- From: Brian Manthos <brianman@microsoft.com>
- Date: Wed, 27 Oct 2010 17:37:24 +0000
- To: Michael Cooper <cooper@w3.org>, www-style <www-style@w3.org>
- CC: List WAI Liaison <wai-liaison@w3.org>
When there is an inset box-shadow on the element in question, would you want it to be above or below "meaningful images" (when box-shadow is enabled)? I would expect below: [top] ... text content meaningful images border inset box-shadow entries non-meaningful images outset box-shadow entries ... [bottom] -Brian -----Original Message----- From: www-style-request@w3.org [mailto:www-style-request@w3.org] On Behalf Of Michael Cooper Sent: Wednesday, October 27, 2010 9:37 AM To: www-style Cc: List WAI Liaison Subject: [css3-background] PFWG Comments: high contrast modes The following is input from the Protocols and Formats Working Group on CSS 3 Backgrounds and Borders. Approval to send this as a WG comment is archived at http://www.w3.org/2010/10/27-pf-minutes.html#item08. PF has concerns about the use of CSS background-image by authors to add images which are not backgrounds to pages. Authors do this to take advantage of positioning and clipping features for CSS which are not available with the HTML img element. One reason for this is to allow multiple graphics to be combined in a single file to reduce server round-trips, which can substantially reduce the monetary cost of maintaining sites by reducing the number of servers needed and the amount of bandwidth used. Because of the monetary costs involved, we do not believe it is realistic to ask authors to discontinue this practice. Accessibility accommodations for users with low vision often disable background colors and images in order to display text more visibly. These accommodations include High Contrast settings in the Windows and Macintosh operating systems, accessibility settings in several browsers, and assistive technologies. Disabling backgrounds is the correct behavior for these tools, but disabling meaningful images is not. With the current CSS implementation, there is no way for authors to indicate which CSS images should be removed in low vision scenarios and which should not. The PF working group feels strongly that there needs to be a mechanism for authors to indicate this. We have three possible solutions to propose: 1) foreground-image: Foreground images sit on top of background images, but are otherwise identical. They are semantically different, in that they are images that are intended to be the only rendered representation of the element. Foreground images MUST NOT be disabled by assistive technology. 2) introduce a property called 'background-noremove' Name: background-noremove Value: <boolean> [ , <boolean> ]* Initial: none Applies to: all elements Inherited: no Percentages: N/A Media: visual Computed value: as specified, but with URIs made absolute This property indicates whether the corresponding background-image(s) must Not be removed for an element as the image conveys important information to the user. Removal of important information is also a failure in web accessibility compliance per WCAG 2.0 (http://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F3). Images referenced correspond to the order specified by the "background-image" property. See the section "Layering multiple background images" for how ' background-image' interacts with other comma-separated background properties to form each background image layer. An example specifying that a background image must not be removed would be: div { background-image: url(tl.png), url(tr.png); background-noremove: true, false } 3. Include no-remove as part of the value set for background image Properties This is a streamlined solution for authors but it would have a greater impact on style sheet processing: background-image: url(test.png) no-remove, url(tr.png) Any of these solutions would be acceptable. In any case, we would also suggest referencing WCAG 2.0 SC1.1.1, specifically F3 (Failure of success criterion 1.1.1 due to using CSS to include images that convey important information).
Received on Wednesday, 27 October 2010 17:38:09 UTC