Skip to content

TODO.md

Latest commit

 

History

History
84 lines (67 loc) · 2.82 KB

TODO.md

File metadata and controls

84 lines (67 loc) · 2.82 KB

TODO notes

Answer questions tl;dr, make reference to longer descriptions

Why?

  • Why jsonnet? What benefit do we gain from it?
  • Why is it worth the inconvenience of people having to learn a new language/paradigm?
  • Why we use jsonnet (vs other solutions/tools)?
  • Consider creating an extensive retrospective design doc/blog post

How? (kind of covered in index page and subsequent lessons)

  • What are the lessons we’ve learned after using jsonnet for years?
  • What are some GrafanaLabs jsonnet idioms that we try to use?
  • Tools/tricks when writing/debugging jsonnet?

Ideas:

  • Consider Jsonnet introduction lesson
  • IDE configuration, guest lessons/breakouts (vim, emacs, neovim, VSCode, IntelliJ)
  • Consider Terminology list
  • Provide good-read list
  • CD process (kube-manifests, jsonnet-libs/k8s github workflow)

Lessons:

  • Write a reusable library

    • Properly use keywords such as self, $, local, super, null
    • Write for extensibility with :: and objects rather than arrays
    • Write object-oriented with 'mixin' functions
    • Apply YAGNI often
    • Know how to avoid common pitfalls:
      • Builder pattern
      • "private" variables

  • Don't write libraries

    • Find existing libraries
    • Vendor libraries with jsonnet-bundler
    • Use a vendored library with JSONNET_PATH

  • Exercise: rewrite reusable library with k8s-libsonnet

    • Vendoring k8s-libsonnet with jsonnet-bundler
    • Understand k.libsonnet convention
    • Use generated documentation

  • Developing libraries

    • Wrapping libraries
      • Pentagon usecase
      • Grafonnet usecase
    • Developing with upstream libraries

  • Documentation & testing

    • Use of docsonnet
    • Test with the use of error, null and prune
    • Investigate testing framework (jsonnetunit seems dead, fork?)

  • Jsonnet use cases

    • Roll out in 'waves' (dev/stag/prod)
    • Come up with a few examples (Grafonnet, GitHub Actions)
    • Write YAML/JSON files (Grafonnet, GitHub Actions)
    • Generate new libraries from specifications (again)

  • Jsonnet features/pitfalls

    • prune is recursive
    • Provide deprecation notices with std.trace
    • Handle dynamic inputs with top-level arguments and /dev/stdin
    • Write arbitrary files (jsonnet -S -m)

Already covered by tanka.dev:

  • Using Tanka

    • Use of inline environments (hint: tanka-util)
    • Keep environments clean
    • Locally inspect with tk eval
    • Continuous Delivery with tk export
    • Know how to avoid common pitfalls:
      • Global variables
      • Smashing libraries together

  • Exercise: Use Helm chart in Jsonnet

    • Vendor Helm charts with tk tool charts
    • Load chart with tanka-util
    • Modify chart beyond values.yaml
    • Patch arrays