systemdlete | Looks like I may have cornered the problem after all: Vbox reports that some recent linux kernels cause problems for at least some hardware. So it seems to be a host-side vbox issue. The problem starts somewhere after about 5.10.0-13; I am running some tests now to figure out exactly which kernel triggers it. | 02:13 |
---|---|---|
systemdlete | I replaced my memory, the board, swapped out the video card... tried multiple kernels in the VMs, etc. Finally, I tested the 5.10.0-13 kernel and that seems to be the fix. (Maybe -14, idk yet) | 02:14 |
systemdlete | I was running at least -14 or lower a few days back. I had left my PC running for about a month, so I missed booting the newer -15 and -16 kernels until now. | 02:15 |
systemdlete | Each year, I lose more and more hair | 02:16 |
DRWhite | Hi folks, tryign to upgrade to Chimaera. Easiest way to do that from Debian would be? | 04:10 |
gnarface | the same way you'd upgrade to any debian release | 04:12 |
DRWhite | So, same way as on the Devuan site? (From bullseye to Chimaera ? | 04:13 |
DRWhite | It's from 9 to Chimaera. It should jsut also work the same? | 04:14 |
DRWhite | I know I have to completely remove SystemD as well though | 04:14 |
gnarface | it should remove systemd for you | 04:17 |
gnarface | there was a problem at one release where it would also remove the kernel, but if you just remember to re-install the kernel package before rebooting it was fine | 04:17 |
gnarface | that problem may have been fixed by now though | 04:17 |
DRWhite | I followed the https://www.devuan.org/os/documentation/install-guides/chimaera/bullseye-to-chimaera listing | 04:18 |
DRWhite | But that will definitely remove all the SystemD stuff that Microsoft put in? | 04:19 |
DRWhite | Is there any light weight GUI? | 04:19 |
DRWhite | XFCE is 2 GB installed | 04:20 |
Unit193 | openbox+tint2? | 04:20 |
DRWhite | It's being put on a server | 04:21 |
DRWhite | But I need to have my remote software on it. | 04:21 |
DRWhite | So I want very very very lightweight | 04:21 |
gnarface | if you want that lightweight you should just not use a GUI at all | 04:22 |
gnarface | but the lightest two are probably blackbox and twm | 04:22 |
debdog | you want a remote desktop? if not ssh + xforwarding is enough, no GUI required at all | 04:22 |
gnarface | twm is next to useless, blackbox at least positions windows automatically | 04:22 |
DRWhite | All I want to do is access a terminal on the GUI. | 04:23 |
DRWhite | I do SSH and all first. But when the SSH has issues when an update changes the configuration... I need a way to get in | 04:23 |
gnarface | the trick is don't install any of them during the upgrade, and just install one afterwards without the graphical login manager and permissions backend/session handling stuff | 04:23 |
debdog | screen or tmux on the server | 04:23 |
DRWhite | Screen will have the GUI how? | 04:24 |
debdog | you said console | 04:24 |
debdog | erm, terminal | 04:24 |
DRWhite | Btu the remote aess uses the GUI | 04:24 |
DRWhite | But the remote access software uses teh GUI | 04:25 |
gnarface | which remote access software? | 04:25 |
DRWhite | You may think of something like Team Viewer type of thing, but something that is actually secure unlike TV | 04:26 |
gnarface | i think you don't need a GUI as badly as you think you do, but whatever | 04:26 |
gnarface | those instructions look legit | 04:26 |
DRWhite | The isntructions have seemed to work fine | 04:26 |
DRWhite | The Remote software talks directly to my other server. | 04:27 |
DRWhite | Meaning I can log in from anywhere. | 04:27 |
debdog | sorry, I don't understand either | 04:27 |
DRWhite | but it needs a GUI to be able to have something to show... | 04:27 |
DRWhite | It's the down side of it all | 04:27 |
DRWhite | GUI GUI GUI, that's all they think about.. | 04:27 |
DRWhite | Trying to get them to build in a CLI version | 04:28 |
gnarface | the point debdog is trying to get at is that SSH alone can forward individual graphical windows even if you don't have the GUI installed on the remote end, this is surprisingly rare knowledge despite being a baseline feature for so many years | 04:28 |
debdog | correct! | 04:29 |
gnarface | you just need some X libs to do it, not the whole Xorg install, or even any desktop window manager whatsoever | 04:29 |
DRWhite | Okay, so how will that interact with the remote software? | 04:29 |
DRWhite | Would SimpleHelp work that way by chance? | 04:30 |
gnarface | well the single graphical window you open can still be the terminal emulator | 04:30 |
gnarface | so it would interact with the remote software in the exact same way, only slightly faster because you don't have the added bandwidth load of forwarding the unused empty desktop space | 04:30 |
gnarface | never heard of SimpleHelp | 04:30 |
DRWhite | That's why I awas referring it to something like TeamViewer | 04:31 |
DRWhite | Becausze most people know what that is | 04:31 |
gnarface | well the scenario you describe is vague but also alien enough that it just sorta automatically flags itself as something unnecessarily complicated to us, i can't be 100% sure you don't really need it, but in most cases like this the problem is merely being framed wrong | 04:32 |
debdog | teamviewer is a remote desktop, right? | 04:32 |
gnarface | iirc it's basically just a web-based vnc middleman that requires flash | 04:32 |
DRWhite | Probably Gnar | 04:32 |
gnarface | not actually using vnc protocols but the same idea | 04:33 |
gnarface | TBH even actually using VNC would be a better choice | 04:33 |
gnarface | but ssh forwarding individual windows is just faster | 04:34 |
gnarface | and saves you a lot of disk space | 04:34 |
gnarface | because like you said... these GUI things take up a lot of it | 04:34 |
DRWhite | SimpleHelp is a remote desktop client/server. The client is installed and it says to the SERVER that it is alive and active. So the connection is from client to server, unlike VNC, TeamViewer and all that the client talks to the server | 04:34 |
DRWhite | TVĀ is the client PC that is talkign to where they want to connect to | 04:35 |
DRWhite | SH is the location I want to connect to, talkign to my server, and me connecting to my server to access the connection then connecting directly to the destination without having to create a new connection and such | 04:35 |
DRWhite | Make any sense? | 04:36 |
gnarface | almost, but if i'm understanding it right, it doesn't sound trustworthy | 04:38 |
gnarface | anyway, not my business | 04:38 |
DRWhite | Why is it not trustworthy? | 04:38 |
DRWhite | TeamViewer is not trustworthy... | 04:38 |
clemz | is mysql compatible with devuan? | 04:38 |
DRWhite | yes | 04:38 |
gnarface | extra middlemen for no apparent reason; a system designed to allow unauditable 3rd party monitoring maybe? | 04:39 |
DRWhite | No Gnar, all my servers. | 04:39 |
DRWhite | Just me on my servers. | 04:39 |
DRWhite | As I said, if I can't get SSH connection, then I have to have access to my own server and data remotely so I can fix it. | 04:40 |
clemz | what is the package name for mysql? | 04:40 |
DRWhite | mariadb | 04:40 |
DRWhite | mariadb-server | 04:40 |
DRWhite | mariadb-common | 04:40 |
DRWhite | mariadb-client | 04:40 |
gnarface | no that's the other one, if you want actual mysql it's still in there as mysql-server, etc | 04:40 |
clemz | the mariadb documentation is telling me to run a msql command to create a database though. | 04:40 |
DRWhite | MariaDB is the server | 04:40 |
gnarface | mariadb is the fork of mysql, though it's widely respected | 04:41 |
gnarface | it should be a drop-in replacement more or less | 04:41 |
DRWhite | it acts directly as mysql. MySQL was sold to Oracle. MariaDB is by the original creators of MySQL. | 04:41 |
gnarface | i believe actual mysql is still present though | 04:41 |
DRWhite | I've been using MariaDB since the sale, and it is a direct drop in replacement | 04:42 |
clemz | Why am i getting this? "E: Package 'mysql-server' has no installation candidate | 04:42 |
clemz | " | 04:42 |
DRWhite | It's also faster | 04:42 |
gnarface | clemz: just use mariadb instead if that's the one that installs easier... | 04:42 |
DRWhite | There is no mysql-server package in Devuan | 04:43 |
gnarface | clemz: oh, sorry, yea it's not in testing or stable right now, i see it in unstable though. just use mariadb i guess | 04:43 |
gnarface | DRWhite: it's there, it must have just been pulled temporarily before the stable release | 04:43 |
DRWhite | I'm in Chimaera repos. | 04:43 |
gnarface | https://pkginfo.devuan.org/cgi-bin/policy-query.html?c=package&q=%5Emysql-server%24&x=submit | 04:44 |
gnarface | present in jessie, ascii, and ceres | 04:44 |
DRWhite | And yeah, only got the mains for standard, updates and security | 04:44 |
DRWhite | But it looks like the update from Debian 9 to Devuan Chimaera worksed perfectly. | 04:45 |
clemz | What other clients can i use with mariadb? | 04:51 |
gnarface | i wouldn't assume the list would be different from mysql, they should all be compatible | 04:54 |
FreelancePolicem | Hello. When using OpenRC, "command_user" in init scripts seems to be ignored or not working properly. No matter what I enter, the daemon remains owned by root. Moreover, the value of command_user is then appended to the command run as an argument, preceded by a colon. | 06:01 |
FreelancePolicem | Is this a known issue? | 06:02 |
gnarface | not a ton of people using openrc, stick around and be patient, someone should show up who does know eventually | 06:06 |
gnarface | maybe check the forum too | 06:08 |
gnarface | i could only guess you might have to change something else to get the daemon to start as a different user? | 06:08 |
FreelancePolicem | Huh, OK. I thought it would have been a more popular system, so I figured plenty of people in chat would have known something about it. I did some googling and couldn't seem to find anything. I was thinking this is a Devuan issue rather than an OpenRC issue, since command_user works as it ought on my Artix box. | 06:10 |
FreelancePolicem | If it's not so popular, that may have explained why I couldn't find anything, though! I'll do some digging on my own and see if I can figure out the issue. | 06:10 |
gnarface | well | 06:12 |
gnarface | one thing that might help to know is that it's set up like Debian's openrc install | 06:12 |
gnarface | which means it relies on sysvinit stuff, so there's some extra caveats that are different from the way other distros use it | 06:12 |
gnarface | i can recall one unrelated caveat about "rc-status" (i think it was called) not functioning as expected because most stuff needs root permission to actually report status so if you call it as a regular user it just gives you false feedback of some sort | 06:13 |
gnarface | and this lead to some other issue with ... postgres i think? | 06:14 |
gnarface | anyway, it was simple to fix though, is the thing | 06:14 |
gnarface | openrc is probably the most popular of the alternate init systems but most of us are still using sysvinit | 06:15 |
FreelancePolicem | Thanks for the advice. I admit I don't know too much about the internals of OpenRC, and where it ends vs. where sysvinit begins. From my attempts at debugging, I believe my problem is with the "start-stop-daemon" program. | 06:15 |
FreelancePolicem | Well, now that I say that, I'm not so sure of that -- at any rate, there's an issue between openrc-run and start-stop-daemon's communication | 06:16 |
gnarface | stick around, someone probably knows but they're probably just asleep right now | 06:16 |
FreelancePolicem | Sure, thanks. | 06:16 |
rrq | FreelancePolicem: fwiw, apparently "command_user" is a variable for opnrc's "start-stop-daemon.sh" and not the "start-stop-daemon" command | 06:17 |
gnarface | also, someone has also made a replacement openrc package that works just like the upstream gentoo one but it's not officially included, but someone around here should have a download for it, i'm not sure if it's in proposed updates or not | 06:18 |
FreelancePolicem | rrq: Where exactly is the .sh file located? | 07:55 |
rrq | try: dpkg -S start-stop-daemon.sh ... useful feature :) | 07:58 |
rrq | looks like the package installs it at /lib/rc/sh/start-stop-daemon.sh | 07:58 |
FreelancePolicem | My god | 08:07 |
FreelancePolicem | I think I fixed it | 08:07 |
FreelancePolicem | Thanks, rrq | 08:07 |
FreelancePolicem | It looks like there's a typo or something in the sh script | 08:08 |
FreelancePolicem | "${command_user+--user} $command_user" should be "${command_user+--chuid} $command_user" | 08:08 |
FreelancePolicem | It looks like it's in the source for OpenRC. I don't suppose there's any use in submitting a bug report to Devuan? | 08:10 |
FreelancePolicem | But if it's in OpenRC, I am really curious why I don't have this issue on Artix. It looks like the Artix package doesn't patch the sh script | 08:10 |
rrq | it'd be an upstream bugreport I suppose; debian pacakges it and devuan includes that package | 08:14 |
rrq | btw it might need --user as well (with some mangling) | 08:21 |
FreelancePolicem | Do you have any reason you'd think that? It looks to me like --user is for something unrelated to what the command_user flag is supposed to be | 08:25 |
rrq | I looked at the first man page example | 08:26 |
rrq | --user applies to recognizing an aleardy running service while --chuid sets up for a new | 08:27 |
rrq | doesn't seem like the strongest argument for having two such flags and it certainly doesn't support the idea of using different formats | 08:32 |
FreelancePolicem | It's a bit late for me and I'm a bit drunk, so I apologize if I'm being slow. How are you thinking the command_user flag in openrc-run would have to do with asking the daemon to do something to running processes? | 08:32 |
rrq | I guess that start-stop-daemon should avoid starting a service if that service is running, and --user is part of recognizing whether that is the case | 08:34 |
FreelancePolicem | Oh. I see where you're coming from. No, that seems to be handled elsewhere | 08:34 |
FreelancePolicem | I'm not sure where elsewhere | 08:35 |
FreelancePolicem | Lol | 08:35 |
FreelancePolicem | But from testing having made this change, openrc can still detect if a service is running | 08:35 |
Guest5976 | Why is there no persistence item in my GRUB menu? link to file grub.cfg https://pastebin.com/kSVH6N1D | 11:20 |
furrymcgee | there should be some build instructions for the iso images | 11:49 |
Guest5976 | furrymcgee, Where? | 11:50 |
furrymcgee | sorry i am not the mainainer | 11:51 |
furrymcgee | i would expect something like this https://github.com/cirosantilli/x86-bare-metal-examples/tree/d217b180be4220a0b4a453f31275d38e697a99e0/multiboot/hello-world | 11:52 |
furrymcgee | look into https://git.devuan.org/explore/repos | 12:00 |
furrymcgee | maybe this https://git.devuan.org/bgstack15/installer-iso/src/branch/master/boot-efi-grub.mk | 12:15 |
rrq | Guest5976: maybe it doesn;t like "linux16" (in the prior entry) | 12:19 |
Guest5976 | I still do not understand what is the connection with the error in the menu? | 12:28 |
furrymcgee | make sure there is no syntax error and it actually uses this file | 12:34 |
Guest5976 | furrymcgee, How can I find syntax errors? With what to compare? | 12:40 |
furrymcgee | compare with a working file, it wont work with syntax error | 12:44 |
Guest5976 | furrymcgee, With what working file to compare, if only one item is not displayed in the menu? (I formatted the text of the question incorrectly at the very beginning.) | 12:52 |
furrymcgee | we may figure it out if you link the image you are using | 12:55 |
Guest5976 | furrymcgee, devuan_chimaera_4.0.0_amd64_desktop-live.iso | 12:58 |
furrymcgee | you can start with link please | 12:58 |
Guest5976 | furrymcgee, Links to what object? | 12:59 |
furrymcgee | devuan_chimaera_4.0.0_amd64_desktop-live.iso | 13:00 |
Guest5976 | furrymcgee, This iso was written to a flash drive using rufus. Now on this flash drive the files that were inside iso. | 13:02 |
Guest5976 | furrymcgee, Among these files \boot\grub\grub.cfg | 13:03 |
fsmithred | Guest5976, did rufus make a partition for persistence and label it as such? | 13:07 |
fsmithred | and did it create persistence.conf in the root of that partition? | 13:07 |
Guest5976 | fs,mithred, yes, i chose this option in rufus | 13:09 |
Guest5976 | Then I decided to add a grub menu item to make the save work. | 13:09 |
fsmithred | oh, "components" in the boot command might cause trouble, but the entry should still show up in the menu. | 13:12 |
fsmithred | Guest5976, do any of those menuentries work? | 13:15 |
Guest5976 | yes | 13:19 |
fsmithred | ok. I don't know what rufus does, so it's hard to know what's wrong. I'm not seeing a syntax error in that entry. | 13:20 |
fsmithred | fwiw, I use this to make live-usb: https://sourceforge.net/projects/refracta/files/tools/refracta2usb-2.4.3.deb/download | 13:20 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!