Skip to main content

Terminal-Native Workflows

CLI-focused development patterns for users who prefer terminal over IDE.

Note: In examples below, replace <project-name> with your actual project name. The $(id -u) part auto-substitutes your user ID.

Philosophy​

Everything via terminal:

  • Edit code: vim/nano
  • Run code: docker exec
  • Debug: logs and print statements
  • Version control: command-line git
  • Sessions: tmux/screen

No IDE required.

Basic Pattern​

# Edit code on host
vim ~/workspace/my-project/train.py

# Run in container (non-interactive)
docker exec <project-name>._.$(id -u) \
  python /workspace/train.py

# Check output
tail -f ~/workspace/my-project/output.log

Vim-Based Development​

Edit on host, run in container:

# Open project in vim
cd ~/workspace/my-project
vim src/model.py

# Test changes
docker exec <project-name>._.$(id -u) \
  python /workspace/src/model.py

# Or enter container for interactive testing
docker exec -it <project-name>._.$(id -u) bash
>>> python
>>> from src.model import MyModel
>>> model = MyModel()

Benefits:

  • No IDE overhead
  • Fast editing
  • Works over slow SSH

tmux Session Management​

Persistent terminal sessions:

# Create session
tmux new -s dev

# Split for editing + running
Ctrl+B %    # Split vertical
Ctrl+B "    # Split horizontal

# Left pane: vim
vim ~/workspace/my-project/train.py

# Right pane: container shell
docker exec -it <project-name>._.$(id -u) bash

# Detach
Ctrl+B D

# Reattach later
tmux attach -s dev

Non-Interactive Execution​

Run commands without entering container:

# Single command
docker exec <project-name>._.$(id -u) \
  python /workspace/train.py

# Multiple commands
docker exec <project-name>._.$(id -u) bash -c "
  cd /workspace
  python preprocess.py
  python train.py
  python evaluate.py
"

# With environment variables
docker exec -e CUDA_VISIBLE_DEVICES=0 <project-name>._.$(id -u) \
  python /workspace/train.py

Log Monitoring​

Follow logs in real-time:

# Container logs
docker logs -f <project-name>._.$(id -u)

# Application logs
tail -f ~/workspace/my-project/training.log

# With grep filter
tail -f ~/workspace/my-project/training.log | grep "epoch"

# Multiple logs
tail -f ~/workspace/my-project/*.log

Command-Line Git​

Version control without IDE:

cd ~/workspace/my-project

# Check status
git status

# Stage changes
git add src/model.py

# Commit
git commit -m "Improve model architecture"

# Push
git push

# View history
git log --oneline --graph

# Diff changes
git diff src/model.py

Process Management​

Background jobs:

# Start in background
docker exec -d <project-name>._.$(id -u) \
  nohup python /workspace/train.py > /workspace/output.log 2>&1

# Check running processes
docker exec <project-name>._.$(id -u) ps aux | grep python

# Kill process
docker exec <project-name>._.$(id -u) pkill -f train.py

File Transfer​

Copy files between host and container:

# Host to container
docker cp local-file.txt <project-name>._.$(id -u):/workspace/

# Container to host
docker cp <project-name>._.$(id -u):/workspace/results.txt ~/

# Entire directory
docker cp ~/data/ <project-name>._.$(id -u):/workspace/data/

Usually not needed - use workspace mount instead.

Shell Aliases​

Speed up common tasks:

# Add to ~/.bashrc
alias dex='docker exec -it <project-name>._.$(id -u) bash'
alias drun='docker exec <project-name>._.$(id -u)'
alias dlogs='docker logs -f <project-name>._.$(id -u)'

# Usage
dex                           # Enter container
drun python /workspace/train.py   # Run command
dlogs                         # Follow logs

See Also​