Unlock the screen orientation when exiting fullscreen

[marcoscaceres]

Closes #202

• At least two implementers are interested (and none opposed):
  • Interested in WebKit (and bug filed)
  • Implemented in Chrome
  • bug filed for Gecko
• Tests are written and can be reviewed and commented upon at:
  • http://wpt.live/screen-orientation/fullscreen-interactions.html
• Implementation bugs are filed:
  • Chromium: already implemented (confirmed on Chrome Android).
  • Gecko: https://bugzilla.mozilla.org/show_bug.cgi?id=1798646.
  • WebKit: https://bugs.webkit.org/show_bug.cgi?id=247250
• MDN issue is filed: Unlock the screen orientation when exiting fullscreen #206 (comment)

(See WHATWG Working Mode: Changes for more details.)

---

Preview | Diff

[marcoscaceres]

Unsure who to ping for comment on for Gecko.

[marcoscaceres]

Checked Gecko (Firefox Nightly on Android with prefs orientation prefs enabled) and it does also exhibit this behavior (it unlocks the orientation lock when exiting full screen). Where it differs is that it doesn't reject the orientation lock's pending promise - but that is minor.

@EdgarChen, given the above, would you be supportive of this change?

[EdgarChen]

Yes, Gecko unlock the orientation lock after document exits fullscreen, see https://searchfox.org/mozilla-central/rev/17349477695facefe7d180d7afc2b74a965c21db/dom/base/ScreenOrientation.cpp#715, so it looks good to me. I filed https://bugzilla.mozilla.org/show_bug.cgi?id=1798646.

[marcoscaceres]

@whatwg/documentation just a heads up about this integration. The tl;dr being that orientation lock is automatically released when exiting full screen (i.e., developer don't need to call screen.orientation.unlock() before exiting fullscreen, as the browser will do the right thing for them).

[marcoscaceres]

I think the CI error is unrelated to this change:

LINK ERROR: No 'dfn' refs found for 'ancestor browsing context'. 

[marcoscaceres]

@annevk any criteria that I've missed that is preventing this from being merged?

[annevk]

This looks good to me, but @foolip is the editor and should probably weigh in.

[marcoscaceres]

@foolip, ping 🛎️