Re: Keyboard events for accessible RIAs and Games

I don't think figuring out how close X is to Y is relevant, at all. And I
did not imply or ask for it.

What is relevant is the following:

#1: That there is a key-code that is reliable, unmodified and can be used
by an application to react to shortcuts. This code would be saved in
shortcut presets and be used to store modified shortcut schemes.
Fortunately Event Level 4 introduces just such a code attribute to the
event, so we're fine there.

#2: A shortcut mapping dialog will need to display to the user a listing of
shortcuts that are configured. Examples of this are found in every shortcut
key dialog in existence such as:
- Blender:
http://codeflow.org/entries/2013/jan/30/keyboard-events-in-javascript-are-broken/blender.png
- Gimp:
http://codeflow.org/entries/2013/jan/30/keyboard-events-in-javascript-are-broken/gimp.png
- Minecraft:
http://codeflow.org/entries/2013/jan/30/keyboard-events-in-javascript-are-broken/minecraft.png

If we where to use the Event Level-4 code attribute for this a shortcut
would be displayed as "ControlLeft+KeyZ" regardless of weather the KeyZ is
actually the key Y (german), the key "<" (Turkish) or no key to be found on
that keyboard at all (Arabic). Besides the obvious usability issue, it's
also not as user friendly as "ctrl+z".

To solve that a suggestion has been made to add a function to translate a
key code (as defined by DOM Event Level-4) to a users locale unmodified
primary keyboard symbol: window.KeyboardEvent.queryKeyCap(event.code)

I think that is a very good suggestions (it's better than my original
suggestion to add it to the event) because it also aides with the display
of presets/configurations of shortcuts regardless of locale.




On Thu, Jan 31, 2013 at 12:26 PM, Bjoern Hoehrmann <derhoermi@gmx.net>wrote:

> * Florian Bösch wrote:
> >I have written a blog post at length about this issue here:
> >
> http://codeflow.org/entries/2013/jan/30/keyboard-events-in-javascript-are-broken/
>
> http://lists.w3.org/Archives/Public/public-webapps/2013JanMar/0170.html
> You seem to have double-posted this to the public-webapps list, where
> most of the discussion seems to be archived now. I note that solutions
> to some of the problems you mention have to account for the fact that if
> non-isolated scripts can know, say, how close "x" and "y" are to each
> other on a keyboard, that information would contribute identifying in-
> formation that is likely to be used to track users. Last I heard people
> wanted to avoid adding such information as much as possible.
> --
> Björn Höhrmann · mailto:bjoern@hoehrmann.de · http://bjoern.hoehrmann.de
> Am Badedeich 7 · Telefon: +49(0)160/4415681 · http://www.bjoernsworld.de
> 25899 Dagebüll · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/
>