AWS CodeBuild vs fastlane

AWS CodeBuild is a paid, cloud-hosted continuous integration service for compiling, testing, and packaging code from build scripts. It provisions and scales build capacity automatically, so projects do not require users to manage build servers. Preconfigured environments cover Java, Python, Node.js, Ruby, Go, Android, .NET Core for Linux, and Docker; teams can also supply Docker-based environments. A YAML build specification defines commands for each build phase and other settings. Builds can start from CodeCommit, GitHub, GitHub Enterprise, Bitbucket, Amazon S3, or CodePipeline, and can work with Jenkins. Artifacts can be stored in S3 or sent to an artifact repository. IAM provides access controls, while KMS keys encrypt artifacts. Build timing, status, commit ID, metrics, and logs are available through the service and CloudWatch, with SNS event notifications. A free plan provides stated monthly allowances; paid usage starts at 0.01 USD and can add charges for data transfer and related AWS services.

fastlane is a free, open-source platform for automating iOS and Android app development and release workflows. Teams define workflows as lanes and can run them with a single command. A lane can manage build numbering, code signing, builds, beta uploads and changelog creation; fastlane also supports store releases, including metadata and screenshot management and submission for review. It can create localized app store screenshots for the languages and devices an app supports. The site lists support for more than 15 beta testing services, including TestFlight, Crashlytics Beta, Play and Hockey. Documentation describes integrations with more than 400 tools and services, including major CI systems, alongside community-built actions and plugins. For iOS signing, signing identities and profiles can be kept in a private encrypted Git repository and synced between machines. fastlane runs on a user's machine or server, and its documentation says credentials and other sensitive information stay there. macOS is officially supported; Linux and Windows are partially supported. The project uses the MIT license.