror? [22:25] https://termbin.com/imzx [22:27] privato89: run software-properties-gtk --open-tab=1 & and remove all lines saying focal-security there, then click on the "Updates" tab and untick (if ticked), then tick, the first two boxes. [22:28] oh my, what is this [22:29] somehow the system lost track of where these packages are from, but they are proper focal packages [22:30] (at least those i looked at) [22:30] There is no one line with focal-security [22:30] Uploaded file: https://uploads.kiwiirc.com/files/51877e084e3c5763aacb5276079f79dc/Screenshot%20from%202021-09-06%2000-29-33.png [22:31] hmm then just untick the upper two boxes on the next tab, and close the window, then open it again, and tick the boxes again, then close the window again. if ti asks about updating apt sources / package cache, let it do so [22:32] did you try running do-release-upgrade after you posted the output of the lentghy command i provided above? [22:32] (the one where the output was https://termbin.com/okno ) [22:33] maybe you mean on the previous tab? [22:34] ravage: is there a way for an isp to override which apt sources will be used for "$release-security" that you know of? [22:34] You can modify the DNS records [22:34] Sure, I tried but it sais I have to update first all packages [22:35] But other than that not really [22:35] privato89: hmm, actually the looks may have changed for focal, i'm looking at the software-properties-gtk version of ubuntu 18.04 [22:36] privato89: you tried what exactly and what said you need to update all packages first of all? [22:36] Maybe he can use the default sources file for focal [22:36] ravage: does that exist on the file system? [22:37] tomreyn: Yes I tried already [22:37] or did you think of unpacking this from some package (not sure which) [22:37] https://gist.githubusercontent.com/ishad0w/788555191c7037e249a439542c53e170/raw/3822ba49241e6fd851ca1c1cbcc4d7e87382f484/sources.list [22:37] This looks pretty clean to me [22:38] points to archive.ubuntu.com for security, which i guess is wrong [22:38] but other than that it could be fine [22:39] Last Update there is a few hours old [22:39] So it should be fine [22:41] privato89: can you show this again: sudo grep -Ev '^([ ]*#.*)?$' /etc/apt/sources.list{,.d/*.list} 2>&1 | nc termbin.com 9999 [22:41] https://termbin.com/s3o1w [22:42] ravage: i think the other reason no to use archive.ubuntu.com for security was to not get "staged updates" (if that'S the right term), where updates are only distributed to some users intially, then mroe and more after a while [22:44] https://dannyda.com/2020/10/03/default-contents-of-sources-list-in-ubuntu-20-04-1-lts-desktop-server/ [22:44] Used the us mirror [22:44] So maybe change that [22:46] privato89: is the system you're working on located in italy? [22:46] yes [22:46] i'm asking to understand which mirror is best for you. [22:46] do you know why you're using this cogent mirror? [22:46] sure, no problem [22:47] No, I don't [22:47] but http://security.ubuntu.com/ubuntu/dists/focal-security/ and http://archive.ubuntu.com/ubuntu/dists/focal-security/ are really identical [22:47] right now :) [22:47] did you check response headers, too? [22:48] no, what should I do for that? [22:48] privato89: not you, sorry, that was to ravage [22:48] haha ok XD [22:48] ravage: this is what i was refering to https://wiki.ubuntu.com/PhasedUpdates [22:49] privato89: wait a second, i'll get you a copy of /etc/apt/sources.list which you can use as a drop-in replacement [22:49] ok I wait, thanks [22:50] from what i read there the only staged=updates are in the updated repo [22:50] *updates [22:51] ravage: that is probably correct then, i didn't read it in a while. [22:58] privato89: close all the graphical frontends for package management / configurations that may be open, then: sudo mv /etc/apt/sources.list{,_old_ubuntuirc} && wget -O /tmp/sources.list && cat /tmp/sources.list | sudo tee /etc/sources.list [22:58] privato89: sorry trhats incomplete [22:59] ok, I wait, no problem [22:59] privato89: close all the graphical frontends for package management / configurations that may be open, then: sudo mv /etc/apt/sources.list{,_old_ubuntuirc} && wget -O /tmp/sources.list https://termbin.com/pzi9 && cat /tmp/sources.list | sudo tee /etc/sources.list [22:59] privato89: close all the graphical frontends for package management / configurations that may be open, then: sudo mv /etc/apt/sources.list{,_old_ubuntuirc} && wget -O /tmp/sources.list https://termbin.com/pzi9 && cat /tmp/sources.list | sudo tee /etc/apt/sources.list [22:59] privato89: hopefully this time ;) [23:00] privato89: did you run it, yet? [23:00] Ok, now? [23:00] yes [23:00] Yes, I did [23:00] so you ran it more than once? [23:00] then we'd need to clean up [23:01] No, just one time [23:01] do I run it again? [23:01] no no [23:01] just run the third variant [23:02] what do you mean exactly, sorry? [23:02] privato89: please run this: [23:02] sudo mv /etc/apt/sources.list{,_old_ubuntuirc} && wget -O /tmp/sources.list https://termbin.com/pzi9 && cat /tmp/sources.list | sudo tee /etc/apt/sources.list [23:02] once only [23:02] none of the earlier versions [23:03] those were bad [23:03] this come out [23:03] !paste [23:03] For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [23:03] https://bpa.st/V7WQ [23:04] so you must have run one of the others, too [23:05] privato89: just do this now: wget -O /tmp/sources.list https://termbin.com/pzi9 && cat /tmp/sources.list | sudo tee /etc/apt/sources.list [23:05] I ran only this before: https://bpa.st/NPHA [23:05] ok [23:06] https://bpa.st/PU7Q [23:07] privato89: okay, so the contents of the file /etc/apt/sources.list should now look the same as https://termbin.com/pzi9 [23:07] can you confirm this? a cursory check is sufficient [23:08] it should start with a "#deb cdrom:" line [23:08] yes it is [23:08] okay now: sudo apt update && sudo apt full-upgrade -V [23:09] this should at most install two packages [23:10] I think you did it! 😍 [23:10] Wait [23:10] please [23:10] yes, we're not fully done, yet [23:10] it is still downloading and installing [23:11] so it's downloading much?! [23:11] yes [23:11] a lot [23:11] hmm, i'm not sure why it would do that [23:12] maybe the cogent mirrors are really outdated, or you were missing something [23:12] I hope the first one 😅 [23:13] not done yet [23:14] please run cat /var/log/apt/history.log | nc termbin.com 9999 once it's complete (not before) [23:14] sure, I will [23:16] oooh, you had "main" missing from sources.list. i don't know why i missed that [23:17] whatever you previously did tro your apt source configuration, it was not healthy [23:17] whatever you previously did to your apt source configuration, it was not healthy. but we should have fixed this now. [23:18] Is that the reason why it is taken a lot of time ? [23:18] yes, that's why it's downloading a lot now [23:19] and this mirror should be a lot faster for you too, now, since previously you've been loading all updates from the U.S.A. [23:19] that's where the cogent mirror is [23:21] We got it! 🤩🤩🤩 [23:21] I think [23:22] let's see that command output before i join the party [23:22] https://termbin.com/3dku [23:25] privato89: so... you actually got security updates now [23:25] privato89: first time since you installed, i assume [23:25] or since you configured this cogent mirror rather [23:26] it should be like you say, but I dont remeber that I changed the mirror [23:27] privato89: how did you install this system initially? [23:27] do you mean Ubuntu? [23:28] yes [23:28] with pendrive, with a normal installation [23:30] okay, then i guess you did something to change /etc/apt/sources.list at some point. if you run ls -l /etc/apt/sources.list_old_ubuntuirc this may tell when this change occurred [23:31] Yes, today [23:31] hmm, no, then that [23:31] is incorrect ;) [23:31] Bot I added a line today at 16.00 pm [23:31] anyways, i think you're done [23:32] you should reboot soon [23:32] I will read the file you told me to check the issue [23:32] and apt list --installed | grep ',local\]$' | nc termbin.com 9999 should be a lot shorter now? [23:33] i'm not sure which file you are referring to [23:33] https://termbin.com/ki8dh [23:33] this one: sources.list_old_ubuntuirc [23:34] the file /etc/apt/sources.list_old_ubuntuirc is now either identical with your newly configured /etc/apt/sources.list or, ideally, it contains what used to be /etc/apt/sources.list [23:34] oh ok, I have missunderstood then [23:35] well, I try to reboot now [23:35] We solved it [23:36] You specially [23:36] https://termbin.com/ki8dh lists packages which you seem to have manually installed at some point. those are packages or package versions which are not available from any of the apt repositories you have configured. this means, these packages may be outdated, will never get security updates, and they may get in the way of resolving package dependencies [23:36] I dont know how to thank you, would like to give you a cup of coffee if you like :D [23:36] you should carefully review this list and uninstall the packages you don't need. and for those you still need, you should fine an apt repository (or snap) providing updated versions of it [23:37] ok I will [23:37] hehe, that's kind, proper italian coffee. i'll rather need to sleep, though [23:37] so, i'll be gone unless you prefer me to wait until you rebooted? [23:38] haha ok I reboot and tell you [23:38] just 1 moment [23:38] really thanks [23:38] you're welcome [23:40] hi, so I'm sure it works [23:40] Which command has I to run to prove it [23:41] tomreyn [23:41] privato89: welcome back. :) what do you mean, what are you meaning to prove? [23:42] that I don't have anymore packages to update [23:42] There is a command to prove it [23:42] !uptodate [23:42] To ensure you have all the latest known patches and security updates for your ubuntu installation, please update with the following command: `sudo apt update && sudo apt upgrade`. See also !upgrades and !security; you may also need to run `apt full-upgrade`. [23:43] so it really is sudo apt update && sudo apt full-upgrade [23:43] and neither should report any errors or warnings [23:43] (errors may be indicated by "E:", warnings by "W:") [23:43] https://bpa.st/X3AA [23:44] that is the output [23:44] privato89: this looks fine. you can remove those packages, if you like: sudo apt --purge autoremove [23:45] Yes, I did [23:45] Finally I can to upgrade the release [23:45] I have 20.04 LTS and wanna update to 21.10 [23:46] the command was: sudo apt do-release-upgrade [23:46] privato89: 21.10 is not released yet [23:47] the last one after focal [23:47] privato89: is there a reason you need to update? [23:48] No, I just wanted to have the last one release [23:48] privato89: right, before you upgrade, go over the https://termbin.com/ki8dh list first. also, for reference, this was probably your old /etc/apt/sources.list )note how it lacks the word "main", which is the very basic ubuntu repository you definitley need): https://termbin.com/uyiv [23:49] privato89: if you don't strictly need newer software versions - personally i prefer to stay on LTS releases. [23:49] privato89: Ubuntu 20.04 is the latest LTS release. It'll be supported for another 4 years. 21.04 will be unsupported next month [23:49] oh I idn't know that [23:50] yes, you'd have to upgrade again next month then [23:50] didn't* [23:50] !LTS [23:50] LTS means Long Term Support. LTS versions are supported for 5 years on the desktop and server. The latest LTS version of Ubuntu is !Focal (Focal Fossa 20.04). Ubuntu !flavors may have different support durations, check their release notes for information. [23:50] ok, Thanks for the info [23:51] Now that I know it I think I stay with Focal [23:51] someone wrote up a longer story on this here https://gist.github.com/tomreyn/8d7675840d7bc7389b32e4d8887ca449 [23:51] (though that's mostly about HWE vs GA) [23:51] so, off to bed, see you, privato89 [23:51] Does the termbin files delete after a while? Or can I save it on my [23:52] bookmarks [23:52] they'll be removed after a while, a month i think [23:52] ok, thank you so much Tom [23:52] Have a good night [23:52] see you soon I hope, saluti dall'Italia :D [23:53] "Life span of single paste is one month. Older pastes are deleted." (Source: https://termbin.com/ ) [23:53] isnt 21.04 the latest LTS [23:53] :) [23:53] srv: no [23:53] 21.04 is not an LTS [23:53] those release every two years, in even years [23:54] * tomreyn afk