Skip to content

Atomic, automated upstream merges #82

@brynpickering

Description

@brynpickering

We should track upstream and attempt a merge on every change, to reduce merge conflicts and make all the latest changes available in our child projects as quickly as possible. This could be an entirely automated process that falls back on human intervention when:

  1. there are merge conflicts
  2. there are test failures
  3. the objective function value of test runs deviate from a tolerance we define (we would need to cache these values to compare against)

The CI could be automatically skipped when a PR is open awaiting human intervention, to avoid a backlog of PRs being generated.

Child projects can then pull in changes at whatever cadence they choose but they have the capacity to pull in latest pypsa/pypsa-eur changes as quickly as possible.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions