Re: [D3E] Gecko will start supporting KeyboardEvent.key from Firefox 23, but only for non-printable keys

On Sun, Aug 11, 2013 at 8:16 PM, Masayuki Nakano <masayuki@d-toybox.com>wrote:

> On 2013/08/11 10:42, Jamie Lokier wrote:
>
>>
>> Before Gecko was changed, this was fine.  Now that event.key ===
>> "MozPrintableKey", the code calculates the wrong value for keyname, so
>> fails to dispatch handlers for the right keys.
>>
>
> Hi, it's unfortunate fact. However, I don't understand why the book
> references .key (and also .char?) now. The .key spec is still discussed.
> Actually, some key names are changed by the latest draft. So, the .key is
> still too unstable for most websites. If web developers need to touch it,
> they need to check all updates on all browsers.
>

It's not unreasonable for the book to recommend "future proof" code like
this.  The assumption is that any browser that provides this attribute will
have a reasonably complete implementation of the current spec (with the
understanding that some details may change).

The current Firefox behavior seems to break this assumption since it is
only a partial implementation that includes non-standard temporary
placeholder values.

The purposes of .key implementation on Gecko are:
>
> 1. Experiments to implement .key on a lot of platforms (IE which was the
> only one browser supporting .key is designed only for Windows).
>

And we *definitely* need this cross-browser/cross-platform experimentation
while we nail down the last bits of the spec.

I'm not very familiar with Firefox's release process, but I thought that
new features were going to "bake" in the Beta channel until they were ready
before being promoted to the Final Release channel.  Is it too late to make
this a Beta-only feature while we work out the spec details?

-Gary