It was agreed that the concept of window state would be dropped from WebDriver L1. We also agreed at the WebDriver WG work day on last Wednesday that the appropriate return type for ElementRectResponse and WindowRectResponse would be signed 32-bit integers because the WebIDL for screenY/screenX/outerWidth/outerHeight mark them as longs. This means that whilst it is technically possible to set a window to a decimal-point value, the serialised return value will always be a long.

I will deal with the ElementRectResponse/WindowRectResponse return types in a separate bug https://bugzilla.mozilla.org/show_bug.cgi?id=1396866.

Comment on attachment 8904591 [details] Bug 1396838 - Drop WindowState. https://reviewboard.mozilla.org/r/176430/#review181358

Comment on attachment 8904617 [details] Bug 1396838 - Set window rect before every test. https://reviewboard.mozilla.org/r/176438/#review181362

Comment on attachment 8904618 [details] Bug 1396838 - Remove window state from wdpsec tests. https://reviewboard.mozilla.org/r/176440/#review181364

Pushed by atolfsen@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/354e76aa5821 Drop WindowState. r=jgraham https://hg.mozilla.org/integration/autoland/rev/3cc131aab5f4 Set window rect before every test. r=jgraham https://hg.mozilla.org/integration/autoland/rev/9eba75195a30 Remove window state from wdpsec tests. r=jgraham

https://hg.mozilla.org/mozilla-central/rev/354e76aa5821 https://hg.mozilla.org/mozilla-central/rev/3cc131aab5f4 https://hg.mozilla.org/mozilla-central/rev/9eba75195a30