Full Linux Desktop with x11docker

A few days ago, I posted a detailed guide on how I got Cursor running inside a container using jlesage/docker-baseimage-gui. And honestly? That setup worked. But “worked” and “worked well” are two entirely different things, and over the past few days, I kept running into issues that pushed me to look elsewhere.

I’m talking:

• Cursor crashing randomly after a few minutes.
• Keyboard input stuck in VNC, had to refresh the page to get it working again (and again… and again).
• Folder sharing with host via bind mounts = permissions hell. Either container complained or the host did. No clean workaround.
• Making the user sudo-ready took extra scripts and hacks.
• Debugging startup issues? Welcome to cont-init.d hell.

It was becoming messy real fast.

Table of contents#

Why I Gave Up on Single App Containers#

All I wanted was a clean setup to run Cursor with:

• Bash shell working inside it.
• Sudo powers.
• Ability to install/update software.
• Folder sharing without pain.

But what I ended up with was me hacking around every edge case, even though the whole point of containers is to keep things simple and reproducible.

Sure! Here’s a rephrased, cohesive version of your section in a smooth, natural tone — nothing important left out, just easier to read and more conversational:

Meet x11docker#

When I first looked at x11docker, it felt like complete overkill. I mean, a full desktop environment? Just to run Cursor?

But here’s the thing: sometimes it’s better to go with something that’s simple, stable, and a bit heavier, rather than wrestling with a lighter setup that needs constant fixing and tweaking. After dealing with all the headaches in jlesage/docker-baseimage-gui, switching to x11docker felt like a breath of fresh air.

Turns out, x11docker isn’t just good, it’s perfect for running full Linux desktop environments inside containers. It’s clean, extendable, and doesn’t rely on clunky VNC or browser-based access. And the best part? It just works.

Why x11docker clicked for me#

Here’s what makes x11docker such a solid choice:

• No lag, no stuck keys, just smooth interaction.
• Offers remote access options (SSH, VNC, or HTML5).
• Clipboard sharing with the host just works, out of the box.
• Run either full desktop environments or single GUI apps (seamless mode).
• File sharing? Super easy. No mount permission hell, just use --home and you’re good.
• Comes with sane security defaults, but lets you enable sudo if you need admin access inside.
• Requires almost nothing on the host, just X and either Docker, Podman, or nerdctl. (For smoother experience, nxagent or Xephyr is recommended. You can also use the prebuilt x11docker/xserver image.)

If you’ve used docker-baseimage-gui before, you’ll feel the difference immediately. No hacks. No scripts. No weird workarounds. Just a fully working Linux desktop in a container, with zero drama.

My x11docker Setup#

Here’s how I set it all up. I’m using podman here, but you can easily swap it with Docker, the commands are basically the same.

# xfce desktop with dev tools
FROM x11docker/xfce

RUN apt-get update && apt-get install -y \
  git curl apt-utils sudo gnupg

# install Node.js if required
RUN curl -sL https://deb.nodesource.com/setup_24.x | bash -
RUN apt-get install -y nodejs
Copy

Save that as Dockerfile, and build the image:

podman build -t x11docker-xfce:1 .
Copy

Running the Container#

Here’s the command that works beautifully for me:

x11docker --desktop \
  --backend=podman \
  -I \
  --clipboard \
  --cap-default \
  --sudouser \
  --home \
  x11docker-xfce:1
Copy

Let’s break this down:

• --desktop: Start a full XFCE desktop session, remove it if you want seamless mode
• --backend=podman: whatever you’re using as your containerization tool
• -I: Enable internet access (no network jail)
• --clipboard: Sync clipboard between container and host
• --cap-default: Use default container capabilities
• --sudouser: Lets the container user run sudo
• --home: Automatically persist home directory between container runs

Persistent Storage#

Thanks to the --home flag, everything inside your home directory sticks around between sessions. That means:

• Whatever you download inside the XFCE session (e.g. Cursor .deb) stays.

• These files are available in host at:

  # default shared folder when --home flag used
  ~/.local/share/x11docker/x11docker-xfce/
  Copy

• Want to copy a file from host → container? Just drop it in that folder.

Clean. Predictable. No permission nightmares.

Installing Cursor#

The only manual step I do each time is installing Cursor, which is no big deal.

1. Download the .deb and place it inside the shared home folder.

2. Inside XFCE, just run:

   sudo apt install -y ./Cursor.deb
   Copy

It takes 5 seconds. And once done, Cursor launches perfectly in a full desktop session.

Compared to docker-baseimage-gui?#

Final Thoughts#

If docker-baseimage-gui was my first real taste of GUI containers, then x11docker was the upgrade I didn’t know I needed.

I’m now running:

• A full XFCE desktop inside a container
• With Cursor, Node.js, Git, bash, sudo, and everything else I need
• Seamless host ↔ container file access
• No browser VNC jank
• No extra init script hacks
• No weird sudo workarounds

I can focus on building, writing, coding, without debugging the container itself every 5 minutes.

If you’ve worked with my previous setup even for a couple of days, you’ll immediately appreciate the beauty of x11docker. It’s a bit heavier, sure, but you get your sanity back.