I believe the time-based tests are falsely failing when the CI machine is under high load. This also speeds up the tests in CI.
I'm not 100% happy with the approach, but this should resolve CI issues in the short term until I can improve things.
- adds a process-compose namespace for running `storybook` and `luakit` together to replicate the `webkit`-based rendering that happens inside of `webview`
- adds some helper scripts for running storybook tests and updating snapshots, with documentation in the README.
- adds a `clan-app-ui-storybook` package which builds and tests the storybook, checking for rendering changes
Currently, we’re only doing markup-based snapshot tests. We’re also using headless chromium for the tests by default as I couldn't get webkit to work in the nix build.
As we’re only markup-based for the time being, this should be ok. But eventually I'd like to get it working with webkit.
