Re: revert request for change made to the HTML5 spec 23/05/2011

On Jun 7, 2011, at 10:42 PM, Tab Atkins Jr. wrote:

> On Sun, Jun 5, 2011 at 1:26 AM, Steve Faulkner <faulkner.steve@gmail.com> wrote:
>> dear chairs,
>> 
>> I request that the following change be reverted:
>> http://html5.org/tools/web-apps-tracker?from=6144&to=6145
>> 
>> It substantially changes the content model of the canvas element, it is
>> claimed that the change is to
>> "Further discourage misuse of <canvas> that cannot be made accessible or
>> usable."
>> 
>> The change does no such thing, it does not stop any author creating
>> inaccessible canvas content, it only discourages authors from attempting to
>> provide accessible alternatives within the canvas sub-dom to inaccessible
>> user interface elements they may create using canvas.
> 
> The restricted elements are nearly impossible to implement in an
> accessible way within a canvas; there's far more to being accessible
> than providing alternative representations.  Assuming an author
> restricts themselves to creating canvas content that can be given an
> alternative representation, this change makes it more likely that the
> canvas content itself will be accessible as well.
> 
> If an author doesn't restrict themselves in this manner, they will
> have both an inaccessible canvas (for people who, for example, need an
> IME to enter text) and either an inaccurate or invalid subtree.
> 
> Sam, if two people can request a revert without filing bugs, what's
> the process for reverting the revert?

File a bug. If the change is made pursuant to a Last Call bug, the Chairs will be more inclined to let any further disputes be settled via the issue process rather than the revert process. The revert process was applied here in part because there was no bugzilla bug to escalate.

Regards,
Maciej

Received on Wednesday, 8 June 2011 06:00:15 UTC