W3C home > Mailing lists > Public > public-webapi@w3.org > May 2008

Re: [DOM3Events] keyCode and charCode data requested

From: Olli Pettay <Olli.Pettay@helsinki.fi>
Date: Wed, 07 May 2008 17:25:15 +0300
Message-ID: <4821BBCB.1060003@helsinki.fi>
To: Andrew Emmons <andrew.emmons@quickoffice.com>
CC: "public-webapi@w3.org" <public-webapi@w3.org>


testing mainly nightly-Gecko/FF3 on WinXP:

"BackSpace", "Tab", "Pause", "PageUp", "End",
"PageDown", "Home", "Insert" "Delete" and F1-F12 dispatch also keypress

With )-", a-zA-Z0-9 and "Space" Gecko uses .charCode with keypress to
indicate the character code (keyCode is then 0)

Gecko has the same keyCodes for '+' and '-' whether the key is used in
numpad or in the main part of the keyboard. Seems like a bug in
In some cases also IE gives the same keyCode for keys in numpad and
keys elsewhere (like with PgDown/Up, Home, End).

On Linux keycodes with non a-zA-Z0-9 keys are quite different, but
my keyboard settings are different than on Windows.
Opera and FF show similar keyCodes on Linux.
.charCodes are the same as on WinXP.

On linux I did some testing also with Konqueror 3.5 (a bit old one, I
know) and there I managed to get keyCodes like 65444 and -92 (that was
with ) and keycode in keyup/down is often the same value as with
keypress. Somehow it is using character code as key code, at least
in some cases.

Ctrl+Shift+(character key) is problematic.
Safari and Opera UI interpret that as a command key, even if
the functionality is bound to Ctrl+(character key)
IE and FF don't have this problem.

If there isn't functionality bound to Ctrl+Shift+(character
key), Opera and FF dispatch similar events as with Shift+(character
key), but event.ctrlKey is set to true. No text is added to <input>.
IE and Safari behave differently. They have different
.keyCode on keypress event than without Ctrl and in some cases
keypress isn't dispatched at all. <input> doesn't get changed.


Andrew Emmons wrote:
> Hello WG,
> Regarding keyCode and charCode, we would like to request data for these
> commonly used event attributes across various implementations and
> platforms. The results should look like the following:
> http://lists.w3.org/Archives/Public/www-archive/2006Nov/att-0047/keyCode-ie.htm,
> except that, where it is supported by the implementation, charCode is
> also tested.
> Please e-mail your results to this list, indicating the implementation
> and platform tested.
> You can use the following tools:
> http://www.w3.org/2002/09/tests/keys-cancel2.html, or
> http://people.w3.org/rishida/utils/keyevents/.
> Regards,
> Andrew
> /Andrew Emmons/
> Architect and Manager, SVG Technologies
> Co-Chair, W3C SVG Working Group
> *Quickoffice, BitFlash Division*
> www.bitflash.com <http://www.bitflash.com/>,
> www.quickoffice.com <http://www.quickoffice.com>
Received on Wednesday, 7 May 2008 14:25:49 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:16:26 UTC