unix

[formal schooner]

gnu make is quite an impressive program, the only bad thing about it really is the makefile syntax

can't vouch for bsd make, never used it

[stark sinew]

finally, it works, thanks for your help @formal schooner

[crude seal]

Hello

My usb device not showing up in my linux computer

[gritty jolt]

Wow vim?

[stark sinew]

ye

[gritty jolt]

Can u give me ur plugins?

[stark sinew]

yeah, 1 sec

so, if you load this .vimrc, it'll auto install

https://github.com/elsoleiro/dotfiles/blob/master/.vimrc

[gritty jolt]

Thank u

[stark sinew]

once pasted into .vimrc, execute :source % then execute :PlugInstall and voila.

your vim is the same as mine.

[gritty jolt]

Yeah i use vim as main Too

😄

[stark sinew]

nice lets see

[gritty jolt]

I would give u tomorrow i am on phone

[stark sinew]

have u mounted the drive?

ok cool

[crude seal]

How?

[stark sinew]

it'll be a few command line arguments

so uhm, what unix are u running?

remove your usb, show me the output of lsblk

then plug in and show the output of lsblk

[crude seal]

i use linux not unix

[stark sinew]

😒

what distro are u using

look if you don't want help then don't ask, simples.

or sort it yourself idk

[crude seal]

i use bodhi distro

it's distro of ubuntu

[crude seal]

this is without usb ```
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 55.5M 1 loop /snap/core18/2074
loop1 7:1 0 77.2M 1 loop /snap/discord/128
loop2 7:2 0 162.9M 1 loop /snap/gnome-3-28-1804/145
loop3 7:3 0 65.1M 1 loop /snap/gtk-common-themes/1515
loop4 7:4 0 32.3M 1 loop /snap/snapd/12398
sda 8:0 0 76.3G 0 disk
├─sda1 8:1 0 100M 0 part
└─sda2 8:2 0 76.2G 0 part
sdb 8:16 0 465.8G 0 disk
├─sdb1 8:17 0 512M 0 part /boot/efi
├─sdb2 8:18 0 1K 0 part
└─sdb5 8:21 0 465.3G 0 part /

[stark sinew]

k and with usb

[crude seal]

ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$ lsblk
NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
loop0    7:0    0  55.5M  1 loop /snap/core18/2074
loop1    7:1    0  77.2M  1 loop /snap/discord/128
loop2    7:2    0 162.9M  1 loop /snap/gnome-3-28-1804/145
loop3    7:3    0  65.1M  1 loop /snap/gtk-common-themes/1515
loop4    7:4    0  32.3M  1 loop /snap/snapd/12398
sda      8:0    0  76.3G  0 disk
├─sda1   8:1    0   100M  0 part
└─sda2   8:2    0  76.2G  0 part
sdb      8:16   0 465.8G  0 disk
├─sdb1   8:17   0   512M  0 part /boot/efi
├─sdb2   8:18   0     1K  0 part
└─sdb5   8:21   0 465.3G  0 part /
sdc      8:32   1   7.2G  0 disk

[stark sinew]

cool, so our device name is sdc ye?

now mkdir /mnt/usbstick

or sudo mkdir /mnt/usbstick/

then, mount sdc /mnt/usbstick/ or sudo mount sdc /mnt/usbstick/

[crude seal]

sudo mount sdc /mnt/usbstick
mount: /mnt/usbstick: special device sdc does not exist.

[stark sinew]

my bad

sudo mount /dev/sdc /mnt/usbstick/

[crude seal]

ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$ sudo mount /dev/sdc /mnt/usbstick
mount: /mnt/usbstick: wrong fs type, bad option, bad superblock on /dev/sdc, missing codepage or helper program, or other error.

[stark sinew]

ok, we need to specify the filesystem type with the -t argument

do u know what filesystem type that usb is

i dunno how we'd check

[crude seal]

i don't know

[stark sinew]

1 mo

what's the output of lsblk --fs

or blkid -p /dev/sdc

is more concise.

[crude seal]

ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$ lsblk --fs
NAME FSTYPE LABEL UUID                                 FSAVAIL FSUSE% MOUNTPOINT
loop0
     squash                                                  0   100% /snap/core
loop1
     squash                                                  0   100% /snap/disc
loop2
     squash                                                  0   100% /snap/gnom
loop3
     squash                                                  0   100% /snap/gtk-
loop4
     squash                                                  0   100% /snap/snap
sda
├─sda1
│    ntfs         BE2CC4812CC435E5
└─sda2
     ntfs         54D8C691D8C670B2
sdb
├─sdb1
│    vfat         3C09-F5DE                               511M     0% /boot/efi
├─sdb2
│
└─sdb5
     ext4         9506d0cf-5cc1-4466-9bcb-21185660512d  420.5G     3% /

[stark sinew]

second one instead pls

sudo blkid -p /dev/sdc

[crude seal]

nothing happens

[stark sinew]

hmmm

[stark sinew]

let's just assume it's fat32

sudo mount -t vfat /dev/sdc /mnt/usbstick/

if not let's try ext4

and then i'm rlly out of ideas

[crude seal]

ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$ sudo mount -t vfat /dev/sdc /mnt/usbstick/
mount: /mnt/usbstick: wrong fs type, bad option, bad superblock on /dev/sdc, missing codepage or helper program, or other error.
ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$

[stark sinew]

and ext4?

[crude seal]

i replace vfat with ext4?

[stark sinew]

wait before that

hmmmm

actually yeah tfry that

i was thinking maybe sdc was partitioned

or something

then try mount -t msdos /dev/sdc /mnt/usbstick

hold on bro

is this usb stick empty?

[crude seal]

?

[stark sinew]

like is it a new usb stick or something?

[crude seal]

it's not empty

[stark sinew]

ok it's got things in it

[crude seal]

last time i used the usb i burned iso file in it

[stark sinew]

it's weird because theres no partition at all

normally you'd have like /dev/sdc has partition sdc1 or whatever

have u tried both -t vfat and -t msdos yet?

[crude seal]

ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$ mount -t vfat /dev/sdc /mnt/usbstick
mount: only root can use "--types" option

[crude seal]

yes

[stark sinew]

so msdos doesn't work?

sudo mount -t msdos /dev/sdc /mnt/usbstick/

run that pls

u need to remember sudo

[crude seal]

ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$ sudo mount -t msdos /dev/sdc /mnt/usbstick/
mount: /mnt/usbstick: wrong fs type, bad option, bad superblock on /dev/sdc, missing codepage or helper program, or other error.

[stark sinew]

ok, pls can u send a screenshot of the output of lsblk -l

[stark sinew]

dmesg | tail -20 and screenshot pls

sudo dmesg | tail -20

oh yeah and finally, this definitely isn't gonna work but can u at least try in sequential order, see if one works.

sudo mount /dev/sdc1 /mnt/usbstick/

sudo mount -t msdos /dev/sdc1 /mnt/usbstick/

sudo mount -t vfat /dev/sdc1 /mnt/usbstick/

let's just guess it does have a partition even though it's not showing on lsblk for some reason.

[crude seal]

ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$ sudo mount /dev/sdc1 /mnt/usbstick/
mount: /mnt/usbstick: special device /dev/sdc1 does not exist.
ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$ sudo mount -t msdos /dev/sdc1 /mnt/usbstick/
mount: /mnt/usbstick: special device /dev/sdc1 does not exist.
ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$
ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$ sudo mount -t vfat /dev/sdc1 /mnt/usbstick/
mount: /mnt/usbstick: special device /dev/sdc1 does not exist.
ahmed@ahmed-HP-Compaq-dx2420-Microtower:~$

[stark sinew]

ok, damn.

i have no idea what else to do now apart from reformat that usb stick.

we'd reformat it, make a partition and then mount it

[crude seal]

How to do that

[gritty jolt]

Distro name?
Theme?

[stoic yoke]

Anyone using WSL v2 Kali, so they can test a quick bug I keep getting?

If I ever double tap the physical escape key twice on my terminal windows it freezes the window temporarily.

[stoic yoke]

Ok well it seems like a Kali/Bash issue with Bash v5.1.4
Edit:
Figured out the issue, double escape in bash does tab complete and then also causes the entire bash window you cache my path. Causing a delay based on the size of my drives and etc....

[quasi warren]

When we do man ascii, we get:

So is it right to say ascii value of A is 65?

and if we wanted to store the numerical value of 65 to a variable instead of using 65 as a magic number, would the following make sense

ascii_of_A = 65

[dark rivet]

yes, but i would probably use ASCII_OF_A because it would be a constant

maybe even omit the "of" entirely

[quasi warren]

Good point!! I will make use of all uppercase! Thanks

[formal schooner]

That foo_of_bar naming style is used in the ocaml stdlib

It's kind of goofy

[high steppe]

hey there, can anybody give me a hand? I'm having trouble with aliases/function definitions and execution/fallback directories for aliases too...

ik I sound like someone who hasn't even read the manual, but I have and I don't understand much

It would be great if anyone knew a tutorial other than the official one for this topic

[heavy patio]

hey there! i hope this is the right channel to ask but how do i install python modules on arch linux? (please ping me)

[crisp furnace]

How do we convert octal o hexadecimal directly? does anyone know?

[lavish storm]

yes

[heavy patio]

yes

but how?

[lavish storm]

echo "obase=16; ibase=8; 17" | bc

example of converting base 8 number 17 to hexadecimal

[lavish storm]

check if you have python (which you should)

[heavy patio]

i do

[lavish storm]

then do python -m pip install XXX

[heavy patio]

ill try that

[captain8771@CapOS ~]$ python -m pip install discord
/usr/bin/python: No module named pip
[captain8771@CapOS ~]$

bruh moment

[lavish storm]

ah

try sudo pacman -S python-pip

[heavy patio]

k

ah thanks!

[lavish storm]

[crisp furnace]

Oh-

I meant without python

Ik this is a python server

[lavish storm]

that is not python

[crisp furnace]

oh

What is it?

[lavish storm]

shell script

Are you familiar with shell? (e.g. bash, zsh)

anyway shell has bc which is a calculator language

so you can use bc to convert any number in any base to a number in a base of your choice

if you want to learn more just type man bc in your terminal

[crisp furnace]

im not....

not unless you're referring to "windows cmd" and "windows powershell"

[lavish storm]

hmm

well

this channel isn't for windows-exclusive users

since windows isn't a UNIX-based OS

[upper prairie]

again, it's not recommended to mess with the built-in Python version :c

Typically you'd use pyenv

or install another version

I don't understand the point you're trying to make

Built-in Python doesn't come with pip

Python is used internally by certain programs

You don't want to install python-pip, and you don't want to use it as a global interpreter

That's up to you, but generally I don't think people should be using the built-in interpreter

it's only there as a dependency, and if something breaks... I don't expect it to be an easy fix

If you look hard enough, you'll see staff making the same recommendation throughout this server

I'd also like to caution against the use of pacman -S, while we're here

Always upgrade (with pacman -Syu) before installing a package.
https://wiki.archlinux.org/title/System_maintenance#Partial_upgrades_are_unsupported

um that's exactly what you're not supposed to do

that's like... the worst thing you could do in this context

That's what you imply

If you say "update the repos", that doesn't imply pacman -Syu because that's a complete distribution upgrade

This is a bit of a fault on the wiki's part, the partial upgrades stuff is not explained here

but they do however expect you to read the system maintenance page

which does explain that you should perform a distribution upgrade before installing a package

and in the case of Arch, it's good habit to always just do sudo pacman -Syu <package>

updating the repos is what sudo pacman -Sy would do

sudo pacman -Syu actually updates packages and upgrades your system

That was not an assumption that I made, I was going solely based off of what you had said

Updating the repos is pacman -Sy, if you asked someone "How do I update the pacman repositories?", they would not respond with sudo pacman -Syu

I don't even use Arch

I... really don't see a need to delete every message

because now this looks incredibly one-sided and spammy

[lavish storm]

It was a one-sided conversation 🙂 you were right all along!

I sincerely apologise, i clearly know nothing

[scarlet totem]

I only saw you use -S, unless you edited it.

[lavish storm]

Clearly my suggestion was wrong, you should know better and use what Phycereal suggested.

Never use pacman -S!

[scarlet totem]

https://wiki.archlinux.org/title/Pacman#Installing_packages -S looks fine, unless the wiki is wrong or I am misreading it. -Sy isn't.

[lavish storm]

According to Phycereal -S is not okay (#unix message). In fact Archwiki is in the wrong for not explaining it properly (#unix message)

so I'm going to listen to the expert advice and not use -S

[scarlet totem]

If the wiki is wrong why hasn't it been updated for something so important

[lavish storm]

Clearly the wiki authors do not know what they are talking about :)

[scarlet totem]

The people who maintain Arch and Pacman

[lavish storm]

ghastly, openSUSE users know better

[scarlet totem]

🙂

[lavish storm]

i guess i was just super lucky using -S for the last couple years

like one in a trillion lucky

https://tenor.com/view/clara-valac-clara-smile-cute-kawaii-gif-22179274

[scarlet totem]

I am an apt user, so not familiar with pacman, but through reading the wiki and an understanding of package managers, my interpretation is -S will install old packages and not the most up to date.

[lavish storm]

depends, if you have run -Syu sometime that day then it may as well be the newest available. -S doesn't imply "download not the latest one"

[scarlet totem]

The most recent version in the cache

[lavish storm]

exactly

i think every arch user knows to do -Syu

very safe to assume so

[scarlet totem]

Is it easier to break Arch due to being a rolling release?

[lavish storm]

i think the main issue is stability, since it heavily prides itself on bleeding edge updates

at least from personal experience

[scarlet totem]

they are normally a trade off

I go for stability

So with Arch installing pip with -Syu and then using a virtual env would be best

[lavish storm]

i use debian for my home server

[lavish storm]

python-pip is one of the dependencies for python in the official repositories

and yes, venvs for project-specific stuff i suppose

[upper prairie]

No, typically you just don't want to acknowledge the preinstalled Python at all... Which is where pyenv comes in handy, letting you use a different Python "globally" without interfering with the inbuilt one

[upper prairie]

I never said -S is not okay, I said

I'd also like to caution against the use of pacman -S, while we're here

and I find that completely fair. A fair bunch of people that I've seen who use Arch don't have a clue what the difference between -S, -Sy, and -Syu might be. I don't understand what you're arguing, when it seems like you agree with me...

ArchWiki says to perform a system upgrade before installing a package, and the easiest way to do that would be sudo pacman -Syu <package>

[lavish storm]

I totally agree with you

I'm not implying you are wrong

I don't know why you are getting that impression, I have totally submitted to your opinions

[upper prairie]

I don't understand the sarcasm

[lavish storm]

It's not sarcasm??

Can't you just accept my defeat for what it is

[upper prairie]

You keep saying things like

ghastly, openSUSE users know better
— I can't help but take offense to that

[lavish storm]

but you do know better, do you not?

That is an empirical fact

[scarlet totem]

Personally I would install pipenv globally with pip and then only use pipenv. They both ultimately get it done. Though this done on Ubuntu so results may vary.

[upper prairie]

I think you're confusing pyenv for virtual environments

or am I not understanding what you're talking about lol

you generally just want to leave the preinstalled interpreter completely alone

[scarlet totem]

nope not confused

[upper prairie]

pipenv requires messing with the base interpreter

and that would also require you to activate a venv for things like IPython

which I mean to have to go around and do

cd ~/Documents/base-venv
source venv/activate.sh
ipython

is quite annoying lol

[scarlet totem]

Stuff like Jupyter I'll run in Docker or pipenv. Dev I'll use pipenv or Docker. I am already in the dir and can run pipenv run dev. Python applications I'll normally install with apt.

pipenv run dev will normally run something like ls *.py | entr -r python main.py

[nova viper]

anybody knows how to run code in VS?

[scarlet totem]

Visual Studio, Python?

[nova viper]

yes

VS code

visual studio code

anybody?

[scarlet totem]

I do

vscode is what I use

[nova viper]

how do you run code

[scarlet totem]

There is a couple of ways, you can user the terminal or tasks

[nova viper]

ok?

because i just left py charm and downloaded this

[scarlet totem]

I use the terminal normally and don't use task.

Ctrl ~ (~ is above tab next to 1) will open a terminal

type python3 and see if it works

Terminal->New Terminal from the top menu will get you one as well

Also what os are you on?

this is probably best in #editors-ides

[nova viper]

i am on mac OS

i love mac because its beautiful

[upper prairie]

I suppose I've not used pipenv enough to understand how this works, but I'd assume it's much easier just not dealing with a virtual environment and the system interpreter to begin with

because with pyenv you can just do like pyenv global 3.9.5 and be on your way, worry free

[nova viper]

my dream was to have a linux system but i din't get one

[scarlet totem]

pipenv is a good tool. It is quite new. It's pip + virtualenv

[upper prairie]

How's that compare to poetry?

might be the wrong channel for this, if you'd prefer we move then you can ping me in a relevant channel

I figure #tools-and-devops

[scarlet totem]

Yeah

[untold abyss]

Used that for piping output of a script. anyone know what does this do exactly? >

>(tee -a stdout.log) \
    2> >(tee -a stderr.log >&2)

[formal schooner]

@untold abyss like this?

cmd \
  1> >(tee -a stdout.log) \
  2> >(tee -a stderr.log >&2)

it emits stdout to both stdout.log and stdout, and it emits stderr to both stderr.log and stderr

>() is "process substitution"

https://www.gnu.org/software/bash/manual/html_node/Process-Substitution.html

[untold abyss]

thanks a lot! 🚀

[main olive]

Im getting the ertot “chroot is running ,command “start” ignored”

[scarlet totem]

@jagged fjord This is the git server I run https://docs.gitea.io/en-us/install-with-docker/

[jagged fjord]

ok, i will check it out

[sick dirge]

any thoughts on this

print("PythonOS [0.1] is booting up...")
print("Importing modules...")
import os
print("Setting keys...")
adminpassword = os.environ['adminpassword']
print("All modules and keys are set up!")
def main():
  print(" ")
  cmd = input("A:/~ ")
  if cmd == "help":
    print("poweroff - Shutdown system")
    print("help - Help command")
    return main()
  if cmd == "poweroff":
    print("Shutting down...")
    print(" ")
    print("This was PythonOS [0.1].")
  else:
    print("Invalid command!")
    return main()
main()

its a basic "opeating system for me but I know its not" what other commands should i add im planning to add a admin "root" command which is being worked on

[dapper musk]

I dont think that it is correct chanell your """"""os"""""" does not have eaven unixish paths

[lavish storm]

lol

[amber garnet]

OS is something more than simple commands console

[fresh saddle]

It is definitely a shell though

[atomic ember]

I have a question it's a good idea to use a recursive function in that "os"?

[formal schooner]

you don't want to implement arbitrarily long loops in python using recursion

because python doesn't have any optimization for recursive functions and you eventually hit the maximum stack size

[atomic ember]

that i was thinking

[lavish storm]

async is probably better

[onyx marsh]

Should i use single quote or double quote to run a python script on linux?
Well, it's not directly on linux terminal. I'm using a run external program on torrent completion on qbittorrent but since it on linux vps, i assume it run bash.

python "/root/nekobot/up.py" -hash "e92f5b4a721c6a5b7f8f257e25e221d0f773c395"

[runic quarry]

If your current setup is just running the command you showed above, it doesn't matter.

Won't make a single difference.

[supple escarp]

Hey I'm writing a deamon in python

[supple escarp]

I've an issue where I've got a class that's in another file

for some reason the daemon starts if the class is in the same file as my driver function. But if I pull it in from another file it fails to start

[supple escarp]

Any ideas?

[lavish storm]

i feel like this is more for #async-and-concurrency

rather than here

[azure portal]

what is the source command for? and how does it work?

[lavish hill]

Source is kind of like import. Allows getting custom variables/settings/functions from different files

[runic quarry]

You're talking about the shell command?

[azure portal]

yes like source ~/.bashrc

[runic quarry]

It just runs the script inside of your current shell.

For example;

Say you have a file named script.sh with this in it:

#!/usr/bin/env bash

echo hi
cookie = cake

If you run source script.sh, it'll print hi to your terminal, and set the value of cookie to cake.

Pretty much what @lavish hill said, just provided an example.

When you run source ~/.bashrc it's just a huge collection of commands and variables entering into your current shell.

[lavish hill]

The cool thing about it is when you write functions in the script. They become callable, and won't be called by source, just to make that clear. When you source bashrc, you'll want your terminal settings stored in there

[runic quarry]

Yeah, like you could put something like this in a file:

print_words() {
  echo "hi"
  echo "hi2"
}

And if you ran source file, you would be able to run the print_words command, which would print hi and hi2.

Is there a certain way you're needing to use it or something?

[azure portal]

no, I just wanted to learn how it works. I'm used to run source ~/.bashrc but simply because it is what on most of the tutorials out there but I don't know behind that command.

@runic quarry @lavish hill ,
how it is different to bash command use to run script?

[lavish hill]

It's different in the way that you're able to use the variables/functions of the file later on.. You're loading it, not directly executing it

[runic quarry]

Great question:

Try creating a file named file.sh with this in it:

#!/usr/bin/env bash
discord=quack

1. First try running bash file.sh. If you then run echo $discord, you'll see it's not set to anything.

2. Now, run source file.sh. If you run echo $discord this time, you'll see that it's set to quack.

Difference is mostly just in that, first one runs in a subprocess, but the second one runs directly inside your shell.

Just fixed the command, still hung up on python variable assignment.

[azure portal]

@lavish hill , you mean with source you directly executing it? and with bash is you are loading it for later use? as what @runic quarry example (if I understand the example correctly)?

[lavish hill]

One actually good example of this:
Inside your bashrc, you can enter this:

PATH=$PATH:$HOME/.local/bin

If you then source ~/.bashrc it tells your terminal that you will later on look for executable files located in ~/.local/bin for your terminal commands. If you just ~/.bashrc, it would only execute, but not be stored in terminal

[runic quarry]

You aren't storing it for later, but rather just running it in a subprocess (does that make any sense?).

So you'd have your current terminal session, and when you run bash file.sh, it spawns a new process to run the command.

When you run source file.sh though, it runs in your current process.

[lavish hill]

Another important point of this is that source does not require the file to be executable!

[runic quarry]

I don't think running bash file.sh does either.

I think it mostly matters just when running something like ./file.sh.

[lavish hill]

True

[azure portal]

yes much better now

[runic quarry]

Great great

[lavish hill]

Completely off topic though, I recommend zsh 😂

[runic quarry]

I'll let you keep your opinion, definitely disagree though.

I'm too lazy to learn zsh. Already stuck learning Python 'cause Bash gets slow as crud real quick when using pipes.

Really liking the speed improvements with Python though.

[lavish hill]

How come? I feel like zsh is the more beginner friendly version of bash, or more easily configurable, but it just took a while to actually get it done

[runic quarry]

I started with Bash, and there's all these tiny differences with zsh that's making me not want to learn it.

In zsh's defense I've spent about five minutes trying to learn it.

Was mainly going to use it to make some spinners for a program I write (Bash was being poopoo at it), but Python looks like it'll work lovely.

[lavish hill]

I mean, you might have the exact functions that I'm talking about in Bash too. I'm just addicted to fzf

[runic quarry]

No clue what that is lol

Just an advanced regex thing or something?

[lavish hill]

That reminds me of why I came on this chat lol! Thanks xD
Do you happen to have a regex pattern to grab between to parentheses?

[runic quarry]

Between a pair of parenthesis?

[lavish hill]

Yea

I got this far: \((.*?)\)

[runic quarry]

Do you want to include the parenthesis, or no?

[lavish hill]

I just want to get rid of them yeah

[runic quarry]

You fine using grep?

I know how to keep the parenthesis, and then remove then with sed, not sure much on how to get rid of them with only regex though.

If you're fine keeping them, something like this would work:

\([^)]*\)

[lavish hill]

Yeah I'm using grep -Eo already

Let me try

It's getting close

[runic quarry]

Just added an escape, had some issues without it.

Should be able to just run sed or something after to get rid of the parenthesis.

[lavish hill]

Hmm yeah, probably need to do that.. Just not sure sed is available where I'm going to run this

[runic quarry]

Will grep be?

[lavish hill]

grep is available

Sed is too 😄

[runic quarry]

Do you need help with the sed statement, or is that covered?

Don't know many places with grep without sed, any recent Linux system would have then anyway.

[lavish hill]

I'm not too familiar with sed, so it would be helpful, but I can google a bit too. It needs to pipe the output though, currently just looking at the default gateway as an example

So my command atm is:

arp -a | grep gateway | grep -Eo "\((.*?)\)"

[runic quarry]

Well if you have a string like hi(123), this would work:

echo 'hi(123)' | grep -Eo '\([^)]*\)' | sed -e 's|(||g' -e 's|)||g'

Just tells sed to wipe out any parenthesis.

[lavish hill]

Kind of nasty looking, but works beautifully

Thanks 😛

[runic quarry]

Joys of regex lol

Yep, np

[lavish hill]

Neat!

#!/bin/bash

InsidePar(){
  echo $@ | grep -Eo '\([^)]*\)' | sed -e 's|(||g' -e 's|)||g'
}

GATEWAY=$(InsidePar $(arp -a | grep gateway))
echo $GATEWAY

[supple escarp]

@lavish storm Maybe, I figured it was a unix issue since I'm using a double fork to start the background process.

[undone pulsar]

You could use | tr -d "()" here rather than | sed -e 's|(||g' -e 's|)||g'

In fact you could probably skip out on the tr entirely by replacing the grep with something like grep -Po '(?<=\()[^)]*'

Worth noting that this particular pattern won't catch unmatched parens though.

[frank coral]

Is it possible to generate a public ssh key without the [e-mail] part in the end?

[amber garnet]

How about -c flag?

[frank coral]

thank you 🙂

[amber garnet]

👍

[nimble panther]

!warn 824602496742719498 Please don't post pictures of random people.

[shy yoke]

:incoming_envelope: :ok_hand: applied warning to @bold fjord.

[shrewd thicket]

Why does my discord bot fail, after a while, with the message "Killed"?

This is a graph of the memory available on my ubuntu machine

The dip is where the process shuts down

[modest cargo]

I have a server and I created a DNS name to point to that server, is there a way that I can check in a script in the server if the DNS name points to the local server?

[main olive]

is anyone willing to help me with an issue i dont see any open help channels. dont really want to get banned by bombing the channel with my problems, if you can shoot me a message, that would be awesome thanks 🙂

[warped nimbus]

There are available channels. The system is explained here #❓｜how-to-get-help

If your question is about unix stuff you can just ask here though

[main olive]

no it wasnt, there were no available channels for a little there i was wondering if it just stays like that or... anyways thanks

[warped nimbus]

There definitely were. We have plenty in reserve right now. If there were not any available then we would have gotten an automated alert.

Unless it was being slow to move the channels over or something

[main olive]

strange twice today i have not seen the available help channel, maybe i missed it cause theres no "#"

[lavish storm]

i see three

[main olive]

yes

[lavish storm]

go ask in one of them now!

[main olive]

i already asked

[lavish storm]

nice! good job!

[main olive]

🙂

[coral moth]

Can we make os with cython

?

[hidden sail]

same question

[amber garnet]

I don’t think so

You need system language like C or Rust

[coral moth]

ig we can still add a little bit of python ?

like 90% is c or c++

[lavish storm]

[white barn]

OS can have a little Python, as a treat

I mean would it count as a bit of Python if you used python like old school 8-bit computers used BASIC?
you didn't use Python for OS development but you gotta do some stuff to make python work with your custom OS

[coral moth]

Oh okay gotcha thanks

[formal schooner]

i believe an OS kernel usually needs a bit of assembly to start itself, and then needs to be written in a language like C (but not necessarily C) in order to have full, precise control over the hardware

once you have a kernel (not a small task) you can write as many python programs as you want... but you will need a python interpreter too

[real swan]

does the base register contain physical address or logical address? when i googled i found that it maybe logical or physical which im not sure if it is right or wrong
any help in this regard would be great

[dapper musk]

logical anddres since CPU is dealing with it. also i wonder what this question has to do with unix or pythron

[pulsar hawk]

I’m here

[gleaming topaz]

I restarted it with sudo reboot and it's stuck like this for several minutes

[pulsar hawk]

do uname -a and tell me what it says

[gleaming topaz]

I can't do anything

[pulsar hawk]

Oh there’s not even a cli ?

[gleaming topaz]

Nothing happens when I type or move my mouse

Also the green LED isn't blinking

[pulsar hawk]

Yeah I don’t know try unplugging it then flashing it again if data gets messed up in the process

I’ve never seen that problem, sorry

[gleaming topaz]

At the top it says Internal error: Oops: 17 [#11] SMP ARM
Wtf

[desert bloom]

Do you have more than one kernel to choose from?

[gleaming topaz]

Nope

Installed the standard raspPi Debian (I think it's called that)

I now unplugged it.

I'll let it cool dow for a bit

Ok it booted regularly. No need to flash it

[coral moth]

alr thanks ^^

[scarlet glen]

Just added ubuntu to the new windows terminal but its pathed as "/mnt/c/Users" when I use it and I cant change drive. The git-bash changes drive with no problems as its not pwd'd as mnt/c and just c/.

[dapper musk]

well thats beacouse WSL is basicly virual mashine and there is no such thing as c/ a/ b/ etc in unix beacouse it uses mount system

and git-bash is just shell like cmd

[olive ice]

I hate wsl just use a gnu/Linux distro instead

[dapper musk]

you can't always do it for example on work computrer

[scarlet glen]

^thats right

I found that I can still path through my disks just differently thankfully, its path'd as mnt/DRIVE etc.

[pastel kestrel]

this chat is MEGA dead

[lavish storm]

Thanks for the insight

Want to tell us a bit more?

[pastel kestrel]

mhm

[pastel kestrel]

yeh

[grand sinew]

genfstab -U /mnt >> /mnt/etc/fstab
say I wanted to execute this (unix like or whatever) command through python with subprocess.call, would i give ">>" as an arg?

[vagrant fern]

no

you'd open /mnt/etc/fstab as a file in append mode and give that to the subprocess stdout

with open('/mnt/etc/fstab', 'ab+') as fstab:
    subprocess.run(['genfstab', '-U', '/mnt'], stdout=fstab)

something like this

[storm trellis]

I have successfully launched a python script using "LXDE-PI" autostart function, however, this script needs to run as root in order to function correctly. How can I do this?

Where would I add "sudo"?

[long dune]

Any way to reboot a remote server via script without losing the current connection or disrupting the script?

[main olive]

Rebooting turns off the machine, so no

[tall bone]

i mean if you are just running a script try serverless and use ur remote server normally

[humble onyx]

you need something like bash -c "genfstab -U /mnt >> /mnt/etc/fstab"

[shy yoke]

Hey @rich kernel!

Uh-oh! It looks like your message got zapped by our spam filter. We currently don't allow .txt attachments, so here are some tips to help you travel safely:

• If you attempted to send a message longer than 2000 characters, try shortening your message to fit within the character limit or use a pasting service (see below)

• If you tried to show someone your code, you can use codeblocks
(run !code-blocks in #bot-commands for more information) or use a pasting service like:

https://paste.pythondiscord.com

[rich kernel]

I am getting this error each time I try to create a virtualenv venv python environment on ubuntu:trusty
https://paste.pythondiscord.com/sipeheweto.coffeescript

On local machine, image works fine, but when I attempt to use same docker image in production I get the errors above.
Any tips?

[short brook]

Hello everyone:

I need some help with a small task.
i want to build a script which will be automatically executed at login to the server via ssh.
this script should accept only a few keywords which are basically {projectnames} and run the underneath laying tasks.
my Problem is that im uncappable of creating the SSH part... the script itself (which just rebuilds some docker containers) is already done.
i need a solution to provide both.. Linux Shell login via Putty or MobaXTerm or similar which gives you a python input inside the script and ssh command execution via Command line like

ssh user1@server1 projectname

[glass remnant]

i know linux bois

[fallow blade]

Hi guys do i really need to go into bios settings to fix the virtualization issue on my VIrtual machine?

[glass remnant]

Yes

you have to turn on virtualization from your bois

bios*

[fallow blade]

AFAIK it doesnt affect performance much right?

[short brook]

Got it working 😄

finally 😄

[woeful dome]

I'm missing memory in /sys/fs/cgroup/ -- what can I do to fix that?

Can I just do mkdir /sys/fs/cgroup/memory?

hm, no, I can't

[opaque ginkgo]

raspi?

[woeful dome]

nope, desktop computer with Manjaro, the kernel is 5.10.49-1

[humble onyx]

$ ls /sys/fs/cgroup/
cgroup.controllers             cgroup.subtree_control         cpu.stat                       io.cost.qos                    memory.stat                    sys-kernel-debug.mount/
cgroup.max.depth               cgroup.threads                 dev-hugepages.mount/           io.pressure                    -.mount/                       sys-kernel-tracing.mount/
cgroup.max.descendants         cpu.pressure                   dev-mqueue.mount/              io.stat                        proc-sys-fs-binfmt_misc.mount/ system.slice/
cgroup.procs                   cpuset.cpus.effective          init.scope/                    memory.numa_stat               sys-fs-fuse-connections.mount/ user.slice/
cgroup.stat                    cpuset.mems.effective          io.cost.model                  memory.pressure                sys-kernel-config.mount/       ```

it doesn't look like it is supposed to be directory

[woeful dome]

🤔

[amber garnet]

@woeful dome

$ ls /sys/fs/cgroup/
blkio    cpu,cpuacct  freezer  net_cls           perf_event  systemd
cpu      cpuset       hugetlb  net_cls,net_prio  pids        unified
cpuacct  devices      memory   net_prio          rdma

[humble onyx]

I ran it on a vm if that makes a difference

[warped nimbus]

It might be something controlled by a kernel parameter

Though I am not finding anything for that for arch or manjaro

You can check if the memory controller is even enabled with cat /proc/cgroups. It will probably show it as disabled.

[woeful dome]

#subsys_name    hierarchy    num_cgroups    enabled
cpuset    0    130    1
cpu    0    130    1
cpuacct    0    130    1
blkio    0    130    1
memory    0    130    1
devices    0    130    1
freezer    0    130    1
net_cls    0    130    1
perf_event    0    130    1
net_prio    0    130    1
hugetlb    0    130    1
pids    0    130    1
rdma    0    130    1

This is what I get, seems enabled.
It was there before -- otherwise I wouldn't get confused

[warped nimbus]

Does cgroup2 have it? Or is that not enabled for you?

I think arch started enabling it by default

Thought it may only enable one at a time

[woeful dome]

Arch Wiki says it's enabled by default, how do I check?

sorry, I'm a giant noob, as you might suspect

[warped nimbus]

If it was enabled, presumably /sys/fs/cgroup2 would be mounted

Well, maybe that's not true. It could be enabled at the kernel level but simply not mounted, perhaps.

I played around with this on my Arch system. It seems like it mounts cgroup2 only under /sys/fs/cgroup, but both v1 and v2 are enabled.

It shows cgroup2 mounted under /sys/fs/cgroup in /etc/mtab

And it shows both fs enabled in /proc/filesystems

Actually, if you look in /etc/mtab, what do you see for the cgroup entry?

This is what I have. You can see "memory" in there ```
cgroup2 /sys/fs/cgroup cgroup2 rw,nosuid,nodev,noexec,relatime,nsdelegate,memory_recursiveprot 0 0

Oh I misread it

Harder to read in my terminal. memory_recursiveprot is not related

So yeah if you check mtab you will see if it's using v1 or v2. v2 has a different structure, which is probably why you don't see memory anymore

[woeful dome]

I get cgroup2 /sys/fs/cgroup cgroup2 rw,nosuid,nodev,noexec,relatime,nsdelegate,memory_recursiveprot 0 0

alright, the reason I needed /sys/fs/cgroup is that that's the default location for nsjail's memory/CPU limiting stuff

[warped nimbus]

nsjail does support cgroup v2, but I only have experience using it with v1

IIRC there is a switch to enable v2 support

[woeful dome]

@warped nimbus Another question: is there any disadvantage in using rlimit_... on nsjail instead of cgroups, if I need to just limit RAM and CPU?

[warped nimbus]

I think the biggest difference is how it works for child processes

cgroup operates on groups of processes. From working on snekbox, I know that the limits set via cgroups are shared by all children. I don't think this is the case for rlimit

By shared, I mean that e.g. if you limit memory to 10 megabytes in a cgroup, then all processes in the cgroup can in total use 10 megabytes, where as with rlimit I think each process would get its own 10 megabytes.

[woeful dome]

oh

so I should 100% use cgroups

[warped nimbus]

Probably, cause I can at least confirm its behaviour.

I am not so sure for rlimit as I don't have as much experience with it

[woeful dome]

Can you recommend some place to read about cgroups?

[warped nimbus]

IIRC redhat had a good guide for them, but I believe it was for v1. I'll try to find it again

Here it is. https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/6/html/resource_management_guide/ch01 but yeah it's for v1

[woeful dome]

thanks, I'll check it out

[warped nimbus]

This is something I planned to read but never got around to it https://lwn.net/Articles/679786/

https://www.man7.org/conf/lca2019/cgroups_v2-LCA2019-Kerrisk.pdf

[woeful dome]

Hm, I enabled cgroups v2:

cgroup_mem_max: 53687091200
cgroup_mem_mount: "/sys/fs/cgroup/memory"

cgroup_pids_max: 1
cgroup_pids_mount: "/sys/fs/cgroup/pids"

use_cgroupv2: true

and I'm getting this error:

[E][2021-07-24T04:08:50+0300][5797] bool subproc::initParent(nsjconf_t*, pid_t, int)():392 Couldn't initialize cgroup 2 user namespace for pid=5804
[F][2021-07-24T04:08:50+0300][1] bool subproc::runChild(nsjconf_t*, int, int, int, int)():448 Launching child process failed

[warped nimbus]

I believe this is the issue where you need to create the cgroup manually before launching nsjail

nsjail doesn't do this for you because that'd require root

[woeful dome]

I did this:

sudo mkdir -p /sys/fs/cgroup/memory/NSJAIL
sudo mkdir -p /sys/fs/cgroup/pids/NSJAIL
sudo chown $USER /sys/fs/cgroup/memory/NSJAIL
sudo chown $USER /sys/fs/cgroup/pids/NSJAIL

[warped nimbus]

That would be correct for v1, but not v2

v2 merges controllers into a single hierarchy, or something like that

You just need to mkdir /sys/fs/cgroup/NSJAIL

Cause the default cgroupv2_mount for nsjail is /sys/fs/cgroup and it appends NSJAIL to that

https://github.com/google/nsjail/blob/master/cgroup2.cc#L42-L44

[shy yoke]

cgroup2.cc lines 42 to 44

static std::string getCgroupPath(nsjconf_t *nsjconf, pid_t pid) {
	return nsjconf->cgroupv2_mount + "/NSJAIL." + std::to_string(pid);
}```

[woeful dome]

So it seems like it will try to create a new directory with the name like /sys/fs/cgroup/NSJAIL.42069 -- wouldn't that require root access?

I tried running it again, still fails, but with a different error

[F][2021-07-24T04:19:18+0300][1] bool subproc::runChild(nsjconf_t*, int, int, int, int)():448 Launching child process failed
[E][2021-07-24T04:19:18+0300][6496] int nsjail::standaloneMode(nsjconf_t*)():256 Couldn't launch the child process
```,and now I have a few directories like

/NSJAIL.5960
/NSJAIL.5961
/NSJAIL.5962
...
```, on which I get Operation not permitted when deleting, even as root

[warped nimbus]

Oh I didn't notice the PID was not a child path

Try configuring cgroupv2_mount to a cgroup that you create beforehand

like cgroupv2_mount = "/sys/fs/cgroup/NSJAIL" and then it should be creating cgroups under it like /sys/fs/cgroup/NSJAIL/NSJAIL.42069

That's basically how it does it for v1.

[woeful dome]

Now I get

[E][2021-07-24T04:25:18+0300][7044] bool util::writeBufToFile(const char*, const void*, size_t, int)():100 Couldn't write '4' bytes to file '/sys/fs/cgroup/NSJAIL/NSJAIL.7050/cgroup.procs' (fd='6'): Permission denied
[E][2021-07-24T04:25:18+0300][7044] bool subproc::initParent(nsjconf_t*, pid_t, int)():392 Couldn't initialize cgroup 2 user namespace for pid=7050
[F][2021-07-24T04:25:18+0300][1] bool subproc::runChild(nsjconf_t*, int, int, int, int)():448 Launching child process failed

[warped nimbus]

Also use -v for verbose output for that error

[woeful dome]

will do

This is the verbose error I get:
https://paste.pythondiscord.com/raw/enuwosifal

@warped nimbus

[warped nimbus]

The weird thing is that it has permission to create the cgroup, but not to write inside it

[woeful dome]

yeah, it actually creates those directories

[warped nimbus]

Are they still there?

Can you try to write to cgroup.procs?

"permission denied" might be a misleading error. It may not actually be a file permission thing but more of a "this operation isn't supported" thing

Which still wouldn't make much sense

[woeful dome]

I tried writing quack in there, wasn't able to

file system is not full, I have 50 GB free

[warped nimbus]

I know that a permission error is given when trying to set memory swap limit but the kernel hasn't been configured for that, for example

What's the ownership/permissions of that file?

[woeful dome]

Access: (0644/-rw-r--r--) Uid: ( 1000/ df) Gid: ( 1000/ df)

[warped nimbus]

Also, you should try writing an actual number in there, it may reject input if it isn't an actual PID (or at least doesn't look like one)

[woeful dome]

huh

but... I am the owner of the file, how can I be denied permission?

[warped nimbus]

Does it let you write with sudo?

[woeful dome]

Yes, it was successful

[warped nimbus]

I'm not sure why

From kernel docs

Delegation

Model of Delegation

A cgroup can be delegated in two ways.  First, to a less privileged
user by granting write access of the directory and its "cgroup.procs",
"cgroup.threads" and "cgroup.subtree_control" files to the user.
Second, if the "nsdelegate" mount option is set, automatically to a
cgroup namespace on namespace creation.

Because the resource control interface files in a given directory
control the distribution of the parent's resources, the delegatee
shouldn't be allowed to write to them.  For the first method, this is
achieved by not granting access to these files.  For the second, the
kernel rejects writes to all files other than "cgroup.procs" and
"cgroup.subtree_control" on a namespace root from inside the
namespace.

The end results are equivalent for both delegation types.  Once
delegated, the user can build sub-hierarchy under the directory,
organize processes inside it as it sees fit and further distribute the
resources it received from the parent.  The limits and other settings
of all resource controllers are hierarchical and regardless of what
happens in the delegated sub-hierarchy, nothing can escape the
resource restrictions imposed by the parent.

Currently, cgroup doesn't impose any restrictions on the number of
cgroups in or nesting depth of a delegated sub-hierarchy; however,
this may be limited explicitly in the future.

Delegation Containment

A delegated sub-hierarchy is contained in the sense that processes
can't be moved into or out of the sub-hierarchy by the delegatee.

For delegations to a less privileged user, this is achieved by
requiring the following conditions for a process with a non-root euid
to migrate a target process into a cgroup by writing its PID to the
"cgroup.procs" file.

• The writer must have write access to the "cgroup.procs" file.

• The writer must have write access to the "cgroup.procs" file of the
  common ancestor of the source and destination cgroups.

The above two constraints ensure that while a delegatee may migrate
processes around freely in the delegated sub-hierarchy it can't pull
in from or push out to outside the sub-hierarchy.

For an example, let's assume cgroups C0 and C1 have been delegated to
user U0 who created C00, C01 under C0 and C10 under C1 as follows and
all processes under C0 and C1 belong to U0::

  ~~~~~~~~~~~~~ - C0 - C00
  ~ cgroup    ~      \ C01
  ~ hierarchy ~
  ~~~~~~~~~~~~~ - C1 - C10

Let's also say U0 wants to write the PID of a process which is
currently in C10 into "C00/cgroup.procs". U0 has write access to the
file; however, the common ancestor of the source cgroup C10 and the
destination cgroup C00 is above the points of delegation and U0 would
not have write access to its "cgroup.procs" files and thus the write
will be denied with -EACCES.

For delegations to namespaces, containment is achieved by requiring
that both the source and destination cgroups are reachable from the
namespace of the process which is attempting the migration. If either
is not reachable, the migration is rejected with -ENOENT.

I figured pasting here is more convenient since I cannot directly link to that section

But it's from here https://www.kernel.org/doc/Documentation/cgroup-v2.txt

[woeful dome]

@warped nimbus what does that mean for my case?

oh

my user needs write access to /sys/fs/cgroup/cgroup.procs, right?

no, that can't be right

[warped nimbus]

If you chown those files then you chould be able to write to them, is what I get out of it.

I'm not sure if the containment section applies, since you're note dealing with adding a process that's already in another cgroup

[woeful dome]

I had done sudo chown -R df /sys/fs/cgroup/NSJAIL, all the files in there already belong to me

[warped nimbus]

Unless it counts a process being in the root cgroup

But that would seem very impractical

@woeful dome maybe you need to do this? https://wiki.archlinux.org/title/cgroups#User_delegation

Honestly this is outside my knowledge at this point

[woeful dome]

did that, doesn't change the error

I feel like this is so much over my head

maybe I should just pick a different project

[warped nimbus]

Maybe mounting cgroup v1 instead would make your life easier.

[woeful dome]

yeah 😄

I guess the issue is that an update switched my system to v2

[warped nimbus]

But everything I have read leads me to believe that what you're doing should work

I don't see why you cannot write if you own the files.

Maybe the "containment" section does play a role here after all. I don't know.

[woeful dome]

it's pretty confusing tbh

everything is a file, except some things don't behave as usual files at all

[warped nimbus]

Yeah they have extra semantics

[main olive]

Is there any distros best for prgramming compared to the others?

[vagrant fern]

no

the "best distro" will be the one you feel the most productive in

[dark rivet]

it will be manjaro though

[arctic saffron]

Ubuntu and Pop os have tools for ML to easily get your graphics card working with tensor flow, Kali Linux has tools for hacking and pen testing, if you want to develop IOS games you should get a Mac. It totally depends on what type of programming.

[shrewd thicket]

Imagine developing for apple

[lavish kettle]

how about what's the best disto to run jupyter notebook 24/7 bots on a raspberry pi 4?

trading bots (crypto/stock) & discord bots

raspbian seems like default choice

[vagrant fern]

raspbian is the most sane, yes

[main olive]

jupyter notebook is not a good fit for bots imo

It's better to have an api of sorts running on the raspi

And a frontend on a pc which gets the necessary info from this api

You can avoid unnecessary overhead and the bot will be more responsive

[smoky harbor]

Mint Linux

Is my favorite of all I've tried if your still looking give it a try with Cinnamon been using as daily driver for a few years

[steep horizon]

having an issue installing python3.9 on Rasbpian buster, read my message in #help-cupcake

message link: #help-cupcake message

[lavish kettle]

@main olive that makes sense. would flask be reasonable choice?
would learning docker be worthwhile for this?

[main olive]

Flask...yes, haven't used it in ages so I don't remember if it has websockets support. Either way, websocket based api would the best option in case of bots and messaging applications. You could also take a look at FastApi.

Docker, in your particular case, most likely won't be required since you'll be only deploying the app on your raspi and don't need it elsewhere

At least not if you need to learn it

[lavish kettle]

that's very helpful. thank you. i'll start looking at fastAPI tutorials. was thinking about docker because there's a chance i'll have to deploy similar bots to different devices but doesn't seem like it'd be super necessary to start with

[main olive]

Yeah, first finalize the code itself before moving to containerize it

[haughty burrow]

Guys idk if the right lounge

but

I delete ubuntu partition on my windows and after extending the partiion

I was abouta go to the Prompt command to do the last command for the boot stuff

my pc got turned off

qfter viewing hella lot of YouTube videos

i was told to download windows 10 on a usb
aka bootable windows 10 usb

after that it led me to this

it's either install now
or repair your pc

what should i do

i want to keep my old files on my windows

[dapper musk]

so you have deletet ext4 partiton and you want data from it back?

[haughty burrow]

no i deleted it for good my problem is when i power the pc It used to give me a choice of what os you gonna use
i had dual boo

boot

[dapper musk]

ah you need to delete GRUB

you are on MBR?

[haughty burrow]

Honestly i dont know and i dont have any means to check right now

im' trying to execute the command
''
bootrec /rebuildbcd

its gives me this

basically means ''add the installation to the boot list ''more or less

it said
''the requested system device could not be found''

[dapper musk]

can you run diskpart

and list vol

and show me the output?

[haughty burrow]

@dapper musk

[dapper musk]

ok then run

bootrec /fixmbr

[haughty burrow]

hmm usually my old c had 322 why is E there

[haughty burrow]

okk

[dapper musk]

and this will remove grub

[dapper musk]

it is fluent

[haughty burrow]

do i run this while being on diskpart?

[dapper musk]

no outside it

[haughty burrow]

exit first then?

[dapper musk]

yes

ctrl-d, quit? or something like this

[haughty burrow]

Yes

it said

operation successful

(roughly translated)

[dapper musk]

great then just reboot

and it should work

[haughty burrow]

just for precaution i saw the video made this commad and then he typed exit

is it okqy to not write it ?

[haughty burrow]

this one *

[dapper musk]

oh as far as im concerned you dont need it

[haughty burrow]

ooo okay then i shall exit now

i got these

[dapper musk]

after reboot?

[haughty burrow]

aaah i should just turn off the pc ?

[dapper musk]

yes

[haughty burrow]

i still haven't

ok

i will do it

got this xc

[dapper musk]

oh ok scan os is nesecary

eh go back to shell

and run fixboot

scanos

bootrec rebuit etc... also to make sure

and dont waste time

[haughty burrow]

oh god it didnt let me go to comand prompt

[haughty burrow]

it sent me to this

[dapper musk]

your's usb is still inside it?

[haughty burrow]

yds

yes

[dapper musk]

ohhh im stupid it was ment to run on windows booted

not on usb

so grub should be still there

f12 del etc

wathever will let you into bios

boot from ssd/harddrive

to windows and run commands on it

[haughty burrow]

f12 where do it type it ? inside bios ?

cuz rn i can only enter bios

grub rescue is gone

[dapper musk]

you use it to enter bios...

ah ok then

you should have

something like this

choose your disk

or just remove usb

[haughty burrow]

my bios is looking like this though

wait

letme redo it

[dapper musk]

you should have boot section here. anyways just remove usb from it

and reboot

it will be easier

[haughty burrow]

these are in the boot options

damn im panicking rn

[dapper musk]

eh anyways

[dapper musk]

you shoudl't your's data are safe

just grub is still there

[haughty burrow]

ouuf im safe

[dapper musk]

just remove usb

[haughty burrow]

i removed it

[dapper musk]

reboot

and grub will show since im stupid

boot into windows

powershell/cmd as admin

[haughty burrow]

still this...

[dapper musk]

wtf. go into bios and check if all drives are listed

[haughty burrow]

according to the drives from this pic it seems like they are they aren't they?

[dapper musk]

ah

you had UEFI and i thought that mbr

but that is strange beacouse you did't had FAT32 prartiton

anyways switch into UEFI

[haughty burrow]

how man ,all i can do is enter this bios or stare at that black window xD

[haughty burrow]

did i f up when i wrote that bootelrec mbr thingy ?

[dapper musk]

it should f up pendrive

beacouse im stupid

but rest was't affected

[haughty burrow]

what does pendrive have

[dapper musk]

it had windows instaler

on it

[haughty burrow]

oh look

it took me back here

i entered the usb

wait

letme screnn it

this thingy

basically ere

[haughty burrow]

here

[dapper musk]

ok then go back to cmd

[haughty burrow]

im on cmd

its clean rn i better not f this up xD

diskpart right?

[dapper musk]

yes

list volumes

[dapper musk]

*list vol

[haughty burrow]

yes look at it

oh bow the c has 322 Gb like it used to

now*

[dapper musk]

ok your windows should be MBR since it does not have 32

and fixmbr acctualy did something

bootsect /nt60 <drive name>: /mbr

[haughty burrow]

oh is it good or bad

[dapper musk]

replace drive name

with your windows drive

[dapper musk]

good

wait a s

to make sure that it is mbr

[haughty burrow]

yes im waitinf

[dapper musk]

list disk

it

if it does not have * here

[haughty burrow]

disk 0 doesnt

[dapper musk]

and it is your main disk?

[haughty burrow]

yes

1tera 300gb

[dapper musk]

ok select it

using select disk 0

[haughty burrow]

done

[dapper musk]

list volumes again it shoud show drive leters

inside disk 0

check what letter is assigned to main windows

[dapper musk]

and use bootsect /nt60 letter /mbr

[haughty burrow]

it's c right

[dapper musk]

it is not

it might be

[haughty burrow]

oh

how to confirm

[dapper musk]

do you remember sie

*size of your main drive?

[haughty burrow]

i remeber Files Being on C

322 gb

[dapper musk]

ok so well it is C

[dapper musk]

.

[haughty burrow]

i had only C and D

ny D didnt have anything important

[dapper musk]

*you saw C and D

[haughty burrow]

ah right

ill do that cmd

[dapper musk]

also maybe it is to late

but you should backup your drive

anyways

run command

[haughty burrow]

on diskpart or outside

[dapper musk]

outside

[haughty burrow]

weird

[dapper musk]

oh letter

you were ment to replace it

[haughty burrow]

ops

[dapper musk]

with C:

[haughty burrow]

C: Not only C ?

[dapper musk]

C:

[haughty burrow]

it says Its successful

[dapper musk]

ok then restart and try to boot into legacy mode

and hopefully it works

[haughty burrow]

just to make sure ill send a screen

nah not this

[dapper musk]

i dont understand french...

[haughty burrow]

i know only how to reboot like ..normally

not legacy stuff

[dapper musk]

your windows was in mbr aka legacy

so yes boot into legacy

[haughty burrow]

i shoudld remove usb forst ?

does it matter ?

[dapper musk]

if you are booting into legacy it does not

if into UEFI it will pioritetize your's USB

and boot into it

[haughty burrow]

it showed me this before

but anyways what shoudl i do now

ooh

i know of a way to fix this

brb

saw it on ytb i should enable legacy something

on bios 'boot options''

[dapper musk]

you should have this enabled but well worth trying

[haughty burrow]

Look at this

@dapper musk a video sqid i should write down the number shown qnd press enter

[dapper musk]

i never saw it and i think that it meant a key

but well type it

[haughty burrow]

ima cry ...

[dapper musk]

ehh it should work. you can do it also other way

copy files into other drive

install windows

copy them again

[haughty burrow]

ia there a way to backup my current drives?

[dapper musk]

there is built in into windows but it sucks. clonezilla

is propably the best way

[haughty burrow]

xc

i really hope i didn't touch any data from the drives

tommorow is shoudl take my pc to fix it

[dapper musk]

ok good night

[haughty burrow]

gnitr

[rich mortar]

Anyone know how to merge 2 bin files into one on a mac?

[dusty basalt]

does anyone know what:
ar -p ifplugd_*_armhf.deb data.tar.xz | sudo tar -xJ -C / ./usr/sbin/ifplugd
of the last param for ./usr/sbin/ifplugd does?

I've been stuck at this for days, the base command works well for extracting the binaries from a package deb file and not installing it

but anything outside that breaks root (thankfully I'm doing this on an rpi)

[formal schooner]

@dusty basalt this is gnu tar, i assume? the man page says:

tar -x [-f ARCHIVE] [OPTIONS] [MEMBER...]

and

-x, --extract, --get
       Extract files from an archive.  Arguments are optional.  When given, they specify names of the archive members to be extracted.

so the ./usr/sbin/ifplugd argument means "extract only usr/sbin/ifplugd from the archive".

not 100% sure what you mean by "anything outside that breaks root", but note that -C means "extract files to /", so maybe you should be using -C ~/ifplugd-extract or otherwise something that isn't /, and then copying files to the right location afterwards

[dusty basalt]

@formal schooner I teased out the problem

[dusty basalt]

you were spot on here @formal schooner

the second argument for tar actually allows you to specify a member of a archive, it wasn't immediately clear that ./usr/sbin.ifplugd refers to the file tree within the archive. One might think that it refers to the file destination path and this was the mistake I made

if you ran ar -p ifplugd_*_armhf.deb data.tar.xz | sudo tar -xJ -C /

this copies the archive files directly to root (bad), and causes all your binaries to be overwritten

I immediately lost access to sudo, ls, basically everything

I couldn't really debug what's wrong when I could not even navigate within the terminal anymore or shutdown ---- this is what I meant by breaking root

I tried writing to a test folder, like -c /home/user/test and it appended to any files within the folder

so I'm not exactly sure what damage is caused by ar -p ifplugd_*_armhf.deb data.tar.xz | sudo tar -xJ -C /

but it was very, very frustrating to learn around this because any small misstep would cause your distribution to break.

[slender rapids]

Hejhej, i dont want to ask a super easy question but i cant find the right thing to google right now.
I have Windows on my main partition (C) and Ubuntu on D. How can i make it so that i get to choose at the start of the pc what i want to use? at the moment i have to restart the pc everytime. Do i need to get into the bios every time i start my pc?

[warped nimbus]

I used to do dual booting on an old macbook and I found the rEFInd boot manager to work well. It can even be spruced up with custom themes, if you're into that sort of thing. https://www.rodsbooks.com/refind/

[slender rapids]

Ill look into it! Thanks mark 🙂

[tawdry fox]

@slender rapids, may I ask what you're using Ubuntu for?

[slender rapids]

Ive been looking into deep learning a bit and ubuntu was recommended. Also i had some weird issues with python and windows

Also i like the idea of an open source product and would use it full time if i didnt like games so much 😅

[warped nimbus]

Well... Canonical (the company managing Ubuntu) is kinda bleh cause they bundled some Amazon app in one of the Ubuntu releases. Sure, that was almost a decade ago, but I have no reason to risk trusting them when there are so many other distros available.

[tawdry fox]

Because I am thinking that WSL might suit your needs.

[slender rapids]

ive tried wsl but i didnt get vim running on it. Not aas nicely as on unix systems

[tawdry fox]

Ohh I see, alright.

[slender rapids]

its probably just my incompetence tbh

[long dune]

Say I have a txt file like :
bobby
daniel
john
shortest bobby
mike
fred
taller bobby

And I want to grep for lines with"bobby" but want to exclude any bobby that is preceded (taller bobby etc). How would I do about doing that?

I am thinking this would be done with regex but I can't quite fathom doing it (perhaps with a negative lookbehind?)

[vagrant fern]

in plain Python i'd go with the best regex trick

http://rexegg.com/regex-best-trick.html

which would bring you to something like \w+ bobby|(bobby)

in grep i would probably just pipe that to another grep -v ' bobby' to exclude lines where there's another word in front of it

[vagrant fern]

✽ It relies on your ability to inspect Group 1 captures (at least in the generic flavor), so it will not work in a non-programming environment, such as a text editor's search-and-replace function or a grep command.

[long dune]

@vagrant fern wow thank you for the website, seems really useful.

I will try piping that to another grep

[pseudo ether]

I'm not sure if this belongs here. But I wrote a program for a not so tech savvy friend using tkinter. But the problem is that he runs Linux Mint on his old laptop and he isn't big on the terminal. So whats the best way to go about making a shortcut that will launch the program when he clicks on it?

[warped nimbus]

I don't know what desktop environment Mint uses, but a general idea I have is to make the file executable with chmod +x and add a shebang at the top. Maybe the DE will recognise that and respect the shebang. If not, maybe you'd have luck doing something similar, but with a shell script that does python3 yourscript.py. Otherwise, you can look up how to create actual shortcuts for whatever DE it uses.

[pseudo ether]

Ah, thanks. Thats what I figured. I was hoping to not have to use chmod. But I'm sure he'd be OK with one line in the terminal.

[warped nimbus]

Hopefully; it's a one-time thing.

[formal schooner]

interesting situation! try tar -t to see what's in the archive

[formal schooner]

if they're using the Mate DE, you might need to create a "launcher" desktop entry https://unix.stackexchange.com/a/580847/73256

it's not unlike the "create shortcut" functionality in windows

https://askubuntu.com/a/286651 possibly relevant for Cinnamon

[woven yarrow]

Can someone recommend a way to redirect the output of a python service with systemd? I wanna do what python3 test.py >> output.log does but for a service

StandardOutput=append:/root/output.log didn't do anything

[tawdry fox]

You should be able to hijack stdout to a StringIO and do whatever you want.

[woven yarrow]

How?

I don't wanna modify the .py I just wanna redirect the output to a file like python3 test.py > output.log

[formal schooner]

normally doesn't it redirect output to journald?

or do you want to also write to a log file? like what tee might do

https://stackoverflow.com/a/43830129/2954547 one possible alternative that sends output to syslog/journald and then configures rsyslog to re-emit the output to a log file

not sure why the StandardOutput=append:/foo option doesn't work

[woven yarrow]

journald was missing the output from the script

[formal schooner]

what does your unit file look like? maybe you need StandardOutput=syslog although afaik that's the default

[woven yarrow]

[Unit]
Description=Python script as service test

[Service]
ExecStart=python3 /root/test_service.py
Restart=always

[Install]
WantedBy=multi-user.target

Am I missing something?

I've also had StandardOutput=append:/root/output.log in previously but didn't do anything other than creating the file

[formal schooner]

does the service emit logs to stdout or stderr?

[woven yarrow]

Uh, no idea to be honest, it just does prints, what does that do?

[formal schooner]

yes that should be stdout

and you put it under the [Service] block? https://www.freedesktop.org/software/systemd/man/systemd.exec.html#StandardOutput=

[woven yarrow]

[Unit]
Description=Python script as service test

[Service]
ExecStart=python3 /root/test_service.py
StandardOutput=append:/root/output.log
StandardError=append:/root/error.log
Restart=always

[Install]
WantedBy=multi-user.target

It was like that

Both were created but stayed empty

All the script does is

from time import sleep

while True:

    print("Hello")
    sleep(1)
    print("Hello again")
    sleep(1)

It works with the > redirect but not as a service

[formal schooner]

huh

honestly no idea, maybe it's a buggy or outdated version of systemd?

[woven yarrow]

Hm idk it's a fresh install on GCP with their Ubuntu distro

[formal schooner]

do you see anything in journalctl from that service?

[woven yarrow]

Yes I see the starts and stops and restarts etc.

But not the prints

[formal schooner]

and i assume you're running this as root with the system-wide systemctl

as opposed to systemctl --user

or maybe systemctl --user but as root is more appropriate? hard to say

[woven yarrow]

I'm running every command as root yes

This is what the journalctl looks like

Jul 29 16:40:19 gcp-us-c1a-f1-micro-01 systemd[1]: test_service.service: Scheduled restart job, restart counte>
Jul 29 16:40:19 gcp-us-c1a-f1-micro-01 systemd[1]: Stopped Python script as service test.
Jul 29 16:40:19 gcp-us-c1a-f1-micro-01 systemd[1]: Started Python script as service test.
Jul 29 16:40:55 gcp-us-c1a-f1-micro-01 systemd[1]: Stopping Python script as service test...
Jul 29 16:40:55 gcp-us-c1a-f1-micro-01 systemd[1]: test_service.service: Succeeded.
Jul 29 16:40:55 gcp-us-c1a-f1-micro-01 systemd[1]: Stopped Python script as service test.

Looks correct just no print outputs

Which is what I want to redirect to a log file not these status updates btw

root@gcp-us-c1a-f1-micro-01:~# systemctl daemon-reload
root@gcp-us-c1a-f1-micro-01:~# systemctl start test_service
root@gcp-us-c1a-f1-micro-01:~# systemctl status test_service
● test_service.service - Python script as service test
     Loaded: loaded (/etc/systemd/system/test_service.service; disabled; vendor preset: enabled)
     Active: active (running) since Thu 2021-07-29 18:28:12 CEST; 2s ago
   Main PID: 102438 (python3)
      Tasks: 1 (limit: 677)
     Memory: 6.7M
     CGroup: /system.slice/test_service.service
             └─102438 /usr/bin/python3 /root/test_service.py

Jul 29 18:28:12 gcp-us-c1a-f1-micro-01 systemd[1]: Started Python script as service test.
root@gcp-us-c1a-f1-micro-01:~# l
Geekbench-5.4.0-Linux/        __pycache__/        distributor/  index/    output.log  snap/    test_service.py
Geekbench-5.4.0-Linux.tar.gz  dask-worker-space/  error.log     monitor/  proxy/      test.py
root@gcp-us-c1a-f1-micro-01:~# tail output.log 
root@gcp-us-c1a-f1-micro-01:~# 

(same with error.log)

Running systemd version 246

I figured it out

I had to add Environment=PYTHONUNBUFFERED=1

[formal schooner]

ah, that's a fun one

so python simply was never flushing the output buffer

good catch

[woven yarrow]

Yeah, I was lucky to find it here: https://unix.stackexchange.com/questions/408929/systemd-how-to-redirect-stdout-to-logfile

Completely unrelated to python but there's that one guy that saved me from a headache lol

[formal schooner]

no, this is a python specific thing for sure

[woven yarrow]

Yeah I know I meant the thread was not related to python

[vagrant fern]

PYTHONUNBUFFERED=1 was such an annoying gotcha for me. i wish python was doing that automatically if stdout wasn't a tty or something

[woven yarrow]

Same, it took me half a day to debug

[summer trail]

You could add -u to the python3 command line, instead. I'd prefer that over the environment variable, since it's not inherited across forks and therefore is a more focused fix

[woven yarrow]

What does that flag do exactly? I’ll try it in a bit

[summer trail]

Exactly the same thing.

It sets Python to unbuffered IO mode, but it only applies to that one run of Python, whereas the environment variable applies to every run of Python from the time you set it onward.

Which could, in theory, break something that expects its IO to be buffered.

[woven yarrow]

Ohh gotcha

[summer trail]

Granted it's unlikely to break anything, but I dislike global fixes to local problems

[woven yarrow]

Yup I agree, testing it now

Yeah works like a charm, thanks man

[past sluice]

hi how to log python print statement output in log file
i try this python3 /root/filedetect.py >> /var/log/filedetect.log 2>&1
but it log other info but not python print statements.
can someone help.

alright the answer is @summer trail -u

[final echo]

hallo

[wise valve]

@final echo can we continue here?

[final echo]

sure

[wise valve]

okay

so i look up some tutorials on how to use my linux?

[final echo]

yeah
ubuntu is mostly similar to windows
i can help you with setting up dev environment

[wise valve]

okay thanks

[final echo]

so u use vscode right?

[wise valve]

but u sure my pc specs dont suck?

[wise valve]

ye

[midnight rain]

I recommend the Debian administrator handbook after you learn some basics

[final echo]

sure they dont
mine is almost same
with no graphics

[final echo]

sure thats a good idea

[wise valve]

where do i learn the basics from

[final echo]

@wise valveok so let me teach you the terminal
https://ubuntu.com/tutorials/command-line-for-beginners#1-overview
you can view them here if u have doubts

[midnight rain]

maybe YouTube videos, on a top of basic commands

[wise valve]

okay

[final echo]

@wise valveyou know how to open terminal

[wise valve]

there is a terminal app right?

ik how to open that

[final echo]

ok nice

so ill show u how to update your system

[wise valve]

okay

[final echo]

sudo apt update && sudo apt update

run this command to update your system

basics

sudo = administrator
apt = ubuntu package manager (the thing where you can install stuff)
update = updating the things
upgrade = applying the update

&& = first do the command on left side then do the one on right

[midnight rain]

only if the first had a successful exit

[wise valve]

kk im downloading

[final echo]

yeah

forgot to tell that

[wise valve]

it downloaded some stuf

and then said command 'sude' not found

[midnight rain]

sudo

[wise valve]

i typed sudo only

[midnight rain]

hmhmm

[final echo]

thats weird

[midnight rain]

Ubuntu comes with it

[wise valve]

should i send you a screen shot?

[midnight rain]

yes ofv

odc

ofc

[final echo]

type sudo and look if its like this

[midnight rain]

it exists

that's the help

[wise valve]

yes this appears

[final echo]

@wise valveyou typed sude at the second time

[midnight rain]

you wrote sude

[wise valve]

oh yea

[midnight rain]

it's sudo

[wise valve]

okay cool now it works

[final echo]

nice

[wise valve]

done

[final echo]

it might take some time

[midnight rain]

I think it's something like Superuser do

[final echo]

ok

[final echo]

its basically root acceess

@wise valvenow would you like to install vscode

[wise valve]

yes please

[midnight rain]

ye but maybe like that he can remember better

[wise valve]

python is preinstalled right?

[final echo]

ok so we will get your hands wet now

[wise valve]

okay

[final echo]

yep
type
python3 --version