26262 - modifier keys should send KeyDown, KeyUp (windows-only)

Status: RESOLVED WORKSFORME

(Core :: Widget: Win32, defect, P3)

Description

[Akkana Peck]

Split off from bug 6649:
We decided when we met regarding the key event spec (see URL field) that
modifier keys by themselves should send the same event sequence as other keys. 
On Windows, pressing shift, alt or control sends KeyDown and KeyUp, but no
KeyPress.

This isn't getting in anyone's way right now -- very low priority -- but it
doesn't fit the spec we agreed on.

Comment 1

[Frank Tang]

make it assign M20

Comment 2

[Frank Tang]

Move it to M16. Should fix with 5716 together.

Comment 3

[Frank Tang]

m16

Comment 4

[Frank Tang]

move it back to M20 since we don't have time, no nsbeta2, and no bug depend on 
it.

Comment 5

[Frank Tang]

reassign to shanjian

Comment 6

[ckritzer (gone)]

Mass update:  changing qacontact to [e-mail]

Comment 7

[Loco]

Updating QA Contact.

Comment 8

[Roy Yokoyama]

Since pressing shift, alt or control sends KeyDown and KeyUp event on Windows,
IME doesn't work well. (eg. 55751 : Mnemonic doesn't work when IME is on,
Alt-Shift combation doesn't work )

Comment 9

[Hixie (not reading bugmail)]

Reassigning QA Contact for all open and unverified bugs previously under Lorca's
care to Gerardo as per phone conversation this morning.

Comment 10

[Frank Tang]

key related issue, reassign back to roy.
roy- do we need this for moz0.9.2 ?

Comment 11

[Kathleen :Brade]

I think because Mac doesn't do this yet and I'm not sure if Linux does this yet, 
that we don't have to have this yet.  I think we should wait to see a new version 
of the dom keyevent spec that joki was working on earlier this year.

Comment 12

[Frank Tang]

base on brade's message. move it to future.

Comment 13

[Madhur Bhatia]

QA contact updated

Comment 14

[Roy Yokoyama]

Accepting.

Comment 15

[William J. Edney]

Attachment: Testcase that shows that Alt fires a keydown/keyup, but no other modifiers do

Since its been a year since anyone kicked this bug, I thought I would do so ;-)
as I've just run into a situation where I need the modifier keys to signal at
least keydown/keyup (I didn't test keypress).

I'm running on Linux (SuSE 7.3, KDE 2.2.1), and the Alt modifier key DOES fire
these events, but neither the Shift or Ctrl keys do. Maybe this is due to some
X windows config keymap thingy that I'm not sure of. The fact that this bug is
still outstanding makes me think that this is not the case.

This testcase sets up two capturing handlers for keydown and keyup on the
window's document and focuses the window in the onload.

It would be great if we could get some progress on this bug.

Thanks!

- Bill

Comment 16

[Kathleen :Brade]

note: this bug is Windows-specific; I believe there is a bug for Macintosh OS's
and another bug for Linux (and perhaps other bugs for yet other platforms)

Comment 17

[Ilya Konstantinov]

Please remove "KeyPress" from the summary. 
  
KeyPress events are not generated for Alt, Shift and Ctrl by design. This is a  
non-standard event (not covered by the DOM Events model) and we're free to  
handle it as we please. In this case, IE compatibility becomes our main goal.  
IE generates KeyPress only for:  
  
Letters: A - Z (uppercase and lowercase)   
Numerals: 0 - 9   
Symbols: ! @ # $ % ^ & * ( ) _ - + = < [ ] { } , . / ? \ | ' ` " ~   
System: ESC, SPACEBAR, ENTER  
  
See:  
http://msdn.microsoft.com/workshop/author/dhtml/reference/events/onkeypress.asp  
  
The intention is to make KeyPress behave more like DOM3's TextEvent..  
  
On the other hand, KeyUp and KeyDown MUST be generated for Alt/Shift/Ctrl.  

Comment 18

[Johnny Stenback (:jst)]

Summary updated.

Comment 19

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)]

Nowadays, keydown event and keyup event are dispatched by all modifier keys. And keypress event isn't dispatched but it's correct behavior on CSS3.
http://dev.w3.org/2006/webapi/DOM-Level-3-Events/html/DOM3-Events.html#events-keyboard-event-order

We can mark this as WFM now.