312882 - Command-arrow gets caught by page arrow key handler

Status: RESOLVED INCOMPLETE

(Camino Graveyard :: General, defect)

Description

[Simon Fraser [no longer active]]

If a page watches for key press events and looks for arrow keys (for example),
their event handler fires for Command-key combinations. This is wrong.

Comment 1

[Simon Fraser [no longer active]]

Attachment: Page to test key events

Comment 2

[Simon Fraser [no longer active]]

It's not clear to me whether the page should get Command-key combinations first.

Comment 3

[Mark Mentovai]

Cross-product, cross-platform

This is less of a problem than I thought it was initially.  The page gets the
event, but can't override back/forward history traversal unless history is
exhausted in the relevant direction.

Are we willing to go so far as to flat-out refuse to fire command events at
content on the Mac?  It might not affect anything significant, since there's no
usable meta key on Windows.  (At the other end of the spectrum, some developers
have asked to be able to provide their own command-Q handlers in content, to
which I say "nyuh-uh.")

Comment 4

[Wevah (Nate Weaver)]

I'm for blocking commands that shouldn't be overridden, for sure (Cmd-Q, and the
like). FWIW, I catch command-click on some popup stuff to make sure I'm not
throwing up a popup window if the user command-clicked on a link that would
spawn a popup otherwise (popup and new tab with identical content = irritating).

Comment 5

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

Mass-reassign of bugs still assigned to pinkerton to nobody; filter on "NoMoPinkBugsInCamino".

Comment 6

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

This was the premise of bug 459744 and bug 569130, right?

It's unclear to me what was considered to be the problem here (the page saw the keystroke, or the page overrode the keystroke?).  Even now, it appears the page can still see (some of) the keystroke, but not eat it (for supported keystrokes, of which neither real Back/Forward nor Cmd-Arrows fake back/forward are); see smaug's testcase which does eat keys (http://mozilla.pettay.fi/moztests/showkeys.html) or either of the all-consuming testcases in bug 459744.

So I'd dupe this to bug 569130, but someone please correct me if the bug was about something else.