* running all the tests
* setting up CI
* fixed the workflow
* workflow on pull request, just to make it appear
* removed the test to run var
* fix archived tests name
* improved the tests, by filtering out some noise
* pr suggestions and added an improvement to the filtering
* improved the interrupt handler
* improved the UI interruption monitor handler
* some more refinement to handle the interruptor + false positive for non human readable labels
* reverted wrong commit
* ready for review, removed the on pull request check
* pr suggestions
* Update our tools package to Swift 6.1
Also improves the package layout with subdirectories 📁
* Update GenerateSDKMocks to be an Async command.
* Add a tool to bump the project CalVer every month.
* Add a workflow to automatically bump the calendar version.
Note: This only does year & month, the patch is handled by the release script.
Because we're now requiring the 'Prevent blocked' check to pass before merging a PR, GitHub Actions apparently expects it to be associated with the latest Git ref of the PR's branch whenever the branch is updated. Therefore we need to re-run the workflow on the 'synchronize' event.
We've haven't really been using these and they're proving to be quite a chore to keep running.
Also, we recently needed an external demo build and we chose to go with custom build on top of testflight instead.
As such it's been decided that we can go ahead and remove them.
* Compile errors 😬
* Update server selection snapshots.
The default footer string has changed.
* Remove the TemplateScreen UI tests.
We use preview tests for snapshots now, UI tests are for flow coordinators.
* Add the same looping on the OIDC UI test (but disable the test in the end).
* Run UI tests separately for iPhone and iPad.
GitHub runners don't seem to have enough memory to handle 2 simulators at once.
possible fix
possible fix
Revert "possible fix"
This reverts commit 6484648153bf8f05485050f544b136e6ca9f0be9.
Revert "possible fix"
This reverts commit abb3a58c630bb0abab7c86a31be022496d4f3b4c.
possible fix
possible solution using ui interrupt monitor
Revert "possible solution using ui interrupt monitor"
This reverts commit ed28c85a37744e615c4bc464b3e8962d8c953be9.
possible fix
Revert "possible fix"
This reverts commit 43039e73c1e1b70493f2e22aa20f925257d2e6de.
fix
* Bump the version for the next release.
* Ignore all generated sources.
* Use Pkl+XcodeGen to inject secrets instead of the project.
* Inject the PostHog/Sentry/Rageshake configuration from the environment.
* Fix bad unicode.
* Fix unit tests.
* Use the new TimelineMediaPreview modifier on the room and pinned timeline screens.
* Use the same presentation logic for all timeline media previews.
* Fix a bug with the detection of the timeline end.
* Send pagination requests from the media preview screen.
* Add SwiftLint to the Danger workflow (it is no longer installed on the runner).
* Put SwiftLint back on all of the GitHub runners too.
* Set the function_parameter_count lint rule to 10.
* Make sure to clean-up any previews when the coordinator is done.
* Handle the viewInRoomTimeline action more appropriately.
* Disable image and document picker integration tests as they randomly fail to load and are flakey.
* Delete any pre-existing log files
* Various tracing tweaks and fixes:
- delete the custom tracing log levels as we can't control their ouput
- implement comparable on them
- change default levels only if the new chosen level increases their verbosity
* Make logging targets mandatory and fix their logging levels
* Switch back to using the `run_tests` reset simulator flag as `fastlane snapshot reset_simulators` was too generic and slow
* Switch all integration test taps to `tapCenter` (nee forceTap) after noticing missed taps on CI.
* Make the logging file prefix explicit, let the main app not use one.
* Rename tracing configuration `target` to `currentTarget`
* Wait longer on authentication flow tests.
* Move default perceptualPrecision value into the snapshot preferences.
* Delay snapshots *before* setting up the test.
* Reset the simulators on GitHub before running?
* Remove a test that is now handled by Rust.
* Fix a test that was yielding.
* Send a message from integration tests and check that it doesn't show up in the app logs
* Increase photo library timeout
* Wait for image tapping to process
* Send a message using the rich text editor too
* Wait for the various pickers to finish loading
* Set all logs to `trace` in integration tests
* Close formatting options after sending the second message
* Make sure logs are actually sent to tracing and get redirected to the right file
* Switch from using a custom `trace` log configuration to the lowest level `TracingConfiguration` we support
