digital haven
130 function dirr_funct() {
131 # I want to see the stack, but not all on one line. This is mainly so that
132 # I can use p5 or something (defined below) to get back to a particular
133 # directory.
134 clear
135 echo $(dirs | sed 's/ /\\n/g') | cat -n | head -n 20
136 }
137 alias dirr=dirr_funct
and this will print out (part of) the stack with number
main olive
that's... incredibly complicated
digital haven
😬
main olive
what exactly are you trying to do
digital haven
print out the stack on different lines with line numbers
pass an arg to popd to pop a given number of dirs from the stack
digital haven
idk how else to explain it
that's that
well ha... from the cdr command the other day
digital haven
i found that i sometimes wanted to get back to where i was, but couldn't easily... so I wanted a command for that - but popd would just go back one
so, it's to address that 🤔
or to make 4 problems instead of 1
main olive
why not just make a "super push d" and "pop d" alias which saves the pwd to a variable and cd's back to that when popped
digital haven
i tried that first actually but the var didn't want to save so i did this
digital haven
i had something like PREVIOUS_DIR=$(echo pwd) or something
main olive
did you just VAR=value or did you export as well
digital haven
yeah - i can't remember exactly, but i couldn't seem to get the pwd to save as a var that i could access later
main olive
also echo pwd will literally just return pwd
digital haven
154 function cdr_funct (){
155 pushd $(git rev-parse --show-toplevel)
156 clear
157 }
i would have put it in there
i guess
main olive
$(pwd) by itself would've returned the current dir
digital haven
130 function dirr_funct() {
131 # I want to see the stack, but not all on one line. This is mainly so that
132 # I can use p5 or something (defined below) to get back to a particular
133 # directory.
134 clear
135 CURRENT_DIR=$(pwd)
136 echo $(dirs | sed 's/ /\\n/g') | cat -n | head -n 20
137 }
so like that?
ah that's in the wrong place
152 function cdr_funct (){
153 DIR_BEFORE_CDR=$(pwd)
154 pushd $(git rev-parse --show-toplevel)
155 clear
156 }
157 alias cdr=cdr_funct
158 alias rdc="echo $DIR_BEFORE_CDR"
so this is what i would expect to work there
cdr - get to the root
rdc - get back to where i was
( it does not work )
'echo' 🤦
ok it works 🌚
thank you
fickle granite
gaah a single global? really? feh
digital haven
@fickle granite is that bad? 🤔
fickle granite
it -seems- bad
like if you do two cdr_functs in a row, you'll only ever be able to get back to the most-recent one
imagine your web browser's history only remembered the last page you were at
digital haven
yeah, that's probably ok for this ha
i have the pops monstrosity for anything further 😄
digital haven
I'm interested to hear others thoughts on functions vs aliases now 🤔
why wouldn't i just use functions instead of aliases
fickle sparrow
I don't think there is anything an alias can do which you could not with a function. There is the alias command that lists all aliasses though..
digital haven
@fickle sparrow yeah, the aliases command is handy actually... I can't think of any other reason to use an alias over a function though
perhaps it's more explicit?
fickle sparrow
I would say so, yes.
digital haven
how can i see the definition of a function at the shell?
so if i have some function f defined, i want to do something like whats f and it will (at the prompt) display the code that defines f
not sure if that's a thing
warm wolf
trishmapow@tpc:~$ hello () { echo "hello"; }
trishmapow@tpc:~$ type hello
hello is a function
hello ()
{
echo "hello"
}
@digital haven
digital haven
-> % x (){ echo 'this is x' }
-> % type x
x is a shell function
ah, @warm wolf thank - seems that zshell doesn't give this information
humble onyx
Is there any difference between
export PATH="$PATH:$HOME"
PATH=${PATH}:$HOME
Why do I have export method anyway?
oak shell
normally, shell variables are for your shell only, but not exported to the environment of child processes
the export command changes that and switches a variable from shell-only to exported
PATH is exported by default though, so in this special case, both lines are equivalent, because once exported variables stay exported
humble onyx
Are they essentially the same?
oak shell
for PATH, yes. For most other variables not.
oak shell
Have a look at this example
humble onyx
shoot
oak shell
$ FOO=bar
$ printenv FOO
$ export FOO
$ printenv FOO
bar
here printenv is an external executable. It does not receive the value of unexported shell variables, but once it is exported, it can be read
glass rock
can i ask linux related question here?
oak shell
yes, just ask. Preferably they would be somehow related to Python development, but small general questions are okay too.
glass rock
i mean im making "auto mount usb" script with python, so its python related too 😄
so about the question, it looks very simple but im stuck and couldn't find any solution
im listing plugged drives, volumes with
gio mount -l
or
from gi.repository import Gio
etc etc
so this gives me the result i want, but when i add my script to cron it only shows the main boot drive and floopy/cdrom etc
gio mount -l >>> the result i want
sudo gio mount -l >>>> the result i don't want
sudo -u myuser gio mount -l >>>> the result i don't want
even though i just use that command without sudo, i don't get what iwant
dusty basalt
anyone around here messing about with the jetson-nano?
lavish badger
(github question) Say a Windows user has core.autocrlf=false and pushes code with crlf line endings. If a Linux user with above set to "input" or "false" downloads it, they would get code with crlf line endings?
fickle granite
dunno. If i had that question, I'd just try it and see.
That'll take some effort, especially if you don't have a linux box handy; but it has the benefit of being authoritative
lavish badger
Ok. Thanks offby1.
digital haven
@main olive i know it was horrendous but those dirr and pops functions have been pretty handy 😛
digital haven
when I display polish chars on my terminal they're mangled by tree, but not by ls, i don't understand why
🤔
tree -f --du -h -C -L 1
ll -lh
those are the two commands
cyan isle
hey guys I'm having probs tryna set a python version to default
anyone know??
nvm got it
haha
jade crater
Would it be better to use a python virtualenv for additional system packages? to prevent messing with python that comes with the OS
carmine ridge
venv starts off with whatever package you have installed right?
its not completely fresh?
daring osprey
it should be a closed system if you don't tell it to hook into system site packages, afaik
spice relic
does cp -a preserve the permissions?
white solar
venv starts off with whatever package you have installed right?
if you dopython3 -m venv venvit starts of fresh
copper sand
Hello, trying to see an example of how you would parse the output of a command and then append that parse to a file on a different machine.
So you would start your script on the machine that has the command you want to do.
You would then parse the output of the command, for example a network address
Then have that captured network address appended to a file on a SEPARATE machine
Any help would be greatly appreciated! I am just trying to learn bash scripting.
warped nimbus
If you have access to the other machine over ssh you can use scp to transfer the file
main olive
from machine 2, you'd do:
ssh user@machine1 'parse_script $(my_command)' >> file_of_things.txt
careful about quoting
I cannot predict whether that $() will run locally or remotely
remotely I guess since it's in single-quotes
main olive
runs the output of my_command through parse_script on machine 1 and appends the output to file_of_things.txt on machine 2
aye, single quotes is important here
fickle granite
I kinda hate the shell
gilded basalt
single quotes = don't expand/evaluate my shit
gilded basalt
double quotes = go wild dog
static mountain
I have a problem, its probably because i run it as root, how can i run python not as root?
cold pagoda
By default, if you just use python with your user account, it should not run as root
static mountain
Weird
cold pagoda
atomic ginkgo
Can anyone help with an issue I'm having packaging a C-based Python package as a manylinux2014 compatible binary wheel?
atomic ginkgo
main olive
is there a vsl flag in curl
main olive
I'm trying to automate some things for the first time due to working with server side coding thats more tediouis than what im used to. So I started very simple.
I discovered this while playing around
woot@wsi:~$ cat start
#!/bin/bash
cd project/
echo "Should Have changed directory but"
pwd
source environment/bin/activate
echo "Should have opened environment but"
which python3
echo "Ready"
woot@wsi:~$ ./start
Should Have changed directory but
/home/woot/project
Should have opened environment but
/home/woot/project/environment/bin/python3
Ready
woot@wsi:~$ pwd
/home/woot
woot@wsi:~$ which python3
/usr/bin/python3
woot@wsi:~$
At first it seemed like it was ignoring me. But it seems its doing everything in a subprocess and then escaping back to the shell I was in before, right? How would I go about something like this. My server is running ubuntu and nginx.
I apologise if my bash/server knowledge is very minimal.
south star
Exactly
One way to achieve what you want, is to run bash at the end, so it'll give you a shell in that env, when you are done, you can exit it with ctrl-d and be back to your original dir/venv
south star
i would just have put /bin/bash but i guess exec would work as well, it'll just replace your current process with the bash one, not sure what happens to cwd and env variables in that case, but it's worth trying.
main olive
This accomplished keeping me in the working directory expected in the script but didnt maintain the venv activation.
south star
gabriel@gryphon:~$ which python
/usr/bin/python
gabriel@gryphon:~$ source /home/gabriel/.virtualenvs/kivywheels/bin/activate
(kivywheels) gabriel@gryphon:~$ bash
gabriel@gryphon:~$ which python
/home/gabriel/.virtualenvs/kivywheels/bin/python
gabriel@gryphon:~$ exit
exit
(kivywheels) gabriel@gryphon:~$ which python
/home/gabriel/.virtualenvs/kivywheels/bin/python
(kivywheels) gabriel@gryphon:~$ exit
the venv is still active, even if not visible in the prompt
south star
probably because bashrc resets the PS1 variable
main olive
yeah, the most confusing thing about bash for me when it gets to complicated redirection is how variables constantly lost. Its the primary reason I dont use bash for anything that involves logic. But I figured a series of commands to the shell makes sense for it
At first I thought I would be using $() a lot to do things that you can't use that to do because it open a subshell and any values you were using are gone.
It reminds me almost of scope only I dont know how to send it values as params
Either way thank you
south star
yeah, things are only available to the subshell if they were exported, and of course if the subshell doesn't overwrites them
the --norc option of bash solves it for me
main olive
I will have to look at the option. It means dont use the config file Im guessing?
south star
yes
as you already have all the settings you want in the parent shell, it'll just inherit them, no need to reset to defaults
main olive
Good stuff.
I see what you mean about it forcing the cwd out.
I didn't realize that display of the prompt was a .bashrc thing.
I should look into that file ive been using linux long enough haha
south star
yep 🙂
you can set your own aliases/functions in there, can be useful to improve your productivity
main olive
for future reference, if you want to change the variables and cd inside your current shell and not a subhselll, you need to source it instead of executing it
ie
$ . myscript
instead of simply
$ ./myscript
or the full source command instead of the dot .
main olive
interesting, I didnt even know about that. I only use source for venvs and otherwise have never checked it
Time to bust out the man page
virufac@box:~$ man source
No manual entry for source
virufac@box:~$ source --help
source: source filename [arguments]
Execute commands from a file in the current shell.
Read and execute commands from FILENAME in the current shell. The
entries in $PATH are used to find the directory containing FILENAME.
If any ARGUMENTS are supplied, they become the positional parameters
when FILENAME is executed.
Exit Status:
Returns the status of the last command executed in FILENAME; fails if
FILENAME cannot be read.
virufac@box:~$
Well then
This led me down the interesting line of thinking...
What is in my activate script. While there is quite a lot in a deactivate function as well as some zsh hacks and making sure you are not running without using the source command...
VIRTUAL_ENV="/home/woot/project/environment"
export VIRTUAL_ENV
_OLD_VIRTUAL_PATH="$PATH"
PATH="$VIRTUAL_ENV/bin:$PATH"
export PATH
# unset PYTHONHOME if set
if ! [ -z "${PYTHONHOME+_}" ] ; then
_OLD_VIRTUAL_PYTHONHOME="$PYTHONHOME"
unset PYTHONHOME
fi
if [ -z "${VIRTUAL_ENV_DISABLE_PROMPT-}" ] ; then
_OLD_VIRTUAL_PS1="${PS1-}"
if [ "x" != x ] ; then
PS1="${PS1-}"
else
PS1="(`basename \"$VIRTUAL_ENV\"`) ${PS1-}"
fi
export PS1
fi
seems to be the bulk of it
export is a command I also dont understand very well
I have never manually adjusted PATH or any environment variables. IF I need a script to be in PATH, I will typically put it somewhere that is already in path.
warped nimbus
It basically prepends the venv's path to the PATH env var, so that it takes priority over any other Python interpreters that may already be on your PATH.
PYTHONHOME is an env var that you may have set to change the location of Python's libs, so this script just makes sure it's unset while the venv is active.
The ps1 stuff at the bottom just changes the terminal prompt so there is a visual indication the venv is active
main olive
thanks.
humble onyx
I have a question. I have a python file which is not by itself an executable. when I try to run with python on command line it runs but when I do ./file.py I get permission denied error(chmod +x file.py works fine that's not the question).
I have she bang line at the top of the file so it should be an executable isn't it? what did I misunderstand?
./file.py # Permission denied. There is a shebang line at the top of the file```the shebang simply tells the shell what to open the file with (it doesn't automatically know it's python)
you still need to actually mark any files with the executable permission to directly execute them
@humble onyx
humble onyx
I got it now thanks
humble onyx
padding = "\x21\x41\xf4\x08"
print(padding)
!A
padding = "\xf4\x41\xf4\x08"
print(padding)
xf4 breaks the hex
what is happening?
main olive
its not a valid ASCII character so it doesn't get printed
it's still there, it just breaks your term a bit
humble onyx
x41 prints A
I have x41 in both strings
@main olive
when xf4 at the beginning it doesn't print A
main olive
like I said, it confuses your terminal and it stops printing it
don't try to print things which aren't real symbols
humble onyx
well I was trying to do something else
I was doing stack overflow
EIP becomes 0x00000000
that is the thing
main olive
no, it should just work
just don't try printing it
if your IP is null then you're doing something wrong
something else, I mean
main olive
Not sure if this allowed here, but when I have a hard drive in my system with files I put on it under Windows, what happens to those files when I boot into Linux? Can I access them on Linux even though they were written to the drive under Windows?
On a dual-booted system
sage solar
Yes
However
Your Linux will most likely have it's / file system formatted as ext4 which is a file system windows won't understand
main olive
I made the mistake when I first got into Linux of using NTFS as my default storage (not OS) drive. It works. But I would recommend against doing this. While you can see into NTFS, it's better in my experience to use Ext while on Linux for youre regular file use. If you need something, copy it over. But work with it on the correct file system.
Do people who are more expertise with this agree that that is best practice?
Ther reason I ask is because in the early days of my Linux use... There were unexplained corruptions and sometimes just disappearances of data that I wish I could go back with the knowledge I have now and figure out why that happened. But it was too long ago now.
I have always attributed it to the file systems tho
I do recall though at that time I was regularly booting into and out and the two OSes and I was sharing the different drives like they were not essentially different.
main olive
the NTFS driver is considered stable, but NTFS itself is horrible from a quality-of-life and performance standpoint
on linux, it doesn't make sense to use NTFS due to the heavy reliance on symlinks and such
on linux, xfs or ext makes much more sense
lavish badger
Will running: (venv) $ pip3 install --upgrade pip from inside a python3 virtualenv affect system pip/python at all?
main olive
no
lavish badger
Thanks @main olive
main olive
also @lavish badger it's good to know that venv and virtualenv are different things
so when googling for things, you won't get conflicting information
main olive
👍
worn aspen
how would I go about creating a sys tray icon and menu as lightweight as possible?
worn apex
there's a library called pystray that seems to handle this @worn aspen
balmy smelt
Hey all!
I want to build a static binary of the Python interpreter, but I also want to include some modules. Let's say I want to have the apt module. Is this possible?
I have followed this: https://wiki.python.org/moin/BuildStatically
But there are only some modules there (I guess they are only some essentials) but I have no idea how, or if even possible, to also include other modules like the apt module?
Do you know if such a thing is possible? If not, what would be the right approach?
fickle granite
if the module is written in python, it's hard to see how you'd get that code into a static library
static libraries contain object code, and the only way I know of to make object code is to compile an Old Skool language like C
the only other thing I can think of is if python has a way to load modules, then dump its memory into a new executable, but I don't recall anything like that in the Python build process. (GNU Emacs does this, but it's weird 🙂)
sage solar
Not only c can do that....pretty much any compiled language will be able to, c, c#, rust, go and what not
digital haven
quite often there are a few directories that i want to navigate to in a project, and cd-ing about is a bit of a chore, is there an easy way around this? I currently have the dir stack printing out as
echo $(dirs | sed 's/ /\\n/g') | cat -n | head -n 30
and a function that'll enable me to jump to a dir there using pop, but that will lose information, and i can only get somewhere I've previously been. Even though it's pretty janky, i make use of it a lot, i just wish there was something a little more robust.
something that could perhaps like
tree -d -L 3
where the lines were numbered, and i could choose a line to go to, that would be nice
-> % tree -f -d | cat -n
1 .
2 ├── ./a
3 │ └── ./a/b
4 │ └── ./a/b/c
5 └── ./x
6 └── ./x/y
7 └── ./x/y/z
8
9 6 directories
so here 4 would take me to a/b/c
oak shell
why that echo? 🤔 You could replace that whole alias with dirs -v
$ dirs -v
0 ~/Pictures
1 /bin
2 /etc
3 /tmp
4 ~
about the tree navigation, I'd suggest you use a tool for that... let me try remember the name
https://github.com/jarun/nnn#readme this maybe
digital haven
why that echo
@oak shell because i don't really know what i'm doing 😄
@oak shell how would i know -v, it doesnt' seem to be in the man page, is this a standard flag?
digital haven
ah, damn, thanks...
these numbers on nnn are super nice 🙂
stray marten
when using stat command
what is the real size of the file_
?
size or io blocks?
size * blocks?
or next power of 2 of size * blocks?
small panther
That is the real size.. Size: 724
It's in bytes however. So bytes * 1024 = kb
And kb * 1024 = mb
stray marten
and the size of the blocks?
u cant know from there, right?
u need %B
which is 512
so how many space does this file occup?
8 blocks * size per block?
@small panther
small panther
I don't understand what you're asking
Size is the actual number of bytes of that file's contents. IO Blocks is the 'chunk size' for the data being read/written from the disk itself.. in your screenshot that would be 32KB minimum per read/write to the disk itself
The file occupies 724 bytes, like the screenshot says
Unless you are asking about lower-level size on the physical layer or related to inodes or something, which i don't know that detailed
stray marten
yes, but on the disk
small panther
Maybe that can be of some help to you
small panther
Probably. I'm not sure. There needs to be some sort of pointer to the memory and the file name, I believe. That would occupy space. I don't know if the size reported in stat is only the file contents, or if it contains any meta-data as well.
stray marten
@small panther one last thing. For linux, day 0 is sunday or monday?
okey 0 is sunday
small panther
That would entirely depend on what you're doing with it
stray marten
with the crontab
small panther
https://stackoverflow.com/questions/18919151/crontab-day-of-the-week-syntax
According to that, yeah sunday is zero. But you can also just type SUN and it will run every sunday
digital haven
does anyone know how i can start vim with a command? I want an alias that will start vim with NnnPicker command (but not always)
vim -c "NnnPicker"
all good
stray marten
@small panther and last last thing XD how can i remove all subdirectories from a given one?
i wanna remove everything inside /user/foo
but i dont wanna do rm -r /user/foo
i wanna delete all inside foo
rm -rf /user/foo/*?
digital haven
Also - re earlier - i think autojump might be pretty much what I was after @oak shell , thought I'd let you know
swift heron
How can i use lolcat -a to animate a whole file rather one line at a time?
main olive
i don't think you can without a script
also that can be a very heavy task for your terminal emulator
main olive
Hi im on linux trying to get pip since it dident come with python idk why
oak shell
what os?
main olive
Mint
oak shell
try sudo apt install python3-pip then
main olive
do i need to reboot?
main olive
ah thanks
Collecting cryptography
Using cached https://files.pythonhosted.org/packages/45/73/d18a8884de8bffdcda475728008b5b13be7fbef40a2acc81a0d5d524175d/cryptography-2.8-cp34-abi3-manylinux1_x86_64.whl
Collecting six>=1.4.1 (from cryptography)
Using cached https://files.pythonhosted.org/packages/65/eb/1f97cb97bfc2390a276969c6fae16075da282f5058082d4cb10c6c5c1dba/six-1.14.0-py2.py3-none-any.whl
Collecting cffi!=1.11.3,>=1.8 (from cryptography)
Using cached https://files.pythonhosted.org/packages/f1/c7/72abda280893609e1ddfff90f8064568bd8bcb2c1770a9d5bb5edb2d1fea/cffi-1.14.0-cp36-cp36m-manylinux1_x86_64.whl
Collecting pycparser (from cffi!=1.11.3,>=1.8->cryptography)
Using cached https://files.pythonhosted.org/packages/68/9e/49196946aee219aead1290e00d1e7fdeab8567783e83e1b9ab5585e6206a/pycparser-2.19.tar.gz
Complete output from command python setup.py egg_info:
Traceback (most recent call last):
File "<string>", line 1, in <module>
ModuleNotFoundError: No module named 'setuptools'
----------------------------------------
Command "python setup.py egg_info" failed with error code 1 in /tmp/pip-build-sozflehp/pycparser/
that's weird, the python3-pip package should depend on python3-setuptools 🤔
main olive
🤔
humble onyx
Hey guys can somebody help me understand what setuid mean? Wikipedia says it gives root privileges to the user but I don't fully grasp it. I have always been the only user on my laptap that may be the reason.
Like how is it different than chmod 777 file. Anybody can do anything with the file already
sage solar
if you execute a binary that has setuid set the binary itself gets executed with admin privileges even if the current user does not have admin privileges
that way you can let users do specific things that would usually require admin without actually granting them admin privileges
however this of course also makes such binaries interesting targets for hackers, because if they can control what they do they essentially have admin privileges on your PC
humble onyx
does it give root privileges or owner's privileges? If I have write permission on a file which has setuid bit. Can I do whatever I want?
main olive
setuid runs the binary as the owner of the binary
setgid does the same, except for the group which the binary belongs to
it's most often used to allow regular users to run binaries as root
humble onyx
What can a binary do with root privileges but it cannot do without setuid bit.
echo laurel
In case of suid,
if the owner is root, you have access to everything root has access to.
if the owner is someone else, you have privilege of that user.
same for giud, but in this case the binary is run with the group the binary belongs to
humble onyx
@echo laurel What if everybody has write permission on that file?
echo laurel
that means everyone can write to the file.
humble onyx
Can anybody manipulate the file to do whatever he wants to do that requires root's privilege?
echo laurel
again, this has nothing to do with root.
ONLY if the owner is root, you will get root permissions
main olive
@humble onyx suid bit is dropped when the file is modified
so if you have write access to a suid binary as a regular user, you can't simply modify the binary to do any malicious thing you want
humble onyx
@main olive Thanks all clear now
swift heron
How can i automatically have a specific user only be allowed to stay logged in for 30 seconds when they login? because from what i have tried so far while its waiting for the 30 seconds to pass they are unable to type or use commands
Emphasis on specific users, I don't want to implement this system wide
main olive
when will an env var set with export VAR="value" be unset?
digital haven
372 project_funct () {
373 pwd
374 }
375 export PROJECT_HOME=project_funct
is there any way to make this work? and why isn't the above working
ah, bc it's run when the script is sourced, ok that's obvious 🤦
amkproject (){
export PROJECT_HOME=$(pwd)
mkproject $1
}
i just went with this 🤷♂️
probably not worth it I guess, though there are some hooks and stuff I might use for mkproject so, eh
digital haven
@fickle granite only pinging because it's something that you'd explicitly mentioned before, but I've only just realised that I did install gnu tools for OSX, but that everything has a g prefix, so rather than setting find to be GNU by default I have BSD find, and also have gfind.
If I run ls -lat | awk '{print $9}' | egrep '^g' then I can see all that start with g, but there's stuff like git-cvsserver that i never use, and am pretty sure isn't a mirror of Gnu and BSD or whatever.
what I'm wondering about, is whether there's a straightforward approach to saying I just want the GNU stuff. I was considering just setting a bunch of aliases, as in alias find=gfind, but I'm not sure if this is a sensible approach
actually alias won't work because then i can't use things like man find, what a load of crap
main olive
You need to, you just make a bash file that has the path to a python file. Let me show you hold on
Oh nevermind
lol
What is the outcome that you want to be different than using bash to call the python?
young sphinx
can u still show me tho, i'm curious how u merge python and unix.
Currently i do subprocess.Popen / os.system ish, and then i created an alias to python myscriptpath.py with my desired cmd name in ~/.bashrc.
But with i only manage to ask for args with input() instead of taking them inline.
Plus, I suppose this is not the easyiest way to make such command.
young sphinx
sure
main olive
virufac@box:/bin$ ll | grep nsr
-rwxr-xr-x 1 root root 58 Oct 4 12:44 nsr*
virufac@box:/bin$ cat nsr
#!/bin/bash
exec ~/Git/nameless-space-replace/nsr.py "$@"
virufac@box:/bin$
so, in the bin directory.... there are the average unix commands that are added to path if you look there. I displayed it showing the permissions and how its necessary to add executable permissions for it to work. You also have to do the same thing with the python file which notice how all the bash script is is executing a path to the python file. That bit at the end $@ just means send all of the arguments to the python file, in case you write a script that takes arguments. So now it works exactly like any other unix command
if you dont know how to add executable permissions
young sphinx
chmod +x
main olive
right
main olive
mm hm
I will add that you should probably only put a bash call for a python file in /bin/ if you are really okay with that script being executed anywhere on the fs.
But sometimes thats what you want
fickle granite
@digital haven I think homebrew gives you an option to install the gnu stuff without the leading g. But I have the vague fear that something somewhere will break, if it was depending on getting the MacOS/BSD variant, and now gets the gnu variant.
digital haven
@fickle granite i think they broke that actually, https://discourse.brew.sh/t/why-was-with-default-names-removed/4405/17
: (
i might have a look at https://github.com/fabiomaia/linuxify at some point
digital haven
🌚
i don't like that it handles some of the shell stuff 🤔
esp as bash isn't even the default in osx anymore afaik
south star
i hoped that linuxify install to ~/.local or something, but it install to the system, using homebrow, seems a bit radical indeed 😬
i would probably try it anyway if i was forced to use osx, i've had some rage about its differences with linux in the cli
digital haven
@south star yeah I think I just ran it then asked questions here after 🙃
however, man find still just returns BSD find, which is super annoying
fickle granite
I've dealt with it by defining a lot of shell functions that run the right program
# Use gnu find instead of the MacOS version.
find ()
{
local find=$(type -ap find)
local gfind=$(type -ap gfind)
if [ -n "${gfind}" ]
then
find=${gfind}
fi
$find "$@"
}
e.g.
that is slightly buggy, now that I think of it: if you change PATH after the shell defines that function, you'll still get the find that it first found
digital haven
@fickle granite if you run man find does this give you the correct man page?
digital haven
🥁
fickle granite
nope
digital haven
😩
fickle granite
I almost never use "man" so I've never noticed
digital haven
god damn osx sucks
digital haven
god damn bsd sucks
fickle granite
not sure I agree there, either. Get Colin Percival's "AMI" for ec2 and give it a whirl; it's different than what you're used to but it's pretty solid
digital haven
no i'm sure people have worked hard on it and that it has merits and reasons, but whatever the reason for this being a faff I'm not a fan of.
fickle granite
well you're allowed to not like it 🙂 I don't like it either
just not used to it, mostly
digital haven
yea - i thought it'd be more straightforward ha... I only realised that I had gfind today though. You and others suggested switching, which i wanted to do bc of the options but installs weren't working. However, they were working, it's just that I now have two versions of everything
this is nice : https://explainshell.com/
fickle granite
I have a small EC2 box running gen-you-wine Linux, for all my Linux needs.
digital haven
i haven't got into this kinda workflow, but would like to
so if you're working on something, you can just zip it over?
digital haven
if there was a seamless way, or very short way, to move between vm and system that'd be great. I only have 8gb of ram 😬
fickle granite
I have the same setup and home and at work: laptop + ec2 instance. So I have a total of four machines I use regularly
I don't use VMs much on the laptop
digital haven
but it wouldn't make a difference i guess - in that you could have that workflow
fickle granite
I have 8GB on my home laptop as well
might have more on the work one; I've never checked
digital haven
example from the other day, i had a couple of 5GB files that i needed to work with
fickle granite
yow
digital haven
if i could have just transferred easily to a vm then that would have been handy
fickle granite
that's not gonna be easy no matter how you do it
digital haven
:<
fickle granite
you can configure virtual machines to share your laptop's file system.
I've played with "virtualbox" and "vagrant", depending on the virtual OS, it's easy
can't remember which OSs are good at that though
digital haven
main thing for me would be able to be in this kind of position and easily move over to VM and carry on, if that was a thing (it must be) then I'd spend a day going through a workflow, I've not had time / justification yet though as I don't know enough i guess
fickle granite
you should try vagrant -- it makes setting up, starting, and stopping VMs quite easy
digital haven
@fickle granite cool thanks - for some reason i thought that vagrant was just local, rather than being attached to a server
probably because i've never really used it 🤦
fickle granite
Can't type. Cat on forearm.
fickle granite
Free at last, free at last, thank God almighty, I'm free at last.
digital haven
i've managed to break my ssh config 🤔
probably shouldn't have blindly run scripts
adding IgnoreUnknown AddKeysToAgent,UseKeychain seemed to fix it 🤷♂️
south star
hm, doesn't that make it auto matically accept host keys the first time you connect to them without prompting?
fickle granite
I like the idea of warning about an unknown host, but in practice it's an annoyance
I have never seen anyone do anything other than just say "yeah yeah whatever" and add it
that includes me of course
blissful sage
could someone please help me out
im freaking out
none of my scripts are working
i recently tried to install ipython
and i think i screwed some thigns up
when i try to run a script i get a Process finished with exit code 134 (interrupted by signal 6: SIGABRT)
when i open terminal i get a
@fickle granite sorry to ping you, but you've always been helpful and I think it's an emergency
digital haven
@south star i've no idea, i've no idea why it suddenly started doing it either 🙃 it worked though
blissful sage
I might have broken references to my virtual environments
digital haven
any nnn users - i thought capital R would let me rename all in a directory, it does nothing, is there something I've missed?
ctrl+r works for a single file
( I'm on OSX )
novel bramble
Is getting CompTIA certified in Linux a good choice?
tepid sphinx
^ same
main olive
I think its about time I learned what the use cases of docker are. I don't think I can even count the number of times Ive mentioned a problem on a more sysadmin oriented group I'm in that they mention docker. At this point I feel like its one of those things im going to shoot myself for not having learned sooner
south star
Docker is nice to know yeah
It's like ligher and more manageable VMs
Easier to replicate and coordinate
digital haven
Is it possible to recreate a directory structure from tree?
I would like some kind of way of being able to do tree -d on my system, then some kinda makedirs the-tree-from-before on another
south star
Rsync probably has a way to only copy the directory structure if that's what you need.
digital haven
@south star without the contents?
maybe I need to just send the directory with all the files deleted 🤔
south star
Hm, can't find an option for that in the man :|
digital haven
i mean - i can just copy teh dir, then find . -type f -delete
i just thought maybe there was a way to send a directory structure as a file that could be recreated 🤔 , maybe it's a daft thing to do though
problem is that we all have data dirs, that are in gitignore, and then they diverge lol
I think I've every dead-end of project planning so far over the past few weeks 🙃
thinking about it - rsync would probably be good if we had the data stored on a server or something, and could rsync to that regularly so that we all had something kinda similar... hmm
fickle granite
find . -type d | ssh "while read d; mkdir -vp $d; done" or something
swift heron
I would like to see who is connecting to my VM. i would like to know either what device they are using or what operating system they are using, for example iOS, windows, linux, mac... etc.
How can i achieve this?
oak shell
are you talking about ssh connections? Getting the remote IP is easy, that's logged and also reported by commands like who.
swift heron
Yes @oak shell
Yeah i know about the who command, and also the /var logs too
But anything to capture OS or device?
oak shell
I don't think that is transmitted with the ssh protocol at all... 🤔
or at least I can not find anything. I believe it's impossible.
swift heron
😦
light elbow
Looked more like user-agent kind of question
I would like to see who is connecting to my VM. i would like to know either what device they are using or what operating system they are using, for example iOS, windows, linux, mac... etc.
How can i achieve this?
@swift heron which problem are you solving?
blissful sage
still desperately need help
none of my scripts are working, I think I messed up by installing different versions of python
now i get this message every time i open terminal
-bash: /usr/local/python3.7.3/bin: No such file or directory
and my ide can't find the image
light elbow
@swift heron i was referring to the general direction of the question and the http "user-agent" header.
If i remember correctly, ssh clients do send something similar. Do ssh -vvv SERVER to see. I don't remember whether it includes the OS. Again, what's the problem you are solving?
@swift heron
I'm asking about the problem because
https://en.m.wikipedia.org/wiki/XY_problem
magic basin
i deleted history with history -c how can i retrieve them back?
main olive
you can't. not easily anyways
magic basin
am i able to stop users bypassing bashrc using paths like /bin/sh or something like ssh -t user@host bash --norc --noprofile
I rely on the bashrc to call for python programs upon startup
main olive
no, but you can change their user shell to whatever is calling your scripts (which finally drops into bash)
so, for example
swift heron
yes i can connect into bash without bashrc
main olive
you set up a script /usr/bin/scripts_and_shell, which has:
#!/bin/sh
python my_python_thing1.py
python my_python_thing2.py
exec bash
But that can be ignored too?
main olive
then you chmod +x that and chsh [username]
and change the shell of that user to the new script you made
when they ssh in, that script is automatically executed with no way to bypass it
swift heron
main olive
(unless they do -N, but that's a different story altogether)
@swift heron yes, that does -N.
i'm assuming this is out of the scope in this case
swift heron
yeah
azure flume
uhh hi everyone. I'm trying to install this https://github.com/Lcorvle/incremental-muticore-tsne in my ubuntu machine by running pip3 install . , but it doesn't seem to work. sorry for the noob question, I'm quite new to both ubuntu and pip.
main olive
what error are you getting?
azure flume
cannot find cmake
do I just clone the repository then copy the files to ~/python3.6/site-packages?
main olive
no, you still need cmake
the readme also says you need cmake
on ubuntu, you'd install it with something like sudo apt install cmake
don't try copying the files like that, that won't work
the setup.py installer does things other than that as well
azure flume
its specifically failing at thiasds block:
os.chdir('multicore_tsne/release/')
return_val = os.system('cmake -DCMAKE_C_COMPILER="C:/MinGW/bin/gcc.exe" -DCMAKE_CXX_COMPILER="C:/MinGW/bin/g++.exe" -G "MinGW Makefiles" -DCMAKE_BUILD_TYPE=RELEASE ..')
if return_val != 0:
print('cannot find cmake')
exit(-1)
looks like the install script is using hardcoded values for windows/mingw... not good
azure flume
yeah. haha. and the readme implies that it works with ubuntu, but the setup.py was obviously made for windows
i thought i might be missing something basic about the understanding of python modules/cmake and stuff
main olive
nah, it just looks like a crappy module, no offense intended
azure flume
thanks @main olive ill probably veer away from this module for now. 🙂
full sky
Hi, im on ubuntu 18.04
Hello, I've been trying to install ndiswrapper for some wireless drivers with make when this error came up: Make[1]: gcc: command not found
when i typed cat /proc/version, the printout said i have gcc 7.04
how can I fix this error
fickle granite
@full sky you need to install gcc
probably a whole C devlopement suite
@full sky what you saw in /proc/version doesn't mean that you have gcc; it means that your kernel was compiled with gcc
twin gust
ndiswrapper :(
full sky
Right thanks, and yes,kill me now
twin gust
personally I'd consider if the component lacking native driver support can be swapped for one that does
fickle granite
hosted on SourceForge, so you know it's good
twin gust
I feel your pain
full sky
yeah, ill mess around a bit more but if that doesnt work ill just change the hardware
thanks anyways 👍
blazing wolf
Anyone knows how to get python3.7 on manjaro?
how to specify package version to pacman?
main olive
manjaro is based on arch and is rolling-release, which means you'll always be on the latest
you'll have to install the older version manually
perhaps the easiest way to do that is to use pyenv @blazing wolf
blazing wolf
Im trying another method, using source code and then
main olive
yes, you can also compile from source
but pyenv might be useful in the future as well, I still recommend looking into it
blazing wolf
and running this
./configure
make
make test
sudo make install```okay
this method is taking ages
main olive
yeah, compiling from source takes a while depending on the hardware
blazing wolf
yea I got a crappy one haha
blazing wolf
it took 8min to run make test
main olive
hi
how do I pass yes to a sh script
right now i'm doing
git clone <gitlink> && myscript.sh
but I want to pass yes to the shell script.. because it prompts to confirm
dreamy wing
is the script yours ?
sage solar
yes | myscript.sh @main olive
yes is a command that just keeps printing a character followed by a new line (default character is y)
full sky
can anyone explain this?
full sky
ah right
dreamy wing
/usr/bin is your binary directory and not where your libraries are stored
dreamy wing
/usr/lib/python*
dreamy wing
full sky
hi again, ive been trying to install numpy with pip but it keeps installing to the wrong location, could anyone help?
actually scratch that
ive gotten it to download to the right location, but it only installs the numpy-dist-info folder
any idea why this is?
ah, so now I've found that the python path is configured as python 3.6 for me
how do I change this?
tame crystal
how can I use the pwd in another command?
I need to pass it as an argument to a unix command
main olive
sharp cove
./my-prog.sh $(pwd)
@tame crystal
that's called a subshell
subshells are executed first, and replaced with the output of the subshell
so, for example, if i did
ping $(echo "example.com")
that would expand to
ping example.com
```Thanks
proper bison
Can i use multiple After= statements in a ubuntu service?
final trench
Good day! I have a problem installing Ubuntu on my laptop. The distributor I bought it from, told me that this model of notebook does not support OS other than Windows 10...
and I am not sure in that. But here comes the problem. When I make a rufus usb with ubuntu 18.04 and plug in the usb, I boot the Ubuntu successfuly, however I get an error in like 1 minute and the notebook shutdowns... When I try to install it, this happens again. I used to research a bit and I found that the error has something to do with the GPU
This is it for now... Down below is more technical info.
When I boot the notebook with the linux usb it shutdowns in like 1 minute, showing that error - nouveau, failed to do smth...
dreamy wing
Probably missing firmware. Would be helpful if you gave the full error and dmesg / lspci
final trench
Ok I am going to send a photo but it is not with the best quality :/
There is also an option next to test boot - install for manifacturers
can it be the right one ?
south star
nouveau is the open source driver for nvidia
you could try installing with the alternate image
and once it's installed, you should be able to install the nvidia driver from command line, if it's not already
@final trench
ah, or with the normal installer, maybe adding the "nomodeset" option to the kernel (in the menu where you can chose "install ubuntu" or "try ubuntu without installing" there is an option to edit the kernel line, on which you can add this option)
final trench
Yes thank you nomodeset worked out for me
full oriole
Get this error when using ftp. How do I log into the server otherwise?
fiery badger
220 GNU FTP server ready.
Name (ftp.gnu.org:willgoldby): anonymous
230-NOTICE (Updated October 13 2017):
230-
230-Because of security concerns with plaintext protocols, we still
230-intend to disable the FTP protocol for downloads on this server
230-(downloads would still be available over HTTP and HTTPS), but we
230-will not be doing it on November 1, 2017, as previously announced
230-here. We will be sharing our reasons and offering a chance to
230-comment on this issue soon; watch this space for details.
230-
230-If you maintain scripts used to access ftp.gnu.org over FTP,
230-we strongly encourage you to change them to use HTTPS instead.
230-
230----
230-
230-Due to U.S. Export Regulations, all cryptographic software on this
230-site is subject to the following legal notice:
230-
230- This site includes publicly available encryption source code
230- which, together with object code resulting from the compiling of
230- publicly available source code, may be exported from the United
230- States under License Exception "TSU" pursuant to 15 C.F.R. Section
230- 740.13(e).
230-
230-This legal notice applies to cryptographic software only. Please see
230-the Bureau of Industry and Security (www.bxa.doc.gov) for more
230-information about current U.S. regulations.
230 Login successful.
Remote system type is UNIX.
Using binary mode to transfer files.
ftp> cd gnu/diction
250 Directory successfully changed.
Never mind. Will use wget. Thanks
main olive
@main olive maybe a simple .sh script which cd's to the correct directory first (you can get the path of the python file with realpath -P file.py) and then calls the python file.
vagrant fern
is installing unattended-upgrades on debian enough to have it automatically upgrade packages? from reading the configuration files supplied it looks like it, and apt-daily-upgrade.timer is running .. or is there something else that needs to be added?
main olive
@vagrant fern I'm pretty sure that's it
I don't remember doing anything else besides that to get unattended upgrades working
humble onyx
Hey guys I have a question. I am the owner of a file but I am unable to change the permissions of the file. How is that possible?
livid verge
sorry this was false information - you actually can change it
main olive
@humble onyx what filesystem? ext4?
humble onyx
Yes
I connected to a remote server and tried to change the permission on a file. It said the the operation is not permitted
Owner is the user that I logged in
humble onyx
bandit21@bandit:/etc/cron.d$ ls -l /etc/bandit_pass/bandit21
-r-------- 1 bandit21 bandit21 33 Oct 16 2018 /etc/bandit_pass/bandit21
bandit21@bandit:/etc/cron.d$ chmod +w /etc/bandit_pass/bandit21
chmod: changing permissions of '/etc/bandit_pass/bandit21': Operation not permitted
main olive
is the root filesystem (or wherever /etc/ is) mounted as rw and not ro?
or rather, can you modify the file at all?
humble onyx
nope
shy yokeBOT
main olive
makes it a bit easier to read
humble onyx
sysfs on /sys type sysfs (rw,nosuid,nodev,noexec,relatime)
proc on /proc type proc (rw,nosuid,nodev,noexec,relatime,hidepid=2)
udev on /dev type devtmpfs (rw,nosuid,relatime,size=2010476k,nr_inodes=502619,mode=755)
devpts on /dev/pts type devpts (rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=000)
tmpfs on /run type tmpfs (rw,nosuid,noexec,relatime,size=403068k,mode=755)
/dev/mapper/vg0-root on / type ext4 (rw,relatime,errors=remount-ro)
tmpfs on /run/lock type tmpfs (rw,nosuid,nodev,noexec,relatime,size=5120k)
pstore on /sys/fs/pstore type pstore (rw,relatime)
configfs on /sys/kernel/config type configfs (rw,relatime)
tmpfs on /run/shm type tmpfs (rw,nosuid,nodev,noexec,relatime,size=806120k)
/dev/mapper/vg0-boot on /boot type ext4 (rw,relatime)
/dev/mapper/vg0-tmp on /tmp type ext4 (rw,relatime)
none on /cgroup2 type cgroup2 (rw,relatime)
somebody told me that i bit was set
i have no idea what i bit means
file could be +i for immutable
humble onyx
well I did man chattr
written to the file. Only the superuser or a process possessing the CAP_LINUX_IMMUTABLE capability can set or clear this attribute.```yep
humble onyx
----i---------e---- /etc/bandit_pass/bandit21
chattr: Operation not permitted while setting flags on /etc/bandit_pass/bandit21```like I said, may need sudo
humble onyx
so I gotta be in sudo group to perform chattr on any file ?
main olive
you just need to have root permissions or be able to execute sudo
humble onyx
I got it thanks 🙂
vagrant fern
thanks @main olive
left ingot
New Unbuntu 18.04 user here, anyone know of a solution regarding low wi-fi signal? I dual-boot windows 10 and see a stark difference in performance.
I've come across a couple solutions, with this one being the most promising. I didn't notice any difference. Is this a catch-all solution? Or is it hardware specific? Am I barking up the wrong tree here? https://askubuntu.com/questions/1058379/wifi-signal-is-weak-in-ubuntu-18-04-with-rtl8723be
main olive
@left ingot could you please post the output of sudo lspci -vmm (assuming your WiFi card is internal and not, say, an USB dongle)
main olive
Typically I can find the cause of 502 responses on my Ubuntu webserver...
But this time I'm fairly stumped. Where in logs might such an error exist? I wasnt sure if this is more suited for web dev or here because its more sysadmin related.
main olive
2020/03/05 15:35:49 [error] 1997#1997: *67 connect() to unix:/run/gunicorn.sock failed (111: Connection refused) while connecting to upstream, client: 192.168.1.152, server: wsi, request: "GET / HTTP/1.1", upstream: "http://unix:/run/gunicorn.sock:/", host: "192.168.1.197" I found it. Not very helpful though. I know that its not the configuration of gunicorn or nginx. Its something in my Django app that Im just gonna have to have to revert recent changes to find.
main olive
does the webserver's account (perhaps www-data) have rw access to /run/gunicorn.sock @main olive
main olive
Good day, I did something silly and delete something from my Linux, I'm running mxlinux that runs on debian. I can't open anything that uses authentication for root, that pop up does not come any longer. Only way I can open in term or through terminal with sudo. Anyone know what I delete? What's the name of the package that pops up asking for root password? Any help would be appreciated
sonic garden
@main olive is it perhaps the gksudo package?
main olive
Should it? I'm fairly new to running this server.
-rw-r--r-- 1 root root 341 Feb 7 19:10 gunicorn.service
-rw-r--r-- 1 root root 112 Feb 7 19:03 gunicorn.socket
@main olive
I did manage to get it working again though. It will sometimes thrwo that 502 response when I have some type of syntax error in Django. I usually just go back and correct it but this time I couldnt find it.
It probably has something to do with not being able to proxy pass.
I probably should be developing on a development test server and not a pretend production server.
main olive
the web server needs to be able to send commands to gunicorn - ie, it needs read and write permissions for the socket
limpid sapphire
Not python related but if anyone knows I might aswell ask.
default acl for a directory : default:user:my_user:rwx
rsnapshot takes a backup of some files everyday, acl for the files : user:my_user:rwx #effective:---
What's the point of setting acl if the effectives permissions don't reflect it ? ...
main olive
i'm guessing they are masked? @limpid sapphire
limpid sapphire
mask::r-x and default😷:rwx for the directory but mask::--- for the file yea
I just don't understand where that mask::--- on the file comes from
dark knoll
[ 5291.078522] ata3: hard resetting link
[ 5296.966378] ata3: link is slow to respond, please be patient (ready=0)
[ 5301.126293] ata3: SRST failed (errno=-16)
[ 5301.126300] ata3: hard resetting link
[ 5302.134319] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[ 5302.141865] ata3.00: configured for UDMA/33
[ 5302.141899] ata3: EH complete
i found this error in dmesg is this a sign of hard drive failure
main olive
it might be, or it might not be. there's no way to know without doing proper diagnostics on the drive
quaint mirage
you can check the health status of your drive using S.M.A.R.T. information
sudo smartctl -H /dev/sda
or whatever drive you want to target
oh, you might need to check if S.M.A.R.T. is supported / enabled on the drive (should be) with sudo smartctl -i /dev/sda
there are also tests you can run using smartctl, check the manual
humble onyx
I have a question about manual page syntax.
git [--version] [--help] [-C <path>] [-p|--paginate|--no-pager] [--no-replace-objects] [--bare] [--git-dir=<path>] [--work-tree=<path>] [--namespace=<name>] [--super-prefix=<path>] <command> [<args>] here there are <command> and [<args>] I am told that angle brackets indicate that the enclosed element is mantadory and as for the square brackets, the enclosed element is optional. What is meant by [<args>] then?
limpid sapphire
Maybe that depending on the command args may or may not be mandatory @humble onyx
main olive
@humble onyx angle brackets do not indicate that the enclosed element is mandatory. Angle brackets instead indicate that it is a "variable" - ie, you should replace the <thing> with something meaningful. Elements without angle brackets mean they should be input as-is
for example
in your help message
[--git-dir=<path>]
the square brackets indicate that the entire thing is optional. The angle brackets around path indicate that it should be replaced with something meaningful - ie, the actual path to the git directory
there are no angle brackets around --git-dir=, which means it should be input as-is
so [<args>] means that it takes optional arguments after <command> - however, those arguments are variable and should be replaced with whatever you're trying to do
i hope this makes sense
humble onyx
@main olive Thanks It makes a lot of sense
candid igloo
diff path/to/file1.txt path/to/file2.txt
[nothing]
diff -y path/to/file1.txt path/to/file2.txt
bunch of output... but as far as I can tell I don't see any differences between the left side and right side of this side by side
meanwhile I'm trying to port over my python web scraping with selenium over to linux and in windows there's a powershell commandlet that pretends to be diff but all my output is straight up empty and I don't know if there's a side-by-side/-y option in the powershell (hence one of the reasons I'm trying to run my code in linux)
my head is spinning. how on earth could the layout of the output change whether or not there's any output at all? that makes no blasted sense.
is it possible there's whitespace changing? spaces, newlines, and tabs could be invisible...wait, nope, I've opened my two files in notepad++ and they look the same and that thing shows me spaces and tabs. OK yup I'm officially lost
does diff have a character limit per line?
warped nimbus
Does sound like a whitespace issue
Have you checked it's not a line ending difference? e.g. LF vs CRLF
diff has some flags for ignoring whitespace changes
candid igloo
Oh! I will read on that. 🤞 fingers crossed that's it
sha256sum a b
same. something is FUNKY with hashlib because EVERYWHERE else things check out
candid igloo
but I do the same in powershell on windows and the hashes are different.
humble onyx
hey guys I have a question how does $0 by itself get executed? I understand that in a script $0 refers to the script name. I tried this in vim :!$0 a new bash process took up the whole screen. It just doesn't make sense
oak shell
this seems vim specific. tbh, I only know how to quit it, but I will just assume that :!whatever will run that command in a shell
then for that new shell process, $0 would be the shell executable itself, i.e. bash.
sturdy fractal
does this apply to python 3.8? didn't seem to work when i tried https://stackoverflow.com/questions/26261080/compile-python-3-4-with-sqlite3
sturdy fractal
tried again, it worked
fickle granite
good
I have a big pile of stuff that I need whenever I build python from source, but I can never remember it all
charred lake
hello guys i wonder if i can find someone that can explain a little bit how can i use docker offline
atomic ridge
Hey so i recently reinstalled Opensuse Leap 15.1 and kept my home directory. Re-downloaded the Jetbrains Toolbox and told it to reinstall Pycharm and Intellij, but even after that the .desktop files dont work. I can start both just fine with the Toolbox and with the .sh directly, but not via the desktop icon. It bounces for a short time (KDE) but then it stops and nothing happens. This is the .desktop:
[Desktop Entry]
Categories=Development;IDE;
Comment=Python IDE for professional developers
Exec="/home/technofab/.local/share/JetBrains/Toolbox/apps/PyCharm-C/ch-0/193.6494.30/bin/pycharm.sh" %f
GenericName=
Icon=/home/technofab/.local/share/JetBrains/Toolbox/apps/PyCharm-C/ch-0/.icon.svg
MimeType=
Name=PyCharm Community
Path=
StartupNotify=true
StartupWMClass=jetbrains-pycharm-ce
Terminal=false
TerminalOptions=
Type=Application
Version=1.0
X-DBUS-ServiceName=
X-DBUS-StartupType=
X-KDE-SubstituteUID=false
X-KDE-Username=
when i run this in the console it works fine: "/home/technofab/.local/share/JetBrains/Toolbox/apps/PyCharm-C/ch-0/193.6494.30/bin/pycharm.sh" %f
quaint mirage
tried removing %f?
primal snow
Howdy folks. I'm on Deb 10 (raspbian) and I seem to have installed python, python3, pip, pip3, and venv without error, but virtualenv foo returns
ERROR:root:failed to read config file /home/pi/.config/virtualenv/virtualenv.ini because OSError(13, 'Permission denied')
and /.config/ does not even seem to contain such a file as virtualenv.ini. Any hints?
main olive
tried making the file in that location?
mkdir -p ~/.config/virtualenv
touch ~/.config/virtualenv/virtualenv.ini
@primal snow
main olive
hey in my ubuntu 18.04.3 m getting this:
@main olive can you post the output of echo $PATH for me?
ubuntu generally doesn't ship without python as far as I know, so unless you removed it yourself it's just missing from your path
also you appear to be using /bin/sh?
its the PATH
odd, can you post the output of ls -l /usr/bin/python*?
@main olive
lrwxrwxrwx 1 root root 9 Oct 25 2018 /usr/bin/python3 -> python3.6
lrwxrwxrwx 1 root root 16 Oct 25 2018 /usr/bin/python3-config -> python3.6-config
-rwxr-xr-x 1 root root 1018 Oct 28 2017 /usr/bin/python3-jsondiff
-rwxr-xr-x 1 root root 3661 Oct 28 2017 /usr/bin/python3-jsonpatch
-rwxr-xr-x 1 root root 1342 May 1 2016 /usr/bin/python3-jsonpointer
-rwxr-xr-x 1 root root 398 Nov 15 2017 /usr/bin/python3-jsonschema
-rwxr-xr-x 2 root root 4526456 Nov 7 10:44 /usr/bin/python3.6
lrwxrwxrwx 1 root root 33 Nov 7 10:44 /usr/bin/python3.6-config -> x86_64-linux-gnu-python3.6-config
-rwxr-xr-x 2 root root 4526456 Nov 7 10:44 /usr/bin/python3.6m
lrwxrwxrwx 1 root root 34 Nov 7 10:44 /usr/bin/python3.6m-config -> x86_64-linux-gnu-python3.6m-config
lrwxrwxrwx 1 root root 10 Oct 25 2018 /usr/bin/python3m -> python3.6m
lrwxrwxrwx 1 root root 17 Oct 25 2018 /usr/bin/python3m-config -> python3.6m-config
so it's there, but there's no symlink from /usr/bin/python to /usr/bin/python3
workaround: use python3 to call scripts instead of just python
solution: create the symlink manually. Generally not recommended unless absolutely necessary
ya i called it python3
and it works, yes?
no same
error
wat
can i see how you're calling the script?
sure on emin
so something inside the script is probably calling python via system calls, which is... very bad
can you do sudo ln -s /usr/bin/python3 /usr/bin/python and try again?
nthing came
yes, that's normal
now try the script again
ty now wrking....
primal snow
@main olive yea I got it straightened out with chown, ty
main olive
alright, cool
magic basin
How do i force a user to run a script without using the bashrc
main olive
you change the user's shell to the script and add exec bash to the end of it
is the most surefire way
@magic basin
magic basin
But they can login without a shell @main olive
magic basin
I told you
To get them to run a script when logging in?
Your solution doesn't fix it
main olive
"logging in"? that's a vague concept
you can't login via tty without invoking the shell
ssh-ing in isn't the same as "logging in"
magic basin
connecting, then.
main olive
connecting to what?
magic basin
you're aware of -F right?
-F configfile
Specifies an alternative per-user configuration file. If a
configuration file is given on the command line, the system-wide
configuration file (/etc/ssh/ssh_config) will be ignored. The default
for the per-user configuration file is ~/.ssh/config.
what?
magic basin
-F /dev/null
magic basin
that's the sshd config
main olive
please read the text again
main olive
mainly where it mentions /etc/ssh/ssh_config and ~/.ssh/config. Those are user configs
the daemon config is /etc/ssh/sshd_config
notice the d
if you could bypass the daemon config, that would be a HUGE security issue
magic basin
So it is not possible?
main olive
what's not possible?
the thing you're asking for? it's very much possible, reference: the link I posted
bypassing the ssh server's daemon config? no, that's not possible.
magic basin
Maybe not the whole daemon config but specific variables in the config
Can be overrun by the client
main olive
dude, no they can't
a connecting client can not change or bypass config values set in the daemon config file
a LOT of options would be meaningless otherwise, such as the chroot options, or anything starting with Allow
can you please show me how you'd bypass ForceCommand?
magic basin
But there is for passwordless authentication
main olive
yes, those are modifying user config variables
not the daemon config variables
ForceCommand is a daemon option
main olive
so i return to my previous question
how would you bypass ForceCommand?
i'm curious what your thought process is
and how i can help
magic basin
Idk?
main olive
because currently I'm not sure what's wrong with the solution I've given you
magic basin
Well for starters exec bash isn't even recognised
main olive
exec bash was for the shell-based solution, not the sshd config one
magic basin
So what is your solution?
magic basin
Doesn't stop them deleting the file with ftp :p
main olive
if your users can delete /etc/ssh/sshd_config, then there's something seriously wrong with your system
main olive
the script doesn't have to be modifyable by the users
you can place it somewhere like /usr/local/bin and make sure that only root owns it via:
chown root:root /usr/local/bin/myscript.sh
chmod 644 /usr/local/bin/myscript.sh
uh
most binaries/scripts on your system are owned by root
do ls -l /usr/bin/ping for example
you'll see that it's owned by root (root root - user and group)
you'll also see rwxr-xr-x
you can split that into rwx, r-x, and r-x - owner, group, and everyone else
meaning the owner can __r__ead, __w__rite, and e__x__ecute it, but the owner group and everyone else can only read and execute it
magic basin
Yeah you're right this solution does actually work
main olive
👍 no problem dog
main olive
please do ping me if you do
craggy hamlet
So trying to install OMV and I get this on boot
http://imgur.com/a/5QBNJFo
twin gust
the "wrong diagnostic page" thing? from what's out there on google for that message it seems to be related to booting from external usb drives. https://www.reddit.com/r/linux4noobs/comments/44i389/getting_info_from_a_usb_port/ ... not sure if that is even the same root cause for that user, but if it's USB, trying other ports is always a good idea.
craggy hamlet
yeah its installed directly to the SSD in the pc
twin gust
I have a NUC running Debian which seems to be incapable of keeping a drive reliably connected on half its ports.
craggy hamlet
yeah mines similar to a NUC, its a lenovo micro pc
once it gets to that blank screen though it stays there
twin gust
probing into it i found other people complaining about the same hardware, and said front vs. back ports are ever so slightly different in how much power is delivered.
moving the drive to the front USB3 port ended my problems.
as for the SSD, no idea ...
probably out of scope for this server? i'd be talking to my distro support in a case like this.
as this is about as far from a python problem as you can get :)
main olive
can you not just do ForceCommand "command1; command2"
spice relic
when i type systemctl edit httpd.service it's a black vim page. Is it supposed to be empty?
blank page*
I thought it's supposed to have sections like [Unit]
twin gust
it's only overrides unless you say --full
(though often overridding is all you'll need, so no need to do --full)
wheat ridge
Got sent here...
How do I go about replacing an old version of python that's within a virtual environment with a new one? I want to switch to 3.7 from 2.7. I thought it WAS 3.7 but apparently it's not.
The virtual environment is on a raspberry pi, raspbian distro, for a django webserver using apache2 and mod-wsgi. Or, that's the plan when I fix this python problem...
I had an error on a super basic Django thing, and was like... What? Which python and python --version told me it was using 2.7 X_X
warped nimbus
The easiest way is to just re-create the venv.
If you want to save the dependencies you could use pip list to export them to a file then install them back with pip install -r requirements.txt once you have the new venv made.
Well unfortunately both pip list and pip freeze output version info. The latter would at least result in a valid requirements.txt file
You probably will end up with some outdated packages
If you don't have a lot of packages installed to begin with it's probably easier to just install them again manually
digital haven
-> % vim --version | grep python
+cmdline_hist +langmap -python +visual
+cmdline_info +libcall +python3 +visualextra
does this mean that I have python 3 support built into vim but not python? i'm trying to setup vim for python stuff at the mo
digital haven
hrm, I can't run
:python import sys; print(sys.version)
from within vim, it gives error:
E319: Sorry, the command is not available in this version
just trying to see what's up atm
(can others run this?)
main olive
use :python3 instead. also see :h python-2-and-3
digital haven
use :python3 instead
🤦
note - if anyone has any python + vim suggestions i'm all ears
currently going through this : https://realpython.com/vim-and-python-a-match-made-in-heaven/#macos-os-x
main olive
neovim has had better integration with python for me personally
magic basin
Hi, i'm trying to ssh to my remote server automatically from my client desktop, using a .bat batch file, but it is not working...
main olive
this seems more windows related, maybe #491524019825278977?
magic basin
Okay
digital haven
@main olive fair, i was getting that impression over the last hour or so and have started setting up nvim instead
currently going through this: https://www.vimfromscratch.com/articles/vim-for-python/
though it seems they use some stuff that's nvim only and some stuff that's vim only, so idk if they've just smashed together random links there 🤔
digital haven
is there a quick way to select 'range' from this drop down in vim? I'm finding that I have to press the down arrow, which seems a bit of a pain. I would have expected either ctrl+j to work or to be able to tab through them, something like that
magic basin
@main olive Hey
main olive
hey yourself
main olive
the prompt "login as" is only displayed by putty itself
the ssh server expects the username as part of the authentication process itself
as such, the username is prompted by putty before connecting
main olive
the password for the authenticating user?
you can set that with passwd once you're in the server as the user
(or passwd [username] if you're root)
main olive
that's entirely a putty thing
or whatever the connecting client is
openssh or whatever
magic basin
Sad
main olive
my honest suggestion? use ssh keys
magic basin
I seen a guy with custom username/password prompts, it was the same on every client
and different colour
but he was on telnet
not ssh
main olive
yeah, telnet behaves differently
magic basin
yeah i use keys, they are cool
main olive
telnet's "password prompts" are controlled by the server
so the server can set them to whatever it wants
magic basin
do you know what i can do to prevent users sharing their ssh keys and sharing accounts?
so that they use their own
main olive
only reasonable thing i can think of is IP whitelisting
but that creates hurdles on its own
main olive
wait til you learn vim ;)
warped nimbus
What's this sort of syntax called in bash? command <(another_cmd)?
runic roost
| pipes (pipelines) the standard output (stdout) of one command into the standard input of another one. Note that stderr still goes into its default destination, whatever that happen to be.
warped nimbus
What would be an equivalent way of doing that if I wanted the output of another_command to be in another variable?
I'm trying to pass some stuff to comm
So it's important that delimiters are preserved properly, whatever properly even means
x="$()"
y="$()"
comm "${x}" "${y}"
was my naïve attempt
Ah I just realise it expects files, so I need to pipe them instead somehow?
runic roost
Something like this maybe:
$ ls | grep test > text.txt
If you want to list all files in a folder starting with the name called test and the output of that will be put into a file called test.txt. Is that something you want to achieve or do I have no idea what you are talking about? If that last one is the case than I’m deeply sorry 😊
warped nimbus
Oh did you just paste the documentation for pipe?
I was asking about <() not |
But maybe you were onto something anyway, I probably do need to pipe somehow to achieve what I want
But comm wants files and I don't have files - I have outputs from two commands.
And I'd rather avoid writing to temporary files.
<() works but it's just kind of ugly, even if I break up the command into multiple lines. I just thought it'd be nicer to have them in separate variables but it seems to complicate things.
runic roost
I think I understand what you are trying to achieve but I really have no clue how to help you with that one. I’m sorry.
oak shell
<() is called process substitution @warped nimbus https://wiki.bash-hackers.org/syntax/expansion/proc_subst - it creates a temporary FIFO file descriptor (also known as named pipes) and substitutes itself with that fifo's path - so your command ends up getting a filename as argument, but it's actually getting the output of the command fed into it.
warped nimbus
Thanks
wheat ridge
@steel verge Getting a 'sys is not defined' error when I try to run the django server via manage.py runserver
Is the issue an issue of permissions, I wonder? I have no idea, I have no idea how to tell, and i have no idea how to fix it if that's the case, or how to know WHEN it's fixed
This is actually a new one, django has not broken like this before
Error # 100 for this project lmao... Oh god what am i doing
steel verge
Now, the files are in a home, did you add that user to wsgidaemonprocess?
WSGIDaemonProcess frontend.wsgi user=wsgi group=wsgi threads=1 like this for example
wheat ridge
Wait, is sys something that needs to be imported by itself? I thought it was a part of os...
wheat ridge
ah heck, sec. That might be my problem
k, server is up, sec...
gonna refresh pg and see what the error is, if any
same, 500 server, checking apache2 error log...
Same ol error as last time
I don't remember where to edit/check WSGIDaemonProcess and that config stuff
uhm... checking notes lool
wheat ridge
Ok, so the .conf files are in /etc/apache2/sites-available, and there is... 000-default,conf, default-ssl.conf, and webserver.conf.
vhost?
steel verge
yeah, the configuration of the virtual host
wheat ridge
I'm sorry, I don't know what that means...
steel verge
It is either one of those files you just mentioned
wheat ridge
I'll show you both.
steel verge
where you defined the django server
steel verge
okay, so one of those must go first
wheat ridge
This is the part of the whole problem/series of problems that I am the most shaky on, I don't know wtf to do with these two conf files. I don't know which one takes precedence, I don't know what they're doing, I don't know what needs to lead to where or if something's broken I don't know what needs to be fixed
steel verge
Each one of those is a so called virtual host
a website if you will
so you got 2 conflicting configurations
are there 2 in sites-enabled as well?
wheat ridge
Can I get rid of 000-default.conf altogether?
steel verge
/etc/apache2/sites-enabled
steel verge
yeah
wheat ridge
Could not remove, permission denied. Sudo it?
steel verge
yeah always sudo in etc
wheat ridge
disabled. Reload as it says?
steel verge
yep
wheat ridge
done
steel verge
ok, now we use the other one as a basis for the web
wheat ridge
reloaded site, URL not found, tail'd error log, seeing this...
oh
Ok, so we're not done.
steel verge
what error did you get?
steel verge
So, the issue right now is that you have the files in your home of the user pi
but apache isn't that user, it probably cannot read the files and not the venv
What you must know, is that Linux has a very strict user policy and very strict rights
it isn't like Windows
well, user desktop Windows
This isn't recommended, but it will do. Addthis: user=pi group=users to the line with WSGIDaemonProcess
wheat ridge
The one in /sites-available/webserver.conf ?
steel verge
you can also edit it in sites-enabled, but yeah
It's the same file, the one in sites-enabled just links back to the other one
like a shortcut in Windows
wheat ridge
Is there a particular place in the line I should add those?
wheat ridge
steel verge
4th from the bottom
steel verge
webserver needs to be in front