[00:00] <friendofafriend> It's 18.04.2 LTS.
[00:00] <tomreyn> hmm, it works for me, i'm also on 18.04.2 with the same shell
[00:01] <tomreyn> apt install bash<tab><tab>   lists 9 packages
[00:01] <friendofafriend> Right, and the same here does absolutely nothing.
[00:02] <ZaZaGX> i got nothing too
[00:02] <tomreyn> friendofafriend: is it in your ~/.bashrc ?
[00:02] <friendofafriend> Is what?
[00:02] <ZaZaGX> apt <tab> works
[00:03] <friendofafriend> Right, apt and tab works here, also.
[00:03] <friendofafriend> But not for the packages.
[00:03] <friendofafriend> I've already done an apt update.
[00:03] <tomreyn> https://paste.ubuntu.com/p/RBK4KHHGHd/
[00:04] <friendofafriend> tomreyn: Yeah, same thing.
[00:04] <friendofafriend> Is there some cache I could rebuild of packages, maybe?
[00:04] <tomreyn> apt update should rebuild apt's cache
[00:05] <tomreyn> if you ran    apt clean    this may explain it.
[00:05] <friendofafriend> That returns nothing.
[00:05] <OerHeks> apt install bash[tab][tab] gives me 6 packages, so it works standard
[00:06] <tomreyn> friendofafriend: what returns nothing?
[00:06] <friendofafriend> apt clean
[00:06] <friendofafriend> Yes, I'm asking what provides that functionality.
[00:06] <tomreyn> that's normal. but i didn't suggest running it
[00:07] <friendofafriend> < tomreyn> if you ran    apt clean    this may explain it.
[00:07] <friendofafriend> That looks like a suggestion, to be fair.
[00:07] <tomreyn> i'm saying if you ran this command recently then this explains why you cannot use bash completion on    apt install
[00:07] <friendofafriend> It's not here or there, I'm just after *what* provides that autocompletion in apt.
[00:07] <OerHeks> maybe true,  run apt update again?
[00:08] <friendofafriend> I've run apt update several times.
[00:08] <tomreyn> i just ran "apt clean" and now havve no output on    apt install bash<tab><tab>
[00:08] <friendofafriend> Cool, can someone help tomreyn fix that?
[00:09] <tomreyn> so my theory that you ran 'apt clean' recently, causing bash completion on apt install to fail could be correct.
[00:10] <tomreyn> (there could probably be other explanations, too.)
[00:10] <OerHeks> what happens after running apt update again, tomreyn ?
[00:10] <OerHeks> resume bash completion?
[00:10] <friendofafriend> So, no, I've only run apt clean after you recommended it.
[00:11] <friendofafriend> And it wasn't autocompleting package names before.
[00:11] <friendofafriend> I've run apt update several times since.
[00:11] <tomreyn> yes, OerHeks
[00:11] <friendofafriend> And it's still no dice.  Where is apt getting that list of packages?
[00:12] <tomreyn> from apt repositories
[00:12] <friendofafriend> What generates it?\
[00:12] <tomreyn> apt does
[00:13] <friendofafriend> When?
[00:13] <tomreyn> when you (or an automated process) run(s) "sudo apt update"
[00:14] <tomreyn> ...then it writes repo / package information into its cache and uses that until another apt update is requested.
[00:14] <friendofafriend> And this is... pkgcache.bin?
[00:14] <tomreyn> ...or until "apt clean" is run
[00:14] <de-facto> Is there any way to delete/destroy gnomes ability to annoy with "mouse battery empty" popup spam? This is extremely annoying...
[00:14] <tomreyn> friendofafriend: /var/lib/apt/
[00:15] <friendofafriend> So, for example.  I run apt-config dump | grep -i pkgcache, I get back Dir::Cache::pkgcache "";
[00:15] <tomreyn> Cool, can someone help friendofafriend fix that?
[00:15] <friendofafriend> I look elsewhere, I see others find:  Dir::Cache::pkgcache "pkgcache.bin";
[00:16] <friendofafriend> Did apt update correct your issue, tomreyn?
[00:16] <tomreyn> i had no issue here
[00:17] <friendofafriend> Wow, Schrödinger's cat in the house.
[00:17] <ZaZaGX> i personally think the cat is alive
[00:18] <tomreyn> de-facto: there's bug 1798166
[00:18] <tomreyn> (note also the link to the upstream bug tracker)
[00:19] <de-facto> tomreyn, yes thats why i need to delete/destroy it, when it shows those messages i loose my mouse focus for a fraction of a second, its seriously harming my workflow
[00:20] <tomreyn> i did not fully read both bug reports, maybe they discuss workarounds
[00:21] <de-facto> i already uninstalled gnome-power-manager but which made it disappear for a few days, but now the popup spam is back (i dont know why)
[00:22] <de-facto> i dont need any gnome notifications ever, they are useless to me
[00:23] <friendofafriend> de-facto: https://askubuntu.com/questions/909771/disable-all-gnome-notifications-gnome-3-20-4
[00:24] <de-facto> yeah i dont think gnome would care about its settings there
[00:25] <de-facto> i think i have to delete an essential binary to make it go away
[00:26] <de-facto> question is just which one?
[00:26] <tomreyn> you apparently haven't read both bug reports, yet
[00:29] <de-facto> what do you mean?
[00:35] <de-facto> I wish i could uninstall upower
[00:35] <de-facto> maybe i could make a fake package so gnome thinks its dependencies are satisfied
[00:36] <tomreyn> i guess you could either mask the upower service or upgrade to a newer ubuntu release which comes with gnome 3.32
[00:37] <tomreyn> 19.04 comes with gnome-shell 3.32
[00:38] <tomreyn> there may also be gnome extensions you can use to filter notifications
[00:41] <de-facto> i think the problem may be even deeper since it makes my mouse loose focus for a fraction of a second. i need to destroy what is at the root of this problem, i guess those popups are just the annoying symptoms
[00:43] <OerHeks> battery power .. we did discuss this before
[00:44] <de-facto> yes again and again i just cant kill it, it came back from itself
[00:50] <de-facto> when I do some gfx work in gimp the mouse looses the tool for a fraction of a second when that popup occurs, maybe its the mouse hw itself sending an battery event? i dont know how its done behind the scenes...
[00:50] <de-facto> i now tried to forbid upower to poll the kernel, silence for now... we will see how long this holds this time
[00:50] <de-facto> i am really getting mad about this
[01:03] <sadlala> if i have 2GB ram, then i can’t use ubuntu desktop?
[01:04] <slide> So when I try to run nginx, it reports that port 443 is in use. But as far as I can tell, nothing is listening on that port. netstat -tulnp | grep 443 shows nothing, fuser -k 443/tcp does nothing as well.
[01:08] <sysifoss> Hello, I'm using 18.04 and realized that the desktop is starting on tty2. If i switch to tty1 there is also a graphical dekstop. How can I set it to only start one graphical interface?
[01:08] <sysifoss> Or is this intended behavior?
[01:09] <de-facto> there its  back again, dang i will throw this mouse on a wall if this continues
[01:14] <de-facto> worst thing is that it always appears in a row, like 3-5 popups and looses of focus
[01:14] <de-facto> then silence for a while then popup spam again
[01:29] <Aleric> omg - a painless upgrade from 16.x to 18.04... Never had that before :).
[01:30] <tomreyn> sadlala: you can try ubuntu 18.04 LTS and 19.04 with 2GB RAM, but it most likely will be a bad experience. with 16.04, it should be ok. also with the current releases of several ubuntu !flavors
[01:30] <tomreyn> !flavors | sadlala
[01:30] <sadlala> thanks tomreyn
[01:30] <Aleric> Thanks everyone for the help, especially tomreyn.  Time for me to leave the channel again for now.
[01:30] <tomreyn> you're both welcome ;)
[01:31] <Aleric> :)
[01:31] <tomreyn> slide: how about    sudo lsof -i :443
[01:32] <tomreyn> slide: another explanation for this message *could be* that it fails to bind to this port because it was not started with root permissions (which are required to bind to this port).
[01:33] <slide> tomreyn, nothing from lsof
[01:36] <slide> tomreyn, hmmm this webserver was running a while ago, and then the host had to restart it and thats when it stopped working I think.
[01:36] <slide> Plus, it's the system install nginx package so I would think it would be able to listen.
[01:36] <tomreyn> sysifoss: yes, it's intended. tty1 isn't a desktop but the graphical login manager.
[01:37] <tomreyn> sysifoss: it is a side effect of switching to user (not root) owned / initialized X sessions
[01:39] <tomreyn> slide: what happens when you try to make something else bind to the port? e.g.   sudo nc -lvvp 443
[01:39] <slide> tomreyn, that works =\
[01:40] <tomreyn> so something is wrong with nginx startup apparently. did you check its logs?
[01:40] <tomreyn> systemctl status nginx    and read up on the logs which may be mentioned there, too
[01:41] <slide> alll i get is "bind() to 0.0.0.0:443 failed (98: Address already in use)"
[01:42] <tomreyn> maybe you just have overlapping or conflicting 'listen' directives
[01:42] <slide> what mechanism does nginx use to be able to bind to low number ports? setcap?
[01:43] <tomreyn> http://nginx.org/en/docs/http/ngx_http_core_module.html#listen
[01:43] <tomreyn> i assume so, but don't actually know.
[01:44] <tomreyn> actually it probably has a root owned worker process which spawns children or drops privs after binding to privileged ports, loading ssl keys etc.
[01:45] <tomreyn> i don't have nginx running here currently
[01:46] <deltab> ss -tlnp sport = 443
[01:47] <slide> ok so I moved all sites out of sites-enabled folder except default and it worked.... I then moved them all back and it still worked... so i dont know
[01:47] <sysifoss> tomreyn, just so I understand, tty1 is only a login screen and then when i login to my user account it starts it up on tty2?
[01:50] <tomreyn> sysifoss: when you login on tty1 a new desktop session will be started on an unused tty
[01:51] <Mechdave> slide, something else may be bound to that address
[01:51] <tomreyn> sysifoss: the next lowest which is unused specifically
[01:51] <Mechdave> slide, have a look here --> https://www.digitalocean.com/community/questions/nginx-is-unable-to-bind-to-443
[01:52] <Mechdave> slide, I suggest you look up fuser manual before running it
[02:08] <oceanquake> Hi all.  So I have a Ryzen Pro laptop with integrated graphics, and am getting lockups that appear to be coming from AMDGPU driver.  Am running 18.04 with hwe stack.  tomreyn and others have been very helpful, and suggested I try hwe-edge (kernel 5.0 family).  Unfortunately, this made the problem drastically worse and it locks up much more quickly.  I presume there are other fixes, e.g., in the 5.2 tree.  At this point, I'm not sure what to
[02:08] <oceanquake> do.
[02:09] <oceanquake> I can etiher live with 2-3 lockups a day (hwe), or lockups within 10 mins of being in desktop (hwe-edge).  Is there a way to try something newer?  Would it help to report hwe-edge bugs?
[02:10] <oceanquake> s/I can either/I get to either/
[02:14] <Grobo> I just installed lubuntu-core to a minimal system. The background and mouse are black. I can't set a background color for the desktop. What package do I need to add a mouse cursor and allow me to change the desktop color?
[02:25] <k_sze> Is there a nice way to set a particular bg colour when I run `less`?
[02:29] <sysifoss> tomreyn, got it thank you!
[02:49] <TuxBlackEdo> Hello, i can't boot into any linux based OS, no matter what live USB linux I use I keep getting "BAD RIP VALUE" and "Kernel unable to handle paging request" https://i.imgur.com/O53lXL5.jpg , I have no idea what's going on. I've tried resetting BIOS to defaults, tries CSM compatiability mode, tried launching in UEFI and legacy bios. I haven't gotten
[02:49] <TuxBlackEdo>  far and i've spent 4 hours on this so I am asking here
[02:51] <ZaZaGX> the imgur pic is loading slowly for me
[02:52] <TuxBlackEdo> is there a better place to upload a picture
[02:52] <ZaZaGX> it must be a great quality one
[02:52] <TuxBlackEdo> oh i guess its 4032x3024
[02:52] <ZaZaGX> did you try using another usb flash drive?
[02:53] <TuxBlackEdo> no? but ive tried ubuntu 18.04 and 19.04
[02:53] <ZaZaGX> i used an old usb flash drive and, it won't boot up
[02:53] <ZaZaGX> try a different one
[02:54] <ZaZaGX> whoa, you got a kernel panic
[02:54] <ZaZaGX> thats like the Blue Screen of Death for Windows
[02:54] <TuxBlackEdo> yea
[02:55] <ZaZaGX> well, i would use another USB flash drive to diagnose it
[02:57] <TuxBlackEdo> ok ill try that
[02:59] <ZaZaGX> is your internet connection good?
[03:00] <TuxBlackEdo> ?
[03:01] <Bashing-om> !md5sum | TuxBlackEdo
[04:28] <TuxBlackEdo> can anyone help me? i can't get any Linux distro to boot up on my computer (already tried different media/updated my bios/reset my bios to defaults/tries legacy and uefi modes) https://imgur.com/a/0m9oO9D these are 3 pictures, Picture 1) Ubuntu 19.04 live cd 2) Arch Linux (latest) 3) Fedora.... Basically theres something hardware related that's not
[04:28] <TuxBlackEdo>  letting me boot linux distros, the only OS that works is windows 10... :/
[04:40] <dars> Hi, I want to fix orientation of some of the images which are rotated 90 degress anticlockwise and want to fix their orientation. Can I do that to many of the images using command in bash?
[04:41] <dars> Please note that all the images are not wrong only some of them have wrong orientation and I want to fix their orientation
[04:46] <EriC^^> dars: you could use convert
[04:47] <EriC^^> https://askubuntu.com/questions/591733/rotate-images-from-terminal
[05:06] <Bashing-om> Hey lotuspsychje - Not a thing to main :(
[05:56] <mahdi_ja> hi all
[05:56] <mahdi_ja> how i can change owner of a folder witch owned by user with root administrator
[05:57] <mahdi_ja> i use chown -R username but owner of a folder not change
[05:59] <Mechdave> mahdi_ja, you might need to use sudo to do that. sudo chown -R username
[05:59] <Bashing-om> mahdi_ja: Show us the complete command and the path to the target directory.
[05:59] <Mechdave> as long as you have sudo priveledges
[06:00] <mahdi_ja> i first use sudo -i
[06:00] <mahdi_ja> and then use chown command
[06:01] <mahdi_ja> sudo chown inanzzz:sftp -R /sftp/mahdi/upload
[06:01] <EriC^^> mahdi_ja: what does "ls -ld /sftp/mahdi/upload" give?
[06:01] <Mechdave> I would recommend against using sudo -i
[06:01] <mahdi_ja> sudo chown mahdi:sftp -R /sftp/mahdi/upload
[06:04] <mahdi_ja> drwxrwxrwx 1 m1 m1
[06:05] <Mechdave> why are you changing the group to sftp?
[06:06] <mahdi_ja> Mechdave, i use this tutorial for createing sftp server
[06:06] <mahdi_ja> Mechdave, http://www.inanzzz.com/index.php/post/ef2z/setting-up-a-sftp-server-and-users-on-ubuntu-16-04
[06:08] <mahdi_ja> this tutorial says a  main folder must owned by root
[06:08] <mahdi_ja> but i can not change folder owner to root
[06:09] <mahdi_ja> and the owner of folder exitst to which user created this
[06:09] <Mechdave> Ok, so it should be sudo chown -R root:sftp /sftp/mahdi/upload
[06:20] <Mr_Cyclops> Hello... struggling with an NFS export issue on Ubuntu
[06:20] <Mr_Cyclops> I have two directories /abc/m1 and /abc/m2
[06:20] <Mr_Cyclops> both are individual mount points
[06:20] <Mr_Cyclops> when I export /abc via NFS, contents of m1 and m2 are not visible on the nfs client
[06:21] <Mr_Cyclops> not sure what I am missing.. plz help, tx
[06:24] <ZaZaGX> i don't know whats a nfs
[06:55] <`whoami`> Good $time_of_the_day, is there an easy way to list all text-based tools ? -- stuff that run in terminal
[06:56] <`whoami`> i mean, search with apt, but only filter stuff that run in console
[06:56] <`whoami`> if that makes sense..
[06:56] <ZaZaGX> sudo apt-cache seach nameofprogram
[06:56] <ZaZaGX> search*
[06:57] <ZaZaGX> is that what your looking for?
[06:57] <fling> `whoami`: no
[06:57] <fling> `whoami`: you can script it with ldd
[06:58] <fling> `whoami`: grepping out binaries linked with gui libs
[06:58] <fling> `whoami`: this will also remove some false negatives :>
[06:58] <fling> good luck!
[06:58] <`whoami`> ah, interesting. tyvm, gonna take a look
[06:58] <`whoami`> thanks :)
[07:01] <`whoami`> fling: but i'll only be able to run ldd against installed stuff, right ?
[07:01] <fling> `whoami`: right
[07:01] <`whoami`> so it means i might have to install a whole bunch of packages to make the test
[07:01] <`whoami`> mhmh
[07:01] <fling> `whoami`: or script it with your package manager to check for deps
[07:02] <`whoami`> ok. sounds more complicated than what i first thought :)
[07:55] <Blushpowder> Hello there.
[07:55] <the_actor> Hi
[08:19] <cjoke> I installed linux-image and header from github , its nessesary for my soundcard to work propherly. I get this anoying message in boot : vmlinuz-* has invalig signature. Is there an easy solution without turning off secureboot? I run ubuntustudio18.04
[08:27] <Blushpowder> @cjoke I don't think there's a way around it.
[08:27] <the_actor> hi
[08:28] <Blushpowder> Without secureboot third party repositories are allowed.
[08:28] <Blushpowder> With secureboot you cannot use them because secureboot sees them as threat.
[08:32] <Bashing-om> cjoke: See: http://blog.cyphermox.net/2017/08/how-to-sign-things-for-secure-boot.html .
[08:32] <cjoke> Thanks guys, Ill check it out.
[10:01] <kk89> Good morning, I connected to my AVM dsl-router, a USB MFP printer to be able to use it on a remote location. The dsl-router has opened port 9100 for the AppSocket / HP JetDirect service. Then I configured, on an Ubuntu workstation with CUPS installed, a new printer by entering the IP address of the router and its port 9100 and I was able to print c
[10:01] <kk89> orrectly. My problem instead concerns the use of the integrated scanner function through the xsane program, I added in the file /etc/sane.d/xerox_mfp.conf the string tcp my-router-ip-address 9100 but it doesn't work
[10:08] <kk89> port 9100 in an MFP printer, should it be used only for print function or can it also be used to access the scanner function?
[10:57] <bf_> I noticed ubuntu 18.04 base installation is a total size of 2 gigabytes. is there a way to reduce it to below 1 gigabyte?
[10:58] <kex> bit of an odd one - I have a HP screen and a Ubuntu laptop, when I plug in the screen the laptop recognises it has a monitor plugged in but the screen doesn't, this is over HDMI, anyone seen something like this before?
[10:59] <Cheez> kex: unfortunately the only time i have ever seen that, the cause was a dead monitor
[10:59] <Cheez> i had to replace it
[11:00] <kex> huh, weird. I just tried the USB C connector on the screen - the screen registers when that is plugged in but the laptop doesn't, doh
[11:02] <ramsub07> Hello, I accidentally pressed  ctrl+<somekey> and I am unable to type anything on my terminal. I am inside a screen at the momen
[11:16] <Dk0dr> close it and reopen it?
[11:21] <BluesKaj> Hi folks
[11:33] <thsnr> ramsub07: you probably pressed ctrl+s which causes the terminal to stop updating. press ctrl+q to continue
[11:39] <tomreyn> bf_: i find 2 GB to already be veryy, very little. maybe you can go even lower with debootstrap and a compressed file system.
[11:39] <tomreyn> bf_: what makes you want this?
[11:40] <bf_> thank you tomreyn, unfortunately I am hitting a hard limit in terms of size due to the rescue system. I am automating the setupof of encrypted hdd with dropbear remote unlocking in the hetzner environment and they don't give me more than 1gb space
[11:40] <bf_> *in the rescue system
[11:41] <bf_> I have learned that debootstrap is an alternative, but I am also considering switching to arch linux if I have to use debootstrap / remote install anyways
[11:44] <tomreyn> bf_: i'm not sure what the hard size limit is on this rescue system (i do know hetzners installation service and would always use debootstrap there).
[11:45] <tomreyn> debootstrap is not realloy 'remote install', you can do it from the recovery system.
[12:10] <Marenz> greetings
[12:12] <Marenz> i am debugging an issue in our project that only happens during the release build for the .deb package. My problem is, I can't find the exact build env. parameters used by looking at the rules file (or any other really). files can be seen here https://github.com/ethereum/solidity/blob/develop/scripts/release_ppa.sh#L157
[12:13] <Marenz> basically, I know the command line causing a compilation error, but I don't know what initial build setup/settings lead to it using those parameters
[12:15] <jujubee> My user is a member of several groups in /etc/group but when I issue $groups command it only shows the group with my username?  How do I fix this?
[12:15] <jujubee> I need to use minicom which requires membership in dialout and it is there, but not working
[12:16] <tomreyn> jujubee: did you logout and login since editing the users' groups?
[12:17] <jujubee> yes, several times
[12:17] <tomreyn> are you using a graphical desktop on this system?
[12:17] <jujubee> when I log in, the groups shows I am member of dialout but I get an error about no permissions.
[12:17] <noudle> i want to run 'apt-get -f autoremove --purge' to get rid of a bunch of unneeded initrd images in /boot/ on my ubuntu 16.04 server, however it throws something like 'Internal Error: Could not find image (/boot/vmlinuz-4.4.0-138-generic)' and so 'linux-image-4.4.0-138-generic' fails and thus the autoremove command fails. any way to fix this?
[12:17] <jujubee> Yes using xfce
[12:18] <tomreyn> this may be bug 1781418
[12:19] <noudle> same error happens on 'apt-get upgrade'
[12:19] <tomreyn> jujubee: see also the duplicate report, which describes the situation form a different perspective.
[12:20] <jujubee> Thanks tomreyn  reading now...
[12:20] <tomreyn> noudle: touch /boot/vmlinuz-4.4.0-138-generic
[12:21] <noudle> lol
[12:21] <noudle> let me try
[12:21] <tomreyn> noudle: alternatively, if you have enough space available now, apt install --reinstall linux-image-4.4.0-138-generic
[12:21] <tomreyn> this would be the better approach
[12:22] <noudle> already did the forst, sec
[12:22] <noudle> *first
[12:22] <noudle> did work, ty
[12:24] <jujubee> tomreyn commenting out in pam.d file worked.  Thanks
[12:27] <tomreyn> jujubee: which ubuntu version are you running there? maybe you could add yourself to "affects me" on top of the bug report, so it gets a bump in priority.
[12:28] <jujubee> tomreyn using xubuntu and I did "affects me".
[12:29] <unimatrix9> hi there, does anyone know this :  ( LTS 18.04 ) running a tkinterwindow as root shows the right fonts, running as normal user its ugly ..
[12:29] <unimatrix9> or how to get around this problem
[12:30] <tomreyn> jujubee: which xubuntu version then?
[12:31] <jujubee> 18.04
[12:31] <tomreyn> jujubee: okay, and it's fully updated? because there were hopes this might already be fixed
[12:33] <jujubee> yes, fully updated
[12:33] <tomreyn> thanks
[12:33] <jujubee> np
[12:36] <tomreyn> Marenz: /join #ubuntu-app-devel   for lack of a more suitable channel, i guess.
[12:38] <Marenz> thanks for the hint
[12:55] <unimatrix9> when i run python -m tkinter , i get ugly fonts
[12:55] <unimatrix9> when running sudo python -m tkinter , the ugly fonts are gone
[12:56] <unimatrix9> any one has tips on how to run as normal user and get "good" fonts ?
[13:00] <lawnmowerman> hey, i have a question. what are the limitations to linux when it comes to reading ntfs filesystems? i have a 1tb hard drive that is my data drive in a dualboot pc.
[13:01] <lawnmowerman> i want to use it with linux and winodws
[13:01] <lawnmowerman> currently it is formatted ntfs
[13:02] <BluesKaj> NTFS will work for both
[13:04] <lawnmowerman> ok thanks, BluesKaj
[13:05] <BluesKaj> just make sure your sharing oermissions are setup for your user(s)
[13:05] <BluesKaj> permissions even
[13:09] <lawnmowerman> ok, thanks
[13:14] <lawnmowerman> i can still reformat the drive as i didn't put any important data on it
[13:14] <lawnmowerman> what filesystem would be better for both operating systems
[13:14] <lawnmowerman> i think i'd want a journal for safety
[13:16] <SwedeMike> lawnmowerman: NTFS is the only journaled filesystem that Windows can read/write out of the box.
[13:24] <dalist22> Hello guys
[13:25] <dalist22> so I had a question
[13:25] <dalist22> is launchpad able to build with python 3 ?
[13:25] <dalist22> because I got this error from build log
[13:25] <dalist22> dh clean --with python3dh: unable to load addon python3: Can't locate Debian/Debhelper/Sequence/python3.pm in @INC (you may need to install the Debian::Debhelper::Sequence::python3 module)
[13:26] <tomreyn> dalist22: /join #launchpad    and   /topic
[13:26] <dalist22> okay, thanks
[13:26] <tomreyn> you're welcome.
[13:37] <artur1> hello
[13:40] <tomreyn> hi artur1
[13:43] <klaus_> hi all
[13:44] <tomreyn> hello klaus_
[13:44] <klaus_> my cpu scaling doesn't seem to work anymore after waking up from sleep, i was wondering if this was a known problem and where to start looking to fix this
[13:45] <klaus_> hi tomreyn
[13:46] <klaus_> i use tlp, my cpu frequency is normaly 2.7ghz (it's a laptop) i scale it to 800mhz usualy. it works fine as long as the computer doesn't go to sleep, then when i wake it up it's always 2.7ghz, no more scaling apparently
[13:47] <TJ-> klaus_: ACPI issue most likely; see https://iam.tj/prototype/enhancements/Windows-acpi_osi.html
[13:47] <tomreyn> klaus_: are you saying that cpu scaling used to work after waking from sleep, but no longer does? if so, which changes have you made which may have caused this to change? and which ubuntu version are oyu on?
[13:48] <klaus_> tomreyn: no as far as i can tell, on this install (ubuntu 19.04 disco dingo) it never worked after coming back from sleep
[13:49] <klaus_> TJ-: thanks for the link, i'm reading
[13:50] <klaus_> tomreyn: i'm just saying it works from initial boot until the first sleep, then after that it doesn't scale no more
[13:50] <tomreyn> klaus_: thanks for clarifying. also consider a mainboard firmware upgrade. but TJ's great article will probably help you work around it.
[13:51] <klaus_> tomreyn: this laptop is a macbookpro (12,1 early 2015) there's no accessible bios or update available of any kind as far as i know
[13:52] <tomreyn> only when you run OS X on it bare metal, i guess.
[13:53] <klaus_> TJ-: so the fix is having passing an argument to the kernel to make linux to pretend to himself he's "windows 2015" ? :)
[13:53] <TJ-> hhhm, not sure if ACPI works the same way on Macs
[13:53] <tomreyn> but they will most likely not have improved the firmware to be more compatible to other OS.
[13:53] <klaus_> -having
[13:54] <TJ-> klaus_: I don't think the strategy will work on a Mac since I doubt, even if it has an ACPI DSDT, it has "Windows" strings in it
[13:54] <klaus_> i run it single OS install, just linux
[13:55] <TJ-> klaus_: the method relies on the way most (non Apple) PCs hard-code Windows version strings into the ACPI DSDT of the firmware
[13:55] <TJ-> klaus_: I doubt Apple does that :)
[13:57] <jameshjacks0njr> Apple dosent do that sadly
[13:59] <tomreyn> klaus_: i'd be interested in the output of this, though:   sudo strings -n5 /sys/firmware/acpi/tables/DSDT | sort | uniq -c | sort -nr | nc termbin.com 9999
[14:00] <tomreyn> it may just report an error about /sys/firmware/acpi/tables/DSDT not existing
[14:01] <klaus_> TJ-: here's the content of  /sys/firmware/acpi/tables/DSDT https://pastebin.ubuntu.com/p/dZXfxt2WZv/
[14:01] <klaus_> tomreyn: i try you suggestion now
[14:02] <klaus_> tomreyn: https://termbin.com/qrcv
[14:03] <TJ-> klaus_: I think we'd make better progress to look at a kernel log that captures the suspend/resume cycle
[14:03] <klaus_> i didn't know about "| nc termbin.com 9999", it's nice
[14:04] <klaus_> TJ-: i woke up the computer about an hour ago, if you tell me what to copy/paste, i'll try ?
[14:04] <tomreyn> klaus_: thanks. so there are "Windows 2009" and "Windows 2012" at least, but keep working with TJ, he knows *a lot* more about ACPI than i do.
[14:05] <TJ-> tomreyn: you spotted Windows strings? huh, I must have missed those
[14:05] <klaus_> tomreyn: thanks for telling
[14:05] <tomreyn> TJ-: yes, those two occur twice.
[14:06] <tomreyn> you also have "_OSI`" and "\_OSI`"
[14:07] <TJ-> tomreyn: well how about that! I wonder why Apple would include those
[14:08] <tomreyn> i think apple supports or did support multi-booting with windows at some point.
[14:09] <TJ-> in which case ... klaus_  it might be worth trying: acpi_osi=! "acpi_osi=Windows 2012" on the kernel command-line, but first lets try to grab a kernel log
[14:09] <klaus_> there's this thing called 'boot camp' i never tried it but it's apple offcialy supporting mac user to install natively windows on their mac. apple provide a package with all the necessary drivers
[14:09] <TJ-> klaus_: "journalctl -k --since='2 hours ago' | nc termbin.com 9999"
[14:09] <klaus_> TJ-: ok i try that
[14:10] <klaus_> TJ-: https://termbin.com/fqz0
[14:13] <TJ-> klaus_: no clues there; try the acpi_osi entries on the kernel command-line
[14:14] <klaus_> TJ-: ok, i will do that, update grub, reboot, then go to sleep, wake up, and come back here doing a "journalctl -k --since='2 hours ago' | nc termbin.com 9999" if that's ok ?
[14:17] <tomreyn> instead of editing /etc/default/grub and running update-grub you can alternatively set those kernel parameters just for the coming boot from the grub menu. that's if you can get there easily.
[14:17] <tomreyn> !kernelparm
[14:17] <TJ-> klaus_: well, yes, but just checking if the governer is letting the CPU 'go slow' would be the best test :)
[14:18] <klaus_> TJ-: yes, i check with cat/proc/cpuinfo |grep -i mhz
[14:18] <klaus_> if it works it reports 800mhz, if it doesn't then it's 2.7gz
[14:18] <klaus_> ghz
[14:19] <klaus_> ok here's my grub line: GRUB_CMDLINE_LINUX_DEFAULT="text acpi_osi=! \"acpi_osi=Windows 2012\""
[14:20] <klaus_> i update grub and reboot, brb
[14:22] <tomreyn> does "text" interfere with UEFI FB handover sometimes? i thought someone said it could, and thus you should just use "nosplash" and not "quiet" instead.
[14:24] <ralf98> hey guys
[14:25] <klaus> TJ-: no dice ..  https://pastebin.ubuntu.com/p/KQT9ZqF4Zt/
[14:25] <klaus> and
[14:25] <ralf98> wow
[14:25] <klaus> journalctl -k --since='30 minutes ago' | nc termbin.com 9999  -->  https://termbin.com/24vo
[14:25] <ralf98> this is the first time i have ever seen an active IRC community!
[14:26] <ralf98> my email ralfbenedictflores@gmail.com
[14:26] <pragmaticenigma> !ot | ralf98
[14:26] <ralf98> ph
[14:26] <ralf98> oh
[14:26] <ralf98> ok
[14:26] <ralf98> sry
[14:27] <ralf98> d
[14:28] <TJ-> klaus: the only thing that *may* be related there are the ACPI errors around "ACPI Error: Method parse/execution failed \_SB.PCI0.WTLT" - not sure what the WTLT is though but it looks to be on the PCI root complex, so possibly won't be related to CPU governer
[14:30] <klaus> TJ-: ok, thanks for the help anyway
[14:30] <klaus> do you think i should try to report a bug somewhere ?
[14:31] <TJ-> klaus: what does this report? "grep . /sys/devices/system/cpu/cpufreq/policy*/*governor*"
[14:32] <klaus> TJ-: https://pastebin.ubuntu.com/p/MqqhRhXkGF/
[14:33] <klaus> it says powersave but won't apply scaling, somehow
[14:33] <TJ-> klaus: how about the actual values: " grep . /sys/devices/system/cpu/cpufreq/policy0/* "
[14:34] <TJ-> klaus: scaling_{min,max}_freq should be a big clue
[14:34] <klaus> TJ-: https://termbin.com/danu
[14:36] <TJ-> klaus: so the min/max are identical and the current is above max. Can you get those same reports from a fresh boot to confirm that _max_ should be higher, I'd expect it to be the same as cpuinfo_max_freq
[14:37] <klaus> yes about the min-max being identical, i think thiis part is from me in tlp
[14:38] <klaus> i specificaly tells it to stays at minimum (tlp-stat says it's 500mhz so i set that, but in practice it's 800mhz as a result)
[14:39] <klaus> i've set in tlp: scaling max frq = 500mhz
[14:40] <TJ-> klaus:  earlier you showed a pastebin with it reporting 800MHz, was that due to a different TLP settings?
[14:40] <klaus> so unless i disable tlp, even from a fresh boot, with scaling working, the output of /system/cpu/cpufreq/policy0/* would still be the same about min-max
[14:41] <klaus> TJ-: no, like i was saying, tlp-stat reports my min freq is 500mhz, so i set that, but i've never seen my cpu at 500mhz, in pratice the minimum is always 800mhz
[14:42] <klaus> hold on, i show you my tlp-stat output
[14:42] <baako> hi guys, what is the quickest way to uninstalled snap packages
[14:42] <klaus> TJ-: https://termbin.com/wura
[14:44] <klaus> baako: snap must have uninstall methods like for apt/dpkg
[14:44] <klaus> baako: yes i just looked: snap command support install, list, remove
[14:45] <klaus> so snap list, to check the package name you want to remove, then : snap remove that_package_name
[14:46] <klaus> i never use snap by the way.. i'm just saying this because it seems logical to me that way
[14:48] <TJ-> klaus: not sure this will help but it seems related (inversely!), comment #9 especially as a workaround there: https://bbs.archlinux.org/viewtopic.php?id=236997
[14:48] <lotuspsychje> baako: sudo snap remove snap-name
[14:48] <klaus> TJ-: thanks for the link, i read
[14:51] <klaus> TJ-: i have no idea what PROHOT is and how to do that: disable BD PROCHOT via msr-tools
[14:52] <klaus> although i want to keep system at minimum, not accelerate it, but still it could be worth checking enabling/disabling those stuff to see what happens
[14:54] <TJ-> klaus: I don't think it applies in your case directly. That acronym means "Bi-Directional PROCessor HOT"
[14:54] <TJ-> klaus: it's a thermal control so hopefully unrelated to your issue
[14:56] <klaus> TJ-: maybe it gets disabled when i wake up (so i get full speed)?  if so i need to re-enable it (to lower speed again), what do you say ?
[14:57] <klaus> i'm trying to look about that "enable/disable BD PROCHOT via msr-tools" action i'd like to try but, it seems a bit complicated, it involve memory addresses and values in hexadecimal
[14:57] <TJ-> klaus: this seems to be a bug in the intel_pstate driver
[14:58] <TJ-> klaus: I'd recommend not messing with the MSR; I highly doubt it will help your situation
[14:58] <klaus> TJ-: ok then i won't touch it
[14:59] <klaus> (thanks for the warning :)
[15:03] <TJ-> klaus: I've just looked at all the commits to intel_pstate.c since 5.0 in case there was an obvious bug-fix, but I don't see anything
[15:04] <TJ-> klaus: I suggest as a final check to test the latest Ubuntu mainline kernel build to see if that behaves the same way. If it is OK then somewhere in the commit history is a fix.
[15:04] <TJ-> !mainline | klaus
[15:07] <klaus> TJ-: good idea, i'm in 5.0, i see here 5.2.4 seems available: https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.2.4/
[15:08] <klaus> TJ-: or do you think i should try directly 5.3-rc2 ?
[15:09] <klaus> i'll go 5.3rc2 first :)
[15:10] <TJ-> klaus: latest first is always best :)
[15:12] <klaus> TJ-: yep :)
[15:13] <klaus> time to reboot and go *kaBoom* latest kernel
[15:17] <klaus> TJ-: https://pastebin.ubuntu.com/p/SZS3WS2ctD/
[15:17] <klaus> at least latest kernel seems to work normaly.. no i try go to sleep
[15:17] <klaus> *now
[15:20] <klaus> same .. cpu MHz         : 2600.401
[15:20] <klaus> at least i'm running latest kernel ;)
[15:21] <klaus> maybe i should try disabling that kernel argument i added earlier
[15:25] <klaus> root@kmbp:~# echo 500000 > /sys/devices/system/cpu/cpufreq/policy0/cpuinfo_max_freq
[15:25] <klaus> -bash: /sys/devices/system/cpu/cpufreq/policy0/cpuinfo_max_freq: Permission denied
[15:26] <TJ-> klaus: wrong node, use scaling_max_freq
[15:27] <nacc> klaus: you can't generally change the max frequency value, it's a read-only file
[15:27] <klaus> TJ-: https://termbin.com/hc40
[15:27] <nacc> I believe that file is just telling you what cpuinfo says is the max (hence is readonly)
[15:27] <klaus> it's the journalct -k
[15:27] <klaus> (since=20minutes ago)
[15:28] <klaus> nacc: ah ok thanks for telling
[15:29] <klaus> TJ-: root@kmbp:~# echo 500000 > /sys/devices/system/cpu/cpufreq/policy*/scaling_max_freq
[15:29] <klaus> -bash: /sys/devices/system/cpu/cpufreq/policy*/scaling_max_freq: ambiguous redirect
[15:29] <klaus> why the * isn't working there ?
[15:29] <jink> Because more than 1 file matches.
[15:30] <klaus> if i do one by one manually, i get no error output, but cpufreq stays at 2.7ghz anyway
[15:31] <TJ-> klaus: when the PC suspends, and when it resumes, is it only on battery, or on AC ?
[15:35] <dwilson5817> is it okay to mount a backup NFS at /mnt/?  the linux system hierarchy says mnt should be used as a temporary mount point but then it says that system software shouldn't depend on it if it is permanent?
[15:35] <nacc> dwilson5817: you can mount whatever you want wherever you want :)
[15:36] <dwilson5817> thanks nacc :)  would it be recommended to do this? might there be somewhere else better suited to mounting it?
[15:36] <TJ-> dwilson5817: I generally create mountpoints under /mnt/ as in /mnt/iso /mnt/tmp /mnt/cttv /mnt/some-remote and so on
[15:37] <klaus> jink: should i do a for loop ? for https://pastebin.ubuntu.com/p/FP3HVS7BPg/
[15:37] <nacc> dwilson5817: i think /mnt is a fine place to put backups, but like TJ- said, i'd use a subdirectory, well-named
[15:37] <klaus> TJ-: in this case it was on AC all the time, but tlp profile says to behave like if it was on BAT
[15:37] <klaus> all settings at minium
[15:38] <dwilson5817> thanks guys, yeah I'm using a subdirectory /mnt/backup
[15:38] <TJ-> klaus: it would be worth testing if being AC is causing a difference
[15:38] <nacc> dwilson5817: i think that's totally reasonable
[15:39] <dwilson5817> that's awesome thanks nacc just wanted to make sure :P
[15:39] <klaus> root@kmbp:~# for x in /sys/devices/system/cpu/cpufreq/policy[0-3]/scaling_max_freq; do echo 500000 > $x; done
[15:39] <klaus> reports no errors but cpufreq is still max
[15:39] <klaus> TJ-:  ok i try unplug and do the sleep/wakeup cycle
[15:42] <klaus>   TJ- same, cpu MHz  : 2600.556 ...  and policy0/cpuinfo_max_freq:3100000  and policy0/cpuinfo_min_freq:500000
[15:43] <TJ-> klaus: must be something to do with how the intel_pstate driver is (not) handling it correctly.
[15:44] <klaus> TJ-: https://termbin.com/6ogfb  i see those lines:   IRQ xxx: no longer affine to CPUx
[15:48] <TJ-> klaus: try booting, and the suspend/resume with these options on the kernel command-line ('debug' is for the kernel - we might get more details): debug systemd.log_level=info
[15:53] <klaus> TJ-: i will try that (debug systemd.log_level=info) but i try this too:   GRUB_CMDLINE_LINUX_DEFAULT="text acpi_osi=!Darwin"
[15:54] <klaus> as detailed here: https://forum.manjaro.org/t/runaway-acpi-interrupt-kworker-cpu-hog/81182
[15:54] <klaus> it might not have anything to do with this problem, but well, i'm trying ..
[15:54] <klaus> brb
[15:58] <klaus> TJ-: i thought i saved your suggestion about systemd debug log kernel parameter, but somehow i didn't
[15:58] <klaus> could you paste it me back please so i can try ?
[15:58] <TJ-> klaus: try booting, and the suspend/resume with these options on the kernel command-line ('debug' is for the kernel - we might get more details): debug systemd.log_level=info
[15:59] <klaus> thanks, i try that now  (the !darwin didn't help)
[16:03] <klaus> TJ-: https://termbin.com/cg29
[16:04] <klaus> (power cable unplugged, systemd debug log in kernel parameters, journalctl -k since 15 minutes boot, sleep, wakeup cycle)
[16:07] <klaus> TJ-: GRUB_CMDLINE_LINUX_DEFAULT="text acpi_osi=!Darwin debug systemd.log_level=info"
[16:08] <klaus> as soon as i put this computer to sleep it's bye bye cpu scaling ..
[16:10] <TJ-> klaus: time to find a relevant bug or create a new one against the intel_pstate driver in-kernel. Take a look at this list, 200759 looks like it could be related.
[16:10] <TJ-> https://bugzilla.kernel.org/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&bug_status=RESOLVED&bug_status=VERIFIED&bug_status=REJECTED&bug_status=DEFERRED&bug_status=NEEDINFO&bug_status=CLOSED&component=intel_pstate&product=Power%20Management&query_format=advanced
[16:11] <klaus> i look
[16:14] <TJ-> klaus: add pstate debugging (after a fresh boot) then do a suspend/resume cycle and report the kernel log again. To add (dynamic) debugging do: echo 'file intel_pstate.c +p' | sudo dd of=/sys/kernel/debug/dynamic_debug/control
[16:18] <klaus> TJ-: i did that command you suggested, but i see no file created ?
[16:18] <VBalazs> hey can i compile source code *(makefile)* on linux to execatuble (.exe file)?
[16:19] <TJ-> klaus: no file is created, you're telling the kernel to write additional debug messages to the kernel log
[16:20] <klaus> ah and i should provide those additional debug log with the output of journalctl -k ?
[16:20] <TJ-> klaus:  'file intel_pstate.c +p' tells kernel to output debug message from the kernel source-code file intel_pstate.c
[16:20] <TJ-> klaus: correct
[16:20] <TJ-> klaus: if you do this it'll show you all the messages that will enable: "sudo grep intel_pstate.c  /sys/kernel/debug/dynamic_debug/control "
[16:23] <klaus> ok i try to do that.. i've never reported a bug to the linux kernel directly i hope they take it easy on me if report in wrong area or submit bad or insufficient data ;)
[16:24] <klaus> ok time to reboot, do the cycle and report then, i'll bbiab, thanks again for all the help TJ-
[16:35] <net|> codeblocks on ubuntu arm64 crashes on startup
[16:35] <net|> tried to use it on jetson nano
[16:36] <lotuspsychje> !arm | net|
[16:57] <klaus> TJ-: https://bugzilla.kernel.org/show_bug.cgi?id=204367
[16:58] <klaus> i hope i submitted that correctly
[17:02] <ncuxo> hello can I receive some information/advice about bridging connections in here or should i search for another channel
[17:03] <sarnold> ncuxo: #ubuntu-server is more likely to know about bridges; but there's more active people in here. tough to guess.
[17:06] <TJ-> klaus: in the attached log, the journalctl towards the end has "kmbp [1415]: Failed to execute /usr/lib/systemd/system-sleep/pre-suspend.sh: Not a directory" ... I wonder if that could be related?
[17:06] <Elliria> What are you trying to bridge, ncuxo?
[17:07] <klaus> ah yes it's a script i made: here's the content: https://termbin.com/4q9u
[17:08] <Gargoyle> tomreyn: Thanks for the PCIe tips. I'll give them a try. It does seem that the problem has gotten much worse with the latest BIOS which is probably changing some things up with support for gen3 ryzen.
[17:08] <klaus> it's because, the computer won't stay asleep, when i close the lid, it wakeup on his own in a few seconds after entering sleep
[17:08] <sarnold> klaus: try namei -l /sys/bus/usb/devices/1-3/power/wakeup and namei -l /sys/bus/usb/devices/1-5/power/wakeup
[17:09] <TJ-> klaus: that script has several syntax errors
[17:09] <klaus> TJ-: ah it does ?
[17:09] <TJ-> klaus: "#!/bin/bash/" -> "#!/bin/bash", and last line is needed: "fi"
[17:10] <tomreyn> Gargoyle: you're welcome. a web search for "1022:1453" also turned up https://bugzilla.redhat.com/show_bug.cgi?id=1616364 and some other potentially useful discussions.
[17:10] <TJ-> klaus: but the error message looks unrelated to that, seems something is expecting a directory and finding a file maybe?
[17:11] <jayjo> I noticed a file system error on a machine of mine, so I restarted it, and then got the error "fsck error on boot: /dev/sda2: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY". I did, and it has been starting up for about 5 mins now, saying "a start job is running for discard unused blocks". What is this - what actually happened to my machine?
[17:11] <sarnold> TJ-: probably ower/wakeup and namei -l
[17:11] <sarnold>                            /sys/bus/usb/devices/1-5/power/wakeup
[17:12] <sarnold> SIGH
[17:13] <sarnold> (hilarious timing, exo cet and gabkd lly joined when I was selecting text, and even thought the selection indicator was on the "/sys/bus/usb/devices/1-3/" part, the next line was selected too!)
[17:13] <TJ-> klaus: this "/usr/lib/systemd/system-sleep/" is not a valid path, if anything, it should be "/lib/systemd/system-sleep/"
[17:14] <TJ-> jayjo: "discard blocks" is an operation on flash devices to release blocks that are no longer required by files
[17:15] <Sven_vB> hi! please help me debug one of my xenial notebooks. it recently started to dim the screen or even turn it black, even while VLC is playing. I thought I had solved this long ago, using puppet to manage all the power saving options, but it seems a new mechanism has appeared that I haven't covered yet. what could it be?
[17:15] <Gargoyle> tomreyn: I must have been really tired and fed up. I didn't even think about googling the actual PCI numbers. I was just confused about why they didn't match "lspci" output! X-)
[17:15] <jayjo> TJ-: is something wrong if it's still working at the 10 min range?
[17:16] <TJ-> jayjo: it sounds like it
[17:16] <jayjo> Can I interrupt it safely to get back to the GRUB menu?
[17:16] <klaus> TJ-: thanks for correcting my errors in the script, but about the path,
[17:16] <TJ-> jayjo: did the 'fsck' you ran manually complete, and then you resumed the system booting?
[17:16] <klaus> TJ-: there's other systemd file in this path, beside mine, see: https://pastebin.ubuntu.com/p/8GqyDdHFTH/
[17:16] <jayjo> all i did in the busybox terminal was run `fsck /dev/sda2` and then hit `a` for always yes. It seemed to go through some unused nodes and said FIXED, and I did `reboot` in that busybox term
[17:18] <TJ-> klaus: what ubuntu release is that? on 18.04 "apt-file search /usr/lib/systemd/system-sleep/" returns no results
[17:19] <TJ-> jayjo: OK... I think it is safe to do do a restart using the kernel's SysRq mechanism, hold down Alt+Ctrl+SysRq then whilst holding them press, in turn, R E I S U B
[17:20] <TJ-> jayjo: it is possible the damage to the root file-system that fsck tried to fix has broken the OS such that required files are missing/corrupted
[17:22] <jayjo> TJ-: after interrupting the 'wait' and restarting, it seems to boot fine. Sort of weary of this, but I've restarted now once and done a full power off and on, seems to be ok. Was it something that I did that caused me to need the manual fsck in the first place? Maybe I can avoid it
[17:23] <TJ-> jayjo: most often those kind of errors are caused by sudden power loss or unexpected system lock-up
[17:24] <Sven_vB> I'm back :) did I miss sth.? the official logs are some hours behind it appears
[17:25] <oceanquake> Hi all.  So I have a Ryzen Pro laptop with integrated graphics, and am getting lockups that appear to be coming from AMDGPU driver.  Am running 18.04 with hwe stack.  tomreyn and others have been very helpful, and suggested I try hwe-edge (kernel 5.0 family).  Unfortunately, this made the problem drastically worse and it locks up much more quickly.  I presume there are other fixes, e.g., in the 5.2 tree.  At this point, I'm not sure what to
[17:25] <klaus> TJ-: it's ubuntu 19.04 and that script is mine
[17:25] <oceanquake> do.  Is there a way to try a newer kernel?  Should I report the problems occurring with the hwe-edge kernel, and if so, would a bugreport to Ubuntu make any difference to AMDGPU devs?
[17:25] <arandomandy> Hey there. I just mounted a USB drive: sudo mkdir /mnt/usb; sudo mount /dev/sdb1 /mnt/usb/. Though I wanted to delete some of the files on it and I get the following message rm: cannot remove 'Test.class': Permission denied
[17:26] <TJ-> klaus: is the script marked as executable as well, couldn't see that in your pastebin
[17:26] <klaus> TJ-: the path wasn't my doing
[17:26] <klaus> yes
[17:26] <TJ-> klaus: looks like paths have 'moved' around a bit
[17:27] <klaus> $ ls -hal /usr/lib/systemd/system-sleep/pre-suspend.sh
[17:27] <klaus> -rwxr-xr-x 1 root root 362 juil. 29 19:11 /usr/lib/systemd/system-sleep/pre-suspend.sh
[17:27] <TJ-> klaus: I suppose the "not a directory" may actually have been referring to the /bin/bash/ shebang
[17:28] <TJ-> klaus: it'd be cool if everything worked fine now that script is fixed and it should actually execute!
[17:28] <oceanquake> arandomandy: if the root of your filesystem is owned by root with no write permissions for other users, then you will have to either change the perms or just modify the files as root
[17:28] <oceanquake> arandomandy: what is the USB drive's filesystem?
[17:28] <sarnold> oceanquake: I can't promise that a bug report would lead to a fix, but if no one files any bugs, the kernel team won't ever know that there's problems
[17:29] <oceanquake> sarnold: Sure, I understand that, but I also don't want to just add a bug to the DB if upstream is not going to do anything about it
[17:31] <TJ-> !mainline | oceanquake: you could try the latest 5.3 release candidate
[17:31] <arandomandy> oceanquake: thanks. i tried modifying the files as root and it worked
[17:31] <sarnold> oceanquake: yeah, makes sense; I know I've seen some bug fixes for ppc64 machines, perhaps only exposed via amdgpu, so it feels like someone cares :)
[17:32] <TJ-> oceanquake: also when the lock-ups occur is it the entire PC? I'm wondering if you have a SSH session to the PC when it locks up locally if you can still investigate?
[17:32] <oceanquake> sarnold: I guess what I'm trying to poke at is I presume upstream will not really care unless I'm on the latest release, but it's not worked well for me when I've tried to use a kernel that doesn't come from the package repo; for example, I've seen firmware files not all being there.
[17:34] <oceanquake> TJ-: I can use Magic SysRq - is there a way to do something useful from there?  I don't like running sshd as this is my "go everywhere" laptop and I don't want that service running, but I could certainly leave it up for debugging purposes
[17:36] <TJ-> oceanquake: I'm thinking for debugging, when you're relatively static, have another PC SSH-ed into with a "journalctl -k -f" running (follows the kernel log) so you might catch reports that never make it to log files for example, and if you're lucky the PC may still remain usable via SSH so you can investigate in-depth
[17:36] <TJ-> oceanquake: if you know a later kernel provokes this more frequently, test with that so you don't have to wait so long :)
[17:37] <sysifoss> Hi I'm trying to start a program on boot without dropping into a user desktop. Is this possible with "startx [program]"? I've tried that command on its own from a tty but I just get a black screen I cant interact with. Is this a problem with my program or am i using startx wrong?
[17:37] <oceanquake> TJ-: kern.log (and corresponding journal) show the error message in full and it's preserved thanks to rebooting with magic sysrq - is there any kind of guide to getting more debug details?
[17:38] <TJ-> oceanquake: there is the dynamic_debug option if the amdgpu driver uses that
[17:39] <tomreyn> Gargoyle: i'd also hoped companies would do actual pre-release testing against linux, apparently this kepps falling under the table due to "fixed release date".
[17:40] <sarnold> sysifoss: what are you trying to do?
[17:42] <TJ-> oceanquake: the kernel log you have may indicate where you could enable dynamic_debug logging within amdgpu. There are rather a LOT of sites, see "sudo grep amdgpu /sys/kernel/debug/dynamic_debug/control"
[17:42] <TJ-> oceanquake: if you create a bug report and attach logs then it would be easier for people to help as they have time and you have newer, more targeted, log info
[17:44] <sysifoss> sarnold, start a unity game when the computer boots. Currently it works fine with desktop icons hidden and "startup applications" setup running the game launcher upon login. Just trying to see if there is a way I can prevent an end user from knocking themselves out of the game by confining them to only running the game with some scheme setup like a daemon or something to check that the game is running in case there is a crash and start
[17:44] <sysifoss> it back up.
[17:45] <sysifoss> Basically trying to take control away from an end user as this is meant to be a seamless installation
[17:46] <sarnold> sysifoss: aha; you could try something like putting the game, and only the game, in the ~/.xinitrc file or ~/.xsession file -- so that it would serve as the window manager. if it does, X11 does, and if you configure autologin, it ought to be able to relaunch the game
[17:49] <sysifoss> sarnold, when you say it ought to be able to relaunch the game do you mean that if it crashes since x11 thinks the game is the window manager it will start it back up?
[17:53] <sarnold> sysifoss: yes; when the last process from ~/.xinitrc dies, X11 stops. so, if you set it to auto-login, it ought to restart once it dies
[17:54] <oceanquake> TJ-: just to confirm, for mainline kernels, I should be downloading the .debs manually - this is not recommended to be done via ppa?
[17:54] <sysifoss> sarnold, cool thanks I'll give this a shot!
[17:54] <sarnold> sysifoss: cool ;) please do report back how it goes, I've never tried autologin before
[18:03] <klaus> TJ-: it would have been nice, but no, problem persists.. well it's not the worst, i'll shutdown instead of sleeping
[18:13] <Sven_vB> looks like something messes with the backlight. I installed xbacklight to monitor it, set maximum brightness via keyboard. "xbacklight -get" showed 100.000…. a few moments, later, it got dark and "xbacklight -get" now shows only 11.1111….
[18:14] <Sven_vB> when I move the mouse it goes back up to 100.000….
[18:18] <TJ-> klaus: something you could try, as an experiment, before suspending reset the _max_ from 500 to 3100000 and then on resume set it back to 500
[18:19] <TJ-> oceanquake: correct, we have to manually download and install the .deb files
[18:24] <manornk> Hey, I am using Ubuntu 19.04 and I need to install php5.4, anyone know how I can do this?
[18:27] <MrCrackPot> manornk you shouldnt really be using php5 any more
[18:27] <MrCrackPot> its too insecure
[18:30] <sarnold> manornk: no supported method exists for php 5.4. You can get php 5.5 via https://ubuntu.com/esm . Or you could look to see if someone else maintains an ancient php ppa
[18:30] <MrCrackPot> sudo add-apt-repository ppa:ondrej/php
[18:30] <tomreyn> manornk: indeed php 5 is not supported on any supported ubuntu release (not counting !ESM here). nor by php.net https://php.net/supported-versions.php
[18:31] <tomreyn> MrCrackPot: i'd recommend pointing people to the website instead, where ondrej explains the support policy (and asks for donations)
[18:31] <MrCrackPot> ah that i dont have sorry
[18:31] <sarnold> TIL onrej's got a website for his ppa :)
[18:33] <tomreyn> manornk: so if you really really must use PHP 5 on 19.04, your least terrible option is probably https://launchpad.net/~ondrej/+archive/ubuntu/php?field.series_filter=disco
[18:33] <tomreyn> see also https://deb.sury.org/
[18:34] <tomreyn> not 5.4, though, just 5.6
[18:34] <TJ-> Sounds like a job for lxd, debootstrap, and old-releases.ubuntu.com
[18:35] <MrCrackPot> sounds like the app website etc needs completely redoing
[18:50] <manornk> MrCrackPot, sarnold tomreyn  Yes, I know, but my client has it, and he doesn't want to invest in upgrade. Thanks
[18:51] <MrCrackPot> id advise him that not investing could potentialy cost him his business and also jail time
[18:51] <MrCrackPot> he is knowingly and willingly using insecure software
[18:52] <manornk> MrCrackPot, jail? What do you mean? GDPR?
[18:52] <MrCrackPot> you are willingly leaving customer information available on an insecure site
[18:52] <Sven_vB> my backlight problem seems to be about xfce4-power-manager-settings, a recent update might have changed its default settings. I've configured it and will observe.
[18:53] <manornk> MrCrackPot, I tried explaining that. I personally tell my clients that I would maintain until Version does not experience EOL, but a company I work for doesn't do the same
[18:53] <MrCrackPot> report
[18:55] <ioria> manornk, can we know why php5.4 it's  so necessary ?
[18:56] <manornk> ioria, old company app is build in it, and I need to do upgrade on features
[19:01] <ioria> manornk, and cannot be rebuilt with current php ?
[19:01] <manornk> ioria, it's old code, so it's expensive upgrade. for example mysql_connect and stuff is still used. And it's not really built by any standards. I am not sure even who built it
[19:03] <ioria> manornk, think twice about it, but , in any case, your only option it's from source: https://github.com/php/php-src/tree/PHP-5.4
[19:04] <manornk> ioria, thanks, but saddly I don't make decisions
[19:04] <ioria> manornk, as you wish
[19:07] <TJ-> manornk: you can find 5.4.6 and 5.4.9 at http://old-releases.ubuntu.com/ubuntu/pool/main/p/php5/
[19:12] <inf009> Hi! I'm new using this channel. I need help regarding dolphin file manager. I want to use gnome terminal as the default terminal, can someone help? I'm using ubuntu 18.04. Thanks.
[19:13] <sarnold> welcome inf009, just hang out for a bit, hopefully someone who knows dolphin will see
[19:14] <oceanquake> isn't Dolphin from KDE?
[19:15] <oceanquake> inf009: what is the thing you need help with?
[19:16] <OerHeks> inf009, dolphin *filemanager*  and gnome *terminal*? not gnome nautilus filemanager?
[19:17] <sarnold> meh, I'm not too surprised someone would want to pick and choose their tools
[19:17] <sarnold> whether or not the tools make that easy or even possible is another question :)
[19:17] <OerHeks> xdg-mime default nautilus.desktop inode/directory application/x-gnome-saved-search  ## to change to nautilus >> https://sites.google.com/site/installationubuntu/tweaking-ubuntu/change-default-filemanager
[19:18] <oceanquake> agree sarnold, just seemed a little odd for a new user to be mixing and matching DE components
[19:19] <sarnold> oceanquake: it certainly soundsl ike a fast way to use twice as much memory :)
[19:19] <inf009> I want to use a more "advance" file manager in ubuntu. I've read dolphin is supposed to be good, so I am trying it out. I've managed to change dolphin to be the default manager, but when browsing to a file location and opening the terminal (Shift + F4) I get an error KDEInit count not launch 'konsole', which I am guessing is the default terminal in KDE.
[19:19] <OerHeks> filemanager <> terminal/konsole... 2different apps
[19:20] <sarnold> OerHeks: I think the filemanager is trying to *launch* a terminal
[19:20] <TJ-> ^^
[19:21] <inf009> I am sorry, the question I wanted to ask is: is there a way to change the default terminal used by dolphin?
[19:21] <TJ-> inf009: you want file-manager and a terminal - "mc" (midnight commander) !! or, are you wanting to execute a command ('file') in a terminal shell ?
[19:23] <inf009> Or is using dolphin in ubuntu a bad practice?
[19:24] <sarnold> no, it's fine
[19:24] <sarnold> but as oceanquake suggested, normally people don't try to mix up their desktop environment; folks who use gnome things or use kde things tend to not use tools from the other environment
[19:24] <OerHeks> dolphin just pulls in some KDE library's, no problem
[19:25] <oceanquake> but you may get more insight from #kde on how to change what dolphin invokes when you ask it for a "terminal here"
[19:25] <oceanquake> at risk of getting beaten up here, I'm actually on KDE Neon :)
[19:26] <ioria> inf009, https://forum.manjaro.org/t/adding-gnome-terminal-open-here-in-terminal-context-menu-in-dolphin-file-manager-in-kde/93392/2
[19:27] <inf009> I don't really care if it is dolphin or something else, I am lloking for a more feature-rich file manager, what would you recommend for ubuntu?
[19:28] <viju> nautilus?
[19:28] <OerHeks> nautilus with some extentions like nautilus-admin is great
[19:28] <Aram> hi. I am trying to help some other person. I am connected through ssh to a live ubuntu image. can I run the CLI installer through ssh?
[19:28] <Aram> I know I can run the GUI installer, with X forwarding
[19:28] <Aram> but I want the CLI installer
[19:28] <ioria> inf009, what features do you need ?
[19:29] <Aram> basically what I am asking is: "what is the CLI installer name/binary?"
[19:29] <ioria> Aram, ubiquity
[19:29] <sarnold> Aram: poke around for a debian-installer kind of name
[19:29] <Aram> ioria: isn't that the GUI installer?
[19:30] <ioria> Aram, sure
[19:30] <Aram> can't run that, need the CLI one.
[19:30] <ioria> Aram, oh, you mean plain cli ? nope, i don't think there is one
[19:30] <Aram> well there's one in the server image, for sure.
[19:31] <ioria> Aram, with ncurses ?
[19:31] <Aram> yes
[19:31] <OerHeks> the mini iso gives a text installer, but has no live option to open ssh-services
[19:31] <DOSfan> question : I am running into an issue where I am getting an error on boot (Ububtu v16) saying ... cannot find output "VGA-0" .. when I look at xrandr I see the minotor on VGA-1 only .. how can I change the settings to look at the VGA-1 output device only?
[19:32] <DOSfan> I have tried doing ... xrandr --addmode VGA-1 1280x1024_60.00 .. but like I said it does not find that device so I cannot add one.
[19:32] <ioria> Aram,  subiquity ? not sure
[19:33] <Aram> sarnold: debian-installer package contains just documentation, it seems
[19:33] <ioria> probably
[19:33] <sarnold> Aram aww :(
[19:34] <DOSfan> opps I ment xrandr --addmode to VGA-0.
[19:34] <DOSfan> or perhaps I can just force add this device VGA-0 to where ever it is getting this information on during boot-up.
[19:36] <Aram> yeah  this is more complicated than hoped
[19:38] <Aram> ok, then trying something else then
[19:39] <Aram> problem is she can't disable secure boot because windows enabled fast boot which disabled the F2/whatever EFI prompt, so the EFI screen is not accesible.
[19:39] <Aram> and she doesn't have windows anymore to disable this.
[19:39] <Aram> can she boot to EFI directly from within the Live image?
[19:39] <Aram> she has to disable secure boot, because unfortunately there's no open source wifi driver, gaah.
[19:39] <Aram> crap laptop.
[19:41] <leftyfb> Aram: she can boot a Windows usb and access EFI settings
[19:41] <Aram> she doesn't have a windows usb
[19:41] <leftyfb> make one for her?
[19:42] <Aram> we're 20,000km apart
[19:44] <Aram> systemctl reboot --firmware-setup
[19:44] <Aram> perhaps this will work
[19:46] <sarnold> Aram: there's links to windows install media at the end of https://rufus.ie/
[19:46] <Aram> or perhaps I can use efibootmgr?
[19:46] <Aram> sarnold: yeah, but then she will lose access to this live ubuntu image as she only has one USB I think
[19:46] <Aram> it's a mess
[19:46] <sarnold> Aram: *ugh* :(
[19:47] <Aram> I think I can use efibootmgr though
[19:47] <Aram> yes
[19:47] <Aram> Timeout: 0 seconds
[19:47] <Aram> maybe I can increase it
[19:48] <pragmaticenigma> Aram: Until you can get windows to release the fastboot lock, all of this effort will be for not. So flashing the USB with the windows installer to get windows back on the machine, to disable the fastboot, then reimage the drive with Ubuntu live seems like a much faster and more simple approach
[19:49] <sarnold> pragmaticenigma: hey :) last week or so puff wanted to say "thanks" :)
[19:51] <ioria> Aram, why don't you use debootstrap if you really want cli-only install ?
[19:52] <pragmaticenigma> ioria: the bigger issue is they are trying to remote install for a friend a significant distance away
[19:52] <ioria> in see
[19:52] <ioria> *I
[19:52] <sarnold> ioria: you'd still have the 'efi boot is kinda busted' part of the problem, I think
[19:53] <ioria> ok
[19:57] <Aram> pragmaticenigma: sudo efibootmgr -t 10 && sudo systemctl reboot --firmware-setup worked!
[20:01] <jujubee> how do I call "service htpdate restart" from a cron job?
[20:05] <pragmaticenigma> jujubee: Why are you using htpdate instead of ntp?
[20:07] <jujubee> pragmaticenigma on a raspberry pi there is no hardware clock.  For a curriculum I use, we need RPIs and sometimes we don't get internet connectivity until it is too late.  I had many headaches setting up ntp.
[20:08] <jujubee> Plus, some places block port 123
[20:08] <jujubee> but not port 80
[20:10] <pragmaticenigma> jujubee: how do you normally call a restart currently?
[20:10] <jujubee> service htpdate restart
[20:11] <pragmaticenigma> jujubee: start with using "systemctrl" as hopefully htpdate has been updated
[20:12] <apxseemax> Hi there
[20:13] <jujubee> pragmaticenigma systemctl you mean?
[20:13] <pragmaticenigma> jujubee: yes ... curse of always using tab complete in the terminal
[20:14] <thebishop> hi friends. i have an hp spectre 13" x360 ('gem cut', late 2018 model), running 19.04.  a couple weeks ago, suspend stopped working when i close the lid.  it definitely worked when i first installed.  even doing systemctl suspend turns the screen off for a sec, but then it comes back on, no suspend.  any ideas how i can debug this?
[20:15] <pragmaticenigma> If htpdate is installed correctly... "systemctl restart htpdate.service" should restart the service form cron
[20:16] <apxseemax> UFW question: (yes I should use iptables, I kinda do not) this: [ 4] 20,21/tcp                  ALLOW IN    192.168.178.0/24, should let me ftp into my machine, but it won't. I am kinda clueless why tho, any hinds into any direction why this might be? (daemon is running, config is setup, I use filezilla, machine is def. reachable)
[20:16] <pragmaticenigma> thebishop: checking system logs might be a good place to start... since you're able to replicate it... note the time, then look in the logs for events occuring during that time to see what might have triggered a wakeup
[20:16] <apxseemax> systemlogs... thats actually a good hint
[20:16] <pragmaticenigma> apxseemax: Does the machine actually have an FTP server installed on it?
[20:17] <apxseemax> pragmaticenigma: yes vsftpd, as written, the daemon is running
[20:18] <jujubee> pragmaticenigma systemctl restart htpdate does work.  so do I need quotes around it in a cron? or just */15 * * * * systemctl restart htpdate ?
[20:18] <pragmaticenigma> no quotes should be required
[20:19] <pragmaticenigma> apxseemax: without qualification, "daemon is running" could refer to anything. helps to be detailed
[20:20] <pragmaticenigma> jujubee: why are you restarting a service every 15 minutes?
[20:20] <jujubee> pragmaticenigma thanks.  I think I was resetting the date/time too close to when the cron was supposed to run.  Patience is not one of my virues ;)
[20:20] <jujubee> seems to be working now.
[20:21] <pragmaticenigma> jujubee: just for reference though... There is a hardware clock. The Rpi platform does not have a Real Time Clock on board, which means when the unit is powered down it doesn't know what time it is. Once the machine is booted it shouldn't require time adjustments that frequently. It may drift a little, but nothing significant that requires checking every 15 minutes
[20:22] <pragmaticenigma> jujubee: If one of your units is drifting that severely, then you have a defective board and it would be a good idea to replace it
[20:24] <apxseemax> pragmaticenigma: [UFW BLOCK] IN=ens192 OUT= MAC=<redacted> SRC=192.168.178.22 DST=192.168.178.34 LEN=44 TOS=0x00 PREC=0x00 TTL=128 ID=226 PROTO=UDP SPT=50854 DPT=8612 LEN=24
[20:24] <apxseemax> looks like filezilla tries udp
[20:25] <apxseemax> here I thought ftp is tcp confirmed
[20:25] <apxseemax> now thats embrassing
[20:25] <pragmaticenigma> apxseemax: It can use both
[20:27] <jujubee> I believe ftp uses UDP for the connection, not the data
[20:27] <pragmaticenigma> apxseemax: I would caution against having those specific ports open though. If you really must use FTP, I would encourage selecting a port higher up. Unless you're on an internal network only, or have gateway firewalls to protect it from the internet
[20:28] <apxseemax> I do have a gateway FW, but I am also restricting the traffic allowed to my private network at home
[20:28] <apxseemax> Its an old vmware HV I am running a small gameserver on
[20:29] <jujubee> pragmaticenigma not drifting as far as I know, just don't know when the user will connect to wifi so the time *can be adjusted...
[20:29] <apxseemax> its running on a deny all, only allow policy for now
[20:29] <sarnold> TFTP uses UDP. FTP uses TCP.
[20:29] <jujubee> pragmaticenigma I am trying to make this RPi image as idiot proof as possible.  Just plug it in and connect to wifi and go.
[20:30] <pragmaticenigma> jujubee: perhaps look into configuring the service instead... cron seems like a taking a bulldozer to move an ant-hill
[20:30] <apxseemax> pragmaticenigma: thats weird, its still blocking the inbound traffic
[20:32] <pragmaticenigma> apxseemax: Don't use FileZilla's quick connect... use the actual connection dialog window to setup the connection. There are more options and it prevents FileZilla from trying it's autonegotiation stuff
[20:33] <thebishop> pragmaticenigma, two things i noticed.  #1: my syslog is being flooded with messages like this: "kernel: [  348.902025] hid-sensor-hub 001F:8087:0AC2.0004: hid_field_extract() called with n (192) > 32! (kworker/7:3)"
[20:33] <apxseemax> Will try
[20:33] <thebishop> #2 not 100% sure but it looks like networkmanager is causing the wake: https://pastebin.com/8ReQx0kP
[20:35] <jujubee> pragmaticenigma do you mean run htpdate on if.up?
[20:35] <pragmaticenigma> jujubee: no
[20:36] <pragmaticenigma> thebishop: It does appear that something with NetworkManager did trigger a wake... I'm not familiar with kernel messages, though "hid" typically refers to Human Interface Device, i.e. Mouse, Keyboard
[20:36] <thebishop> yep
[20:37] <thebishop> not sure if that's connected, i found a couple references to it on the kernel mailing list and not much else
[20:37] <godfuture> hi guys, I am fighting with my old Samsung Ativ Book 7 and linux backlight. This notebook has Intel and AMD graphics integrated. The problem? My fn keys change the intel backlight, but nothing happens. Only if I write to acpi_video0 backlight, the screen reacts. How can I redirect my fn keys to acpi_video0 backlight?
[20:37] <pragmaticenigma> jujubee: did you configure htpdate to run as a system service?
[20:40] <pragmaticenigma> thebishop: I'm not sure what might be the busy body with NetworkManager :-/
[20:41] <thebishop> pragmaticenigma, i'm still seeing the issue after service stop network-manager
[20:41] <jujubee> pragmaticenigma not sure, someone else installed htpdate I handled other items, but he is not longer working on this with me.
[20:43] <pragmaticenigma> jujubee: If you could, I'd suggest reaching out. If not, it would appear that someone created a service for it. I don't have a machine to test with at the moment. Once concern I have is if that is a custom service script, restart might actually create a new process each time
[20:43] <jujubee> pragmaticenigma I don't see antyhing in /etc/system.d/system/ for htpdate
[20:43] <pragmaticenigma> jujubee: leaving the old one in running
[20:44] <pragmaticenigma> jujubee: path should be /etc/systemd/system
[20:45] <jujubee> pragmaticenigma nah, only one process running, I checked
[20:45] <jujubee> fat fingerd the .  sorry
[20:45] <pragmaticenigma> jujubee: I guess if it were me, I'd disable the service, and just have cron call up htpdate directly
[20:46] <pragmaticenigma> jujubee: then it can run as often as you need it to, and you can more readily troubleshoot it for errors or hanging processes
[20:46] <apxseemax> pragmaticenigma: Okay I checked but its still blocked
[20:47] <apxseemax> also checked the fail2ban jail just in case it locked the user for sftp
[20:47] <apxseemax> but it didn't
[20:47] <jujubee> pragmaticenigma what do you think of the first answer to https://unix.stackexchange.com/questions/166473/debian-how-to-run-a-script-on-startup-as-soon-as-there-is-an-internet-connecti
[20:47] <jujubee> pragmaticenigma then it only runs on if.up
[20:49] <pragmaticenigma> jujubee: that probably is the most appropriate method to do so
[20:50] <pragmaticenigma> jujubee: Then you can disable that service, and not worry about the units making extra connections and stuff just to maintain the time
[20:52] <pragmaticenigma> apxseemax: I think the probablem is that FileZilla is trying to connect via SFTP and not FTP
[20:52] <apxseemax> I used the prompt like you said and defined ftp
[20:52] <apxseemax> instead of sftp
[20:52] <sarnold> ftp is terrible
[20:52] <sarnold>  only use it if you have no choice
[20:53] <jujubee> pragmaticenigma turns out there was already a script in if-up.d (what a dolt).  Problem is that it ran when eth0 was up (which is pretty much right away.  I need to wait for wlan0 so I put in an ignore for eth0
[20:54] <apxseemax> sarnold: UFW is already setup to accept ssh connections, I am kinda failing to see why it would still block the approach
[20:54] <TJ-> jujubee: it sounds like all you need for that (starting htpdate.service after network is online) is an over-ride file (/etc/systemd/systemd/htpdate.service.d/override.conf) with something like "[Unit]",  "Requires=network-online.target", "After=network-online.target"
[20:54] <sarnold> apxseemax: I missed most of the conversation but the one log line I remember was a UDP packet being blocked, not ssh..
[20:56] <jujubee> TJ- I think you are right.  Once the interface comes up, we may still have to use http-auth page but the OS will think the interface is up anyway
[20:57] <apxseemax> I am
[20:57] <apxseemax> such
[20:57] <apxseemax> an
[20:57] <apxseemax> idiot
[20:57] <apxseemax> omg
[20:57] <apxseemax> fucked up the to IP
[20:58] <EoflaOE> !language
[20:58] <apxseemax> 33 instead of 34
[20:58] <TJ-> apxseemax: your "20,21/tcp  ALLOW IN" is wrong; FTP protocol in active mode, for the server, *listens* for incoming connections on port 21 but makes *outgoing* connections *from* port 20 to the ftp client's listening port
[20:58] <apxseemax> sorry
[20:59] <EoflaOE> apxseemax: Ok. Make sure you don't swear.
[20:59] <pragmaticenigma> TJ-: Shouldn't there be a pre-made UFW rule for enabling FTP connections?
[20:59] <apxseemax> *should*
[21:00] <pragmaticenigma> Similar to how one enable SSH ?
[21:00] <TJ-> pragmaticenigma: no idea, I don't touch UFW, but IN for both is certainly not correct (though it won't hurt) but if the 20 OUT isn't also specified and the default policy on OUTPUT is DROP... it won't work
[21:01] <TJ-> apxseemax: the other option is to have the FTP client switch to PASV mode (passive) where everything is done on the client > server:21 stream
[21:01] <apxseemax> TJ-: So basically I need an outbound ok for the cmdport 20
[21:01] <pragmaticenigma> apxseemax: that is part of it, yes
[21:01] <TJ-> apxseemax: well only if all OUTPUT is default DROPed
[21:02] <TJ-> apxseemax: I think in most cases OUTPUT is ACCEPT to prevent causing all manner of hassle for local processes making outbound connections
[21:02] <apxseemax> let me check that real quick, not sure if I am dropping everything w/o allow right now, but I would say no
[21:02] <TJ-> apxseemax: try "sudo iptables -nvL OUTPUT" see what the policy for the chain is
[21:03] <TJ-> apxseemax: e.g. mine is default ACCEPT: "Chain OUTPUT (policy ACCEPT 949K packets, 80M bytes)"
[21:04] <apxseemax> Chain OUTPUT (policy ACCEPT 0 packets, 0 bytes)
[21:04] <apxseemax>  pkts bytes target     prot opt in     out     source               destination
[21:04] <apxseemax>   47M 3119M ufw-before-logging-output  all  --  *      *       0.0.0.0/0            0.0.0.0/0
[21:04] <apxseemax>   47M 3119M ufw-before-output  all  --  *      *       0.0.0.0/0            0.0.0.0/0
[21:04] <apxseemax> 19682 1271K ufw-after-output  all  --  *      *       0.0.0.0/0            0.0.0.0/0
[21:04] <apxseemax> 19682 1271K ufw-after-logging-output  all  --  *      *       0.0.0.0/0            0.0.0.0/0
[21:04] <apxseemax> 19682 1271K ufw-reject-output  all  --  *      *       0.0.0.0/0            0.0.0.0/0
[21:04] <sarnold> wow, drone's nice and quick :)
[21:04] <sarnold> apxseemax: a pastebin site please, the pastebinit tool can make that really easy :D
[21:04] <OerHeks> !paste
[21:05] <dax> Drone also tells them about pastebins in PM iirc, which is handy
[21:05] <dax> (and I think Sigyn yelled at them too, lol)
[21:05] <sarnold> dax: oh sweet :)
[21:05] <TJ-> dax: it's teaching apxseemax about firewall rules and blocking :D
[21:07] <TJ-> apxseemax: it's possible that one of those supplemental ufw chains has a DROP at the end. you'd need to check if UFW has the concept of applying a default policy though
[21:07] <apxseemax> https://codeshare.io/GAzqyv
[21:10] <apxseemax> TJ-: posted both outputs into the bin https://codeshare.io/GAzqyv
[21:12] <TJ-> apxseemax: didn't you say you've another system doing firewalling upstream of this system? Is that allowing incoming 21 TCP and outgoing 20 TCP ?
[21:13] <apxseemax> TJ-: it does as long as the request come from inside the /24 network the VM is in. Both machines are in the same net.
[21:14] <apxseemax> thats why I whitelisted the /24 of the server
[21:15] <apxseemax> so basicaslly anything 'inside' is cool, but internet is nono
[21:15] <TJ-> apxseemax: from the remote FTP client, if you switch it into PASV mode, does it connect and work correctly?
[21:15] <apxseemax> standyby
[21:19] <apxseemax> TJ-: oh wow... I tried to connect to the wrong IP, interesstingly enough my server reacted to the wrong IP as well, but rejected it, which is now confusing me even more
[21:19] <apxseemax> If I connect it says no encrypted traffic is supported
[21:20] <apxseemax> which is weird
[21:20] <apxseemax> this is with connection type default tho
[21:21] <apxseemax> passive is giving me the same message
[21:21] <apxseemax> all on pure ftp that is
[21:21] <apxseemax> havent tried sftp, but it should work as well due to the openssh rule
[22:05] <jrg> So i did a fresh install of ubuntu. And when i did it i had to disable modereset because it’s an A10
[22:05] <jrg> And i guess the install reverted to default grub options for boot. And i can’t seem to get a menu. Am i missing something here? Is there a key you press during boot to bring up a way to edit on the fly?
[22:09] <jeremy31> jrg, try the left Shift key or ESC early in boot
[22:10] <jrg> Ok. Actually that was my fault. I’ll just pretend i didn’t do what i just did.
[22:10] <jrg> I pulled out the keyboard usb IR instead of the usb drive that i used to install... oops.
[22:11] <jrg> Although. Now it booted.. and all i see is started gnome display manager and ... no gnome heh. Maybe it’s just taking a minute since i have a platter connected and not an ssd
[22:14] <jrg> Weird. Just sitting here. Maybe something to do with it being connected to the hdmi port?
[22:24] <mithrison> hi
[22:24] <mithrison> where can I download Ubuntu Server Aeone from?
[22:25] <tomreyn> mithrison: "Aeone"? do you mean !eoan
[22:25] <mithrison> ah yes!
[22:26] <tomreyn> !eoan
[22:26] <mithrison> tomreyn can I upgrade a bionic to eoan?
[22:26] <sarnold> server eoan install images http://cdimage.ubuntu.com/ubuntu-server/daily-live/current/
[22:27] <sarnold> you can upgrade bionic through cosmic and disco to eoan, yes
[22:27] <OerHeks> that is live, this is the regular http://cdimage.ubuntu.com/ubuntu-server/daily/current/
[22:27] <tomreyn> "Join #ubuntu+1 for support and questions."
[22:27] <sarnold> OerHeks: yikes, thanks
[22:27] <OerHeks> you must have a reason to run non LTS
[22:28] <mithrison> I'm on Raspberry pi. any notes on that? sarnold OerHeks
[22:28] <OerHeks> there is a armv8 iso
[22:28] <mithrison> OerHeks ya, bionic has problems with Raspberry pi. the new eoan has full support
[22:28]  * dax throws a U at ubottu
[22:28] <sarnold> mithrison: I used the server image for rpi3 no trouble on my rpi3: https://ubuntu.com/download/iot/raspberry-pi-2-3
[22:28] <mithrison> OerHeks is that the one with rpi3 support?
[22:29] <mithrison> hmm interesting
[22:29] <tomreyn> thanks dax
[22:30] <mithrison> sarnold which version of ubuntu is that?
[22:30] <sarnold> mithrison: that one is bionic
[22:30] <mithrison> sarnold hah so is that a custom kernel and modules for rpi?
[22:31] <OerHeks> mithrison, read the whole wiki, you end up with https://wiki.ubuntu.com/ARM/RaspberryPi#Packages with a lot of tips
[22:33] <sarnold> mithrison: probably; there's at least a linux-raspi2 source package http://ports.ubuntu.com/pool/main/l/linux-raspi2/ -- though the 'standard' linux source package also builds arm64 packages http://ports.ubuntu.com/pool/main/l/linux/
[22:33] <mithrison> sarnold what does that mean?
[22:34] <sarnold> mithrison: it means I don't know if the linux-raspi2 kernel packages are intended only for the rpi2 exactly or if they can also do newer rpis; and I don't know if the binary packages built from the linux package are suitable for rpi or only big arm64 hardware
[22:46] <padarc> hm ... i have a weird issue. When there is much going on (alot of hdd writing/reading) and for example a video or music is playing in the browser, sometimes it stutters or stops for ~1sec. Is there an easy way to fix that? It does not happen on windows (so its not an hdd error)
[22:47] <CodeMouse92> padarc: (1) how much RAM does your device have, and (2) are you using swap? (Also, what version are you running)
[22:47] <CodeMouse92> *version of Ubuntu
[22:48] <CodeMouse92> (3) What is your CPU (speed, cores)?
[22:50] <Sven_vB> seems like my power management settings are now fixed. however, each time I reboot, the clock is about 10 minutes late, until I sync it again via ntp. what kind of effect could that be? an empty mainboard battery should set the clock to a fixed time, shouldn't it? and the timezone offsets for my region (CEST) are multiples of one hour, so it shouldn't be that either.
[22:51] <padarc> CodeMouse92: i have 8gb ram (most of the time barely 4gb are used), swap is 1gb (didnt change anything, it was created while installing that way). CPU should be more than enough i guess (ryzen 2600)... running on 19.04
[22:51] <padarc> CodeMouse92: oh, drive is encrypted (checked it while installing)
[22:53]  * Sven_vB tries whether syncing hwclock might help
[22:55] <jrg> Yeah i don’t get this. It just stays stuck at starting gnome display manager
[22:55] <jrg> And it never starts
[22:55] <jrg> But you can tell it’s running because it’s not a normal terminal
[22:58] <Sven_vB> jrg, did you verify the boot command line has the options you suspected might have been omitted? also I'd remove quiet and splash.
[22:58] <sarnold> padarc: when the video or audio skips happen, does anything stand out in top or vmstat 1 ?
[22:59] <Sven_vB> yay hwclock fixed it. I should probably add that to my ntp sync cronjob.
[23:03] <tomreyn> jrg: i didn't understand this part:  "i had to disable modereset because it’s an A10". are you certain this is needed?
[23:04] <tomreyn> jrg: did you try amdgpu.dc=0 instead?
[23:07] <padarc> sarnold: ok i googled a little. When i set the priority higher on ksysguard (im on kde atm, but it also happened on the standard desktop) and run "stress-d 1" on the terminal, it never skips except 1time in the beginning while playing videos. Without changing priority audio/video stops every 5sec for 0.5-1sec
[23:08] <sarnold> padarc: what's stress-d? I don't see it via apt-file search
[23:09] <padarc> sarnold: Its "stress" .. the -d option is for "stressing" the hdd
[23:09] <sarnold> ah!
[23:09] <padarc> sarnold: while running it, it takes up to 25% cpu .. i guess it goes so high because my drive is encrypted
[23:09] <sarnold> padarc: yes, that's quite likely :) heh
[23:17] <jrg> Hm. Let me take a look at that. Would that be a kernel option around splash?
[23:19] <jrg> I’m still getting hung up at just a blinking cursor at the top left and no gnome. I managed to get it to start once and thought i had changed what i needed but that doesn’t seem to work anymore.
[23:21] <tomreyn> !kernelparm | jrg
[23:21] <jrg> tomreyn: yeah that’s what i’m doing now
[23:22] <jrg> I’m editing grub on boot
[23:22] <jrg> To find out where the problem is.
[23:22] <tomreyn> okay, just wanted to make sure you know how
[23:22] <tomreyn> amdgpu.dc=0 is a kernel parameter, yes
[23:23] <tomreyn> well one for the module really but you pass it on the "linux" line
[23:24] <jrg> I’m trying just ro ACPI=off noapic nomodereset and seeing what happens.
[23:25] <jrg> Nope. Still just a cursor on top lol
[23:32] <tomreyn> jrg: when you just boot to multi user mode, can you login and get / post a log?
[23:32] <jrg> Seems so. I haven’t tried single user yet.
[23:33] <jrg> I managed to get it to run once. Changed the default grub parameters to match what worked the first time. Now I’m stuck at it not working at all heh
[23:34] <tomreyn> maybe doa full power off, power on cycle
[23:38] <jrg> Oh my. Finally got it to get back after disabling almost everything lol
[23:40] <tomreyn> i don't think "nomodereset" is a valid kernel parameter. there is "nomodeset", disabling kernel mode setting.
[23:41] <tomreyn> but i don't think you need this on any amd GPUs
[23:43] <tomreyn> the systemd.unit=multi-user.target kernel parameter should boot to non graphical login (multi-user mode)
[23:44] <jrg> Yeah. Sorry. I meant nomodeset
[23:44] <jrg> Hm. If this boots now i think i figured it out.
[23:45] <jrg> Ok. Guess not. It’s so random.
[23:46] <jrg> Disabling the load_video option in grub... then waiting for the blinking cursor.... then alt+F2 to swap to a console.. then i get a login.
[23:46] <jrg> I’m going to try that with no kernel options.
[23:48] <jrg> Yup. That did it.
[23:49] <jrg> But i don’t see where you get rid of that in /default/grub
[23:50] <tomreyn> get rid of what exactly?
[23:50] <tomreyn> it's /etc/default/grub
[23:53] <jrg> Forgot the /etc
[23:54] <jrg> But yeah.. seems like it’s the load_video portion that throws it off.
[23:54] <jrg> And setting grub to console is what fixed it.
[23:54] <jrg> So now i get an ugly large boot... but at least the desktop starts without an issue
[23:55] <jrg> Let me go ahead and update it now that i figured it the booting problem. Thanks for the help.
[23:56] <jrgilman> wow that was weird seeing jrg
[23:56] <jrgilman> uh anyways
[23:57] <jrg> Haha. Well... small world i suppose.
[23:57] <jrgilman> my ubuntu desktop shat the bed yesterday, I have the hard drive and I'm on mac os x at the moment, how can I go about mounting a cryptfs drive so i can read the contents in mac os x
[23:57] <jmg> hi all, suddenly my Xorg isnt starting and dmesg isnt showing anything with edid
[23:57] <jmg> it does scroll text before trying to start slim though, so its definately plugged in
[23:59] <sarnold> jrgilman: install ubuntu in a VM, and pass the hard drive through to the VM instance