Configure lincbrain PyPI releases

[kabilar]

• Update lincbrain PyPI releases. I would suggest removing previous WIP releases, and starting at 0.1.0. In future releases, let's only increment the minor version since the dandi package has not yet hit 1.0.
• Understand semantic versioning for autorc package releases
• Update release docs
  • Create labels for semvar bump in GitHub CLI repo

  versionLabels: Map(6) {
    'major' => [ 'major' ],
    'minor' => [ 'minor' ],
    'patch' => [ 'patch', 'performance' ],
    'skip' => [ 'skip-release' ],
    'release' => [ 'release' ],
    'none' => [
      'internal',
      'documentation',
      'tests',
      'dependencies',
      'released',
      'prerelease'
    ]
  

  • For releases, manually label each pull request with release tag and major, minor, or patch label.

[aaronkanzer]

Update: Seems there have been some silent errors regarding GitHub CI/CD pushing an updated .whl package to PyPI, thus we are currently pushing from a local build via the following steps:

Bump the version in setup.py

then

python3 -m build. # builds `.whl` in `dist/`

then

twine upload lincbrain_cli-*.whl

# where the username is __token__ and the password is the API Key

[kabilar]

Hi @aaronkanzer, could you please update dandi/handbook#104 with our understanding of how the releases are handled (some notes on this process are above)? We can then close this issue. Thank you.

[kabilar]

Closing as we are working to replace the lincbrain-cli with the dandi-cli. See dandi#1527.