Control Center

The Control Center is a web-driven interface that launches by default upon invoking Iris. It serves two main purposes:

1. View past runs.

   • The index page contains a list of past runs with basic information, and a link to a detailed run page.
   • The run page provides all of the metadata of a given run, as well as details of each executed test. This can be used to debug a failed run, as the page provides links to debug images, console logging, and other saved output.

2. Configure and launch a new run.

   • A list of all tests is provided, with the ability to enable/disable tests on a package or individual basis.
   • Desired characteristics of the test build, such as channel and locale, can be chosen, as well other runtime parameters.

---

To view past runs:

1. Launch the Control Center by simply typing iris in the console. The feature is enabled when Iris is invoked, either with no parameters, or with the -k flag.
   • Note: If you use any other combination of runtime flags, the Control Center is not launched. In this case, it is assumed that the user does not require a visual interface and prefers to work with command-line arguments.
2. The topmost element on the opening page is the runs table. If this is the first time you have run Iris, it will be empty.
3. With each run, this table will be populated with an item identified by its run ID, equal to the timestamp when it was triggered. Basic information, such as number of tests run, failures, build channel and locale are displayed. In addition, visual icons on the right allow one to either view the test run's page or delete the run.
   • The trash can icon, used to denote the delete functionality, removes the run's data from the disk. This action cannot be undone.
   • A run that indicates number of tests as * is a run that was interrupted before completion. It is an invalid run.
4. The table can be sorted using the column headers. The default sort is newest run to oldest.
5. Clicking on either the run ID link or the eye icon in its row will open a new page, the run page.
6. The run page is divided into two main sections - metadata and tests.
   • Metadata is a list of properties and values that describe the environment of the test run, including the browser, Git repo, runtime flags, and other identifying data.
   • Tests are listed in their own section below, in the order that they were executed. Clicking on a package name expands all tests present in that package, and further expansion of test names reveal more data about each executed test.
     
     Each test displays its own data, including description, Firefox parameters, profile, and length of time to execute. In addition, you can view all associated debug images generated while the test was run, to recreate a visual timeline of its image detection logic. This is very useful for debugging a failed test, to verify quality of test images, as well as to inspect screen captures for any details that could have bearing on the test.

---

To launch a new run:

1. On the initial index page, view the test package list.
2. Click on packages or tests included within to add to the run.
3. Select a build and locale of Firefox, or leave the defaults.
4. (Optional) You can also enable other runtime features using the additional controls on the page. Documentation on this section is TBD.
5. When you have chosen your desired test set and build, click the "Go" button.
   • If you instead wish to exit without running tests, click the "Cancel" button.