Introducing score-compose 0.13.2

[sujaya-sys]

Release announcement

We're excited to announce that a new iteration of the score-compose reference implementation has been released. This update marks a major milestone for Project Melody, aimed at achieving a smoother and more consistent developer experience with Score.

Tip

Follow the installation instructions for score-compose to download the latest version 0.13.2

Introducing score-compose 2.0 with Project Melody

Project Melody aims to make Score highly accessible by default, sparing developers from having to deal with intricate configurations, while still allowing for customization in more advanced scenarios.

In response to feedback received on our score-compose reference implementation, we’ve focused on improving the resource management experience. While defining dependencies like databases in the Score Spec is straightforward, configuring them to run with score-compose has required additional manual steps. In light of this, we've rolled out a series of improvements to score-compose, where the heavy lifting is handled directly within the CLI.

Change overview: Say hello to init & generate

We’re excited to introduce the init and generate commands to the score-compose CLI. While we've tidied up a few bugs and implemented improvements from our backlog, these two main bits of new functionality are the stars of the show.

• Init: The init command makes it significantly easier to get started with score-compose, preparing the current directory and any necessary local files or configurations needed to get your workload running. It also generates a default score.yaml file to help you get started.

• Generate: With the generate command, Score files in the your score-compose project are converted into a consolidated Docker compose manifest. This includes resolving and provisioning all resources and links between workloads as needed.

Check out the quick demo below to see these two commands in action. In this example, we’ll be working with a Score file, which defines a dependency on a Postgres database and a Redis cache:

# Specification reference: https://docs.score.dev/docs/reference/score-spec-reference/
---
apiVersion: score.dev/v1b1
metadata:
  name: example
containers:
  hello-world:
    image: nginx:latest
    variables:
      EXAMPLE_VARIABLE: "example-value"
      volumes:
        - source: ${resources.one}
          target: /foo
service:
  ports:
    www:
      port: 8080
      targetPort: 80
resources:
  one:
    type: volume
  two:
    type: redis

Now, score-compose can generate resources without developers needing to configure their own resource definitions or determine desired outputs manually:

In addition to postgres, volume, redis and s3 we’re planning to support resource types such as DNS and route, with additional options planned for the future. And you can customize and extend with your own project, team, or organization-specific resource types from day one.

If you have any feedback or questions, please feel free to reach out to us at any time on the Score community Slack.

Reference

• https://score.dev/blog/score-at-kubecon-eu-in-paris
• https://docs.score.dev/docs/score-implementation/score-compose
• https://github.com/score-spec/score-compose/tree/main/examples