It was a crisp December morning at the North Pole, and Pixie the Elf (whose parents had weird ideas about names) was already hard at work in Santa’s Workshop. Her job was critical: maintaining Santa’s Good/Bad Child Tracking App, affectionately nicknamed "NaughtyOrNice". The app, a sprawling Perl application, kept records on every child’s behaviour throughout the year. But as Christmas drew nearer, the workload multiplied, and Pixie was feeling the strain.

The Problem: Manual Testing Mayhem

For years, Pixie had been managing NaughtyOrNice with a mix of manual testing, haphazard deployment scripts, and a good deal of holiday cheer. But this year, the app had grown exponentially, with new features like integration with Smart Chimney Sensors™ and support for multi-timezone sleigh operations.

Every code change required hours of testing—on Pixie's local machine, of course—followed by an error-prone manual deployment process. One too many late-night debugging sessions left Pixie wondering if she was on Santa's Naughty list herself.

"I need help," she muttered to herself during one particularly long debugging session. She decided to take a break, grabbed a coffee from the machine and started idly browsing r/elftech. Immediately, an article about GitHub Actions caught her attention.

What Are GitHub Actions?

Pixie had vaguely heard of GitHub Actions before, but she’d always assumed it was for the tech-savvy elves in charge of JavaScript or Rust projects. The article made it seem that GitHub Actions were just as useful for Perl developers, and that it could automate much of her workflow.

Intrigued, Pixie decided to dive in.

A First Workflow: Automating Tests

Pixie started with the basics: automating the test suite for NaughtyOrNice. The app already had a robust set of tests written with Test::More , but running them manually was tedious.

Pixie created her first GitHub Actions workflow file, .github/workflows/test.yml:

name: Perl Test Suite

on:
  push:
    branches:
      - main
  pull_request:

jobs:
  test:
    runs-on: ubuntu-latest

    steps:
    - name: Checkout code
      uses: actions/checkout@v4

    - name: Set up Perl
      uses: shogo82148/actions-setup-perl@v1
      with:
        perl-version: '5.40'

    - name: Install dependencies
      run: cpanm --quiet --installdeps .

    - name: Run tests
      run: prove -lr t/

Pixie was impressed. With just a few lines of YAML, she could trigger her tests automatically whenever someone pushed code to the main branch or opened a pull request. No more manual testing! She could focus on writing code instead of running it.

Adding Deployment: Automating the NaughtyOrNice Rollout

With her test suite humming along, Pixie turned her attention to deployments. Previously, deploying the app involved copying files to the production server with a combination of scp commands and festive hope.

She created another workflow file, .github/workflows/deploy.yml:

name: Deploy to Production

on:
  workflow_dispatch:

jobs:
  deploy:
    runs-on: ubuntu-latest

    steps:
    - name: Checkout code
      uses: actions/checkout@v4

    - name: Deploy to server
      env:
        RSYNC_PASSWORD: ${{ secrets.RSYNC_PASSWORD }}
      run: |
        rsync -avz --delete ./ naughtyornice@production-server:/var/www/naughtyornice/

By adding a secret to the GitHub repository for her server credentials, Pixie automated the entire deployment process. The workflow_dispatch key in the on section of the workflow added a "Run" button to the "Actions" tab in the GitHub repo, and just pressing that would deploy the application to the production server. No more late nights. No more missed sleigh rides. Pixie even dared to start dreaming about creating a Docker container for NaughtyOrNice and deploying the application that way. The time she would save would allow her to look into long-awaited improvements like that.

The Real Magic: Pre-Written Workflows

Pixie was excited. This was going to make her life much better in so many ways, And then she remembered the main point of the article she had been reading. A group of Perl programmers on GitHub had created a set of reusable workflows. They were aimed at CPAN modules, but as the NaughtyOrNice codebase followed all of the CPAN packaging conventions, she thought she could probably use these workflows for her code.

Pixie read the instructions for using these workflows (they were online at https://github.com/PerltoolsTeam/github_workflows/) and, in a few minutes, she had a new workflow which looked like this:

name: CI

on:
  push:
    branches: [ main ]
  pull_request:
    branches: [ main ]
  workflow_dispatch:

jobs:
  test:
    uses: PerlToolsTeam/github_workflows/.github/workflows/cpan-test.yml@main

  coverage:
    uses: PerlToolsTeam/github_workflows/.github/workflows/cpan-coverage.yml@main

  perlcritic:
    uses: PerlToolsTeam/github_workflows/.github/workflows/cpan-perlcritic.yml@main

  complexity:
    uses: PerlToolsTeam/github_workflows/.github/workflows/cpan-complexity.yml@main

Pixie was very happy. In a few lines of code, she had replaced and improved on her existing manual processes. The application was now being tested automatically on several versions of Perl and on Linux, Windows and macOS. She was also getting test coverage reports on Coveralls.io. But the things she was most excited by were the PerlCritic and Cyclomatic Complexity reports. Both of these gave her a huge number of ideas for improvements to the code. And now that the process was so much more efficient, she might actually have time to implement some of them. Happily, she started adding the suggestions into the NaugthyOrNice issue tracker.

The Result: A Smoother Christmas

On Christmas Eve, as Santa prepared for his big journey, he stopped by Pixie's workstation.

"Pixie," Santa said with a twinkle in his eye, "thanks to your hard work, this year the app is the best it's ever been. You’ve truly outdone yourself."

Pixie blushed. "It wasn’t just me, Santa. I had help—from GitHub Actions!"

Your Own Holiday Automation

Pixie's story shows that even the busiest developers can save time and reduce stress with GitHub Actions. Whether you're maintaining a legacy Perl app or building a shiny new one, these workflows can simplify your life.

To get started, check out the pre-written workflows at https://github.com/PerltoolsTeam/github_workflows/. Who knows? They might just bring a little magic to your own development process.

Happy holidays, and happy automating!