Last updated: June 1st, 2021
Draft of the v1 release slated for Summer 2021.
kpt has added a declarative way to customize and validate configuration.
This allows you to run several mutation and validation
functions in a pipeline alleviating the need to create shell scripts that do
the same thing. Further information can be found in the
declarative function execution section of the The Kpt Book.
Setters used to be a special entity without sufficient differentiation from KRM functions. In kpt v1 setters become just another function and drastically simplify syntax. Configuring 4 setters used to take 20 lines of yaml and now takes 6. Setters also get recursively applied to sub-packages by default. For further information on the setter function please visit: apply-setters documentation.
kpt package updates now default to the resource-merge strategy which allows you to edit configuration with an text editor of your choice and still be able to get updates with upstream changes.
kpt live used to use ConfigMap to store inventory information. This was convenient as it didn't require any CRDs, but it had challenges around encoding of the GroupKind, name and namespace, and it didn't allow us to easily add additional metadata about a package, such as the package version. This is is now migrated to use ResourceGroup CRD.
The majority of the code for kpt live is in the cli-utils repo. Our goal is to keep this code independent of kpt and provide the apply logic as a library that can be used by other tools. The API is still going through some changes, but we are actively working to stabilize it.
Documentation was a major area of investment, including the The Kpt Book. The book is a methodical way to introduce some unique kpt concepts like in place editing and hydration. It's meant to be a hands on guide where the user configures and deploys wordpress and nginx while learning about the kpt concepts.
All of the hydration and validation logic has been moved from the kpt binary to functions allowing for flexibility and security. This enables new scenarios like limiting the customization and validation to a subset of allowed functions. The function catalog has received additional functions, examples and help. Please visit the function catalog for further information.
Since this is a draft of the release notes you should be aware of the ongoing work. At the time of writing the team is in Milestone 3, and we are planning to complete the work in Milestone 4. For instance as of 6/1/21 out of place hydration support is in the works and will land in O(week) timeline.
The list of all the future milestones can be found here: kpt milestones. As dictated by semver, post v1 releases will be guaranteed to be backwards compatible. Major enhancement in release 1.1 will be around OpenAPI schema extensability.
There are a number of breaking changes that had to be done to clean up the CLI and the data format for kpt. Please visit the migration guide for your existing kpt content.
- File a new issue on Github, but please search first.
- kpt-users@googlegroups.com