Apache tooling

Clone this repo:
  1. 017ef24 Update the proposed packaging conventions by Sean B. Palmer · 6 days ago main
  2. ad1bf3d Update Tooling action commit hashes in workflows by Sean B. Palmer · 7 weeks ago
  3. 15e4ebb Remove checkouts where not required by Sean B. Palmer · 4 months ago
  4. 215e16e Bump the version for testing GitHub actions by Sean B. Palmer · 4 months ago
  5. e4c3a55 Update dependencies by Sean B. Palmer · 4 months ago

asf-example

This package does not have official Apache Software Foundation (ASF) releases.

This is an example test package issued by Tooling at the ASF. It contains no stable user facing functionality. We do not make guarantees about any of the interfaces contained herein, nor the continued existence of this package.

The purpose of this package is to test:

  • Proposed package naming conventions and version schemes within ASF Infra and ASF Tooling
  • GHA workflows for trickle down use in other ASF Tooling projects
  • Interaction with the ATR platform presently under development by ASF Tooling

Proposed conventions

The proposed packaging conventions are:

  • Use asf${NAME} or ${NAME} for PyPI package names, depending on whether the package is for ASF use only or not
  • Ensure that package code is in ${NAME}/ only
  • Use 0.0.N for versions, starting with 0.0.1
  • Increment N for each released version
  • Optionally use 0.0.(N+1)-devM for non-release main commits
  • Start with -dev1 if using non-release main commit versions

A package having three commits on main, then a release, then two more commits to main, then another release, will for example use the following versions if opting into -devM versions:

0.0.1-dev1
0.0.1-dev2
0.0.1-dev3
0.0.1
0.0.2-dev1
0.0.2-dev2
0.0.2

But there is a question of what to do about rebasing.

These packaging conventions are proposed only for ASF Infra and Tooling packages which will be published to PyPI. Other packages, for internal use, may use any packaging conventions. Aligning with these conventions, if adopted, would, however, make it easier to promote internal packages to PyPI.