In the lifecycle of vX.Y.Z, all tests are run against vX.Y-test
before it is published, including when the tag is pushed because the
automated release process now runs end-to-end before pushing the
release to forgejo-experimental.
Running end-to-end against vX.Y-dev is therefore redundant with at least
two other runs with exactly the same SHA (the one before the tag is
pushed and the one when the tag is pushed). There would be value in
doing that if it allowed to detect race conditions in Forgejo. But
such races were not found in the past six months and there is a lot
more scrutiny on commits merged in Forgejo which makes it even less
likely than it was before.
Running the tests on vX.Y instead of also including the built version
provide the same coverage and reduces the workload.
/repos/{owner}/{repo}/actions/workflows/{workflowname}/dispatches
* fails if the required inputs are not provided
* sets defaults to inputs that are not provided
The type of inputs are only used for building the web UI interface,
not for validation. All values are strings.
Do not force the runner version, use whatever default setup-forgejo
provides.
Forcing the version will break the cascading-pr workflow. It will
attempt to download a version that does not exist instead of building
it from sources.
This reverts commit dca3641cf3.
Instead of hardcoding "forgejo" as the base for the daemon to store
the PID and the logs, use the base of the WORK_PATH so that a given
work path can run a dedicated forgejo instance by the same name.
* replace the high level test running actions tests with end-to-end.sh
* set DOMAIN to the IP instead of 127.0.0.1 for runner <-> forgejo communications
* move forgejo_cli from a function to a file so that it can be used by forgejo-runner.sh
* keep the documentation updates workflows separate because they need to open one PR per version
