The WebDriverCommand::FullscreenWindow in geckodriver maps to the non-existent Marionette command "fullscreenWindow". It should be "fullscreen".

Comment on attachment 8894491 [details] Bug 1388036 - Map WebDriverCommand::FullscreenWindow correctly. https://reviewboard.mozilla.org/r/165654/#review170738 ::: testing/geckodriver/src/marionette.rs:1044 (Diff revision 1) > SetTimeouts(ref x) => (Some("setTimeouts"), Some(x.to_marionette())), > SetWindowRect(ref x) => (Some("setWindowRect"), Some(x.to_marionette())), > GetWindowRect => (Some("getWindowRect"), None), > MinimizeWindow => (Some("WebDriver:MinimizeWindow"), None), > MaximizeWindow => (Some("maximizeWindow"), None), > - FullscreenWindow => (Some("fullscreenWindow"), None), > + FullscreenWindow => (Some("fullscreen"), None), When it wasn't working correctly so far, why don't we use the `WebDriver:FullscreenWindow` command directly?

Comment on attachment 8894494 [details] Bug 1388036 - Add WPT tests for Fullscreen Window. https://reviewboard.mozilla.org/r/165660/#review170756 ::: testing/web-platform/tests/webdriver/tests/fullscreen_window.py:26 (Diff revision 1) > +def test_handle_user_prompt(session): > + # step 2 > + session.url = alert_doc > + response = fullscreen(session) > + assert_error(response, "unexpected alert open") > + We need to add more tests like in `get_title.py` for prompts ::: testing/web-platform/tests/webdriver/tests/fullscreen_window.py:47 (Diff revision 1) > + rect = response.body["value"] > + assert "width" in rect > + assert "height" in rect > + assert "x" in rect > + assert "y" in rect > + assert isinstance(rect["width"], float) This should test against `(int, float)`

Comment on attachment 8894492 [details] Bug 1388036 - Add client.Session#fullscreen API to WebDriver client. https://reviewboard.mozilla.org/r/165656/#review170758

Comment on attachment 8894493 [details] Bug 1388036 - Clean up fullscreen state after wdspec test. https://reviewboard.mozilla.org/r/165658/#review170760

Comment on attachment 8894491 [details] Bug 1388036 - Map WebDriverCommand::FullscreenWindow correctly. https://reviewboard.mozilla.org/r/165654/#review170762

Comment on attachment 8894491 [details] Bug 1388036 - Map WebDriverCommand::FullscreenWindow correctly. https://reviewboard.mozilla.org/r/165654/#review170738 > When it wasn't working correctly so far, why don't we use the `WebDriver:FullscreenWindow` command directly? WebDriver::FullscreenWindow is only available in Firefox 56 and greater. By using "fullscreen" it will work with earlier Firefoxen.

Comment on attachment 8894491 [details] Bug 1388036 - Map WebDriverCommand::FullscreenWindow correctly. https://reviewboard.mozilla.org/r/165654/#review170738 > WebDriver::FullscreenWindow is only available in Firefox 56 and greater. By using "fullscreen" it will work with earlier Firefoxen. Makes sense. Thanks.

The new fullscreen tests appear to cause too many intermittents with Linux debug builds. I suspect I need to revisit the actually Marionette implementation and make it more reliable.

Comment on attachment 8894494 [details] Bug 1388036 - Add WPT tests for Fullscreen Window. https://reviewboard.mozilla.org/r/165660/#review171362

Comment on attachment 8898821 [details] Bug 1388036 - Restore window when setting window rect. https://reviewboard.mozilla.org/r/170194/#review175896

Pushed by atolfsen@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/d8f04b1a073f Restore window when setting window rect. r=automatedtester https://hg.mozilla.org/integration/autoland/rev/8efd7bd68e91 Map WebDriverCommand::FullscreenWindow correctly. r=automatedtester https://hg.mozilla.org/integration/autoland/rev/c0f5b257ba6b Add WPT tests for Fullscreen Window. r=automatedtester

https://hg.mozilla.org/mozilla-central/rev/d8f04b1a073f https://hg.mozilla.org/mozilla-central/rev/8efd7bd68e91 https://hg.mozilla.org/mozilla-central/rev/c0f5b257ba6b