Re: html5 editor responds to Canvas accessibility related bugs

On 3 Oct 2010, at 19:04, Charles Pritchard wrote:
> SVG support is not consistently implemented; IE9 will not have "fixed" the problem.

Performance issues aside, can you give any examples of SVG inconsistencies you expect to see between IE9 and (say) HTML5-parsing Gecko that would make it difficult to implement an app like Bespin in SVG?

> What kind of experimental implementations do you have in mind?

I meant experimental implementations of the required features or performance improvements to either replace or accessify "canvas" UIs.

> Most of the accessibility proposals need not be directed at canvas. Providing details about the position of the mouse cursor and focus area can still be useful for items outside of canvas, as a worthwhile extension to the concept of tabbed indexes.

I think that's a good point: if we can design APIs that can be reused beyond "canvas", that would be a win over designing APIs that only be used within "canvas".

> I don't understand why there is a stand being taken against the current use of canvas.

I think it's because a key plank of the "web standards" movement of the past decade involved protecting users' ability to consume content as suits their needs and desires (for example, offering a seamless experience that finds them the content they're looking for, presents it to them in their chosen media and theme, and allows them to interact with exposed services using widgets that behave consistently with the rest of the platform), and building UIs in a bitmap makes all that way harder, perhaps ultimately impossible.

> reasonable evidence to acquit us canvas enthusiasts from wrong-doing.

Isn't this about strategy more than ethics? Given generally agreed requirements like enabling accessibility and internationalization and given web app developers want to make things like Bespin, how can user agent developers best serve end-users with their limited resources?

--
Benjamin Hawkes-Lewis

[resent via an account the W3C lists system will recognize; apologies to anyone who gets this message twice]

Received on Sunday, 3 October 2010 21:10:02 UTC