1. Build the Apptainer .sif from the def file here

ml purge
ml Apptainer
apptainer build quarto-code-server.sif quarto-code-server.def

2. Create an interactive session with Screen or something similar
3. Allocate an interactive node. Using salloc is preferred over grabnode due to the automatic assignment of a tmpdir
4. Launch the apptainer with sbatch launch_quarto_server.sh
   • Make sure /var/tmp is writeable for unpacking vs-code server (needed for first time installations)
5. Follow the in-terminal prompts to setup and connect to the vs-code server in your browser.
6. Navigate to your project directory in the browser vs-code.
   • When you first open the server in your project, give the app a minute to set up extensions. These extensions will allow jupyter environments to be auto-detected and loaded when opening terminals.
7. Run quarto check in the vs-code terminal to ensure all langauges and extensions are properly loaded.
8. You can add extensions as needed (e.g. syntax help or linters). These should not affect the ability of other users to work with the code in a project.
9. If you want to preview a quarto document, you will have to set the host to 0.0.0.0 and manually navigate to the address of the host-node with the port given by the quarto preview output. The built in viewer doesn't seem to play well with the 'local host'

quarto preview mydoc.qmd --no-browser --host 0.0.0.0

Then you can go to <$(hostname)>.fhcrc.org:<port> to preview the live output