cross-posted from: https://programming.dev/post/14680192

> I have a VPS, but no root access so I can't use apt, or even read a lot of the system files. I would like to get jellyfin (or any media server, really) running on it. Jellyfin has a portable installation option, so I followed the instructions in the docs to install it from the .tar.gz. > > But it says I have to install ffmpeg-jellyfin, and I can't find a portable installation of that. My VPS already has ffmpeg installed on it. Will jellyfin work if I just point it to that instead? Or, how can I go about installing ffmpeg-jellyfin without root access?

I have a VPS, but no root access so I can't use apt, or even read a lot of the system files. I would like to get jellyfin (or any media server, really) running on it. Jellyfin has a portable installation option, so I followed the instructions in the docs to install it from the .tar.gz.

But it says I have to install ffmpeg-jellyfin, and I can't find a portable installation of that. My VPS already has ffmpeg installed on it. Will jellyfin work if I just point it to that instead? Or, how can I go about installing ffmpeg-jellyfin without root access?

I have been using sway (basically i3 for Wayland) instead of a traditional desktop environment because it really makes a difference in my laptops performance.

But apparently sway ignores .desktop files which was how i was autostarting things on KDE.

Is the best way to handle this by going through the sway config? If not, how would you do it.

Bonus points if you can tell me how to get the autostart programs to also open in specific workspaces.

I'm working on a little gui app that will eventually (hopefully) add a watermark to a photo. But right now I'm focused on just messing around with tkinter and trying to get some basic functionality down.

I've managed to display an image. Now I want to change the image to whatever is in the Entry widget (ideally, the user would put an absolute path to an image and nothing else). When I click the button, it makes the image disappear. I made it also create a plain text label to see if that would show up. It did.

Okay, time to break out the big guns. Add a breakpoint. py -m pdb main.py. it works. wtf?

python def change_image(): new_image = Image.open(image_path.get()).resize((480, 270)) new_tk_image = ImageTk.PhotoImage(new_image) test_image_label.configure(image=new_tk_image) breakpoint()

with the breakpoint, the button that calls change_image works as expected. But without the breakpoint, it just makes the original image disappear. Please help me understand what is happening!

edit: all the code

```python import io import tkinter as tk from pathlib import Path from tkinter import ttk

from PIL import ImageTk from PIL import Image

from LocalImage import Localimage from Layout import Layout

class State: def init(self) -> None: self.chosen_image_path = ""

def update_image_path(self): self.chosen_image_path = image_path.get()

def change_image(): new_image = Image.open(image_path.get()).resize((480, 270)) new_tk_image = ImageTk.PhotoImage(new_image) test_image_label.configure(image=new_tk_image) breakpoint()

TEST_PHOTO_PATH = "/home/me/bg/space.png" PIL_TEST_PHOTO_PATH = "/home/me/bg/cyberpunkcity.jpg" pil_test_img = Image.open(PIL_TEST_PHOTO_PATH).resize((480,270))

why does the resize method call behave differently when i inline it

instead of doing pil_test_img.resize() on a separate line?

root = tk.Tk()

root.title("Watermark Me") mainframe = ttk.Frame(root, padding="3 3 12 12") mainframe.grid(column=0, row=0, sticky="NWES")

layout = Layout(mainframe)

image_path = tk.StringVar() tk_image = ImageTk.PhotoImage(pil_test_img) test_image_label = ttk.Label(image=tk_image)

entry_label = ttk.Label(mainframe, text="Choose an image to watermark:") image_path_entry = ttk.Entry(mainframe, textvariable=image_path) select_button = ttk.Button(mainframe, text="Select", command=change_image) hide_button = ttk.Button(mainframe, text="Hide", command= lambda x=test_image_label: layout.hide_image(x)) test_text_label = ttk.Label(mainframe, text="here i am") empty_label = ttk.Label(mainframe, text="")

for child in mainframe.winfo_children(): child.grid_configure(padx=5, pady=5)

entry_label.grid(column=0, row=0) image_path_entry.grid(column=1, row=0) hide_button.grid(column=0, row=3) select_button.grid(column=0, row=4) test_image_label.grid(column=0, row=5) empty_label.grid(column=0, row=6)

image_path_entry.insert(0,TEST_PHOTO_PATH) image_path_entry.focus() breakpoint()

root.mainloop() ```

Sometimes my CMUS will randomly stop playing a track or it won't play the next track until I manually go in and hit c (for resume) or otherwise initiate playback. I would like to be able to see what happened before these instances so i could either fix my config or, if its a problem beyond the scope of my local environment, get info to write up a proper bug report with. Where can I find such logs? Would they be in the systemd journal or somewhere in /var ?

So I just had an issue where my shebang lines weren't parsing properly for a python script I was attempting to execute. A quick google revealed that it was probably because I wrote the scripts on windows and now I was trying to run them on Linux (so happy i finally made the switch btw. using the computer is fun again!). So i took the advice I found and tried to run :%s/^M/ (using C-v, C-m to insert the escape character), and it failed to find any matches. I tried the same command in vanilla vi and it worked.

Is there some setting I don't have configured properly? I would prefer to be able to do this sort of thing within neovim.

If anyone could help me out with this, I'd greatly appreciate it. Basically, I can't ctrl + v to paste anything I pipe into the clipboard selection in xclip, and i can't xclip -o anything I copied with ctrl-c or ctrl-shift-c.

Maybe I want to paste a path into a neovim file. echo $(pwd) | xclip -i -sel c Now the path is in my clipboard, right? It sure shows up if i xclip -o -sel c!

But when I go into neovim and paste from the unnamedplus register, instead it pastes the last thing I copied in my browser.

if I want to copy the output of something from my terminal and google it, ctrl+v in the browser completely ignores my xclip selection.

i am forced to use the mouse and ctrl+shift+c in order to paste it into a search engine, like a caveman.

I hope I've done a decent enough job explaining the problem. It was most apparent earlier, as I was making a cronjob and I had to be explicit about file paths because i couldn't assume the working directory would be the directory of the script I was calling. I really wish I could have just echo $(pwd) | xclip -sel c; open neovim; hit p; see the path appear in my file.

I have a little clipboard icon in my system tray with my copy history, except none of the things I put in the clipboard selection with xclip -sel c or xsel -b appear there. I think that program is klipper, but I'm not sure.

I know there's a number of work arounds but still this kind of frustrates me. I think it has something to do with wayland and xclipboard not talking to each other. I am running wayland, KDE Plasma 5.27.5, and Debian 12.

Is there a simple configuration setting I can tweak, or do I need to find something to replace klipper or xclip? I have tried toggling the keep selection and clipboard the same and always save text selection in history settings in the plasma clipboard, no change. I tried two terminal emulators to no avail.

I will happily provide any more information if it would be helpful.

Lately ive noticed that i was wanting to do certain things on Windows that just seemed much easier and more intuitive on Linux, based in the OS specific solutions i would see to problems i encountered. And i was more frequently using software where Windows support seemed like an after thought.

A couple days ago i finally sat down and tried to install Mint. The installer didnt recognize my windows partition so it didnt offer any assistance. And a stroke of fate saw my internet connection dieing at the exact same time. Yes, i cant believe it either.

So i decided to live dangerously and just try to wing the installation with no outside help. It seems like creating a second EFI partition was not the right call. The install failed, and I couldn't get back onto windows.

I wound up just using a live ubuntu image for a few days while i wrestled with repairing the boot loader. I didnt succeed. Eventually i just made a windows recovery disk from my Desktop with an intact copy of windows, and had to reinstall windows.

Then i did manage to successfully install Debian, and ive been having such a great time with it so far. I feel like i probably didnt even need to keep a windows partition, especially since i could have just used my desktop if i REALLY needed windows. I havent had this much fun just using the computer since i was a kid.

I'm new to vim and getting the hang of the configuration. I've gone through a couple of youtube videos giving a tutorial of setting it up with lazy.nvim. Now I am following ThePrimeagen's config, and UndoTree doesn't work.

The plugin is loaded and my config is properly sourced, but when I run run :UndoTreeToggle, I get this lovely error:

(E5108: Error executing lua: function undotree#UndotreeToggle[11]..<SNR>37_new[2]..41, line 6: Vim(echoerr):"diff" is not executable. stack traceback: [C]: at 0x7ff77ae89570)

I am running windows (currenty in the process of backing up my personal data so I can try out Linux without feeling too disconnected from my daily computer user). I did a little bit of google searching and checked the issues on github, unfortunately I couldn't find anything related to my problem. Any help would be much appreciated.

I am working on user authentication in Flask. I have my User class, which inherits from db.Model (SQLAlchemy) and UserMixins (flask-login): python class User(db.Model, UserMixin): id = db.Column(db.Integer, primary_key=True) email = db.Column(db.String(100), unique=True) password = db.Column(db.String(100)) name = db.Column(db.String(1000)) and I create a new User object during registration: python new_user = User( name=request.form["name"], password=generate_password_hash(password=request.form.get("password"), salt_length=8, method="pbkdf2:sha256"), email=request.form["email"]) Since I inherited from UserMixins, I started to get an "unexpected arguments" warning from pycharm when I create new_user. can someone explain to me why that is? If I don't inherit from UserMixins, the warning goes away.

So, it used to work just fine. Then jerboa became basically unusable due to some bug. That was a few weeks ago. I saw an update was available, so I thought to give it another try. It's much more stable after the update, and my lemmy.one account works just fine. But when I try to log in with this account on jerboa, I get an incorrect login error. I set the instance to "programming.dev" and I know I used the right credentials because my password manager filled them in, just like it does in the browser.

Any ideas on a cause or fix? It might be a jerboa issue but I don't get why it seems to only impact this instance.

I am trying to create a playlist with spotify and the spotipy library in python. However, I keep getting a "No token provided" error when making my API request. However, if I use the same token with a curl request, it works! Can someone please help. This is my code:

``` auth_manager = SpotifyOAuth(client_id=CLIENT, client_secret=SECRET, redirect_uri="http://example.com/", scope=SCOPE, username=spotify_display_name ) token = auth_manager.get_access_token( as_dict=False, check_cache=True )

sp = spotipy.Spotify(auth_manager=auth_manager, auth=token ) user_dict = sp.current_user() user_id = user_dict["id"] print(f"Welcome, {user_dict['display_name']}")

SEARCH

QUERY FORMAT: "track: track-name year: YYYY"

spotify_search_endpoint = "https://api.spotify.com/v1/search/" test_query = "track:Hangin'+Tough year:1989"

search_parameters = { "q": format_query(test_query), "type": "track" }

results = sp.search(q=search_parameters["q"]) print(results) ```

output: {'tracks': {'href': 'https://api.spotify.com/v1/search?query=track%3AHangin%27%2BTough%2520year%3A1989&type=track&offset=0&limit=10', 'items': [], 'limit': 10, 'next': None, 'offset': 0, 'previous': None, 'total': 0}}

{ "error": { "status": 401, "message": "No token provided" } }

This is really frustrating! The authentication is working, otherwise the token wouldn't have been valid for the curl request. I must be doing something wrong with spotipy.

I was looking over the first kata i did on codewars, and I thought it would be fun to try and solve it in C. The object was to return a string based on a boolean input. it took a lot of trial and error, googling, chat gippity, but I eventually got it to work. I am still focused on learning python, but I've had it in my mind that I should branch out once I've reached a competence plateau in python. I'm nowhere near that plateau yet, but this seemed simple enough to warrant the necessary investment in time to accomplish it.

// C: #include <stdbool.h> // FIRST EVER C PROGRAM const char *bool_to_word (bool value){ // you can return a static/global string or a string literal if (value == 1){ return "Yes"; } else{ return "No"; } }

I realize this is pretty trivial, but still, it's a milestone for me and I wanted to do my part to get the ball rolling on this community.