W3C

Site Navigation

Nearby

Results of Questionnaire ISSUE-74: canvas-accessibility and ISSUE-105 canvas-usemap - Straw Poll for Objections

The results of this questionnaire are available to anybody.

This questionnaire was open from 2010-10-13 to 2010-10-21.

8 answers have been received.

Jump to results for question:

  1. Objections to adding a boolean nonav attribute to the canvas to hide its contents from AT and keyboard navigation
  2. Objections to removing the ability to use the children of canvas as an accessible dom entirely, in favor of other techniques, such as usemap?
  3. Objections to adding advisory text explaining how to use the focus rect API for efficient caret position notification, and change nothing else?
  4. Objections to adding image map / usemap support as a possible way to represent focusable regions on the canvas?

1. Objections to adding a boolean nonav attribute to the canvas to hide its contents from AT and keyboard navigation

This option is:

If you have strong objections to adopting this option, please state your objections below.

Keep in mind, you must actually state an objection, not merely cite someone else. If you feel that your objection has already been adequately addressed by someone else, then it is not necessary to repeat it.

Details

Responder Objections to adding a boolean nonav attribute to the canvas to hide its contents from AT and keyboard navigation
Richard Schwerdtfeger I have no issue with removing -nonav as long as the text regarding the canvas subtree states that there is a 1 for 1 navigational mapping to UI elements drawn on the physical canvas as it states today. In short whatever is in the subtree must be a a representation of what you see on the physical canvas at all times. This is much different from having a fallback that is an alternative for the canvas used as say a long description. This would not constitute a one for one mapping.
Ian Hickson This is already supported by the simple means of removing the contents of a <canvas> element using script. This handles all the use cases that might apply. (Bear in mind that we know script is enabled if the fallback is unnecessary, since otherwise the <canvas> wouldn't work in the first place.)
Cynthia Shelly I see nonav as a "nice to have" feature. I have no objection to removing or keeping it.
Steve Faulkner
Martin Kliehm
Eric Eggert
Frank Olivier the attribute seems superfluous; canvas authors can correctly handle fallback and a11y considerations with the method that Ian describes.
Eric Carlson

2. Objections to removing the ability to use the children of canvas as an accessible dom entirely, in favor of other techniques, such as usemap?

This option is:

If you have strong objections to adopting this option, please state your objections below.

Keep in mind, you must actually state an objection, not merely cite someone else. If you feel that your objection has already been adequately addressed by someone else, then it is not necessary to repeat it.

Details

Responder Objections
Richard Schwerdtfeger This proposal is currently irrelevant. The canvas subtree as the spec. currently reads allows for an accessible DOM subtree by default. No special -adom attribute is required -now.

Here is the text from the HTML 5 spec regarding the <canvas> element:

"When a canvas element represents embedded content, the user can still focus descendants of the canvas element (in the fallback content). This allows authors to make an interactive canvas keyboard-focusable: authors should have a one-to-one mapping of interactive regions to focusable elements in the fallback content."

Ian Hickson Image maps have not been demonstrated to be usable to create accessible image maps.
Cynthia Shelly The accessible DOM sub-tree is implemented in IE9, and is currently in the HTML 5 spec. usemap is "nice to have" in addtion to the accessible DOM sub-tree, but does not cover all of the use cases, and is not a replacement. This issue seems to be overcome by events.
Steve Faulkner I object to removing the ability to use the children of canvas as an accessible DOM, this feature is an important aspect of making canvas conetnt accessible. This should be provided in comjunction with other techniques.
Martin Kliehm
Eric Eggert
Frank Olivier This is the best current solution to solving canvas a11y; it is a familiar solution for authors - build an interface using HTML elements.
Eric Carlson

3. Objections to adding advisory text explaining how to use the focus rect API for efficient caret position notification, and change nothing else?

This option is:

If you have strong objections to adopting this option, please state your objections below.

Keep in mind, you must actually state an objection, not merely cite someone else. If you feel that your objection has already been adequately addressed by someone else, then it is not necessary to repeat it.

Details

Responder Objections to adding advisory text explaining how to use the focus rect API for efficient caret position notification, and change nothing else?
Richard Schwerdtfeger This is inadequate. I completely object to this proposal. What is in the specification is inadequate if people do something like rich text editing.

The poll does not include our proposal for adding an API to support caret and selection. This proposal should have been included in the survey.
Ian Hickson
Cynthia Shelly This approach does not solve the problem for more complex applications. While the editor may feel that it is a mis-use of canvas to create a text editor in canvas, it is technically feasible, and and seems a likely scenario given the desire of web applicaiton developers to have greater control over the presenation of their applciations. Unless there is a way to prevent this use of canvas (and I don't think there is) then we need to provide a way for it be made accessible to a variety of assitive technologies.

I think there may be an error in the survey, as question #1 was supposed to be an item relating to "Change proposal to provide accessiblity info for driving screen magnifier tracking..." which I support. It appears instead that items 2&3 are duplicates.
Steve Faulkner I object to ONLY adding advisory information without adding features, the focus ring on its own is an inadequate tool for providing accessibility information to ATs. More information is required to be exposed to ATs in order for canvas accessibility features to be flexible enough for current and future usage scenarios. If not provided in the browser, developers who wish to provide accessible canvas content will be forced to hack around the lack of accessibility support, as experience with developer based accessibility hacks tells us, this is not a desirable or generally psotive outcome and will do a disservice to users with disabilities.
Martin Kliehm
Eric Eggert
Frank Olivier I've been looking into what one would need to do to have a full-featured text input / canvas text drawing solution - and there is a very significant amount of work that a UA/author would have to do to make this work without any loss in functionality (compared to the native text entry elements).
Eric Carlson Implementing a text editor in canvas is clearly ill advised, and the spec *should* contain cautionary text. The text proposed by Ian will make an excellent addition to the spec, although it should not be the only change for canvas accessibility.

4. Objections to adding image map / usemap support as a possible way to represent focusable regions on the canvas?

This option is:

If you have strong objections to adopting this option, please state your objections below.

Keep in mind, you must actually state an objection, not merely cite someone else. If you feel that your objection has already been adequately addressed by someone else, then it is not necessary to repeat it.

Details

Responder Objections to adding image map / usemap support as a possible way to represent focusable regions on the canvas?
Richard Schwerdtfeger I would support providing this as a way to make canvas accessible but it should not be the only vehicle. I would categorize this as a nice to have feature. Also, someone needs to drive this forward as it would require, in many cases, having a full DOM subtree with support for pop-up menus, etc.

This approach has some benefits as well as limitations over the DOM subtree. A a nice to have is allowing the map to automatically draw the focus ring without a specific drawing call to achieve it. This would not work for things like selected items in menus or listboxes. So, this would also require a caret/selection API to drive magnification.
Ian Hickson Image maps have not been demonstrated to be usable to create accessible image maps. We should not add features that are hard to use, especially when they are only useful for limited (simple) cases.
Cynthia Shelly This is a "nice to have" feature, in addition to an accessible DOM sub-tree. I have no objection to its inclusion, but think it is a lower priority work item than an accessible DOM sub-tree with magnification and caret support.
Steve Faulkner There has been no technical arguments against the addition of an image map feature to canvas. There has been a willingness shown by implementors to add the feature. There has been examples (http://www.paciellogroup.com/blog/misc/canvas-pie/pie.html) of how it would work provided (via hacking a transparent image map over the top of a canvas). There has been no indication that the addition of this feature will cause an undue development burden on implementors. There has been no change proposal objecting to this addition. While not a complete solution it could prove to be a worthwhile method in cases where complex interaction is not required, but clickable/focusable canvas areas are. The major adavantages of this method are that it is familiar to authors, many html editors have image map editors built in, the creation of a clickable area using image map also provides a programmatically determinable focus ring and it responds to keyboard input by default.
Martin Kliehm
Eric Eggert
Frank Olivier no objection, but it is a lower priority work item
Eric Carlson

More details on responses

  • Richard Schwerdtfeger: last responded on 13, October 2010 at 14:48 (UTC)
  • Ian Hickson: last responded on 13, October 2010 at 18:57 (UTC)
  • Cynthia Shelly: last responded on 13, October 2010 at 19:41 (UTC)
  • Steve Faulkner: last responded on 19, October 2010 at 09:00 (UTC)
  • Martin Kliehm: last responded on 21, October 2010 at 18:10 (UTC)
  • Eric Eggert: last responded on 21, October 2010 at 21:15 (UTC)
  • Frank Olivier: last responded on 21, October 2010 at 23:43 (UTC)
  • Eric Carlson: last responded on 22, October 2010 at 01:40 (UTC)

Everybody has responded to this questionnaire.

Compact view of the results / list of email addresses of the responders

WBS home / Questionnaires / WG questionnaires / Answer this questionnaire

Report issues on GitHub project w3c/wbs-design or by mail to sysreq.