tools-and-devops

[rapid sparrow]

they should teach themselves how to code stuff like that

from simple to more obscure encrypting algorithms

on a more low level

[wet dune]

Yeah. Tht machine will have skme rev shells n privilege escelation stuff. The double base64 is just to annoy em tbh lmao. Imagine the code obfuscated. They will have to figure how to decrypt the rsa key. I tried to make it as simple as possible

May i ask wdym

[rapid sparrow]

ergh. The code just has too many times repeated base64 encoding decoding. 9 times. It should happen only two times in the code.
too much of repeated open file/read code as well.

And most importantly the code in part two is just terrific student level interface, Could u use normal tool to make CLI interface? Like inbuilt argparse, or preferably click which provides the best CLI interface experience, with automated most comfortable... standards similar to other CLI tools in Linux, and inbuilt help support to navigate which commands exist and which parameters they take
Then usage of program will be up to Linux CLI comfortable standards, which will make the code really easy automatable for any tool

Like program.py key generate

[wet dune]

Ofc i could do tht.

That code is meant to be for ppl with almost 0 exp

Thts why there are so many file opennings and calls to those base64 methods

To not make the user jump back n forth

Now, applying argparse or any cli lib isnt hard so i can do tht

[rapid sparrow]

click is the best, but requires third party library to install, really easy to eye due to decorators usage, quite pleasant code in result
argparse is default inbuilt one, not requiring any library, that's the advantage it has.

[wet dune]

Ik used argparse lots of times before

[rapid sparrow]

i like click because it allows easily nested commands

gathered in groups

[wet dune]

Never used it. Remember using another kne but idk its name

[rapid sparrow]

click is part of flask package, i think

[wet dune]

Got it installed for an example iss tracker project

I will remakecit using 2 methods. Encrypt / decrypt

Functions*

[brazen forge]

it's an indepedent package. flask uses it for its cli

[arctic bridge]

Anyone know where I could ask some docker/jupyter questions?

[tawny temple]

Here

[arctic bridge]

I am setting up a docker container with the JupyterLabs and trying to set up an accessible drive, using windows 11 & WSL

I can’t seem to get anything to work and suspects it’s because I am a moron

[tawny temple]

By accessible drive, do you mean that you want to set up a mount/volume for the container?

[arctic bridge]

Yeah and that be accessible from the desktop

I can navigate to the WSL volume but no data related to my container there either

[tawny temple]

In other words, a directory that can be accessed both on your host machine (Windows) and inside the container?

[arctic bridge]

Yes

[tawny temple]

Then what you need is a bind mount. You can use the --mount option with docker run to specify the mount. Is that what you've been trying?

Or --volume, either way. Just different syntax

[arctic bridge]

Ok I don’t think that’s what I was doing, I will look into that

[tawny temple]

Like docker run --volume C:/local/path:/path/in/container imagename

https://docs.docker.com/storage/bind-mounts/#choose-the--v-or---mount-flag

[arctic bridge]

Thanks mark

[olive pine]

does anyone know how to use the tweepy v2 api streaming class?

[wise cave]

hi guys. how do i change the package install to a custom enviroment on mac os? So that everytime I !pip install package it goes to the custom enviro

[left girder]

I like to use pyenv to manage environments on my mac

but there's other options too, and people seem to be opinionated about it 🙂

there was a recent talk at pycon on pyenv https://www.youtube.com/watch?v=-YEUFGFHWgQ Talk - Calvin Hendryx-Parker: Bootstrapping Your Local Python Environment @wise cave

[uneven python]

Hi everyone!
I was just offered a promotion from Support into a DevOps engineer role. Anyone have any solid resources for getting started? I'd like to have an edge once I begin my training for the new role.

[heavy knot]

Does anyone know any online sprite cutting tools like Alfred’s sprite sheet unpacker?

[heavy knot]

invalid syntax error help me ?

[thorny shell]

@heavy knot ask in a help channel, not here #❓｜how-to-get-help

[heavy knot]

shut and up ?

[graceful light]

Is there an existing tool to look at older Python files (like ones for 3.4) and find what they need to have changed in them to work on newer Python versions like 3.9?
Not looking specifically for those version stuff, but a tool that tells you what you need to fix when going from version 3.x to 3.y

[brazen forge]

!pypi pyupgrade

[rancid schooner]

pyupgrade v2.34.0

A tool to automatically upgrade syntax for newer versions.

[past onyx]

hi all, i have a question regarding Github. On my page, it shows that I have 16 repos but I actually only have 7, where did the rest of them come from?

[open forge]

Is there an issue with pip rn? I'm getting

WARNING: There was an error checking the latest version of pip.

In my terminal, but that's all I'm getting. I have an internet connection and I can install packages just fine, and pip is already up to date, so I can't think of what else would be causing this?
I'm on windows on py3.10.4 if that helps

[brazen forge]

how are you running pip? py -m pip?

[brazen forge]

I only see 7

[open forge]

pip and py -m pip 🤷

[rancid schooner]

Hey @open forge!

You either uploaded a .txt file or entered a message that was too long. Please use our paste bin instead.

[open forge]

https://paste.pythondiscord.com/ikohawayes this is pip debug if its worth anything

[past onyx]

on your side? It's 16 on mine...

[brazen forge]

debug output looks fine. you could try increasing the verbosity of pip install to see if that gives anything meaningful

[open forge]

alright, one sec

[brazen forge]

could be private repos?

[open forge]

When I increased the verbosity it managed to upgrade pip what the hell

one sec let me see if the issue is still there

[past onyx]

no i dont have private repos

[open forge]

oh I'm a dumbass I forgot to include -U when I did py -m pip install -U pip but I did include it when running pip install -U pip

My bad 🙏

[brazen forge]

have you contributed to any repos?

other than yours I mean

[open forge]

also since when did pip use and vendor rich?

[past onyx]

contributing to repos counts as 1 repository towards my page?

[brazen forge]

apparently. mine shows 14 but I've only got 7

[brazen forge]

https://github.com/pypa/pip/commit/7510fb04b6d12487a7de2d49cad8cb29366a6ae2

[open forge]

ah sweet

[past onyx]

so...it's not a glitch or?

[brazen forge]

a cursory search on the webs indicates it's a bug? not sure though

[past onyx]

hmmm...alright well as long as everything's there, it should be fine, though I'll try to find out about it still

Thanks so much for the help too 😄

[wet dune]

lmfao bro you dont know basic python syntax and you ask help for that?

[topaz aspen]

Is it possible to use pre-commit in a nox session without installing the tooling separately ?

Looking at this for example : https://github.com/cjolowicz/cookiecutter-hypermodern-python/blob/main/{{cookiecutter.project_name}}/noxfile.py#L113-L134

the linters etc used in pre-commit are installed with session.install - but this is typically something that pre-commit does itself if it's not been setup, so I'm wondering why this is necessary, seems as though it's duplicating what's already in the pre-commit config.

ping me if you respond please

ah maybe it's because this is set to local only https://github.com/cjolowicz/cookiecutter-hypermodern-python/blob/main/{{cookiecutter.project_name}}/.pre-commit-config.yaml#L2 , hrm, tho i'm not sure why that would be done

[rancid schooner]

%7B%7Bcookiecutter.project_name%7D%7D/.pre-commit-config.yaml line 2

- repo: local```

[tawny temple]

Indeed if it's local then pre-commit won't use its own venv for that hook. The only reasons that come to mind are:

1. They prefer to manage dependencies here. Maybe consolidating them into a single location.
2. The Python version may affect the versions that get installed. Not sure how flexible pre-commit config is regarding dynamically specifying a Python version.

Number 2 indeed seems to be the explanation. I skimmed through the article that repo is based on

There is a problem though: The sample configuration pins Black to a specific version, and so does Poetry’s lock file. This setup requires you to keep the versions aligned manually, and may result in failed checks when the environments managed by pre-commit, Poetry, and Nox drift apart.

[rancid schooner]

.pre-commit-config.yaml lines 1 to 2

default_language_version:
    python: python3.10```

[tawny temple]

But that's not dynamic

[dull kiln]

I'm usijng pyinstaller with a .spec file
and i got these hiddenimports

a = Analysis(['entry.pyw'],
             pathex=['C:\\Users\\ME\\PycharmProjects\\pythonProject\\projectFolder'],
             binaries=[],
             datas=[],
             hiddenimports=["mss", "numpy", "win32gui", "cv2", "win32con", "win32ui", "winsound", "tkinter", "requests"],

but i'm getting ModuleNotFoundError: No module named 'requests'
and i did pyinstaller --clean entry.spec

what am i missing?

[graceful light]

So while using black, isort, and flake8 with a project containing Cython *.pyx files with Cython syntax for some commands, black and isort will seemingly work fine, but flake8 will throw false positives since it's not built for handling Cython code
Is there a flake8 alternative I should be using for checking Cython code?
And on a similar note, is there anything I can use with VS Code for handling Cython? I found this VS Code plugin called Language-Cython found on the Visual Studio marketplace, but it makes me apply the language to the file every time I open a *.pyx file

[graceful light]

That sounds more like an issue with your global or virtual environment not having the requests library installed

[dull kiln]

it turned out that requests just doesnt work with pyisntaller

i changed it to urllib and it worked fine

[long raft]

Wait

So with RubyGems

Bundler and Gems will manage package versions AND the dependencies for the package version seemlessly

unlike with venv, if I downgrade a package in my gemfile.lock in Ruby

all the dependencies all the way up the graph are downgraded by Bundler/RubyGem

seemlessly

but my global Ruby Version will have the latest dependency

is that right?

Is it just me or is that much better than Python?

[tawny temple]

Not necessarily. In Python, I believe pip gives an error if constraints cannot be resolved. For example, your program depends on a feature in library A>=2.0, and A depends on B>=2.0. If you downgrade B to 1.0, then A>=2.0 would no longer be satisfied. If A were to be downgraded to work with B==1.0, then your app would break because it relied on a feature in A>=2.0.

[dry vale]

Hey guys, I'm in a selective process for a company, and in the test instructions they say Use of Docker is NOT required but *greatly appreciated*.. The project is mostly a jupyter notebook using pandas. What do they mean by this?

Like, have a container for the app? How would that even work?

I get it if it was supposed to be made in Django or Flask or whatever, but this makes me confused

[tawny temple]

I am not familiar with jupyter but doesn't it host a local webpage that you can access? Maybe they mean that you should set up a container to host that. https://jupyter-docker-stacks.readthedocs.io/en/latest/index.html

[dry vale]

But what reason is there to do that?

[thorny shell]

to prove you know Docker? 😕

or more seriously, to prevent jupyter from modifying your host machine's existing stuff?

[tawny temple]

Maybe it's easier for them to run what you worked on if it's containerised (less work for them to set up a suitable environment with dependencies)

[dry vale]

Maybe

I think I'll just upload the .ipynb file to Google Collab after I'm done 😆

Much easier, same result

But thanks guys 😄

[topaz aspen]

thanks for this btw... I'm not sure how i feel about what they've done there I guess, re using nox in that way. I've not seen it done so much - perhaps it's fine.

[dry vale]

Colab can't handle Tkinter unfortunately

https://towardsdatascience.com/how-to-create-a-docker-image-with-jupyter-notebook-kotlin-2e8bbf212f81 anyone know how to convert this from Kotlin to Python?

[tawny temple]

Have you tried reading through the link I sent? Here it is again https://jupyter-docker-stacks.readthedocs.io/en/latest/index.html

[dry vale]

Yes, couldn't find an example Dockerfile

[tawny temple]

I don't think you necessarily need to write your own image. Do you have additional dependencies that the base images don't supply?

[dry vale]

Yes

I think so at least

Don't think the base images support Tkinter

[tawny temple]

I'm not sure either.

It seems to use conda to install Python

I suggest you test if the provided images come with tkinter before going through the effort of writing your own.

The article you linked for Kotlin doesn't state whether tkinter is supported by that setup.

[dry vale]

Hm

Is it a bad practice to upload a venv to GitHub?

Instead of a Dockerfile

[tawny temple]

Yes because it won't work

venvs are not designed to be relocated to different directories let alone different machines

Best practice is to include a requirements.txt with all the dependencies or equivalent if you use a tool like Poetry.

I don't know how conda users deal with it though.

If that's what you're using instead of pip

[dry vale]

I use pip actually, I was gonna include requirements.txt anyway but since they specifically mentioned Docker I don't know

Anyway, thanks 😄

[dry vale]

Since I'm using jupyter, should I install all required libs with !pip install [...] on the first cell instead of doing this?

[tawny temple]

I have never used jupyter so I don't know what the best practices are for dealing with dependencies.

In general though, first a venv is created, then activated, then pip install -r requirements.txt. Once all that is done, the actual program can be started

[dry vale]

No worries, thanks!

Can anyone explain what the ~= does in a requirements.txt? If I'm planning on the code being usable for both Python 3.8 and 3.9, should I use that operator or <=?

[tawny temple]

https://peps.python.org/pep-0440/#compatible-release

Upper version constraints for dependencies are a contentious topic. Here is one lengthy discussion if you want to dive into the rabbit hole https://iscinumpy.dev/post/bound-version-constraints/

It's more of a problem for libraries than for applications though.

[dry vale]

Thing is, technically it should've been written in Python 3.8

But I'm on 3.9, since there's virtually no reason for me to be in 3.8

So my generated dependencies are for 3.9, therefore I don't know what to do

[tawny temple]

What do you mean by "generated" dependencies?

You mean what you get when you pip install a package without a version specifier?

[dry vale]

pip freeze on my venv

[tawny temple]

Oh

Yeah locking dependencies with an approach like pip freeze doesn't necessarily work out if you want to support multiple Python versions

You'd need a separate pip freeze output for each python version you want to support

[dry vale]

Hm

[tawny temple]

Though you may pick the lowest version and it will probably work fine on later versions

[dry vale]

If I leave just the libs on requirements.txt, will it install the most recent compatible version?

[tawny temple]

If you put the package names without any version constraints then yes, it will install the latest (or use the existing version if the venv already has it)

[dry vale]

Great, thank you! 😄

[tawny temple]

But using the latest version isn't guaranteed to always work. A newer version may introduce a breaking change

So some people set upper constraints to prevent that. But then that causes other issues. So, it's a complicated topic that the post I linked goes into.

[dry vale]

That's rough

[dry vale]

In your personal opinion should I set upper constraints or no then?

[tawny temple]

What I've been doing for applications is essentially a pip-freeze that's generated from some initial constraints (that typically only contain lower bounds)

I haven't had to deal with supporting multiple Python versions recently so I'm not sure how I'd handle that.

[dry vale]

Okay

Thanks for the help 🙂

[tawny temple]

You're welcome.

Sorry I cannot offer more concrete advice on version constraints. It's a daunting subject.

[dry vale]

No worries 😄

[dry vale]

Hey guys, not sure if this is the correct channel, but is there a way to check which of my 3.9 libraries have been ported to 3.10 and "auto install" them?

[gentle solstice]

You can use pip list -o to list outdated packages

[hybrid rampart]

im trying to make a ig tool.can any1 help me..

[slate sable]

Im getting this 405 response, i want to be posting to port 5000, is it not doing that?

[left grove]

Guys I need to use another library in my public project. My project is licensed under Apache 2.0 and the third party library is under the GNU General Public License v3.0. From what I read they are compatible.
I do not want to add this library to the project dependencies, but simply add the folder with the module to my project (so as not to depend on it on pypi, suddenly HE will remove the package).
My question is: what should I do so that I can do everything as described above? I saw some people call such folders "vendor" and add a license to it, is that enough?

[latent canyon]

How to choose a cloud provider for an edtech startup ?

[low trench]

in jupyter lab, can i attach a console to the currently running notebook kernel?

[low trench]

nm, found my issue. .. was starting a new 'terminal', .. i should've started a new 'console' and attached it to the currently running kernel

[wooden ibex]

You pick one.

[latent canyon]

There's a lot to choose from. Digitalocean , Linode , heroku , google cloud platform , azure , aws.

[latent canyon]

I've narrowed down to aws , azure and gcp ? Any personal recommendations on these 3.

[wooden ibex]

AWS has first mover advantage, Azure is good but bit enterprisey, GCP is cheapest generally but A) It's Google and B) Has least features.

Pick one, whatever decision you make will be wrong

[flat path]

You can't. Vendoring that code means your project must also be licensed under GNU GPL v3.

It's debatable whether GNU GPL v3's viral nature also applies if you're using it as a standard dependency, but vendoring is 100% going to cause you to run into the copyleft clause

[dry monolith]

I need to deploy (zip files) AWS Lambda functions written in Python to an S3 bucket. Does anyone have a great open source tool to make that easy? I am familiar with Docker and Gitlab CI (in fact we already have Django build and deployments, but I'd need retrofitting and some changes).

[urban bay]

Hi everyone! Can anyone recommend any decent courses/workshops for furure SDETs? CI/CD, performance testing(jmeter/gatling), any additional stuff expected from ex QA moving towards SDET career? :)

[old lion]

https://python-poetry.org/ anyone used this and know how i can utilise it with pycharm?

[left grove]

thank you!

[brazen forge]

you can configure PyCharm to use Poetry using Settings → SDKs → Poetry environment

[old lion]

neat, thanks!

[warped latch]

Ive built my docker project having two entirely different configs for prod and dev, and i realize now that this is not the way to do it. It makes prod and dev too different.

the django-cookiecutter config is too divided between prod and local for no reason.

[rapid sparrow]

Add third config: staging/preprod
And it should be having almost everything similar to prod
Including deployed to cloud

[waxen gull]

It complies with the 12 factor development approach https://12factor.net. This helps to test what works in local will also work in production. And whatever is in development in local will not affect production until it's ready.

[lusty sigil]

can someone help me with parallelism in python i'm stuck. I've a project and i need multiple instances running at the same time bot i don't know how to do that can someone explain it to me please ?

[waxen gull]

You mean like multiprocessing or threading?

[lusty sigil]

yup

but for variables lengh of process

[waxen gull]

idk what that means but if you just want parallel processing, just use either threading or multiprocessing.

[waxen gull]

I like how this guy explains it https://youtu.be/ecKWiaHCEKs

[lusty sigil]

yup but once i start my loop it didn't start the other process

[lusty sigil]

thx

i'm looking it

[heavy knot]

hey

I wanna make a tool which will take values from a few live graphs

and depending on the values execute action 🎬

how much time will it take me tto complete the project

give your answer in hours

[thorny shell]

45 minutes, plus or minus 10 minutes

[heavy knot]

thank you

[thorny shell]

any time

disclaimer: that was 100% sarcastic; I haven't the slightest idea how long it'll take; I don't even understand what you're trying to do.

Other than that, however, I stand by my answer.

[heavy knot]

lol I too was sarcastic

[thorny shell]

thank God

of course I need to adjust my sar-cas-o-meter

[warped latch]

this is what i was thinking

[warped latch]

isnt the 12 factor approach to make prod and deveopment environments as close to the same as possible so that you can make many pushes rather than few?

[waxen gull]

yes and no

[warped latch]

its number 10 lol

[waxen gull]

yes that its supposed to be as close as possible so you can replicate what you have at prod to local and vice versa

But its also set as a way to decouple local and prod so whatever you are working on local and not sure to push to prod yet, don't put in the prod configs until its bug free

[warped latch]

its definitely possible the blame isnt on the config or structure but on my implementation of it... not focusing on treating development like staging by default

and to be honest it has local/prod files for too many things.

[waxen gull]

One clear example is the debug value for local. You wouldn't want debug enabled in prod

[warped latch]

right i get that but in the grand scheme, these should be deploying the app to a container on docker that is treated like a production environment. the dev step should use all the same steps as prod except a few things.

[waxen gull]

exactly

[warped latch]

you should just start building for production in development i thnk instead of working on a dev config that you cant use

[waxen gull]

What do you mean by "can't use"?

Why would it be in dev config if I can't use it?

[warped latch]

im just saying, spending all this time getting things to work in a dev environment is a waste of time. You dont need twenty billion files to first build the dev app, then build the prod app, almost a completely different app in some ways.

Instead, just build the production app. In the dev environment, flip a few switches, change this to that. But you shouldnt configure an entire series of "For the development version" files from requirements to django settings to different evn files to to different dockerfiles to different compose files.

its excessive

how they do it in that cookiecutter structure

[waxen gull]

So you would have to remember to always flip the switches back before deploying to prod, right?

The cookiecutter structure was supposed to not have the switches flipped or be reminded to flip between prod and dev since the configs are separated out from the beginning

[warped latch]

Im just saying, some of the templates ive seen are not in their perfected forms is all.

and thats probably true of every docker config ive looked at...

it just took me going ove rthem to learn this

[waxen gull]

Not saying you have to follow the 12 factor approach.

[warped latch]

yeah.

[waxen gull]

But it is one of the approach

Especially working in teams

Where everyone has the same development env set up

But working solo, we can do whatever we want

[warped latch]

but even in 12f

https://12factor.net/config

the configs are tied to the env and everything else falls into place. There is no need for... look at this

why not all this in... less files?

i like the overide yaml system

[waxen gull]

Not sure what the problem is. It's just extra files.

[warped latch]

its a mess

it makes you uncertain what effects comes from where

[waxen gull]

I guess we're different. When I open a linux CLI, I don't see files until I ls them. Otherwise, it's just another blank black screen to me.

As long as I know where local and production files are, that's good enough

And whatever is supposed to be in production is supposed to be there, is there. local files should stay in dev.

[warped latch]

hm. I mean this is just my first time doing this and the way i did it wasnt perfect.

thats really all it is]

and i learned some things from it

[waxen gull]

Well worry less about it

[warped latch]

yeah. i just need a break haha

[waxen gull]

I mean

You can try and build a django site/app from scratch, the traditional way

[warped latch]

i been at this for like 24 hours lol

[waxen gull]

without the cookiecutter

And then just build and extend Django app/site over time

[warped latch]

well nah i think the cookie cutter is a great idea, i just think i might consider some better organizational skills than i did this time

[waxen gull]

Then you'll find the reason behind 12factor approach

[warped latch]

really i think the fault is with me, not the structure

[waxen gull]

it was to solve that struggle

[warped latch]

well the one thing that is making the most sense to me at this point is number ten

whcih i definitely did not follow

at all

[waxen gull]

but it made sense to you?

[warped latch]

it does make sense to me but I wish i had prepared for production configs sooner and wasted less time working only on the dev environemtn.

[waxen gull]

Well that's what the cookiecutter did. It made the prod configs and the dev configs. So its easier and faster to deploy

[warped latch]

yes but as I built the app, I neglected to work on prod and worked on development configs. In the end, I have to rewrite the app. It was a waste of time. In the future, my dev config will be ....as per number ten of your 12 factors.... essentially a prod config but staged to a stage server/container

that is otherwise a prod config

and that is where i will develop

it wasnt a waste of time...

there is no way i would havel learned all the things i learned if i didnt try doing it

[waxen gull]

Cool

[pine fern]

has anyone tried sagemaker serverless? how is it compared to traditional lambda + API gateway architecture?

[hearty finch]

aaah

now that rings a bell

[heavy knot]

Hello, im trying to get priority speaker in discord group chats , or to make me louder in discord Group chat ( just a way to lower people and be louder ) Can anyone help me ?

[heavy knot]

How do I merge my master branch to main branch?
I'm a beginner and pushed my project onto github.

[rapid sparrow]

It sounds confusing. Let's have master branch as main one

Delete .git, git unit it locally, dump everything into initial commit, and force push to remote

Or tinker in github settings online, if they still allow choosing master as master, and just delete main

[heavy knot]

Yeah I mean people use only one branch either main/master but just wanted to know how can I merge my master into main

[rapid sparrow]

git checkout main
git merge master

[brazen forge]

you just want to rename the branch, right?

or do you have both branches simultaneously?

@heavy knot

[heavy knot]

It's more like I wanted to try make a pull request but then I wasn't able to.
Though then I changed my default branch to master and deleted the main branch as @rapid sparrow suggested.

[brazen forge]

okay. you can just rename your branch though

[heavy knot]

I tried that too but changes were not made on main branch on GitHub.

[brazen forge]

git branch -m master main
git push --set-upstream origin main

[rapid sparrow]

uh, git checkout master -> git pull master -> git checkout main -> git merge master then

[heavy knot]

Aight okay

[heavy knot]

Is there a way, by chance, to make the On branch whatever when you do git status show up in all caps or a different color or something? I'm always looking for that tiny line of what branch I'm on but it gets lost in the other output

[thorny shell]

maybe just type "git branch" instead

[brazen forge]

and check for *

[thorny shell]

or use some sort of UI -- many shells have git status stuff built in (ohmyzsh and/or powerlevel10k e.g.)

e.g.

[brazen forge]

homedir on git

[thorny shell]

sho nuff

that way I can have it everywhere

[heavy knot]

can those themes be applied to powershell perchance 👀

[thorny shell]

heh

but I think powershell has plenty of customization too. I put a timestamp into my powershell prompt, but never bothered with git status since I don't git much on winders

https://git-scm.com/book/en/v2/Appendix-A%3A-Git-in-Other-Environments-Git-in-PowerShell e.g.

https://www.google.com/search?q=powershell+prompt+git+status generally

[brazen forge]

IIRC oh my posh for Powershell

but never really used it

[thorny shell]

"oh my posh"? Heh

[heavy knot]

posh-git works and was easy to set up thanks offby

[thorny shell]

💐

[latent canyon]

hey , I have this django project deployed on a local server running on docker containers
I'm using nginx to serve media files. It works on my machine , but when deploying to the local server the mediafiles are not loading. Any Solutions ?

[mystic void]

• have you verified these files are accessible?
• have you verified the URLs you generate are valid?

[latent canyon]

The things that surprises me is that , it works on my machine and not on the local server.
The folder and the files exist in the docker container.

[mystic void]

then the obvious answer is to dive into the differences

all else being equals, it should work

[latent canyon]

How do i set django's STATIC_FILES and MEDIA_FILES location as a directory in another docker container ?To be precise i want to set the STATIC_FILES and MEDIA_FILES location to be a folder in the nginx container ?

[shadow crow]

Is it possible to download the activity data from github as numbers?

this thing:

[tawny temple]

I think your only option is to scrape it

[brazen forge]

perhaps https://docs.github.com/en/rest/activity/events#list-events-for-the-authenticated-user ? though that contains a lot of events

[subtle quarry]

I've read and seen examples that the past data can be written to show usage (or not)

[narrow slate]

Is anyone aware of a marketplace AMI which embed a CMS and allow to use a remote postgresql backend using tags or whatever which avoid to do anything from within the instance ?

[eternal flicker]

the v4 api has that information

https://docs.github.com/en/graphql/overview/explorer

query {
  user(login: "kosayoda") {
    contributionsCollection {
      contributionCalendar {
        totalContributions
        weeks {
          contributionDays {
            contributionCount
            weekday
            date
          }
        }
      }
    }
  }
}

[south marsh]

How would I resolve this issue when running pip install -e .? ```
Obtaining file:///home/mart/git/kaudio-python
Preparing metadata (setup.py) ... done
Installing collected packages: kaudio
Running setup.py develop for kaudio
error: subprocess-exited-with-error

× python setup.py develop did not run successfully.
│ exit code: 1
╰─> [79 lines of output]
    <...>
    running develop
    /home/mart/.local/lib/python3.10/site-packages/setuptools/command/easy_install.py:144: EasyInstallDeprecationWarning: easy_install command is deprecated. Use build and pip and other standards-based tools.
      warnings.warn(
    /home/mart/.local/lib/python3.10/site-packages/setuptools/command/install.py:34: SetuptoolsDeprecationWarning: setup.py install is deprecated. Use build and pip and other standards-based tools.
      warnings.warn(
    error: ("Can't get a consistent path to setup script from installation directory", '/', '/home/mart/git/kaudio-python')
    [end of output]

note: This error originates from a subprocess, and is likely not a problem with pip.

error: subprocess-exited-with-error

× python setup.py develop did not run successfully.
│ exit code: 1
╰─> [79 lines of output]
running develop
<...>
/home/mart/.local/lib/python3.10/site-packages/setuptools/command/easy_install.py:144: EasyInstallDeprecationWarning: easy_install command is deprecated. Use build and pip and other standards-based tools.
warnings.warn(
/home/mart/.local/lib/python3.10/site-packages/setuptools/command/install.py:34: SetuptoolsDeprecationWarning: setup.py install is deprecated. Use build and pip and other standards-based tools.
warnings.warn(
error: ("Can't get a consistent path to setup script from installation directory", '/', '/home/mart/git/kaudio-python')
[end of output]

note: This error originates from a subprocess, and is likely not a problem with pip.

[south marsh]

Here's my setup.py: https://pastebin.com/HHYhvrZu

[shadow crow]

🤔 is there a way to not let it do stuff to Pydis?

or does is not change anything?

thanks, I'll take a look at it

I have no idea what GraphQL is so I'll need to do some digging

[rapid sparrow]

GraphQL is just an alternative view to REST API
instead of having multiple endpoints, they are merged into one that can be queried with graphql language to get different outputs

[shadow crow]

yeah I get the idea

so I can use the REST API?

[rapid sparrow]

as a pros of this approach, client can query precisely info which is needed. no less, no more
and theoretically more stuff can be queried in parallel in one request

[rapid sparrow]

if they provide rest api endpoints, why not 🤷
or if u wish to provide rest api in your app

also small benefit of graphql, that it can potentially decrease amount of endpoints 🤔 i mean it would be just one endpoint with less duplicated logic perhaps

[copper basalt]

Hello. 👋
I have a slight case of dependency hell which has led me to ask the following question, to which I have not been able to find a suitable answer:
How do you deal with a situation where your dependencies have conflicting dependencies in the same project?
The information I've found for dealing with dependency issues was all about separating into different projects, virtual environments, etc. - but what do you do when you need these things in the same place?

[shadow crow]

you mean, two versions of the same library?

I don't think there's a good solution

well, you can always run two separate Python processes

[rapid sparrow]

So far I was just resolving those conflicts. And that's it

with having requirements.txt only with core deps, and all secondary deps in constraints.txt, it makes easy to resolve conflicts

as an option there is also to use pipenv, which sorts of helps with this too, but I still struggle how to solve problems in a nice way in it xD

Sometimes i have situation when library is not updated and just can't be used in newer versions

in this case I just take this library out of deps and put into my source code folder

and update the necessary stuff in it

optionally I could have deployed a fixed version to pip though

[copper basalt]

Yeah, I've been doing that, but it's a painful process and I find myself thinking there must be a better way.
I had to update a certain package, which forced me to update another, which forced me to update another, ...
Each one I upgrade breaks some other package dependency, and I don't even want to update the other things that use them - if X and Y were using v1 of Z and now I upgraded Y which makes it need v2 of Z, I still want X to use v1 of Z.
Is there no way to accomplish that?

[rapid sparrow]

1. just try to add less packages to your project. xD

u will have less problem then

2. split the service into mciroservices, each one will be having need for less packages to use

3. optionally clean architecture can be used in order to contain implementation details in certain places for easier changes

[shadow crow]

Introducing microservices or any other IPC sounds like a lot of extra complexity if you just need two versions of a package

[copper basalt]

Good advice, but doesn't help with a project that's been around for years already. 😅
I'll be working on reducing the dependencies there, but my question is more general and not specific to my project.
Is there a way to make a package use a specific version of some dependency it has while another package uses a different one, all in the same virtual environment?

[shadow crow]

Is there a way to make a package use a specific version of some dependency it has while another package uses a different one, all in the same virtual environment?
No, that is not possible in the general case.

I guess if the library uses relative imports (like from .bar import baz instead of from foo.bar import baz) you can rename one of the packages into foo2 and still have it work. But it's a hack, and it won't work if it's a transitive dependency

Maybe justuse can help in some way? https://pypi.org/project/justuse/
its README claims it can have two versions of the same library in the same venv, somehow

[copper basalt]

Thanks for the suggestion. I checked it out just now and found this:

np = use("numpy", version="1.21.0rc2", hashes={"3c90b0bb77615bda5e007cfa4c53eb6097ecc82e247726e0eb138fcda769b45d"}, modes=use.auto_install) inline installation of packages and importing the same package with different versions in parallel in the same code - most people wouldn't even dream of that!
This is quite interesting and potentially promising (assuming it actually works).
Now, if there was a way to take a module and redirect its import statements to use a similar mechanism to take the right version, that'd be great.

One thought I had was to map the dependencies in some file, then wrap imports in code which temporarily places you in an environment where the expected packages have the right versions installed.
That is, continuing my example from above, you'd have both v1 and v2 of package Z installed somewhere, and you replace import X with X = myimport("X"), which would then point the module search so that importing Z would lead to your installation of v1, just for the duration of performing the import, in order for the loading to get the right thing.
And similarly it would point it to v2 when doing myimport("Y").
But that would only work on the top level, as I can't go and replace import statements in external packages.

[brazen forge]

you could ask the authors, which are regulars here. @greyblue92 and @amogorkon

[shadow crow]

cc @next spruce

do you know if justuse can help in this case?^

[rapid sparrow]

Is there a way to make a package use a specific version of some dependency it has while another package uses a different one, all in the same virtual environment?
@shadow crow it is possible actually

just copy the package into your source code folder

and rename it being a different one

woala. U have two packages of same thing that can be even different versions

A bit of a WET solution, but if we will explicitely forbid changing code in this package, I think it is sort of a solution

although technically it is not WET, since both packages will still not be containing any of your code 🤔

it will be just a different way to install package

slightly dangerous due to possibility for you introduce changes into it, that will be hard to track if you decide to upgrade the package

[brazen forge]

did you mean voila

[thorny shell]

I think what you're describing is called "vendoring"

[rapid sparrow]

Exactly!https://c.tenor.com/sw4GnyyBTREAAAAC/voldemort-smiling.gif

[rapid sparrow]

yeah

[copper basalt]

I was planning on doing just that; didn't know they were regulars here, so that's a time saver. Thank you. 🙂

[copper basalt]

That would work for having multiple versions, yes, but it wouldn't solve the issue of your dependencies needing to use said versions, as they only have "import whatever" rather than whatever name you gave them.
But that's the basis of what I said, sort of. Having multiple versions available, then redirecting the imports to the right ones.
It doesn't need to be in your project though, just some folder you create when you do your setup and install your dependencies.

[rapid sparrow]

Oh yeah, it would mean u would have to copy not just deps

but also the parent dep

[copper basalt]

Yeah, it gets complicated when you take multiple levels into account. =\

[rapid sparrow]

Anyway. Your situation starts to sound similar to this funny interview https://www.youtube.com/watch?v=Uo3cL4nrGOk

[rapid sparrow]

"My job is to keep our code running while other packages are changing theirs." (quote from video)

https://jjquery.io/ bwhaha. Someone had a humor and created web site in honor of the video above

[shadow crow]

won't work if the package uses absolute imports

[copper basalt]

Love people with good humor. 😄

[sonic elk]

Hey guys I have a question about GIT and VSC :
I just installed Windows. git is also installed. i am cloning a repo then making a change. nothings come up here :

It was not like this before my windows formatting

[narrow slate]

Does someone retrieve secrets from AWS secretsmanager through ansible ? I have a hard time achieving to do it, the lookup plugin don't find it even if I force all envs while I can retrieve it with the cli 😦

in fact the lookup return not found if the secret exists but is empty 👌

[gentle solstice]

Did you click the refresh button? Does running git status manually in the terminal yield anything?

[ivory bobcat]

Is this for like libraries and ide extensions we make

Cuz like I’m not terribly great at making stuff but I’m pretty good at making stuff to make stuff with

[sonic elk]

here:
and yes i tried to click refresh

[gentle solstice]

reloaded the window? (Ctrl+P, "reload window")

[sonic elk]

I tried. nothing changed

[gentle solstice]

Could it be that the "publish to github" prompt is blocking any git operations?

press esc until it goes away

[misty sorrel]

so I deployed a flask application on nginx (ubuntu vps) and connected a mysql database to it
its up and running but every other functionality except login is working
I used flask login for saving the user session
everything works properly on my local server

but here the session is not being saved

[rapid sparrow]

Why people use pipenv?!

it is not showing dependency errors in detail! This tool suggests me only to install stuff with --skip-lock and checking manually what the hell went wrong in pipenv graph

in regular requirements.txt+constraints.txt it is much easier to handle

[thorny shell]

what is "constraints.txt"?

https://pip.pypa.io/en/stable/user_guide/#constraints-files
TIL

[rapid sparrow]

when you have everything installed, you throw it to constaints: pip freeze > constraints.txt
it locks secondary dependencies but will not install them

you point stuff necessary to install in requirements.txt without secondary deps

you can even point requirements.txt without versions at all xD they are locked in secondary deps anyway

makes relatively easy to handle deps i think.
we just regulate requirements.txt file, while having a free reign in constraints

it is technically really similar to how pipenv works

we just use pip default functionality

[thorny shell]

I started using pipenv because it's from Kenneth Reitz, who made requests, so I figured it'd be equally awesome. It isn't. 😐

[waxen gull]

Is a Github Token the same as Personal Access Token?

[brazen forge]

GitHub does use PATs. or do you mean a deploy token?

[thorny dome]

In Mac OS 12 terminal, python commend cannot be recognized, but python3 can be recognized, anyone know how to fix this?

[thorny shell]

get used to typing python3 instead of python is probably the best way

it's not broken; why do you want it "fixed"?

[brazen forge]

well, you could also symlink or alias python to python3

or install python from homebrew, which uses python and you can put it ahead in the path

[thorny dome]

Thanks~

[waxen gull]

Kinda. I basically wanna use it for GitHub Actions but not sure of which token is which and where.

[brazen forge]

apparently it's just a special token

https://dev.to/github/the-githubtoken-in-github-actions-how-it-works-change-permissions-customizations-3cgp

[heavy knot]

can someone help me with this after i convert it into exe ?

please

[finite fulcrum]

you need a terminal window if you're using input

[heavy knot]

how ?

please

[finite fulcrum]

don't pass the flag that disables it

[heavy knot]

ai need something put

[finite fulcrum]

It's from the line you used to build th app, not in the app iteslf

[heavy knot]

?

like this ?

[finite fulcrum]

don't do the -w

[heavy knot]

how can i add more script into one exe ?

bcs it open onlY one script after that when it will go to menu list it will close

i send it friend and he only got opened login

[prime grail]

So I am trying to build a Dockerfile and I get this error...

Here is my Dockerfile

FROM python:3:10-slim-buster

WORKDIR /api

RUN pip install poetry

COPY pyproject.toml /api
COPY poetry.lock /api

RUN poetry install

COPY . /api

CMD ["sh", "launcher.sh"]

This is the traceback

Sending build context to Docker daemon  23.38kB
Step 1/8 : FROM python:3:10
1 error occurred:
        * Status: invalid reference format, Code: 1

[brazen forge]

3.10, not 3:10

[prime grail]

ah tysm

[steel sundial]

Question: In Sphinx, where is the builder_name added to the build directory?

Eg: buildir /docs becomes /docs/html when building with the html-builder.

————————
Using the recommended make-mode aka -M.

I have combed through the source code without finding it where it actually modifies the outdir.

The reason is I am building for Github-Pages and the trick with the redirect inside an extra index.html seems to be breaking some stuff (mostly css for some reason).

I wrote a little sphinx wrapper because I also am doing other things not possible with base sphinx. This currently builds to an temp dir and then copies it to the docs folder directly with no html subfolder.

It works but if I could just fix that by sub-classing/custom-builder/extension/… I would rather do that, as the current solutions is quite duct tapey and also makes sphinx console output non-sensical (your files are in <temp-dir>/htm/ for example)

Anyone has some knowledge about sphinx internals, and can help me?

feel free to ping/@ me when answering!

[fickle bough]

https://www.reddit.com/r/Python/comments/vk872e/sort_3_000_000_000_lines_by_most_repeated_one_via/?utm_source=share&utm_medium=mweb3x

Umm if links are prohibited, I'll copy paste the whole post over here

[spice pulsar]

I am creating a python library. I have put detailed docstrings with examples in all functions. Which tool can I use to generate documentation from these docstrings

I want this documentation to be available/readable on Github

So I guess it needs to export it to .md files? I'm not really sure about this part

[loud bear]

Not at all use mkdocs

With mkdocstrings

Very ease and simple

Make sure you adhere to a doc style like google or numpy or whatever else there are

[spice pulsar]

Thanks. I'll check it out.

[brazen forge]

that's something you should ask in an off-topic channel

[warm hinge]

ok

[pure narwhal]

Hello, guys!

Can anybody recommend some solutions for monitoring cron jobs?
I saw some paid options but prices were too high.

Looking for free self-hosted or cloud solutions with moderate price for small amount of jobs (up to 100 cron jobs/up to 1000 runs daily)

From my perspective, there is a free self-hosted option of grafana + timeseries db, when every job contain formatted output with result of job result (datetime, duration, name of process, high level exec result, low level exec result, etc) and this output will be redirected to our timeseries db. after that we put this data to grafana. We getting dashboards for cron jobs + option of alerting on unsuccecful runs or too long runs.

What do you think about that? What do you use?
Appreciate any advice or thoughts 🙏

[rapid sparrow]

My first thought would be to recommend migrating from Cron jobs to Celery Beat, and just getting its monitoring system Flower

[pure narwhal]

will check this option in details, looks promising. Thanks!

[stuck citrus]

what is the best way to use all the linux commands in the windows terminal?

[thorny shell]

WSL

for some value of "in the windows terminal"

with WSL, you can run "bash" exactly the same way you'd run "cmd" or "powershell", and it's either 100% pure Ubuntu Unix, or else so close that I haven't noticed the difference

[not that I've spent a lot of time with it, but it's pretty impressive]

[red locust]

sudo chown myuser /usr/local/var/postgres/ is this location for mac?

i saw this on stacksoverlfow, and unable to resolve issue

this is the error

i am using ubuntu on windows

i have to urgently install an application, as asked by my professor

[thorny shell]

could be, or it could be for any of a number of Linux distros

you have to figure out which package includes postgresql-setup, install it, and figure out the directory into which that file goes

I just tried apt-file search postgresql-setup and it returned nothing which suggests there's no Ubuntu package that includes a program of that name 😦

I think the instructions you're reading are

• for Fedora, not ubuntu; and
• pretty sloppy

[red locust]

basically as a someone who just downloaded ubnutu, i am ...., right

[thorny shell]

https://www.prisma.io/dataguide/postgresql/setting-up-a-local-postgresql-database says

Once the MySQL package is installed, initialize the database by typing:

sudo postgresql-setup initdb

so you can tell this is not super-professional documentation

[red locust]

i am following this to reinstall:

https://www.cherryservers.com/blog/how-to-install-and-setup-postgresql-server-on-ubuntu-20-04

is it ok?

[thorny shell]

how would I know?

if you are following their instructions and you've got a working postgres setup, then they're OK; if not, not.

I glanced at it and it looks reasonable; that's all I can say

[red locust]

according to professor instruction:

is this cosistent with alternate aproach to installation

[thorny shell]

🤷

[red locust]

?

[thorny shell]

ask your professor then

[red locust]

😢

[thorny shell]

the postgresql12 package contains a file named /usr/lib/postgresql/12/bin/initdb. That probably does what postgresql-setup initdb does. Try it and see.

if that advice turns out to be useful, the lesson is: Ubuntu's "apt-file" program is super-handy; ask me for details.

[red locust]

ok trying

[red locust]

what to choose?

shows installed in 14.4 version of postgresql

[thorny shell]

I haven't the slightest idea

I don't use postgresql. and I don't know what your instruyctor had in mind

[red locust]

https://github.com/jkonglab/DigitalPathology

if i ask the professor, he might be dissappointed

[thorny shell]

well no offense but I don't want to study that repo in order to answer your question

[red locust]

my pc+my brain is going hay wire

[thorny shell]

well, your brain will cure itself if you take a rest. Your PC, though, I can't help with

[red locust]

havent cried in years, seems like record gonna break

[tawny temple]

postgresql-setup seems to be specific to the way rhel/centos/fedora packages postgres. Since you're using Ubuntu, you won't have that. initdb should be sufficient. https://www.postgresql.org/docs/14/app-initdb.html

[thorny shell]

yeah we've been through that

[tawny temple]

Sorry, I suppose I don't understand what the current problem is then

[thorny shell]

OP got initdb to work, but it's interactive and they don't know what to click on

[tawny temple]

Weird. I guess Ubuntu does things differently too

Normally it's a CLI command and you just need to give it a directory. https://www.postgresql.org/docs/current/creating-cluster.html Seems like Ubuntu has pg_createcluster so maybe try that https://manpages.ubuntu.com/manpages/trusty/man8/pg_createcluster.8.html

[red locust]

ok please tell me one thing

what type of location is that

is it mac?

how does it translate

i have located initdb

[thorny shell]

can't tell for sure

[tiny jungle]

sudo and the /usr/ path format are unix/posix-based only, so mac and linux

though it appears for you that postgres follows similar paths in windows

[brisk kettle]

fwiw on my mac my ~ is /Home/username

[red locust]

So do i need to enter from c: uptill bin for windows?

[tiny jungle]

that would make sense, yes. to make that job easier, note that you can right click the address (the This PC > Windows (C:) > blah blah blah) and press Copy address instead of typing it all in manually

[red locust]

Ok thank you,

[velvet spire]

I'm trying to connect to a python language server from python, are there any lsp clients written in python?

also how can i run a language server independently 😅

[thorny shell]

sounds like something google could answer

[velvet spire]

sounds like i asked here intentionally since someone might have an answer

(i already googled it and spent quite a while but didn't find a result)

[thorny shell]

best to say that up front, to head off smartasses like myself 🙂

how to run it independently: the one that I use is just a python module, which you just run from the command line, and iirc it just sits there either listening on an IP port, or else reading from stdin

I bet pylance works similarly

[velvet spire]

which one do you use?

[thorny shell]

I actually have to check! Hold on

python3 -m pip install 'python-lsp-server[all]'

my notes also say Beware: there's also a "python-language-server" but it's flaky and abandoned

[velvet spire]

ah and which editor do you use to interface with that?

[thorny shell]

Emacs

not recommended for normal humans

pretty sure pylance is what you want; it's the default for visual studio code, and I've never encountered, nor heard of, any problems with it

[velvet spire]

what i'm doing is making an api that interfaces with a language server

this is not for an editor :P

[thorny shell]

when you say you "didn't find a result", did you mean you didn't find anything that actually worked, or ... ?

because

[velvet spire]

I couldn't find a client

that was maintained*

[thorny shell]

ah

that is easy to believe 🙂

[velvet spire]

might have to roll my own

[thorny shell]

I assume 99% of the interest in LSP is for editors, and I assume 0.0001% of editors are written in Python

[velvet spire]

my goal is an api which can return introspection on provided code and other tools

it won't be a public api, I'll be using it with another project

embracing lsp means I don't have to worry about the hard parts and can switch it out as well

i found this so post

https://stackoverflow.com/questions/55312875/guidance-to-writing-lsp-client

[jaunty wren]

https://medium.com/geekculture/recognize-speech-convert-it-to-text-491706fc1d24

[warped latch]

Lets say ive been working on the same project in two repos.. and im now tired of switching back and forth. I would like to add one repo into a dir called old and merge that into that old one. But im having trouble getting it to "merge two unrelated histories"

Advice/guides on how to go about this?

or some other way i could slowly go over both in the same project and keep the best of both. keep in mind, these projects have entirely different structures despite solving the same problems

[brazen forge]

ideally the old and new should have been separate branches instead of repos

git does have a way to merge unrelated histories IIRC

https://git-scm.com/docs/git-merge#Documentation/git-merge.txt---allow-unrelated-histories

[misty ice]

Guys anyone familiar with configuration management in DevOps, having a hard time getting my head around what it does for the process?…also do we use tools for this too? Since we want everything to be automated?

[thorny shell]

awfully broad question

that's like asking "Hey anyone here a doctor? How do you do brain surgery?"

[misty ice]

😭😭😭😭

[thorny shell]

just ask a more specific question

[misty ice]

You’re hilarious man

I’m trying dude 🤣

CI/CD is hard enough

Then IaC - this stuff never ends 😂

[thorny shell]

"what does it do for the process"? Automates it, obviously. That's good because it reduces the risk of accidents caused by humans typing the wrong thing.

"DO we use tools for it"? Of course. Ansible, Azure Dev Ops, Github actions, AWS surely has something, &c &c

[misty ice]

Thanks man 🙏

[thorny shell]

I don't do all devops every day, but sometimes it feels like it (spoiler alert: I much prefer regular coding)

[misty ice]

Trying to get my head around the term ‘pipeline’ as I’ve been reading it a lot. I hear it in many things. Like DevOps pipeline and CI/CD which is also apart of the micro services architecture too. I read just under this page with the title ‘CI/CD pipeline’

The first sentence is

[thorny shell]

yeah it's pretty vague

[misty ice]

‘Your pipeline is the path your code travels from development to production.’

Is this the case for all pipelines?

Like when I read the word, pipeline, is this what I should be visualising ?

[thorny shell]

I'm familiar with Azure Dev Ops pipelines, which is basically a super-complex process whose point usually is to take code you've checked into a git repo, and get it running on a machine, or machines, "in production", often with intermediate stops on "development" and "integration" machines

yeah that's roughly how I use the word pipeline too. But I bet you'll find people using it to mean slightly other things as well; it's a relatively new concept (I don't recall people talking about pipelines 20 years ago e.g.) and the definition hasn't quite settled down yet

[misty ice]

Thanks for your help man. Really helpful good answers

[thorny shell]

I think the best way to learn about a tool like this is to not use it, and suffer; eventually you'll say "geez I wish there were a way to automate all this crap". At that point you'll be receptive to the idea 🤣

... and if you never find yourself complaining like that, then pipelines probably aren't for you

I'm 100% serious here

[misty ice]

Alright Thanks man 🤞

[thorny shell]

if it helps any -- if you're good at this you can get paid a lot of money

[misty ice]

👌✍️😆

[thorny shell]

that's another way of saying it's hard --

If it were easy, everybody'd be doing it

• A. Berg

[red locust]

so, finally was able to start initdb by sudo service postgresql start
but what do i do about systemctl
a little search indicate it require extensive hacking for working on windows?
are there alternative to systemtcl steps

[thorny shell]

if you have "service", that's roughly equal to systemctl

[red locust]

wininit i have heard, and wwould scm

would it suffice? also the second step is done right ? if it says:

i think i have installed 2 version and it start both

right?

[strong dew]

how would you ban a specific string (TODO in this case) using pre-commit? It seems the fail language can only ban commits with filenames that don't match rules?

I think I will try a local hook with the 'pygrep' language

[wheat furnace]

https://github.com/guedesfelipe/pls-cli

[brisk kettle]

hi there, please read our #rules 🙂

is this a self-promotional post?

is this a tool that fits in the scope of this channel for #tools-and-devops ?

[leaden lotus]

i use pips and imports so i don’t have to copy paste raw package code

[subtle quarry]

That's smart!

[brazen forge]

you should open a help channel. this doesn't seem relevant to this channel

[civic grove]

is there a tool for jupyter note book to play a sound when running a code cell finishes

dont wanna waste my time having to keep checking on it

[indigo zenith]

Have you seen this? https://stackoverflow.com/questions/17323336/automatically-play-sound-in-ipython-notebook

[civic grove]

that looks like what im looking for at first glance. thanks!

[red locust]

help

^

[thorny shell]

have you tried adding the exception like it says?

disclaimer: I never use phusion; I'm just reading the error message

[latent canyon]

How to mount directories in docker-compose file ?

[latent canyon]

What does this mean in docker-compose.yml file : - ./static:/etc/nginx/static ??? 🤔

[full bronze]

Looks like it does something from ./static and applies that to /etc/nginx/static. Probably a copy or perhaps mounting a directory

[brazen forge]

yes, that does a bind mount by default

[tender current]

I'm running conda 4.12 on WSL2-Ubuntu20.4 and trying conda env export > environment.yml, which fails with InvalidVersionSpec: Invalid version '1.12.0<2': invalid character(s)
This seems to be a popular solution, but I don't have a nb_black at that location https://github.com/conda/conda/issues/9624#issuecomment-737107961
And this one provides a workaround for not having nb_black, but it's for windows 😕
https://github.com/conda/conda/issues/9624#issuecomment-777160270

Does this command work for you? Please ping me if you reply

[steady wraith]

hey can anyone help me on which Plugs should i use in neovim for python development

[brazen forge]

that question's better suited for #editors-ides I think

[subtle quarry]

Check Lunarvim or Lunarvim docs for plugins

[sterile oracle]

hi

how to connect thonny and QT Designer

[stuck citrus]

can someone help me with pyautogui

[latent canyon]

Any ideas on solving this error :
Refused to apply style from '<URL>' because its MIME type ('text/html') is not a supported stylesheet MIME type, and strict MIME checking is enabled.
Tech stack :
web server (reverse proxy) : nginx , application : django , application server : gunicorn & Docker.

[abstract cape]

Does anyone here have a working docker setup that can run GUI applications? I'm on Ubuntu 22.04 and none of the SO answers/blog posts that seem to have worked in the past work for me. Can't get tkinter to find my X server even after stripping away all security layers.

[rapid sparrow]

Docker can run only CLI, TUI, and web UIs

the most close to GUI is obviously web UI at localhost somewhere available

[abstract cape]

It's been done in the past. You can share your display server socket with docker. If you don't believe me: https://www.youtube.com/watch?v=cMsIT2otEjA

[rapid sparrow]

wow. That's some dark magic

[heavy knot]

Hey! I'm currently using Pyautogui to automate a program. I'm running my script on a vps, but when I close the connection to my vps it stops working. I'm using .write(), .moveTo(), etc.

[heavy knot]

I've seen some post saying I should use some vnc stuff, but I didn't quite understand howto

[subtle quarry]

Could you start your script in tmux?

[heavy knot]

Maybe, but what difference would it have?

[thorny shell]

tmux prevents your program from stopping when you close your VPS connection

now, if it's a gui program I'm not sure it'll work

is the "connection to your VPS" a web page, or some program you run at the command line, or what?

[subtle quarry]

It will not stop when you disconnect

[heavy knot]

My script doesn't stop when I disconnect from my VPS, but Pyautogui doesn't work because there is no active connection or something like that.

[heavy knot]

I'm using the .rdp file I downloaded from Vultr, but I can also connect using a web page

[thorny shell]

ah gotcha

well ...

I don't know why your program stops when your RDP connection drops

[heavy knot]

Pywinauto's docs explains everything and a solution, I just don't understand the solution
https://pywinauto.readthedocs.io/en/latest/remote_execution.html

[past hound]

can anyone tell me if my understanding of docker is correct

so ive been trying to learn docker for a while now and I get the basics of docker and docker-compose

so if I containerized my app using docker and want to deploy these are generally the steps I would have to follow

• create a docker image that would contain my application
• push the docker image to some docker repository
• on my cloud service (digitalocean, aws, gc, whatever...) I would pull the docker image from wherever I pushed my docker image to and simply just start the docker container?

[tawny temple]

Yes

[past hound]

would I have to do any additoinal config relating to networking side?

also would it be better if I developed my application first and then dockerized the app or should I just dockerize the app from the start?

[tawny temple]

Depends on how exactly you deploy it I suppose. Like, if you deployed it just to a VPS then you would likely need to configure the firewall, among other things. But a VPS is typically not the most cost-efficient way to deploy containers anyway.

[past hound]

which process is better?

whats the best way to deploy a docker app? I should use something like AWS?

I see lots of people recommending DigitalOcean

[tawny temple]

Cloud providers typically have services that let you deploy an image directly

[past hound]

I just want to build a dead simple app, dockerize it and deploy it just to understand the basics

cloud providers you mean like AWS, Azure, GC etc...?

if I deployed to AWS, would I still need to configure firewall and additional other security measure?

[tawny temple]

Again, depends on how you deploy it. There are multiple ways to do it.

For example, AWS has Elastic Beanstalk which can be used to deploy containers. They also have AWS App Runner, which is a simplified version of Beanstalk. These approaches require less configuration than if you were to just use EC2.

[past hound]

is EC2 a VPS just like having DigitalOCean server?

[tawny temple]

Yes

[past hound]

I remember using AWS a while ago but it was confusing as hell

even to do something very basic

[tawny temple]

AWS is just an example. Other providers have similar offerings

I'm just more familiar with AWS so I used it as an example.

[past hound]

would I have to push my docker image to the AWS docker registry or can I push my docker image anywhere?

[tawny temple]

IIRC for AWS App Runner it only supported automated deployments for images on ECR, their container registry.

Not sure about Beanstalk. You'd have to check their docs.

[past hound]

ok no owrries. appreciate the info. much love

[tawny temple]

In any case I believe ECR is free if you push the image to the public repo. The cost is when you want it to be private.

Well, free as in up to a certain limit

[past hound]

if I push my docker image to ECR and make it public then I can host the dockker image for free on EC2?

[tawny temple]

nope

ECR and EC2 have separate billing

And if you use EC2 directly you don't necessarily need to use ECR

[past hound]

oh so you're saying its free to deploy my docker image to ECR for free if i make it public but I still have to pay for EC2?

oh right that makes sense

that means I can just deploy to dockerhub or wherrever and use EC2 separately?

[tawny temple]

Yes.

Beanstalk and App Runner have more integrations for automated deployment and at least for App Runner I am sure only ECR was supported.

[past hound]

My only concern with AWS is unexpected high costs... I read some horror stories where some users are racking hundreds of dollars unexpected bills

[tawny temple]

I suppose that can happen with any cloud provider.

[past hound]

like if you dont know what you're doing you can be hit with some gnarly usage fees

[tawny temple]

In AWS you can set alerts and limits for your billing

[past hound]

thats what scares me to be honest

I dont think DigitalOcean has that issue. at least I dont' thiink so

[tawny temple]

Well AWS only charges you for what you use, so if you unexpectedly use a lot, then you will be charged a lot.

It's not like a flat rate

[past hound]

well I wouldn't expect my app to receive high traffic but once you share it publicly to users, there will be someone who will try to exploit your application

which is what happened to someone

someone ran an infinite loop which was constantly hitting their api which racked up some hefty costs

[tawny temple]

That's why one should implement rate limits for their API

[past hound]

yea but for someone who might not be too experienced this could be a very costly mistake

😬

[past hound]

im trying to open my app at localhost:8000 but keep getting Unable to connect to localhost:8000

CONTAINER ID   IMAGE                        COMMAND              CREATED         STATUS         PORTS                                                 NAMES
f8ed60d8a39b   prakhar1989/foodtrucks-web   "python3 ./app.py"   6 seconds ago   Up 4 seconds   5000/tcp, 0.0.0.0:8000->8000/tcp, :::8000->8000/tcp   foodtrucks

but should be the correct port? what am I doing wrong

[heavy knot]

my process memory scanner

[dusky mantle]

there are a lot of values with "5" :P

[heavy knot]

ye

[serene notch]

hey guys, wonder if anyone can help me maybe guiding me or providing me some inputs on this. My end goal is to implement a pipeline in Azure DevOps that executes usual python steps (code validation, tests, etc) and deploy or make the updated version available to be deployed to my server. As my environment is still based on servers, not containers, I am wondering how I can accomplish this.

What´s still not clear to me is the actual "code deployment" to the server. That´s what I have in my mind for the pipeline:
code quality validation (flake8, black, isort), tests, etc
once this all passes, it would either move the code to artifactory and then deployed to the server ( it´s not container )
Should I have something on the server itself being triggered to download the latest version of the artifactory?

Appreciate any tips, articles or recommendations.

Thanks in advance

[tawny temple]

How about simply SSHing into the server and then using git to pull changes?

[serene notch]

the issue in this case might be that my server is not in the cloud. I think my azure devops environment wouldn´t be able to reach it. Should I have something on the server end looking for code updates and then trying to pull it from my repository or artifactory?

not sure if this is an unusual setup. This is a new environment to me, so tyring to see how I can make this a good solution considering devops good practices

[tawny temple]

Is the server connected to the internet? If so, then you can set it up to run sshd and connect to it over ssh via Azure Pipelines.

[heavy knot]

Could someone tell me what devips is

Devops

[serene notch]

I will check it out. I´ve heard ansible might be a good option for it?

[tawny temple]

Yes. In fact, Ansible ultimately operates using SSH. If you simply need to move some files then Ansible is overkill. But if you have a lot of other things you need to provision on the server, then check it out.

[serene notch]

cool. I will check more into it and will post here what I was able to test. appreciate your time and answers @tawny temple

[finite crest]

hey guys i have a lot of python files which writes to a log files
and these python files mostly talk to a elasticsearch db
i was wondering about what would be the best approach to monitoring these logs
like getting an email or sms when i get a error or debug log level
I checkout out the logstash
but grafana does sound like a good option as well
or maybe a self hosted sentry
any ideas please?

[mystic void]

In general, I don't use logs for alerting if I can help it. This means:

• Metric/events for any quantifiable information such as golden metrics
• Logs/Traces for insight and debugging

Regardless, you can use some services like New Relic or Datadog, which have some free tiers and provide everything you need.

[heavy knot]

`print("1. Tool Help")
print("2. Tools List")

import keyboard
keyboard.on_press_key("1", lambda _: print("Test"), suppress=False)
while True:
pass`

I want that u only can type the "1" one time
to open the "Tool Help"

[carmine dirge]

Hi, given project structure (name-spacing):

root/
    sub1/
        namespace/
            __init__.py
        pyproject.toml
    pyproject.toml

Can I specify in the root/pyproject.toml somehow the local sub1/pyproject.toml as dependency?
(Couldn't find anywhere in documents. If someone can point out doc that says that, I'm all ears)

Additional question: Can I specify somehow sub*/pyproject.toml as dependency?

Essentially, what I want is to be able to do is pip install -e . in the root folder

[hallow gull]

Does anyone know how to cross compile a Python executable? I want to cross compile for aarch64-pi while on an amd64

[past hound]

➜  server sudo docker-compose up
WARNING: The PWD variable is not set. Defaulting to a blank string.
Removing server_caddy_1
Recreating e18c48c71aed_server_caddy_1 ... error

ERROR: for e18c48c71aed_server_caddy_1  Cannot start service caddy: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error mounting "/Caddyfile" to rootfs at "/etc/caddy/Caddyfile": mount /Caddyfile:/etc/caddy/Caddyfile (via /proc/self/fd/6), flags: 0x5000: not a directory: unknown: Are you trying to mount a directory onto a file (or vice-versa)? Check if the specified host path exists and is the expected type

ERROR: for caddy  Cannot start service caddy: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error mounting "/Caddyfile" to rootfs at "/etc/caddy/Caddyfile": mount /Caddyfile:/etc/caddy/Caddyfile (via /proc/self/fd/6), flags: 0x5000: not a directory: unknown: Are you trying to mount a directory onto a file (or vice-versa)? Check if the specified host path exists and is the expected type
ERROR: Encountered errors while bringing up the project.

version: "3.7"

services:
    caddy:
        image: caddy:latest
        restart: unless-stopped
        ports:
            - "80:80"
            - "443:443"
        volumes:
            - $PWD/Caddyfile:/etc/caddy/Caddyfile
            - $PWD/site:/srv
            - caddy_data:/data
            - caddy_config:/config

volumes:
    caddy_data:
        external: true
    caddy_config:

[brazen forge]

have you tried ./Caddyfile:/etc/caddy/Caddyfile?

[past hound]

yes tried that also

didnt work

[brazen forge]

because from the error message it seems like Docker Compose doesn't recognize $PWD

[past hound]

https://hub.docker.com/_/caddy

im following the exact steps in the docs

if i remove the volumes in docker-compose file then everything works

[brazen forge]

what OS? what shell?

[past hound]

if i change $PWD to just a .

then the WARNING: The PWD variable is not set. Defaulting to a blank string. goes away

but container still doesnt start

im using linux

[brazen forge]

hmm. $PWD should be set automatically by the shell

[brazen forge]

does it still throw the same error?

[past hound]

how do i know which shell im using?

yes same error

[brazen forge]

echo $SHELL

[past hound]

im using zsh

[brazen forge]

hmm. straaange

[past hound]

/usr/bin/zsh

[brazen forge]

can you run

echo $PWD

[past hound]

this is like the 10th time im trying to learn docker

and its stuff like this that meakes me want to giveup

ok

$PWD => /home/rickyjw/odinproject/test-app/server

[brazen forge]

hmm

[past hound]

yes hmm

[brazen forge]

could you try

PWD=$(pwd) sudo docker-compose up

[past hound]

k one sec

same error

i dont think its a $PWD issue

cause even if i replace $PWD with a . its the same error

some it must be something with the way im creating the volumes

you know what im trying to say

[brazen forge]

with the ., what exact error do you get?

[past hound]

but like i said im just following the docs

the error that I pasted above

➜  server PWD=$(pwd) sudo docker-compose up
[sudo] password for rickyjw: 
Removing server_caddy_1
Recreating e18c48c71aed_server_caddy_1 ... error

ERROR: for e18c48c71aed_server_caddy_1  Cannot start service caddy: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error mounting "/home/rickyjw/odinproject/test-app/server/Caddyfile" to rootfs at "/etc/caddy/Caddyfile": mount /home/rickyjw/odinproject/test-app/server/Caddyfile:/etc/caddy/Caddyfile (via /proc/self/fd/6), flags: 0x5000: not a directory: unknown: Are you trying to mount a directory onto a file (or vice-versa)? Check if the specified host path exists and is the expected type

ERROR: for caddy  Cannot start service caddy: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error mounting "/home/rickyjw/odinproject/test-app/server/Caddyfile" to rootfs at "/etc/caddy/Caddyfile": mount /home/rickyjw/odinproject/test-app/server/Caddyfile:/etc/caddy/Caddyfile (via /proc/self/fd/6), flags: 0x5000: not a directory: unknown: Are you trying to mount a directory onto a file (or vice-versa)? Check if the specified host path exists and is the expected type
ERROR: Encountered errors while bringing up the project.

[tawny temple]

What is Caddyfile?

Note

are you trying to mount a directory onto a file (or vice-versa)?

The mount paths must both be directories, not files.

[past hound]

they are directories

[brazen forge]

Caddyfile is a directory?

isn't it a file?

[past hound]

I can cd into it

[brazen forge]

[past hound]

so doesnt that mean its a directory?

[brazen forge]

what are the contents of Caddyfile?

[tawny temple]

cd into it where?

On the host or in the container?

[past hound]

in my host

its in the same folder as my docker-compoes file

[tawny temple]

I believe the error is stating that the destination (i.e. within the container) is not a directory.

[past hound]

• ./Caddyfile:/etc/caddy/Caddyfile

[brazen forge]

Caddyfile is supposed to be a file AFAIK

[past hound]

so how would I fix that line

[brazen forge]

could you answer this?

[past hound]

like I said, im just copy/pasting frm the caddy docs

its empty

[brazen forge]

uhh

[tawny temple]

Sorry I misspoke. They can be files but they have to both be files. Or they have to both be directories.

For caddy I think it should be a file

[brazen forge]

yea

make Caddyfile a file

something like https://github.com/caddyserver/dist/blob/master/config/Caddyfile

[past hound]

im so confused

make a caddyfile where?

[brazen forge]

could you show your directory structure as it is currently?

tree .

[past hound]

➜  server ls -l
total 44
drwxr-xr-x. 1 rickyjw rickyjw     0 Jul  6 15:31 caddy_config
drwxr-xr-x. 1 root    root        0 Jul  6 14:59 Caddyfile
-rw-r--r--. 1 rickyjw rickyjw   381 Jul  6 15:40 docker-compose.yml
-rw-r--r--. 1 rickyjw rickyjw     0 Jul  6 14:47 Dockerfile
-rw-r--r--. 1 rickyjw rickyjw    12 Jul  6 14:54 index.html
-rw-r--r--. 1 rickyjw rickyjw   268 Jul  5 17:52 index.js
drwxr-xr-x. 1 rickyjw rickyjw  1646 Jul  5 17:51 node_modules
-rw-r--r--. 1 rickyjw rickyjw   347 Jul  5 17:51 package.json
-rw-r--r--. 1 rickyjw rickyjw 27956 Jul  5 17:51 package-lock.json
drwxr-xr-x. 1 root    root        0 Jul  6 14:59 site

[tawny temple]

First, do you need or want to override the default Caddyfile? The mount is optional. So if you don't need it, don't do it. If you do need it, then for -v $PWD/Caddyfile:/etc/caddy/Caddyfile, $PWD/Caddyfile should point to a file on the host.

[brazen forge]

basically, delete the Caddyfile directory and make a file named Caddyfile

[past hound]

@tawny temple I just want to follow the official caddy docs

to get something working

im not trying to do anything advanced here

[tawny temple]

Well on Dockerhub they imply that overriding the Caddyfile is optional. But I know nothing about Caddy.

[past hound]

@brazen forge ok give me a sec

[brazen forge]

well, mounting a custom Caddyfile would be the advanced thing to do here

[past hound]

ok so something interesting happened

so i deleted the Caddyfile folder and created a new Caddyfile file

and now I get this

➜  server PWD=$(pwd) sudo docker-compose up
[sudo] password for rickyjw: 
Removing server_caddy_1
Recreating e18c48c71aed_server_caddy_1 ... done
Attaching to server_caddy_1
caddy_1  | {"level":"info","ts":1657137329.5013137,"msg":"using provided configuration","config_file":"/etc/caddy/Caddyfile","config_adapter":"caddyfile"}
caddy_1  | run: adapting config using caddyfile: EOF

does that mean its working?

[brazen forge]

probably yea

but like Mark said, you probably don't need to mount a custom Caddyfile for basic usage

[tawny temple]

I kind of suspect it isn't working properly if you created a file without putting anything inside it

[past hound]

its not working

cause when i got to localhost, it gives me the page not found

normally it would bring up the default caddy page

[tawny temple]

Then don't overwrite the caddyfile

Until you know what you want to put in there

[past hound]

you told me to create a Caddyfile?

[tawny temple]

#tools-and-devops message

No

#tools-and-devops message

[past hound]

#tools-and-devops message

am i wrong?

[tawny temple]

That instruction was likely given in the context of you wanting to override the config. But as I suggested, you don't have to do it; it's an optional step.

[past hound]

i dont want to override anything

i jsut want to get it to work

if there is anyone in here that knows anything about docker please just send me a DM. thanks

[tawny temple]

We already gave you a solution.

Don't overwrite the config. That mount is optional

[past hound]

dude im just followingg what i was being told

he said to delete the directory and to create a Caddyfile

thats what i did

i dont know if it overrides some other file

im obviously new to all of this

[tawny temple]

That was said to you before it became evident that you didn't want to override the config. It was (perhaps mistakenly) taken for granted that you understood that it was an optional step, so they were following with that assumption and just giving you instructions on how to correctly overwrite the Caddyfile. After that was given, I put into doubt whether you actually need to overwrite it, and evidently you don't; you didn't realise it was optional. It was a miscommunication but I hope that clears it up.

[past hound]

ok so then what was the solution?

cause im still lost

[tawny temple]

This option creates a bind-mount which overwrites the image's Caddyfile with the file on your host -v $PWD/Caddyfile:/etc/caddy/Caddyfile. Just remove that and it will revert to using the default Caddyfile in the image.

[past hound]

so i delete the Caddyfile?

[tawny temple]

On your host, you can delete it for the sake of tidiness but it doesn't matter really. The important part is that it's not mounted i.e. that the option I just mentioned is not used.

[past hound]

yea but i really want to just follow along with the docs says to do

version: "3.7"

services:
  caddy:
    image: caddy:<version>
    restart: unless-stopped
    ports:
      - "80:80"
      - "443:443"
    volumes:
      - $PWD/Caddyfile:/etc/caddy/Caddyfile
      - $PWD/site:/srv
      - caddy_data:/data
      - caddy_config:/config

volumes:
  caddy_data:
    external: true
  caddy_config:

[tawny temple]

This is all the docs have to say on the matter in the "Basic Usage" section

To override the default Caddyfile, you can mount a new one at /etc/caddy/Caddyfile: ...

[past hound]

yea sure I can delete the volumes to make things work

but id like to understand also how to get it to work using volumes as well

[tawny temple]

That instruction is basically saying that if you have a reason to override the Caddyfile, here is how you do it. It's not saying you should unconditionally do it. In your case, you just want the basic set up so there's no reason to overwrite it. Just remove this line - $PWD/Caddyfile:/etc/caddy/Caddyfile.

If you did want to override it, then you basically had the mounting part figured out. The only problem was that your Caddyfile was empty. You'd have to read their instructions to know what to put inside. But as you said, you're not trying to do anything advanced so why make this harder for yourself?

[past hound]

ok let me make sure im understanding this corectly

i only specify that line if I plan to overwrite the original Caddyfile?

[tawny temple]

Correct

[past hound]

if not then I simply just dont include that line?

[tawny temple]

Correct

[past hound]

ok one other thing

i was following this tutorial

s

https://www.youtube.com/watch?v=2oNsjyaCIrI

at 4:54 you can see that includes that - $PWD/Caddyfile:/etc/caddy/Caddyfile`

in his docker-compose file

is that because is overwriting the Caddyfile?

because late ron in the video he creates a Caddyfile

to do some reverse proxy stuff

[tawny temple]

Yeah, if you want to follow this tutorial you will end up overwriting the Caddyfile. So you'll want to keep that line.

And it also shows what to put inside the Caddyfile, so that solves the problem of having an empty one

[past hound]

ok makes sense now

the docs was confusing me but it was probably just my lack of knowledge

anyway thakns dude. i really appreciate your help

i dont know where the other dude went but thank you too

[tawny temple]

You're welcome

[finite crest]

Oooh, I'll take a look at those
Thanks man

[dry vale]

What is the best way to create a simple API with Python and publish it?

I already have the "base" for it

The base being this code: https://github.com/gtkacz/nba-headshot-api/blob/main/source_code/crawler.py, that will crawl the NBA.com website for every player in the league and store their ID, and therefore can generate an URL for the official headshot for every player using their ID

[buoyant quarry]

Can someone help me, I succesfully made a list generate a random text. But it only regenerates it when I start my script, I want it to do it every second or so. Does anyone know how I can get this?

These are some parts of my code:

random_rose = random.choice(random_rose_list)```
```    client.set_sounds({
        "rose": random_rose,```

[rapid sparrow]

Has anyone was able in vscode to point as a python venv to be inside docker container? 🤔 (docker-compose)

UPD: Nevermind; Attached. It is possible and quite good. I think better than dealing with venv ;b

[velvet elk]

I want to learn some programs

I can make

I am a basic user

[thorny shell]

!kindling

[rancid schooner]

Kindling Projects

The Kindling projects page on Ned Batchelder's website contains a list of projects and ideas programmers can tackle to build their skills and knowledge.

[wooden ibex]

Why would you?

just install everything into the docker container

[rapid sparrow]

oh, i mean to say, how to point for vscode to use python environment from docker container xD

so that vscode would apply syntax recognition from installed libraries and etc

nevermind, it all works

[woeful goblet]

can someone pls code a youtube downloader in Py for me?

[brazen forge]

nope because 1) that's against YT's ToS 2) we don't just write some code for people here

[woeful goblet]

okay but thx for the reply

[rapid sparrow]

also there is already youtube downloader https://youtube-dl.org/, it is already even having python integration

hail FOSS

[olive path]

Any tips on enforcing unspoken and perhaps somewhat weird style rules? I have a project of mine I want to make sure upholds a set of guidelines I try to follow, but I am not sure I can easily find a linter which does this.

Essentially, I am wondering what would be the easiest way to define very custom styling rules. Is making my own linter very difficult?

[thorny shell]

I'd expect so 😐

[olive path]

😔 To give an example... I want to ensure that the order of __all__ is the same as the definition of the names inside of it

[thorny shell]

that, by itself, should be easy; at least for function definitions

each function has some attribute that tells you the file and line number.

[olive path]

Right, this will only really be for functions and classes

[iron basalt]

I'd probably look at what it would entail to write a flake8/pylint plugin - I imagine it will do most of the work for you and you just hook in your checks at some point, but it's something I've never dabbled in

[rapid sparrow]

Considering that python gives away a lot of information in Dunder methods.
And that we need just throw away single scripts for one thing only
I would expect it should be easy to write your own single time linting scripts

Most of desired linting should be within the grasp of some regex

[thorny shell]

🤣

[finite fulcrum]

is there some pure python line profiler?

[clear sorrel]

Any python developer here knows how to use autodocs in sphinx?

https://stackoverflow.com/questions/72909680/how-to-use-autodoc-in-sphinx-to-show-docstrings

[brazen forge]

!pypi pprofile

[rancid schooner]

pprofile v2.1.0

Line-granularity, thread-aware deterministic and statistic pure-python profiler

[wet dune]

I made docs in sphinx so yeh

[pearl goblet]

https://medium.com/@lminhkhoa/how-to-build-a-fully-automated-web-scraping-pipeline-for-dashboard-742b6dce9f0f

[umbral jay]

anyone used pycharm community ver for Django? Is it good enough? Pycharm pro version has special support for Django. Is it worth going for it, in your experience?

[neon swan]

Anyone familiar with Poetry? I've gone from poetry just working without manually activating environment's and being able to call one project from another to Poetry stropping all the time and having to delete, recreate and reinstall environments. I'd like to go back to seamless usage. Any tips?

[tardy island]

hey any recommended advanced git tutorial?

Like for people who have comfortably clone/pull/commit/push/merge for years and now want to learn about more messy stuff like branching, rebase etc?

[brazen forge]

by branching, do you mean branching strategies?

[brazen forge]

also, have you been through the Pro Git book? https://git-scm.com/book/en/v2

[tardy island]

Nope

This looks good

So basically I know stuff up to chapter 2

I'll start reading from chapter 3 i guess

thx

[brazen forge]

I think the more advanced stuff is in Ch. 7

[rapid sparrow]

go through https://learngitbranching.js.org/?locale=en_US interative tutorial and it gives visual understanding what happens under the hood
Fun and fast, and highly useful

[thorny shell]

It's neither "advanced" nor a tutorial, but I found https://eagain.net/articles/git-for-computer-scientists/ super useful at first

[rapid sparrow]

I am trying to run script `python3 src/scrappy/manage.py
https://github.com/darklab8/darklab_darkbot/blob/master/src/scrappy/manage.py

But i get this error

darklab_darkbot$ python3 src/scrappy/manage.py 
Traceback (most recent call last):
  File "/home/naa/repos/pet_projects/darklab_darkbot/src/scrappy/manage.py", line 1, in <module>
    from .database import engine
ImportError: attempted relative import with no known parent package

how to how to resolve those importing problems? xD

file has code like

from .database import engine
from .players.models import Base

Base.metadata.create_all(bind=engine)

I hoped relative import will work

i altredy tried absolute importing too

how can I debug it even?

Well. I was able to solve it no idea how 🤔

import database as database
import players.models as models

models.database.Base.metadata.create_all(bind=database.engine)

changing imports to this apperently solved issue

i wonder if i can use somehow imports starting from folder src

[full bronze]

You might be able to run it as a module with the -m flag

[heavy knot]

I constantly forget to pull after I've pushed to one of my own repos with a github action that commits changes to the repo

Is there a tool that maybe reminds me to pull or pulls for me

It makes merge commits after I forget with isnt great

[thorny shell]

not that I know of

you can tell git to prohibit merge commits, though, iirc

so you'll get an error rather than a merge commit

[heavy knot]

That could work ill look into that

[thorny shell]

put this in ~/.gitconfig

[merge]
        ff = only

I have written (in python!) a tool that fetches from all my repos more or less at once, and I run that a few times per day

nice and fast

https://gitlab.com/offby1/homedir/-/blob/master/local/bin/fetch-em-dano.py

[rapid sparrow]

when I try launching program from root level of folder src/scrappy with uvicorn main:app.py
i get error with relative import ImportError: attempted relative import with no known parent package

how to resolve it?

https://github.com/darklab8/darklab_darkbot/tree/master/src/scrappy

that's how stuff is there

so far as solution i was only able to launch the folder from parent one. But I wish to try launching it from the root of this scrappy folder

[tawny temple]

That does not seem appropriate

scrappy is a package since it has __init__.py so really it should be scrappy.main:app. And in main.py you should be doing from scrappy.players import views as player_views

It doesn't make sense to launch from within scrappy because it's a package

Also I don't think src should have __init__.py

And definitely not the root of your repository

You would have to launch with src as the cwd. But typically the intent behind using the "src" directory structure is that you plan to install your scrappy package. If you do install it, then you'd be able to run from the project root, or from anywhere really.

[rapid sparrow]

That's helpful

this explained everything

I just deleted __init__.py file from scrappy folder

and it started to work under different importing rules

which resolved everything correctly now

[tawny temple]

That is not the appropriate solution

scrappy should be a package

so it should have __init__.py

[rapid sparrow]

why? Scrappy is a microservice, it is not a package

it makes sense for it to be not a package

Scrappy is a fully fledged application

with its own dockerfile

launched as uvicorn main:app

[tawny temple]

It's a package in the sense of the Python import system not in the sense of publishing something to pypi

[rapid sparrow]

urgh. my intent for usage of src was to be having absolute imports along the project, which are less messy than this situation xD

[tawny temple]

Well I don't quite follow your intent or directory structure

[rapid sparrow]

I mean...

[tawny temple]

It looks strange to me

[rapid sparrow]

picture in architecture folder explains

it shows future plan of infrastructure

src is having 4 different microservice parts

which are going to be in the project

[tawny temple]

Usually all your Python files are within a package (i.e. a folder that has an __init__.py) and everything else like dockerfile, requirements.txt, etc are outside in the project root

This is quite standard regardless of whether it's a library or an application

[rapid sparrow]

erm. The confusion is because, technically each folder of src should be having its own repository

I simplified situation and made them all in one repo

each src folder, would be having (potentially) its own dockerfile

as they are all different apps to pack

therefore they have their own requirements/constraints files

[tawny temple]

I'm not familiar with all the files (like what manage.py and databases.py) are for, but I would structure it like this

src
    scrappy
        scrappy
            alembic
            bases
            core
            forum_records
            players
            __init__.py
            main.py
        tests
            __init__.py
            conftest.py
            test_whatever.py
        .dockerignore
        alembic.ini
        dockerfile
        requirements.txt

If you have some python files that are more like scripts for development/deployment rather than a core part of the app then those can live in a "src/scrappy/scripts" or "src/scrappy/utils" directory

[rapid sparrow]

Oh yeah, I tried already to do that but encountered doubled initialization of ORM model that lead to error

Hopefully now since it is not a package, I will be able to structure it

[rapid sparrow]

I kind of half agree with your location of test files though

[tawny temple]

I'm not familiar with sqlalchemy so I don't know why that would happen for you

[rapid sparrow]

I think relevant test code should be in same sub package where it belongs

[tawny temple]

Yeah, that's fine. It's suggested by pytest as an alternative.

https://docs.pytest.org/en/7.1.x/explanation/goodpractices.html#tests-as-part-of-application-code

[rapid sparrow]

kind of nice that I encounter such problems. Good that I finally went to FastAPI

in Django i remained blissfully ignorant to such problems xD

everything worked out of the box

[tawny temple]

Yeah it generates a project structure for you. You could deviate from it but most stick to defaults

[rapid sparrow]

i will have better understanding of the stuff that happens not in the box now

i wish to make a full comparison of stuff that is better or worse in fastapi

and seeing things manually happening in fastapi and effort i need to make in order to make them good again