- From: Charles Pritchard <chuck@jumis.com>
- Date: Tue, 05 Oct 2010 13:08:29 -0700
- To: David Singer <singer@apple.com>
- CC: public-canvas-api@w3.org
On 10/4/2010 4:16 PM, David Singer wrote: > I wonder if there are some conclusion jumps being made here, or at least extreme positions (these are both upcoming olympic sports, by the way). ... > Can we make it possible to write interesting interactive accessible applications in canvas, without doing harm to the ourselves or the specification? If so, we have broadened the scope for invention. For what it's worth, we started with a need in our web app: to stylize our "select" elements. As everyone is aware, "select" does not have much flexibility in presentation. It's widely known that "li" or "div" tags may be used instead of "select", if more control is needed. We thought it made more sense to use canvas. I'd imagine that Ian accepts this usage, as he has given the nod to implementing checkboxes and radio boxes in canvas. Most of the input types of html 5 are not implemented by current browsers: http://dev.w3.org/html5/markup/input.html All of them can be implemented, per the standard, within canvas and/or a combination of DOM/JS. As we were also developing a drawing application, we did need a string editor. We wanted the user to be able to interact directly with the text they were entering; though we still use an input form type apart from the display. An editable <select> element (<input type="text" list>), with possible suggestions, is a desirable element to have in a web app. To make that list of <select> options editable, we needed basic text editing. We already wanted that, for our drawing application, when the user places text. Those are the reasons why we started working with text editing in canvas. It was necessary for our application to function at the level that the market expects. Now, Ian may say: Why would anyone create a drawing program in Canvas... why would anyone create an image manipulation program in canvas... Why, dear god, why! Shouldn't they use C++ for that, and maybe some cross platform tool kit. Shouldn't they wait until 2012? It's just not feasible, Ian might say. So anyway.. I can't dispute his cries for sanity. But I can say, that having spent over three years on such applications, it is feasible to create them, and they do run satisfactorily, in the current browser generation. -Charles
Received on Tuesday, 5 October 2010 20:35:12 UTC