tools-and-devops

[upper oyster]

Yeah you can't use git reset

That won't work because the history after the reset commit still exists and is still shared

[sand thistle]

so the only way is to sort through these changes

the problem is that there are so many, literally it was a branch dedicated to refactoring

that got pulled into the test branch prematurely

and i accidentally commited over with this pulled version

my master branch is up to date

so id like to just replace the current test branch, with the content of the master branch

ive done so locally

by deleting the test branch

and pulling from the remote master branch, after recreating test

[upper oyster]

But the test branch is also a shared branch, right?

[sand thistle]

shared, yes other people have access

and pull/push to it

[upper oyster]

So you can't just delete it if others are using it

[sand thistle]

i tried to reset the head

to the prior commit

[runic patrol]

this is what happens when I use git push --all in the pycharm terminal```py
To https://github.com/ChiliPy/fuck-me.git
! [rejected] master -> master (fetch first)
error: failed to push some refs to 'https://github.com/ChiliPy/fuck-me.git'
hint: Updates were rejected because the remote contains work that you do
hint: not have locally. This is usually caused by another repository pushing
hint: to the same ref. You may want to first integrate the remote changes
hint: (e.g., 'git pull ...') before pushing again.

[sand thistle]

the last line

[upper oyster]

And you can't do that as well because the other commit history after that commit is shared

You can use something like git cherry-pick instead

If you cherry pick a commit

It should recreate any commit you specify but at the tip of the branch

Which is what I think you want. You want to basically undo all the changes in the branch, but you need to do so in a way that doesn't fuck up existing shared history

I think cherry-pick would do it

[sand thistle]

its only 1 commit ahead

[upper oyster]

Yeah but that doesn't matter

If it's shared history you need to keep it

Otherwise you'll cause issues for the next person who wants to commit on top

If you cherry pick your original commit, that should do it

[sand thistle]

ok

https://www.atlassian.com/git/tutorials/cherry-pick

The --no-commit option will execute the cherry pick but instead of making a new commit it will move the contents of the target commit into the working directory of the current branch.

i guess i do have to commit though

[upper oyster]

Yeah, you can use that option if you are unsure of the final result

But obviously you will have to commit it eventually

[sand thistle]

         \
           e - f - g Feature```

the problem in the example

is not equivalent

         \
           e - f - g Feature```

[upper oyster]

So you don't have to cherry pick from another branch

You can cherry pick any commit, I believe

[sand thistle]

so you're saying i could

e - f - g Feature move f ahead of g

[upper oyster]

Yes I believe so

[sand thistle]

but wouldnt the ideal solution be to get rid of g

[upper oyster]

No lol

There is a golden rule of Git

One that you must never break if you are working with others

Never change shared commits!

If commit g was only on your machine, it would be no problem

You could git reset to e and get on with your life

[sand thistle]

but that only matters if anyones pulled

[upper oyster]

But if commit g is shared, it must stay there

Well eventually someone will want to commit on top of the branch no?

[sand thistle]

so then i recommit

ahead of the bad one

[upper oyster]

No but the history is messed up

It's not the same commit object

[sand thistle]

but cherry picking messes up the history anyway?

[upper oyster]

No it doesn't

e f g still exist

[sand thistle]

but how, in this situation im literally pulling an old commit ahead of a new one

[upper oyster]

f is still there in history

You create an f'

So e f g f'

[sand thistle]

i see

[upper oyster]

f is still in Git history

f' is a brand new commit object, but with contents of f

[sand thistle]

ok cool

mind if you walk me through this?

[upper oyster]

Lol sure

[sand thistle]

i found the sha

i need

[upper oyster]

Ok

I would suggest reset to the tip of the branch first

And then do the cherry pick

[sand thistle]

how

git reset sha for the tip?

[upper oyster]

I think you can reset to HEAD

I think lol

If you do git log

Is HEAD pointing to the tip of the branch?

Heck, even if it isn't, you can just reset to the tip of the branch using the SHA

[sand thistle]

uhh

[upper oyster]

The idea is that you want to be equivalent to the GitHub state of the branch

[sand thistle]

doing git log

[upper oyster]

Actually scratch what I said

You want to git reset to the last commit from GitHub

For your branch

[sand thistle]

yea

so just git reset sha

[upper oyster]

So the origin equivalent of your branch

Yeah

[sand thistle]

ok

done

went onto the test branch on my github

found the commit from that branch that i want to go back to

copied the sha

git reset sha

[upper oyster]

Yeah, that should bring you to the tip of the test branch from GitHub

And then you can cherry-pick the commit and place it on top

[sand thistle]

ok

just fyi

im cherry picking the same sha

[upper oyster]

Ok

Btw

[sand thistle]

You are currently cherry-picking commit 057b4df.

nothing to commit, working tree clean
The previous cherry-pick is now empty, possibly due to conflict resolution.
If you wish to commit it anyway, use:

    git commit --allow-empty

Otherwise, please use 'git cherry-pick --skip'

[upper oyster]

Huh

Wasn't expecting that lol

[sand thistle]

uhh

idk how to proceed

[upper oyster]

did you reset to the correct commit?

Perhaps you resetted one commit too far back

[sand thistle]

uhh

i looked at the commit i want to move to

i copied the sha

i did git reset sha

[upper oyster]

oh you don't want to reset to the commit you want to move to

if you have e f g

and you want to go to e f g f'

Then you reset to g and then cherry pick f

[sand thistle]

ohh

[upper oyster]

Btw did you say it was just one commit behind?

[sand thistle]

so

lets say

a b and c

im at c

my commit that i want to be the new tip of my code

is a

[upper oyster]

right

so if you are already at c you don't need to do any reset

and then you just cherry pick a

to create a b c a'

Btw, how many people are working on the test branch?

[sand thistle]

a few

why

[upper oyster]

so, if you wanted to make it so that master and test were the exact same commit object

then you could do git reset a and then force push it

However everyone would have to delete their test branch and recreate it

because you rewrote shared history

[sand thistle]

ok

so

hold on

[upper oyster]

so that is an option if you communicate that to people who are working on test

[sand thistle]

lets stick to one thing

[upper oyster]

lol ok

[sand thistle]

error: could not apply 057b4df... working on db issues
hint: after resolving the conflicts, mark the corrected paths
hint: with 'git add <paths>' or 'git rm <paths>'
hint: and commit the result with 'git commit'

seems some merge conflicts

i resovled them

but its talking about a path

it means just git add the filename im assuming

[upper oyster]

yeah

[sand thistle]

yeah but like..

this version that im on

the tip

it still has all the code and stuff that im not trying to commit

[upper oyster]

What code? The code on top of the cherry pick?

[sand thistle]

i have a b c

i reset to c

c has code i dont want

i cherry pick a

fix merge conflicts

but now there's still the content from c

[upper oyster]

are you talking about stuff you didn't commit previously?

[sand thistle]

no

im talking about stuff

[sand thistle]

wait what do you mean

use the symbols a b c

so i know what you mean

[upper oyster]

so, if you are at c, you may have also started developing on top of c

[sand thistle]

no

its not uncommitted changes

i reset to c

that means i have the code that exists at c

[upper oyster]

Not quite

because by default a git reset does not wipe away all uncommitted changes

you can have soft, mixed and hard git reset

by default, git reset is mixed, which means that all staged changes are unstaged

but they aren't completely removed

they still exist in an unstaged state

[sand thistle]

i see

[upper oyster]

if you really want to get rid of those changes, you can do this

First do a git stash to save a snapshot in case you screw things up

then, you git reset --hard to c

That will reset to c and remove all changes, even those ones that were unstaged previously

then you can cherry pick again

if you commit the cherry pick, then you could reapply your unstaged changes by doing git stash apply

if you wanted to

[sand thistle]

ok so

i did a hard reset

to the tip

then tried to cherry pick

and ok fixed the merge conflict

but what i was saying, and trying to explain

is

since i am on c

the garbage commit

i still have the code from c

even if i fix the merge conflicts

literally the code im looking at in my ide, is that of c

which is not what i want to add

[upper oyster]

so how are you fixing the merge conflicts? When you have a merge conflict you basically want to take the a version completely

you don't want to keep anything from c

[sand thistle]

yes

but i only have merge conflicts in one .py file

my models

the system is flagging only a merge issue on a

[upper oyster]

ok

Once that's fixed though, everything should look like a

[sand thistle]

it does not

[upper oyster]

hmm, not sure how that's happened then

This is literally what cherry pick is meant to do, recreate the contents of a

[sand thistle]

alright thank you

i have to go now

[upper oyster]

OK

Sorry I couldn't help completely there

[sand thistle]

9ts ok appreciate the time

[tight moss]

Hi everyone, I have a question which I think falls under tools-and-devops : after uninstalling poetry, does this mean that all the dependencies from the apps/venvs made using poetry also got deleted? or do i now have to hunt around for them? i think i pulled the trigger too soon ^^;

it does say here https://stackoverflow.com/questions/50267732/what-will-happen-to-packages-installed-when-the-virtual-environment-is-deleted that YES they do get deleted, but this is my first time with poetry. i have previously used virtualenv and virtualenvwrapper 🙂

[leaden ruin]

||url = 'https://accounts.spotify.com/login/password'
PostAtma = {"username": data[0]
"password": data[1]}||

    "password": data[1]}
    ^
SyntaxError: invalid syntax``` how can i fix it?

[night quest]

Use comma , like {"username": data[0], "password": data[1]}

[leaden ruin]

ty

url = 'https://accounts.spotify.com/login/password'
PostAtma = {"username": data[0], 
            "password": data[1]}```
Traceback (most recent call last):
  File "/storage/emulated/0/Download/spotify-checker/spotify.py", line 36, in <module>
    "password": data[1]}
IndexError: list index out of range

[candid kayak]

Heyo, was trying to poetry add more-itertools and got this error - https://mystb.in/FirewallHolyGonna.sql
What is going on?

[heavy knot]

hello when every i open my python file it just closes rightaway

[pearl jungle]

that means your getting an error

[heavy knot]

how do i fix

its a txt file

@pearl jungle

[pearl jungle]

your python file is a txt file...?

[heavy knot]

idk

how do i change

[rancid schooner]

You are not allowed to use that command here. Please use the #bot-commands channel instead.

[sly sleet]

@heavy knot um just rename it?

and if you double-click from file explorer or whatever

the window is going to close when the program ends

so instead run it from command prompt/powershell

[vital epoch]

hey folks, is there a way or a tool to inspect what's occupying memory at any point in time? I can use trace malloc to identify what's allocated during a certain period of time, but I expect much of that to be garbage collected after the work is done. Ideally I could find a tool that let's me snapshot exactly what's taking up memory at one point in time so that I can verify that I'm not hanging onto stuff that should have been GC'd

[rugged hare]

Are you facing a memory leak? That is, is your process continuing to grow?

[vital epoch]

appears to be yea

it looks like I'm probably not dropping some references that I should be dropping

but I'm not sure how to track them down

[rugged hare]

@vital epoch what does your program do?

[vital epoch]

generally coordinates a few independent processes via socket communication and sends some REST request up to the cloud

leans very much on asyncio

I'm wondering if maybe, when I asyncio.create_task if I need to be explicit about how and when to clean that task up

is there a way more generally to just see where I'm holding references?

interesting, when I use python's gc module to print stats, I often see this: gc: done, 460 unreachable, 0 uncollectable, 0.0596s elapsed

very curious to see exactly what is unreachable

[frozen anchor]

I have a Github action that connects to a database. I don't want the DB information to be public.
For my dev and prod environments, I use env variables and would like to be able to do that on a github action as well.

I have set up an environment in github, lets call it myenv. How do I tell my github actions worker to select that env for the action?

postgres:
        image: postgres:13
        env:
          POSTGRES_USER: ???
          POSTGRES_PASSWORD: ??? 
          POSTGRES_DB: ???

what do I have to enter here?

Does this also allow for os.getenv or are there more steps to take?

'NAME': os.getenv('DATABASE_NAME'),
'USER': os.getenv('DATABASE_USER'),
'PASSWORD': os.getenv('DATABASE_PASSWORD'),

Full github action code: https://paste.pythondiscord.com/ewuwiraquw.yaml

[rugged hare]

@frozen anchor you can create secrets in your repo, and use them in actioins

[uncut quartz]

How can I get the all time downloads of a pip package?

I have been using https://pypistats.org/ but it is not enough

[frozen anchor]

I was already that far. Question is - how do i access them? How does actions know which of the X environments to pick? I found nothing about that in 4 different docs

[rugged hare]

i guess you'll need differently named secrets for each environment? How many wilil your action be using?

[frozen anchor]

for now the 3 env variables for the DB but once I implement more test cases that number will increase.
Different names for each env sounds off tho - lets say I have 2 github actions workflows that need the same env variables named differently - for example access to the DB with different permissions - that would not be possible. I doubt that this is the case tbh

[rugged hare]

why would more test cases need more environemnt variables?

[frozen anchor]

for example to check if the email validation for a newly created account works

[rugged hare]

Generally, the db credentials don't change for each new user of the application

[frozen anchor]

found it tho, it is the environment keyword, overlooked that I guess

[frozen anchor]

not talking about DB here, my dev and prod env have EMAIL_NOREPLY_USERNAME and EMAIL_NOREPLY_PASSWORD that are used to send emails from [e-mail]

[rugged hare]

i guess i don;t understand what your X different environments are.

or why your REPLY_USERNAME needs to be a secret for your tests

you can set environment variables in your tests, not in actions

[frozen anchor]

well, it does not have to be secret for those - but as the project is public I prefer to not give away that much information of how my backend looks like

[rugged hare]

is your code public?

[frozen anchor]

yes, hosted on github

[rugged hare]

what information are you trying to not give away?

if your code is public, why do you need any secrets for the tests?

[frozen anchor]

using env variables for the tests is perhaps overkill, you are right about that

but I don't want to make another conditional in settings.py.
I want to have it managed via env variables in all cases like this:

    DATABASES = {
        'default': {
            'ENGINE': 'django.db.backends.postgresql',
            'NAME': os.getenv('DATABASE_NAME'),
            'USER': os.getenv('DATABASE_USER'),
            'PASSWORD': os.getenv('DATABASE_PASSWORD'),
            'HOST': '127.0.0.1',
            'PORT': '5432',
        }
    }

and not

if DJANGO_HOST == 'github_action'
....
    'NAME' = 'postgres'

[rugged hare]

ok, so you will have a half dozen variables.

[frozen anchor]

I have these variables for dev and prod either way, so I don't think it matters that much if I use env vars for actions as well.

[rugged hare]

right, it's fine to have some secrets for actions, but you were talking about needing more and more as you added tests. i dont see that happening.

also, btw, a db password for tests? what db are you connecting to?

it's very common to have a tests settings file

[frozen anchor]

no, I create a new DB, run migrations and then run my tests. This was just an example of why you might want secrets for actions.

[rugged hare]

i see

[frozen anchor]

So, you are right and I should not use secret env variables. Instead I went for

    env:
      DATABASE_USER: postgres
      DATABASE_PASSWORD: postgres
      DATABASE_NAME: github_actions

[pallid cove]

hi, i have a data engineering question. i am storing and updating many pandas dataframes

hdf5 looks ideal, but im concerned with corruption

is something like storing each dataframe as a sql blob a more reasonable strategy?

[weary parcel]

Hello, I'm looking for someone who knows a lot about the selenium python's tool and especially about loading DRM-protected content in headless mode

gonna sleep asap feel free to dm or ping me btw

[rugged hare]

what are you testing?

[weary parcel]

@rugged hare some browser automation stuff, i used to test my program in "classic" selenium mod because i thought that headless mode was working the same, but apparently headless mode doesn't have feature that classical one have that's why i want to learn more about how to make things like protected contents works

[rugged crescent]

Guys, is there an open source for http://pythontutor.com/visualize.html#mode=edit ? I can't find it. Woo-hoo.

[spiral valve]

someone got tips for pycharm

[buoyant atlas]

Hi

did anyone installed application that has setup.py in docker image?

[night quest]

Just call python setup.py install

[buoyant atlas]

i figured this out, somehow run pip install .

did not install requirements that i specified in setup

[night quest]

Can you show your setup.py and setup.cfg?

[buoyant atlas]

from setuptools import setup, find_packages


def read_req():
    with open("requirements.txt") as req:
        content = req.read()
        requirements = content.split("\n")


setup(
    name="task",
    version="0.1",
    packages=find_packages(),
    include_package_data=True,
    install_requires=read_req(),
    entry_points="""
        [console_scripts]
        task=task.cli:cli
    """,
)

[night quest]

Your read_req doesn't return anything

[buoyant atlas]

you are right 😄

thanks

[night quest]

Your welcome!

[proven arch]

hey does anyone know the best coding software to use

[night quest]

I like PyCharm, many other use VS Code for example

[deep estuary]

Hey folks! Just a note that we've added a channel for editors and IDE discussion! #editors-ides

[steep sonnet]

Hi guys , i'm trying to deploy an app to apache2 with mod_wsgi but i'm getting an ImportError , and I don't know why ? May I get some help?

And tha'ts the apache conf

[abstract roost]

hello

i need bit of a help

how can i create a future order takeprofit

[ember canopy]

Hey all, got a question about paring down a requirements.txt file automatically but my reddit post got automodded: https://www.reddit.com/r/Python/comments/lpftn3/paring_down_pip_requirements/

[graceful patio]

well the text has been removed

from the post

[candid kayak]

What are your opinions on pipenv vs poetry?

for whatever reason pipenv seems to error out for me with an odd error (says path not found pointing at an old install of python that I've since deleted), and have been using poetry only

Am I missing out on anything great that pipenv offers that poetry doesn't?

[short peak]

I tried poetry in one project and I liked it. The only thing I'd love to see is to handle run as npm/yarn does so we can be able to simply run regular cli commands through poetry

[candid kayak]

🤔

I've liked poetry so far too fwiw, just wanna know if I should be using pipenv for some cases or if they're comparable on all fronts

[fossil finch]

Hello, I would like to ask something if what I'm doing is correct, this is basically a like a insert and get items list with redis caching

POST item in the list
-get the data
-insert in database
-insert in redis cache (with key of user_id_items_list)

GET all items list
-check if user_id_items_list exists
-if it exists, returns the cache value
-else, do query
-save the result of list in the redis cache with key of user_id_items_list

[stray pine]

Hi, is there an extension in vscode that automatically generates getter and setter methods?

[ember canopy]

Ah it's still there for me sorry

Basically I'm trying to automatically prune my requirements.txt down to just the immediate dependencies of the app

I don't want to list all of their dependencies as well

So for example, I want to list flask but not werkzeug

But this is an old project I inherited with a lot of deps so I'm wondering if there's a way to automate it

[heavy knot]

A good Calculator

Program make a simple calculator

This function adds two numbers

def add(x, y):
return x + y

This function subtracts two numbers

def subtract(x, y):
return x - y

This function multiplies two numbers

def multiply(x, y):
return x * y

This function divides two numbers

def divide(x, y):
return x / y

print("Select operation.")
print("1.Add")
print("2.Subtract")
print("3.Multiply")
print("4.Divide")

while True:
# Take input from the user
choice = input("Enter choice(1/2/3/4): ")

# Check if choice is one of the four options
if choice in ('1', '2', '3', '4'):
    num1 = float(input("Enter first number: "))
    num2 = float(input("Enter second number: "))

    if choice == '1':
        print(num1, "+", num2, "=", add(num1, num2))

    elif choice == '2':
        print(num1, "-", num2, "=", subtract(num1, num2))

    elif choice == '3':
        print(num1, "*", num2, "=", multiply(num1, num2))

    elif choice == '4':
        print(num1, "/", num2, "=", divide(num1, num2))
    break
else:
    print("Invalid Input")

[ember canopy]

Umm is this the channel for that?

[heavy knot]

I do not know

I am new here

[ember canopy]

This is for python related tooling and devops style tasks so no

[sly sleet]

this is python, not java ™️

if you're writting getters/setters for ordinary attrs, you're doing something wrong

[graceful patio]

ah yep it gets shown for you i think, no worries

and for your actual question, i've often wondered about this

never found a good way other than just starting with a fresh venv and installing everything again

doesn't seem to be a way to tell pip freeze to do that

[ember canopy]

yeah I guess I could iterate over the requirements and see which ones are listed as deps of others

[sly sleet]

@ember canopy hm checkout pipdeptree

[ember canopy]

😮

That looks like it'll be hella useful, thanks @sly sleet

[sly sleet]

np

[tired sluice]

Hey all. To learn Python i'm working down a list of 50 project idea's. I was planning on using GitHub to store them and keep track of everything. Is there an easy way to organise my learning projects, i.e, folder directories, or should I just prefix my repo's with python-learning or something?

[night quest]

I recommend to split your work into multiple repositories. When you are going to find a job then you can select few of them and send as attachements. If someone is going to be interested then is able to check all of them. When you have all in single repo then you have fewer repos but you should write some guideline what your repo contains because directories like game-2d or example-gui-calculator don't have enough information like used technologies.

[tired sluice]

So bundling all my little learning projects into one repo isn’t an issue aslong as I am descriptive? I’m not planning on sending any of this off with job applications, I’m just learning to be more productive in my current job as a server admin

[night quest]

Single repo has this advantage that you have all your learning stuff in one place so I think that it's not the problem in this case

However being descriptive is also good for you because after some time you can easily find interesting part of code even if you don't remember directory name

[tired sluice]

Very helpful. Thanks mate

[night quest]

Your welcome

[stray pine]

aha, lol. Thanks for that, completely forgot, Java is my first language.😩

[candid bobcat]

hey, i need some help. I have something to do in turtle, you can call it the sun. it consists of 120 elements and must be in a square. I can make this so-called sun, but I don't know how to make this 'trace' from the right side of the square. can someone help me?

[ivory perch]

Is this where we discuss about sphinx?

[night quest]

I guess yes

[ivory perch]

It looks like I did not get help on #help-avocado

I have an issue regarding the creation of documentation using sphinx. Sphinx runs fine on my local machine and generates API documentation using autosummary, but not on docker (Gitlab CI). It throws a warning saying failed to import (and no autosummary API documentation appears). Here is my project: https://gitlab.com/rebornos-team/fenix/libraries/analyzing. Since the documentation is generated fine on my computer, I think it may be a problem with Gitlab's package versions. I am not sure

I run a similar thing (with the same Gitlab script) on a different project: https://gitlab.com/rebornos-team/fenix/fenix-installer and for that project, there are no issues on Docker while generating documentation

On Gitlab CI/Docker:

On my local machine:

In order to help me troubleshoot, you can clone the project locally and try the documentation generation yourself (I even have a script there in the documentation directory). This is open source

[night quest]

Have you tried to run it inside Docker container manually?

[ivory perch]

That is a good idea. Let me set it up

[night quest]

You can run Docker container in interactive mode and see which components are missing

[ivory perch]

I just realized I have no idea how to set it up. I downloaded Docker and I pulled the Python image

[night quest]

docker run --name mycontainer -it python:3.9-alpine bash

Add something like -v $PWD:/code to create volume from current directory

[ivory perch]

Thanks. I will run it. I have never used Docker by myself. Always through Gitlab CI

[night quest]

I am always testing locally before I push changes to CI

[ivory perch]

$ docker run --name FenixTest -it python:3.9-alpine bash
Unable to find image 'python:3.9-alpine' locally
3.9-alpine: Pulling from library/python
ba3557a56b15: Pull complete 
ffd49e023448: Pull complete 
46ec0d464620: Pull complete 
44e992cdfdac: Pull complete 
7c71687af071: Pull complete 
Digest: sha256:95217fdd1431fe6230e033513670673c81a50debdb0065e54cbda261f3d76528
Status: Downloaded newer image for python:3.9-alpine
docker: Error response from daemon: OCI runtime create failed: container_linux.go:367: starting container process caused: exec: "bash": executable file not found in $PATH: unknown.
ERRO[0010] error waiting for container: context canceled 

Lol. I have bash. I am on Arch Linux

[night quest]

There are no bash inside container

[ivory perch]

Ah..

[night quest]

You are trying to run Alpine

Use sh

[ivory perch]

Too late. The container is running. Can I still enter a shell?

[night quest]

-it is shorthand from --tty and --interactive

[night quest]

Kill it docker container kill FenixTest

[ivory perch]

It is apparently not running. It has to be removed

[night quest]

You can check status with docker container ps --all

[ivory perch]

It says created.

[night quest]

Try docker run --name FenixTest -it python:3.9-alpine sh

[ivory perch]

$ docker run --name FenixTest -it python:3.9-alpine sh
docker: Error response from daemon: Conflict. The container name "/FenixTest" is already in use by container "cc9dbb7fcf2f3c142eecef8ab50e072cae47a824de096b1e9b951129c2647a27". You have to remove (or rename) that container to be able to reuse that name.
See 'docker run --help'.

Nevermind. Removed and restarting

Will test. The container is up. Thanks @night quest 🙂

[night quest]

It's not the end of this story!

Have you attached volume?

[ivory perch]

I did not. I downloaded my project from git and ran everything in the CI script.

Same issue as in Gitlab. I am testing it on a virtual machine now

[night quest]

You can use -v $PWD:/mycode to mount your currect directory to /mycode directory inside Docker container

[ivory perch]

Ah...How can I mount another directory?

[night quest]

-v /path/to/my/dir:/my-dir-inside-container

[ivory perch]

Good idea. That way I will know if it is the environment or the files that are at fault

[night quest]

I am not sure but is the project in your working path?

[ivory perch]

I do not know what a working path is, but I don't think so

[night quest]

If your project is in /path/to/my/project and you have fenix_library/analyzing here it can cause error when you try to generate docs from /path directory

[ivory perch]

I changed to the project directory and used PWD

So I generated the web pages for documentation. I need to test them. How can I get those files back on my host system?

[night quest]

From mounted directory?

[ivory perch]

Yes

[night quest]

It's already in

[ivory perch]

Oh

[night quest]

When you mount directory to Docker container you can easily interact with them from host or container

There are docker cp command too

[ivory perch]

Thanks. It seems like the problem only happens in that docker image. Even if I use the same project directory on my computer

I will have to see if there is an unstated dependency that I have installed on my local machine that docker image won't have

[ivory perch]

So. The problem is solved. There were two things: (1) pipenv,was not installing all the dependencies (especially the ones that are needed to check imports) from the lockfile because the lockfile was outdated. (2) The public folder in Gitlab CI was not empty and some files would not be updated

Thanks @night quest !

[night quest]

Nice!

Well done

[ivory perch]

Thanks for all your help 🙂

[night quest]

Your welcome!

[stable orbit]

I can install a private repo from Gitlab with pip install -e git+https://gitlab.com/user/project#egg=project. I've heard that with using tags, I can get a specific version. Is it possible without tokens, like this? git+https://${GITLAB_TOKEN_USER}:${GITLAB_TOKEN}@gitlab.com/user/project.git@{version}, but without token just like I did before?

[sand thistle]

!paste

[heavy knot]

why is it telling me "position" is not accessed?

[leaden tartan]

can you share the exact error?

[leaden tartan]

can you also include the line numbers in the screnshot?

[heavy knot]

nvm it works

[candid juniper]

@toxic bough Do you happen to use mac?

[toxic bough]

No

i dont

[candid juniper]

Linux or Windows?

[toxic bough]

pyenv works on mac
use homebrew to install it

Im on linux

[candid juniper]

Ok cool so a similar platform at least, usually the tools work mostly the same between linux and mac

Are there any other tools you use on linux you suggest I look into?

[muted notch]

Anyone familiar with Step Functions and choice conditions? Need to validate something real quick.
Is this the correct way to satisfy 2 conditions using the And operator?

"check-job-complete": {
    "Type": "Choice",
    "Choices": [
      {
        "Variable": "$.training_job.status",
        "StringEquals": "Failed",
        "Next": "training-job-failed"
      },
      {
        "Variable": "$.training_job.status",
        "StringEquals": "Completed",
      },
      "AND:"
      {

      {
        "Variable": "$.training_job.model_name",
        "StringEquals": "web-annotation-extractor-${environment}",
      },
      "Next": "training-job-completed"

[toxic bough]

pipenv

for managing dependencies and virtual environments

[candid juniper]

I am a bit confused with all these similar sounding tools 🤯

So pyenv is for managing my python installs and pipenv is for managing libraries and virtual environments? So pipenv replaces virtualenv?

[toxic bough]

yeah, iirc pipenv uses virtualenv

[toxic bough]

you need not worry about these tools as a beginner tbh
these tools will probably be overkill for learning basic syntax of the language
you can use pyenv to change the global python and not bother with pipenv until you start a project

[candid juniper]

thanks

it is a bit overwhelming 😂

[lucid tangle]

ok frogive me here, first time using heroku and flask at a much more advanced level

here is my directory

and in my procfile i have ```bash
web: gunicorn decaturSite.api:api

but heroku dumps errors, give me a sec

idk what im doing wrong.... ive been trying different combinations with no luck :((

[lucid tangle]

nvm

[full heart]

is there a library that can recognizee screen colors?

[heavy knot]

i think he's doing a series on like open csv or something rn

that can do image recognition

[thorn geyser]

Is PyDatalog still being maintained now...?

[wooden ibex]

Check their github for commit activity?

[smoky fog]

How I can run my github repo in heroku?

[frigid light]

You'll have to link your github and select the repo

assuming it has all the necessary files you should be good to deploy

@smoky fog

[smoky fog]

Idk I just have config.json and main.py lol

@frigid light

[frigid light]

Cool

[smoky fog]

I need anything else?

[frigid light]

https://devcenter.heroku.com/articles/getting-started-with-python

read the docs

[smoky fog]

My thing doesn’t clone tho

@frigid light

I did got clone and the link

But it doesn’t clone

‘git’ is mot recognized as an internal or external command, operable or batch file.

[frigid light]

Do you have git installed on your pc?

[smoky fog]

I have git setup idk lol

[frigid light]

Try running it then

or perhaps you never added it to path

[smoky fog]

Ah idk how to fix :/

[frigid light]

Well, you can always learn

[smoky fog]

I installed github

Still have the same problem idk if its in the path :/

[frigid light]

Yeah, maybe this might help
https://git-scm.com/book/en/v2/Getting-Started-Installing-Git

[smoky fog]

It aint even letting me uninstall github to install again and add to path @frigid light

[frigid light]

You do not need github installed to use git

git is different from github desktop

[smoky fog]

Ok

@frigid light why it aays bash: cd: too many arguments? Any idea?

[frigid light]

Are you trying to create a folder with space in it's name?

then put that in quotes : cd "new folder"

next time an error comes up try googling it

[smoky fog]

Its a folder already tho

Ok

[frigid light]

I see

*I meant change directory not create (my bad)

[smoky fog]

Ok

[smoky fog]

I thought heroku login was a command

git add . command says fatal: not a git repository

[smoky fog]

@frigid light another gir process seems to be running in this repository............ remove the file manually to continue

Yeah I saw something crash

[steel sundial]

I hope this is the correct channel, if not I am sorry and would appreciate it very much if you could point me to the correct one.

My Question:

Does anyone have a link to some good documentation or tutorial, regarding packaging EXTRAS? How to structure the project, where to put the extras, how to declare them. Generally how to create a package, that has extras.

I can't seem to find really find anything that explains it from the ground up. I found a lot of info, how to declare extra dependencies and how to install packages with extras, but sadly not much else and I want to check if I am just bad at searching 🙂 .

[hushed orbit]

How to setup a git config file and remote urls for multiple github accounts / repos.
I have a seperate account for work and personal github. Tried setting up a config like

# githubPersonal
Host personal
        HostName github.com
        User git
        IdentityFile ~/.ssh/id_rsa_personal

# githubWork
Host work
        HostName github.com
        User git
        IdentityFile ~/.ssh/id_rsa_work```
and an remote origin `git@work:companyname/repository.git`
However it ended up creating a new branch on a repo for my github user, on a random project.

[night quest]

My GitHub remote is [e-mail]:nickname/repository.git so maybe try companyname/companyproject.git?

[hushed orbit]

If you're referring to the .git part, is just forgot to add that to the message. My bad

[night quest]

I am referring to nickname/repository, I never used nickname/branch

[hushed orbit]

Oh yeah, I am using repository, not branch. Just brainfarted

For some reason set my remote url to my own repo.. changed it, let's see how that goes

It's trying to push to git push origin branch:branch 🤔

manually writing git push origin branch will work, so i guess it's a vs code issue now

[night quest]

~/.ssh/.config:

Host github
    User git
    HostName github.com
    IdentityFile ~/.ssh/keys/github_rsa

$ git remote -v
gh    github:nickname/project.git (fetch)
gh    github:nickname/project.git (push)

Works perfect

git fetch gh

[hushed orbit]

yeah i think it's something vsc has setup

or cached

idk tbh

[jade dagger]

any idea to login google account using .chrome() selenium?
i've try cookie method but it doesnt work for me

[inner pollen]

Do check this out and provide feedback. https://github.com/Agent-Hellboy/Adorn

[thorn geyser]

Thx, their up-to-date commit seems to be in 2017

[fickle finch]

I am looking for a way to publish my project to github. As the project is a command line tool (with the possibility of future gui) I was wandering should I distribute it with PyPi or require users to cone and run? Another question is how exactly should I figure out all of the dependencies for requirements .txt if I go with the latter?

[vale wedge]

anyone here aware of a json pretty printer that allows to compact display certain objects (im trying to get diff friendly but compact serializations of openapi specs)

[inner pollen]

I think clone n run will be good here. And you can create a virtual environment then install the dependencies in that environment and then freeze the dependencies using pip freeze > requirements.txt

[fickle finch]

@inner pollen wow thanks a lot!!!

[vale wedge]

@fickle finch it tends to be a good sport to publish via pypi if its a command so people can install in their own environments, if it turns out to be a gui, publishing as installable/startable artifact as well

[fickle finch]

One note, one main audience will be of no technical knowledge so i want it to be the easiest way for them possible

[vale wedge]

whats the use-case the tool gaters towards?

[fickle finch]

Basically I combine 2 existing tools to help language learners work with local media files in order to extract audio and sentences

And send them to the other learning program

[vale wedge]

then its important to know about the tools/systems of the target audience, linux/windows, cli savy or not

[fickle finch]

I thinks it shouldn't be a problem as long as python3 is installed

I used path lib so I think I am covered

Apaer from that ther should be no issues

[vale wedge]

so if a recent python 3 + pip is the only requirement, then having something that can be pip installed easily might be nice

[fickle finch]

I see

Yea

They would also need git if I go with github

As the only option I mean

[vale wedge]

if you know for sure that most of the users use something like windows, or a specific linux distro, it may also be a good idea to package for that in addition

[fickle finch]

Thats is something I am conflicted about

Almost all windows users use windows 10

They can get wsl running in 10 minutes as long as they know how to use a mouse

And it would take me a lot of time to figure and maintain the packaging thing

[vale wedge]

in that case having a pip package + a wsl howto is a fair starting point

later on if a gui is being added it may be sensible to bundle it as a app

[fickle finch]

Yea I doughty about using gooey for that as its a nice wrapper around argparser

But wsl doesn't have X running

Command line options are not that complicated I will wait for the feedback before messing with gui

[vale wedge]

yup start simple, then expand

[fickle finch]

Thanks a lot for the insight! Really helped me a lot

[vale wedge]

happy to help, good luck with helping your future users

[wanton haven]

Hello i'm trying to find a good programming environment for Python

does someone maye has tips for something easy to start with for a beginner

[hearty wind]

hi is there a way to display hindi characters in vscode integrated terminal?

[sly sleet]

@hearty wind you might have to change the terminal font

[ionic jungle]

how to detect motion gesture in python using OpenCV

[cobalt ibex]

can anyeone with selenium expereince pm me?

[turbid hollow]

Hi @ all

i can't find the switch statement, does it have another name

it could look something like:

switch b:
  case a:
  case b:
  case c:  

[sly sleet]

@turbid hollow python doesnt have switch

just use a dict or something https://stackoverflow.com/questions/60208/replacements-for-switch-statement-in-python

there's also pattern matching coming in 3.10, which is arguably more powerful than switch

[turbid hollow]

Thank you this is good

[foggy haven]

Depending on how you structure a if elif else ladder, it can look pretty much the same as a switch case.

[floral prism]

is this the right channel to get help about installing pynput and/or activating conda enviroment

[marsh folio]

is there a reasonable way to install pyenv system-wide so that we can all share a common set of virtualenvs? is it a terrible idea? my boss wants me to set it up this way but i can't google up any authoritative discussion about it and i'm starting to suspect it's a bad way to operate

[azure hornet]

you typically want venv(s) per person, or people will stomp on each other's dependencies

[reef latch]

i'm trying to setup a private repo to be pip install-able--my setup.py looks like

from setuptools import setup


setup(
    name='Foo',
    description='A learning experiment',
    url='[e-mail]/myOrg/foo.git',
    packages=['src.foo'],
    install_requires=[],
    extras_require={},
)

and i'm able to pip install it, but none of my files wind up in the target virtualenv--am i missing something? fwiw my directory looks like:

├── Makefile
├── README.md
├── requirements.txt
├── setup.py
└── src
    └── foo
        ├── __init__.py
        └── external_requests.py

but when i pip install it, the install succeeds but the import fails. anything obvious?

[tawdry needle]

@reef latch ```python
from setuptools import find_packages, setup

setup(
name='Foo',
description='A learning experiment',
url='[e-mail]/myOrg/foo.git',

packages=find_packages('src'),
package_dir={'': 'src'},

install_requires=[],
extras_require={},

)

1. use find_packages to automatically resolve all package names, there's almost no reason not to use this

2. if you really wanted to manually specify, the package name here is foo, not src.foo. that is what find_packages('src') returns. if you don't believe me, cd to your project top-level and run this:

python -c 'from setuptools import find_packages; print(find_packages("src"))'

3. you forgot package_dir

packages is a list of package names, package_dir is the location of each package

[reef latch]

thanks, very thorough! appreciate ya

[tawdry needle]

you're welcome and bless you for caring about packaging

if you want to be really thorough, add this pyproject.toml file alongside setup.py:

[build-system]
requires = ["setuptools", "wheel"]
build-backend = "setuptools.build_meta"

this tells the pip that is installing your package how to install it; e.g. if you need cython at install time then you can specify it here. this also lets you specify a minimum version of setuptools if you need a more recent feature or if you want to blacklist a certain version with a bug, etc.

related reading, if you are interested:
https://www.python.org/dev/peps/pep-0517/
https://www.python.org/dev/peps/pep-0518/
https://setuptools.readthedocs.io/en/latest/userguide/quickstart.html#basic-use

@reef latch ^

[reef latch]

thank you so much! amazing

[ember wren]

cool idea, am gonna spread the word, i use vs so i cant test it out...

[round berry]

As my first introduction to containerization, I'm putting together a simple website using docker. I'm not entirely clear on what the structure of one of these instances should be. I've got Nginx, Node.js backend, PostgreSQL, and Letsencrypt I want to use, generally, would I be deploying all of these to a single docker instance?

[inner pollen]

I will try to incorporate it in vs also. Actually people were not showing interest in it that's why I have made extension for gedit and sublime only. Thanks man. If you like it do give a star, this will be the source of motivation for me.

[short peak]

as introduction, you could. But usually it is a good idea to split components in different container, e.g. a container for a web server, one or more for your application, a container for your db, a container for background tasks. It really depends on the needs of the requirements and perhaps the budget of the project

[round berry]

What kinds of things impact the budget?

[ember wren]

i like to push button instead of running commands, im sure im not alone

[short peak]

high availability, traffic load?, size of your team?

[inner pollen]

Exactly.

[short peak]

e.g. app can't be down, you might want to implement replicas. Need to handle thousands of requests per second, you might want to have several instances running. Your team requieres different environments for QA the app

[round berry]

Makes sense to me, I certainly don't foresee such large demand

And in this case, Nginx and the web server should be separate containers?

And what is the benefit of having separate containers? At least, aside from spreading out load

[short peak]

you can manage deployment separately

split point of failure

makes easy to manage each piece independently

I don't think main reason for that is the performance though it also helps to scale

[round berry]

Hmm, fair enough

And regarding Nginx and the web server? Should those be separate? I don't really have an application layer separate from the webserver atm

[short peak]

nginx is a web server

if you don't have a need to have your app and the web server in different boxes/containers, then don't 😄

[round berry]

Ahh, i see

Heh, this is good practice, i at least don't foresee needing to scale but it will be a learning opportunity. Thank you!

[short peak]

sure 🙂

[inner pollen]

I don't think I have tested the sublime extension for windows, if someone will test it for windows and add the installation instruction then that will be a great help for me.

[foggy haven]

I like to run commands instead of pushing buttons

[heavy knot]

it dosnt let me convert py to exe

any1 can help

[final swallow]

Like already mentioned the point of docker is to split containers up into separate services. If just running vanilla docker look into docker-compose. https://docs.docker.com/compose/

[final swallow]

Docker-compose will let you define each service in the stack in one file. Then bring up everything at once with one command

[primal basin]

if i have two services that require a postgresdb, do i make one or two postgres services in my docker-compose?

and how should i map out hte ports so they dont collide?

[final swallow]

Ports are defined in docker-compose too

Are both services running their own db? If you are just deving on your local id put both dbs in the same PG instance

[primal basin]

yeah they'd be running hteir own db

metadata store for both services

so run docker-compose --scale postgres=2 ?

wha about the ports

[final swallow]

in your compose file you will define each service with what ports you want them to run on

[primal basin]

ahh ok so the default port is 5432 (i believe) for hte second one i can make up any port ,say 5433?

[inner pollen]

Can you elaborate more? There is a library named cx_freeze that can be used to generate exe binary from python source code.

[final swallow]

correct is this just for local dev?

if so id stick both dbs in the same pg instance

[primal basin]

hyeah just poking around on my local

[final swallow]

yeah in that case just run one pg container and create two dbs

[primal basin]

thanks! , will try this now

[final swallow]

in your compose file in the pg service bind a local folder mount into the container. This way your data persists

Example:

services:
  db:
    image: postgres
    ports:
      - "5432:5432"
    volumes:
      - /local/path:/var/lib/postgresql/data
    environment:
      - POSTGRES_DB=postgres
      - POSTGRES_USER=postgres
      - POSTGRES_PASSWORD=postgres

[primal basin]

thx, i am reading up on creating the multiple db script

[round berry]

awesome thank you!

[devout temple]

after running git add ':!<folder>', I still see a few directories that are untracked (under Changes not staged for commit). The idea is to go into each folder and run git add . manually?

[austere brook]

I'm looking for a software development project management tool that helps me estimate the total man hours required for our backlog of issues. Perhaps even slotting them into a calendar, where they do not overlap, so that I can see how far out they extend.

Does anyone know of a tool like that?

[austere brook]

I would just git add . . Are there folders in the root directory you don't want to stage?

[short peak]

like these https://alternativeto.net/feature/project-management/ ?

[austere brook]

slack doesn't do that

but thanks for the help

[short peak]

you're right, but you can tweak the filter in that page according your needs

[blissful ice]

Question: setuptools has a mechanism for defining optional features / extras via extras_require, which doesn't affect the package content, just it's dependencies, right? Is there a way to declare a module in a package as optional? E.g. include namespace/module_a.py only if extra-feature "a" is selected for install?

[vale wedge]

intentionally not - the extras turn availiable/enabled practically when the other stuff is installed and the installed package doesnt change

if you want to split up a package for features, split it into multiple packages that you can put into the feature requirements of the main one

[blissful ice]

I was afraid of that. It feels like splitting up packages is a bit overwhelming though. For instance, I have a project with a protocol definition, client code, and server code. In order to distribute a client package I'd need to separate and maintain 3 different code bases? I hoped I could get away with something like pip install mypackage[client].

[final swallow]

git add -A

[fickle finch]

I tried to add a GUI to my otherwise terminal tool. I tried Gooey but got disappointed by the lack of the documentation and the flexibility I need compared to standard arg parser. What alternatives do I have? I would like to have some predefined widgets like radio buttons, file inputs, dropdowns... and be able to combine them. I also want to be able to monitor the state of one widget in order to turn on/off the other ones

[vale wedge]

@blissful ice just ship it together, and have the parts that lack dependencies fail ?

@blissful ice its entirely ok to have the parts that lack dependencies fail, just add a little extra that adds a extra note to the import errors

[carmine briar]

Hi,
When I try to clone a repository from internet to my HDD it copies all the files contained in the master Repository I want to just clone a subfolder of a repository Is this possible? If yes, then how?

[heavy knot]

With git? Seems possible even if its a bit involved https://stackoverflow.com/questions/600079/how-do-i-clone-a-subdirectory-only-of-a-git-repository/52269934#52269934

You can also do a shallow clone if you are not interested in the commit history

[carmine briar]

Thanks for reply : D How do I do that?

[heavy knot]

For shallow clone i think its just adding --depth 1 to the git clone command

[carmine briar]

Im using desktop app & sorry i'm new to github 😅

[carmine briar]

How do I do with the desktop app?

[heavy knot]

No idea sorry I only use the cli 😅

[carmine briar]

Can I DM you?

[heavy knot]

If you want but I wont know anything more about the desktop app

[carmine briar]

Okay Thanks

[tiny dove]

Hello all

I have a challenge getting my API built using Django-REST framework to run on Docker

On the local machine, once I run manage.py runserver, the server runs and the app works accordingly

But from docker I get this error when I try to use the app with postgres in docker

I get this error -

django.db.utils.OperationalError: could not translate host name "postgresql" to address: Temporary failure in name resolution

PS - I am working on Windows OS

[night quest]

Are you using Compose?

Paste some code

[tiny dove]

My docker-compose.yml

https://pastebin.com/uceZLcrW

My project structure -

The dockerfile for the API -

FROM python:3

RUN apt-get update \
    && apt-get install -y --no-install-recommends \
        postgresql-client \
    && rm -rf /var/lib/apt/lists/*

WORKDIR /app/api
COPY requirements.txt /app/api
RUN pip install -r requirements.txt

EXPOSE 8000
CMD ["python", "manage.py", "runserver", "0.0.0.0:8000"]

Dockerfile for the frontend -

FROM node:13.12.0-alpine

WORKDIR /frontend

COPY package.json /frontend
RUN npm install 
RUN npm install react-scripts@3.4.1 -g 
COPY . ./
EXPOSE 3000

[night quest]

Your containers are in different networks

You can inspect it using docker inspect ...

[tiny dove]

I got this -

NETWORK ID     NAME                                DRIVER    SCOPE
c8f74f76e832   bridge                              bridge    local
5e01270c53c9   host                                host      local
070efbf04ddf   indie-tour-app_db-net               bridge    local
ae9cddca6a4f   indie-tour-app_default              bridge    local
7d9e9753173f   none                                null      local

..when I did docker network ls

@night quest hello 👋

[night quest]

Easy

[night quest]

This is not the case

As you can see you have two bridge networks in your project

default is default network - each container is attached to it

[tiny dove]

Yes.

[night quest]

Unless you change configuration in docker-compose.yml

Then you must explicitly set

  networks:
    - one
    - default

[tiny dove]

So what can I do to have a single bridge network bridging all parts of the app

[night quest]

You can create internal network between API and database container

[tiny dove]

So would that be? -

networks:
  - one
  - default
  db-net:
    driver: bridge

I'm sorry, I'm a beginner at docker, so a lot of things are confusing

Forgive my ignorance

[night quest]

Rather:

services:
  postgress:
    ...
    networks:
      - database
  ...
  api:
    networks:
      - database
      - default
networks:
  database:
    internal: true

[night quest]

No problem at all

You don't need to create different bridge network I think - because you have default

[tiny dove]

@night quest Man, It now works

I'm so elated!! 😂

[night quest]

Yay, nice

[tiny dove]

Been battling this since Monday 😂 😭

[night quest]

Glad to help

Remember that internal containers don't have access to the Internet

[tiny dove]

Yes.

I have a question though

[night quest]

What's up?

[tiny dove]

Apologies, power cuts and I had to get power up and running

[night quest]

Don't worry

[tiny dove]

I went to the swagger doc for the api on docker

[night quest]

I don't know what swagger is

[primal basin]

whats the general rule for building a docker-compose file with regards to number of services within a single docker-compose/
are there diminishing returns or confusion as it gets larger and larger?
most of the examples i see online have only a handful of services, mine is becoming like 15 services...i am wondering if this is bad practice

[tiny dove]

swagger - basically puts the API documentation on the web.

[hexed oasis]

Not sure if there are general rules, but yes I think it would get rather confusing as you add more and more services. If you can, I'd recommend splitting them into grouped docker-compose files and then calling -f multiple times when building/running.

Other than that, have you considered moving to helm? I think 15 services would justify the move

I have a question for the channel: have you ever/do you currently use a monorepo? We're looking into the possibility of converting from multirepo at work. If you have tips, or can advise on any common pitfalls to avoid that would be awesome to hear

[final swallow]

Yeah id say docker-compose is great for a single app stack. Anything larger or complex you'd prob want to use Kubernetes

[primal basin]

hmm ok and just to confirm, some of the images that i am using just have multiple instances (ie. services) so, im not saying that there are 15 different images i am using as services... for example; apache druid has 5 services using the same image (ie. broker, coordinator, middlemanager, router)

i will look into kubernetes, i havent touched that or helm before

[final swallow]

helm is a tool to template and deploy yaml

[hexed oasis]

He's right, learn Kubernetes first if you've not touched it before 😄

[tiny dove]

Please how do I migrate models (Django) to docker?

I have the db and api linked now but I have to migrate my models to the dockerized postgres db

[final swallow]

exec into the django container and run your migrations

docker exec -it <container name> bash

[worldly drum]

Hey, not sure if it's the right channel but i'll try
I'm working on a micro-service within few other micro services
I'm trying to manage dependencies correctly, right now using requirments.txt
the problem is that requirements.txt is only handles top level requirements and i also want to handle sub dependencies (sort of equivavlent to package-lock.json with npm)
I made a little research and read some about pipenv
does any one have experience with it? can recommend? have other recommendation?
Thanks in advance

[tiny dove]

Greetings brethren

So I found a way to migrate models to docker

But I want a situation where when I do a compose build I run migrations to check if there are changes to the model

Currently I was able to create ta bash script, run it in my dockerfile, then call that bash script in my docker-compose

It builds as expected

However, when I attempt running docker-compose up -d and I get the container in question, I see the particular image I'm trying to run migrations on restarting severally

In my migrations.sh file -

#!/bin/sh

python manage.py migrate
python manage.py makemigrations

Docker file for api -

FROM python:3

RUN apt-get update \
    && apt-get install -y --no-install-recommends \
        postgresql-client \
    && rm -rf /var/lib/apt/lists/*

WORKDIR /app/api
COPY requirements.txt /app/api
RUN pip install -r requirements.txt

EXPOSE 8000

COPY migrations.sh /migrations.sh
RUN chmod +x /migrations.sh
CMD ["python", "manage.py", "runserver", "0.0.0.0:8000"]

In my docker-compose.yml file for the service api -

api:
    restart: always
    container_name: api
    entrypoint: /migrations.sh
    command : 
      - python manage.py runserver 0.0.0.0:8000   
    build: ./api
    links:
      - postgresql:db
    volumes:
      - ./api:/app/api
    ports:
      - "8000:8000"
    depends_on:
      - postgresql
    networks:
      - database
      - default

[night quest]

Why you use command both in Compose file and in Dockerfile?
Why you use both command and entrypoint?

[final swallow]

Agreed. Best to leave your migration commands out of the Dockerfile. If you want to automate migrations you could add them as a service in your compose file.

https://stackoverflow.com/a/44283611

[vale wedge]

Ideally migration happens in separate containers and their deployment is planned so that they enable deployment to the services without migration disruption at code update time

[pure galleon]

import smtplib, ssl

port = 587 # For starttls
smtp_server = "smtp.gmail.com"
sender_email = "[e-mail]"
receiver_email = "[e-mail]"
password = input("Type your password and press enter:")
message = """
Subject: Hi there

This message is sent from Python."""

context = ssl.create_default_context()
with smtplib.SMTP(smtp_server, port) as server:
server.ehlo() # Can be omitted
server.starttls(context=context)
server.ehlo() # Can be omitted
server.login(sender_email, password)
server.sendmail(sender_email, receiver_email, message)

Everything works on my computer, but when I give the program to a friend, he gets an error. hostname, aliases, ipaddrs = gethostbyaddr(name)
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xf7 in position 7: invalid start byte

[frozen anchor]

I think this is better suited for a help channel. Also use a code block and tell us which line the error is thrown at

[pure galleon]

I thought like auto mail is like devops automation.

[frozen anchor]

but it is not related to automail, it is a unicode decode error.

[pure galleon]

Probably fair

[graceful moat]

Hi guys I'm quite comfortable with Python Flask API but was wondering how I could deploy a python backend in the cloud (AWS etc.)

I'm assuming I can process data and have backend python data processing methods

so I can call the api from another place in my ReactJS app

[frozen anchor]

does this help? https://flask.palletsprojects.com/en/1.1.x/deploying/

[graceful moat]

can anyone point me somewhere

[graceful moat]

thanks, I'll check it tout

[inner mesa]

Do I have to cite the copyrights of a gohugo theme that has an MIT license? In other words - can I just change that text on the bottom?

[vale wedge]

@spark fjord im preparing the next iteration of setuptools_scm - dropping support for eol python versions - wondering what to use as minimal setuptools version supported

[round berry]

I'm again using docker for the first time, with docker-compose specifically. I've got a github repository of the form

docker-compose.yml
web/
   server.js
   Dockerfile 

I've got several services running through docker compose, if I want my node stuff to be synced with my github repository, should I do that within the web service, or should I be doing that outside/between my docker-compose up?

[tawny temple]

Do it outside Docker. You'll have to rebuild images in order for them to get the changes. For local development, you can specify a bind mount in the compose file. It will then read from the host's files instead.

[round berry]

In which case, each time I update should I be stopping docker compose and restarting it after pulling the changes?

[tawny temple]

Yeah but maybe node has something that supports auto reloading your files, like the Django dev server does

[spark fjord]

@vale wedge I'm not aware of any constraints you need to apply. Projects can select minimum Python versions based on the features they need. I'd only specify a minimum Setuptools version if setuptools_scm has a requirement on a specific feature.

[primal basin]

any idea why my postgres service wont docker-compose down? i keep getting: ERROR: for postgres UnixHTTPConnectionPool(host='localhost', port=None): Read timed out. (read timeout=70) ERROR: An HTTP request took too long to complete. Retry with --verbose to obtain debug information.

docker logs returns database ready to accept connections

[spark falcon]

does anyone know the FEN string for chess board before any moves are made?

[sinful fox]

have you looked at this? https://python-chess.readthedocs.io/en/v0.24.1/core.html#chess.Board.fen

[vale wedge]

I want to stop the egg building, im experimenting with setuptools 45 as minimum atm

[heavy knot]

Hello, does anybody know how python applications can make files ?

[vale wedge]

More context please, open for writing should do already

[heavy knot]

basically creating a file

[vale wedge]

Just open it for writing?

[heavy knot]

that also, but mainly creating a new file

[vale wedge]

Opening for writing is how you create a new file in general

[heavy knot]

ok, but how can application make that ?

[vale wedge]

Now please explain what you want to achieve, not just a small piece of the issue

[heavy knot]

well i basically have a problem with app not working

i runned in shell or something like that, and it worked fine, but when i compiled it into app, it basically didnt worked

does app need some special perms to do that ?

should i also post the app ?

[vale wedge]

my time for helping is unfortunately up, and it seems like you keep asking incomplete questions and leave out the real issues and key details for whats actually wrong, so good luck

[heavy knot]

eh, ok, thanks for help anyway 🙂 , bye bye

[tiny dove]

I just followed what I saw in a tutorial

[tiny dove]

I tried to mimic what was in that stackoverflow link -

api:
    restart: always
    container_name: api
    #entrypoint: /migrations.sh
    command : python manage.py runserver 0.0.0.0:8000   
    build: ./api
    links:
      - postgresql:db
    volumes:
      - ./api:/app/api
    ports:
      - "8000:8000"
    depends_on:
      - migrations
      - postgresql
    networks:
      - database
      - default
  
  migrations:
    container_name: db_migrations
    command: python manage.py migrate --noinput
    build: ./api
    links:
      - postgresql:db
    volumes:
      - ./api:/app/api
    depends_on:
      - postgresql
    networks: 
      - database
      - default

That was what I have currently in my docker-compose file

But each time I build and do compose up, I notice for the db_migrations image I get this error -

https://pastebin.com/8L7L9Z7L

[tiny dove]

Hmm, interesting.

So I made changes to the migrations service -

migrations:
    container_name: db_migrations
    command: python manage.py migrate --noinput
    build: ./api
    links:
      - postgresql:db
    volumes:
      - ./api:/app/api
    depends_on:
      - postgresql
    networks: 
      - database

And I got the expected message from the migrations image -

Operations to perform:
Apply all migrations: admin, auth, contenttypes, indieco, sessions
Running migrations:
Applying contenttypes.0001_initial... OK
Applying auth.0001_initial... OK
Applying admin.0001_initial... OK
Applying admin.0002_logentry_remove_auto_add... OK
...etc

[ionic pagoda]

Hey there. Anyone aware of a way to have commands like docker build and docker push on a windows PC without having to install full docker desktop?

[tiny dove]

So I did a compose-down to rebuild this docker container and compose-up and I got a repeat of this error -|

https://pastebin.com/8L7L9Z7L

I checked the mount for db_migrations and it was mounting to the right path. But checking the port, I noticed this -

Port
8000/tcp
Not binded

[tiny dove]

More observations. Turns out the problem is not with the migrations. I removed the migrations service and spun the container

I checked the api service and got this -

https://pastebin.com/BKXiaQcq

It's a connection issue. Not sure why it connects, then I do a compose-down and rebuild and compose up and get the errors I'm seeing

Do I need to restart Docker?

Lemme try that

[tiny dove]

Restarted. And everything (without the migration service) works as it should

Added the migration service and we are back to the connection issue

I observed that when I decided to add -
restart: always to the migration service, I first got an error, then I got the successful migration message after a restart. But the service just kept on restarting

[tiny dove]

Smh, just decide to work alone with the api service and add a second command

So I removed the migrations service

And in the api service, just added

command : bash -c "python manage.py migrate && python manage.py runserver 0.0.0.0:8000"

*facepalm why didn't I think of this before?

[fierce sun]

Hi, we are working off of a manufacturer provided repository, but would also like to setup a local repo to track our changes to the manufacturer rpo

what is the best way about this? I read something in regards to manifest files. Is that better than adding multiple remotes to the repo?

[primal basin]

can i get some docker help? this is my error ERROR:flask_appbuilder.security.sqla.manager:DB Creation and initialization failed: (psycopg2.OperationalError) could not connect to server: Connection refused Is the server running on host "postgres" (172.18.0.5) and accepting TCP/IP connections on port 5433? my port is currently port:5433 is the error msg telling me to change it to ports: "172.18.0.5:5433" instead?

[analog kiln]

So, as I was digging around the interwebs, is there no better way to enforce the order of containers in docker-compose? I have a rabbitmq container that takes some time to start. I found healthcheck, that seems to have no effect. Only solution I found was using sh scripts in the containers.

[tiny dove]

The most feasible I've seen is to use the depends_on property while declaring your services.

[analog kiln]

That exists, but only starts the containers in certain order. But doesn't wait for them to be available.

So I need combination of depends-on and my own scripts?

[visual ridge]

Hi please help. I was using github desktop and I accidently switched branches. It changed like almost every file. How do I revert?

[spark fjord]

SGTM.

[vale wedge]

That one is new to me

[spark fjord]

Sounds good to me.

[vale wedge]

Doh, so obvious when written out

I strongly recommend taking one of the help channels for your case, i can't help atm as i gotta run now

[sly sleet]

@visual ridge just switch branches again back to what it was before

https://docs.github.com/en/desktop/contributing-and-collaborating-using-github-desktop/managing-branches#switching-between-branches

[visual ridge]

i switched back and now its gone

[sly sleet]

so it worked?

[visual ridge]

no the changes arent there

[wooden ibex]

That should be handled in code

You could put retry in there or just crash the container and let docker restart the container

Network issues happen

[primal basin]

can i get some docker help? getting this error:

    Is the server running on host "superset_db" (172.22.0.5) and accepting
    TCP/IP connections on port 5433?``` 


my ports for it: `ports: - "5433:5432"`

i have two postgres services, the error is happening on my second postgres service

[wind salmon]

I'm trying to package a CLI app as a standalone executable using PyInstaller. I don't have a Linux OS so I used Vagrant to package it, taking the CentOS 7 box since the server where the executable will run is also CentOS 7.

I was able to create a standalone executable and it ran successfully with a fresh CentOS 7 Vagrant machine (no python3 dependencies installed). But when I run the executable in the server, it gives me this error:

[user@server ~]$ emails_automation/emails --help
[22874] Error -3 from inflate: unknown compression method
[22874] Error decompressing _asyncio.cpython-38-x86_64-linux-gnu.so
Failed to write all bytes for _asyncio.cpython-38-x86_64-linux-gnu.so
fwrite: Bad address

Both VM and actual server has same glibc version:

VM:

[root@localhost emails]# ldd --version
ldd (GNU libc) 2.17

Server:

[user@server ~]$ ldd --version
ldd (GNU libc) 2.17

Any pointers?

[vale wedge]

The unknown compression errors seems relevant

[split stag]

Guys, what is the best python library you know for animation??

[vale wedge]

What type of animations, there is a number of different best in their field

[split stag]

I'm just looking for a python library for a typical 2d animation that is easy to learn

[split night]

would anyone know why git is ignoring json files that arent in gitignore?

[blissful ice]

Hi, I'm trying to package a compiled library for easy installing with pip / setuptools but I don't know where to start. In Conda there are non-python packages for installing data to local /lib /include /bin folders. Does pip allow to do something like that or is this a conda-only feature?

[dusty maple]

I believe this is typically done by publishing wheels for supported platforms, although I'm personally not knowledgeable on packaging non-pure-Python packages

[random dawn]

I know about https://docs.manim.community

[random dawn]

You should be creating wheels something like https://cibuildwheel.readthedocs.io/en/stable/ would help :)

[vale wedge]

pip is primarily concerned with python packages, and not at all like conda in that regard

[blissful ice]

Allright, I knew conda was being advertised as being more 'open' for distributing non-python packages in comparison with pip. But I faintly recall someone saying that this isn't the case for "modern" pip anymore, given that there are wheel packages like cudatoolkit etc.

[dry knoll]

Btw, what's devops and where can I learn about It? Thx a million

[vale wedge]

Pip simply follows the python standards, wheels with binaries are a python standard, random non pythin blobs for non python folders are not

[arctic flicker]

is it at all possible to log the shell in pycharm?

i.e. both commands and output

[arctic flicker]

i.e. saving those into a file after each run

[arctic flicker]

ok, rephrasing: Is it at all possible to include the debug console input when saving to log?

[arctic flicker]

or else, is there any IDE that can do this?

[heavy knot]

Hi, I'm working on tool that checks MD5 hashes .. and I wonder if it's possible in python to make update just the variable with hashes on screen without having to clear whole screen

[random dawn]

have a look at https://docs.python.org/3/library/hashlib.html

[night quest]

You need to take a look at something like ncurses

[heavy knot]

well my problem is not connected with hashes but with "visualisation" of that program

[heavy knot]

ty I'll check it out

[silver marlin]

Hello everyone.
I just finished my first tool in Python today (I started learning Python recently). It works, but I don't know if there are a lot of begginer mistakes, or things that can be improved. Do we have here in the community any type of code reviewing? There are any online tools to check the code?
Thanks 🙂

[heavy knot]

well I guess there is some web pages where you can paste your code and it will edit it to look better

you can try that

[silver marlin]

I don't need to format the code to look better, for that I have tools in VSCode. It's just that I don't know if what I did it's the most efficient way of doing it in Python. My background is PHP and lately a propriety language

[silver marlin]

Let's call it a "Senior" review 😄

[heavy knot]

I see

[vale wedge]

if you have it on github, you can just send a link and ask for suggestions ^^

[silver marlin]

Thanks a lot. That is what I will do

I made the following project as part of my Python learning. This is my first tool. Can someone give me feedback about it?
https://github.com/gh-rboliveira/fillnewcopy

[short peak]

I'd suggest to don't version your python environment (your env folder)

[silver marlin]

Thank for the input... I was wondering if someone cloned the repository how would they know the packages that they need to install. What is the alternative? Just don't commit them? 🙂

[short peak]

I know a couple of options for that:

1. Create a requerirements.txt file - https://stackoverflow.com/q/29938554
2. Use poetry - https://python-poetry.org/

though 1st one is probably the most used one

[silver marlin]

I just made the change with requirements.txt 🙂 Thanks a lot for the help here.

[ebon swift]

Is it possible to improve the flow of coders by designating one person to carry out all merge conflicts?

[wooden ibex]

most companies have developer opening the pull request fix the merge conflicts

though merge conflicts can indicate poor branching practices

[silver marlin]

I would say that all merge conflicts should be addressed to the person that owns the code/change creating the conflict since it's the person that knows better the code. If you don't know the code, you don't know what to keep, what to remove from the merge conflict

[crystal agate]

hey, does anyone have an experience connecting to a docker mysql image from a django app in the same container ? I tried with different hosts from the app but it keep giving me:

2003: Can't connect to MySQL server on '0.0.0.0:3306' (111 Connection refused)

[ebon swift]

@wooden ibex do you have any recommendations if we were to look to improve our branching practices?

[wooden ibex]

We do Main -> Development -> Feature branching

[ebon swift]

@silver marlin I thought this too, but I'm trying to be open to the other point of view because I understand the value in staying in the zone when coding rather than coming out of it to fix conflicts. But I worry we're going to hit problems if only one person is fixing conflicts

[wooden ibex]

main is stable tagged prod code

development is test stable (hopefully) code

feature branches are developers currently working on something and may break at any time, generally two developers do their best not to touch same part of code but push into each other branches

[ebon swift]

At the moment we're doing Main -> dev -> subtask of story branching

[wooden ibex]

in that case, how are you having conflicts, are devs not talking and touching same parts of code?

[ebon swift]

Yes

[wooden ibex]

seems like the fix is easy then

devs should talk

[ebon swift]

This is one of the suggestions I've put forward yep.

I feel like integration is a pain rather than easy, but also that trying to avoid doing it regularly and ourselves probably leads to worse pain. It seems like that's the general feeling from what I've read, but is there something I can show as evidence?

Also if anyone has arguments to support the other side they're equally welcome

And by anything I can show as evidence I guess I mean something particularly well regarded or compelling

[wooden ibex]

I'm sure there is some management consultant book like "BETTER DEVELOPER MANAGEMENT"

but in general, I doubt there is hard and fast rule you can point to

people problems rarely have that

[ebon swift]

Ok, no worries. Thanks 🙂

[narrow folio]

I'd like to work on this issue, I don't think this the contributors of this repo decided on a CI platform to use? Where do I start? https://github.com/softwaresaved/lowfat/issues/604

[wooden ibex]

They are on github and github actions are free for x amount for open source projects, so start with that?

[vocal quiver]

Hi, I installed docker today with the defaults settings and after this installation I get this message. I don't get what they are saying in the message, Thanks

[vocal quiver]

I am using windows OS

[tawny temple]

Install wsl2 by following the link they provide

I guess Docker uses wsl2 as its back end by default but it's complaining cause you haven't installed wsl2

[silver marlin]

Hello,

I created from a script an executable file with pyinstaller:
pyinstaller --onefile --windowed myscript.py

It runs on my computer, it runs on another colleague's computer, but not in a third colleague's computer. We all have Windows 64bits.
Anyone is aware of this problem and how to solve it?

[sinful kettle]

Who uses replit btw

[brittle mist]

i did for a while
..

hello,
i am trying to settup docker to be used to automatically update my bot from my repo on github.
what kind of packaging should i use for it?
are the depensiencys automatically added to it?
and how can i automatically start the wb.py file?

[raven forum]

I want the server to transmit sound from the browser or from the video player, how can I do this? (I already have a server and a client ready, I just do not understand how I can catch the sound and play it on the client, I use the socket library for the server)

[heavy knot]

How do i begin creating devOps tools?

[north reef]

What is Git Credential Manager? it is in my github authorized apps and it says it was last used within last week. Should I revoke its authorization?

[tawny temple]

It's what saves your git credentials. I believe it's a Windows-only thing that comes with the standard Windows git installer.

Well, it may not exactly be that, since I don't believe it communicates with GH. But maybe it is something related.

Maybe it does... https://github.com/microsoft/Git-Credential-Manager-Core

I haven't used this new "core" version, since by the time it was released I had already moved on to using SSH keys.

[primal basin]

stupid question but when i am writing config files inside a compute engine VM, do i leave localhost as is or do i replace localhost with internal or external IP?

the default config file example
druid.host=localhost

[visual ridge]

Hi, I used a script to auto commit to save my work. I want to merge all the commits to the latest commit I make, how can I do that?

[tawny temple]

Use an interactive rebase to either squash or fixup the commits. The command for an interactive rebase is git rebase -i <commit>, where the commit specified is the oldest commit that you want to squash/fixup.

If these commits have already been pushed, then you'll need to do a force push. This is because a rebase will rewrite history, and such pushes are normally rejected.

That is why this kind of git workflow is generally not advised. But if you're working alone on your branch then you do you.

[visual ridge]

They have not been pushed yet. Is there no way to squash like 60 commits into the LATEST one I make? I want to sqaush all into the one I have sqaured

[tawny temple]

Yes, there is a way and I just explained how.

[visual ridge]

Oh I thought you said oldest

[tawny temple]

The command takes the oldest commit, so it knows where to start from

It starts from the oldest and goes until the newest

[visual ridge]

Ah ok

[tawny temple]

If you have so many commits, a rebase may be too tedious

You could also do it with git reset

[visual ridge]

Alright, ill go with rebase since I want to be safe. So for example a SHA for a commit is 5c4ee8549644a3cf041b30f0245894c2674679c8 which is the one from where I want to start sqaushing from. So I would do git rebase -i 5c4ee8549644a3cf041b30f0245894c2674679c8

[tawny temple]

When you rebase, you have to individually mark every commit you want to squash. With a reset, you just specify the oldest commit and it will remove all those commits and then keep the changes unstaged (or staged with git reset --soft). So you'd lose the commits but not the actual changes within them.

Yeah, that is correct

[visual ridge]

Ah

[tawny temple]

Thought I always forget if the commit hash is exclusive or inclusive 😄

[visual ridge]

So if I specify a commit 3 days ago (5c4ee8549644a3cf041b30f0245894c2674679c8). I made 10 commits from then. Then it would erase the 10 commits and that commit I specified with the --soft flag but I would not lose any changes. I would use git reset 5c4ee8549644a3cf041b30f0245894c2674679c8 --soft

Thanks alot

[tawny temple]

The hash may have to be at the end of the command

That's how I always do it anyway

[visual ridge]

Ah alright, so it will do what I intend it to? (by that I mean would it would erase the 10 commits and that commit I specified with the --soft flag but I would not lose any changes )

[tawny temple]

Yes. If you're worried you can create a backup branch: ```py
git checkout -b mybackup
git switch originalbranch
git reset ....

[visual ridge]

Alright, last time I did that it pulled the backup branch and I lost all my changes, thats why I used that auto commit script lol. Ill just do git stash and switch

[tawny temple]

Can you stash committed changes?

[visual ridge]

Dont think so. the commits that have not been pushed will persist if I switch or they will be destroyed

sorry for being so paranoid but this is how i lost 1 week of code

And I have just rewrote

[tawny temple]

Well if you're ultra paranoid, copy and paste the .git folder

But yes, the commits should persist. They'll just be on another branch.

[finite fulcrum]

your reflog should keep track of almost everything you do in case you mess it up

[tawny temple]

Well. You'll have them on both your backup and original branch.

[visual ridge]

Ah alright

\Thanks alot

Also, one more general question that I couldnt find the on the web. I usually push my code with varaibles that the user should configure when they git clone it. However, when I pull it I lose the the variable that I defined. So basically what I want to do is when I git pull I want some of the part of code remain unchanged. How to do that?

[tawny temple]

Don't check in that file. Add it it .gitignore.

Instead, add in your documentation that users need to create the files themselves. Tell them where to create it and what the format is.

If you think it helps, you can commit an example or template file with a different name so users can copy and rename it to the right thing then go from there.

[visual ridge]

No, I will be changing the code but one part is token = jwd8y7wed7qyw7y7dy7qaydjsad I wont want to reveal my token so I would just write toke = pasteyourtoken however every time I git clone or pull my token will be changed to pasteyourtoken and I would need to copy it again and again. How would I do that?

It is in the readme.md to paste your token there

[tawny temple]

Don't hard-code your token. Read it from a separate file that solely contains the token

Or read it from an environment variable

[visual ridge]

and add that to .gitignore?

got it

[tawny temple]

Or read it from command line arguments

Yes

[visual ridge]

Also I used the reset thing which squashed 60 commits into one commit that I made 3 days ago. However, that commit isnt descriptive so I would like to make one more commit and squash the already sqaushed commit into the one I make. So the normal squash command can do that correct?

[tawny temple]

Use git rebase -i HEAD~1 and then replace "pick" with "r" for the commit shown

Then it will open your text editor and let you enter a new commit message

[visual ridge]

Crap, i squash 60 commits to one and switched branches. Now changes are all messed up

I opened reflog and I want to go to

the one i squared

[tawny temple]

What do you mean it's messed up?

Switching branches is a harmless operation

Just switch back.

Switching branches doesn't mutate commits on the branch you're leaving.

[visual ridge]

I switched to backup branch and did git reset.
I switched to main and now its messed up.

I brought the changes with me instead of leaving them on backup branch. Just how do I revert the commit I squared

the changes messed up main

it tried to merge them or something

[tawny temple]

Did you use git switch or git checkout?

[visual ridge]

git checkout

[tawny temple]

That shouldn't carry over changes. In fact, it complains if you have uncomitted changes.

sorry, I'm just having trouble understanding how this could happen

[visual ridge]

I do not know, I used the github desktop thing..

[tawny temple]

!!!!!

Okay, that's a significant detail

I do not use that app so I cannot help you with it. I am only familiar with behaviour of git via the command line.

[tawny temple]

Is your backup branch still intact?

[visual ridge]

think so

ill copy it and delete the .git folder

just to be safe?

[weak tendon]

any good resource to learn about openstack for cloud apis? Thankyou

[tawny temple]

Don't delete the folder.

[visual ridge]

Okay

[visual ridge]

Unlike last time I commited alot. Even If I lose my code now could I just revert to a commit?

[tawny temple]

Yes it probably stll exists somewhere

[visual ridge]

i did git reflog

i did git reflog

it says this:

when i try to revert

[tawny temple]

Revert is probably not the right command. You'd need to rebase. What are you trying to do right now?

[visual ridge]

Nothing, I just want to revert it to that command

commit*

[tawny temple]

And discard everything after it?

[visual ridge]

yup

[tawny temple]

git reset --hard <commit> would do that.

[visual ridge]

just to make sure

If I do git reset --hard 1020f4c it would revert to that commit and discar everything after it?

[tawny temple]

Yes

It only affects the current branch though

[visual ridge]

it worked

thanks

[tawny temple]

You're welcome

[visual ridge]

should i use switch or checkout?

to be safe

[tawny temple]

Either works but switch is a relatively newer command designed to have a more narrow focus, and therefore a simpler interface, than checkout

[visual ridge]

how does git stash work? if i change for example greeting.py Hi to Hello and I run git stash it would save Hello and i can just run git stash pop?

Also if I have changes on the main branch but I want them to be pushed to feature branch then how would I do that?

[visual ridge]

alright thanks

[heavy knot]

hi

[vocal quiver]

Thanks, I installed it and now docker is not complaining

[steep temple]

Is it possible to make a webinterface to remotely execute code on a Rasberry pi 4?

[night quest]

Yes

[steep temple]

My Plan is to host a local Webserver on a Rasberry pi 4 and then control it trough that webinterface

[steep temple]

How would make that? Could you give me a short explanation?

[night quest]

Do you want to have access to something like shell via browser or just few buttons like "Power off", "Reboot" etc?

[steep temple]

Access to shell

Like just remotely access the shell

[night quest]

You can install Jupyter and there is access to shell remotely

[steep temple]

But can I put that on a Webserver?

[night quest]

Jupyter is available via browser, is it what are you asking for?

Here you have installation guide: https://jupyterlab.readthedocs.io/en/stable/getting_started/installation.html

[steep temple]

The plan is that the Webserver automatically starts and I get shell access on startup

And if possible I want to create a WiFi from the Rasberry pi so that I can connect anywhere

[night quest]

You can set up any program to start with system