From Automated WCAG Monitoring Community Group
- What does the following mean exactly? : Retrieve the text from its child nodes of type text or containing text before or after the selected element
Ready for review
- What about image sprites? - should be covered as all CSS-styles are applied to the presented item. also the background-position
- Step 1: What about "backgroundRepeat: inherit"? In the current logic this passes the test. - now checking for the computed backgroundRepeat property
- Step 3: I think we need to be a bit more specific about what is presented to the user. For example, if the element with the css background image also contains an <img> element. The images must be presented in such a way that a user (without html knowledge) can distinguish foreground and background images. - steps 3 and 4 updated by preparation of the element
- Can step 3 and 4 be merged? - Done
- I don't quite understand the following sentence "Retrieve the text from its child nodes of type text or containing text before or after the selected element and assign it to variable T1" - Clarified in last Call?
- Step 4 has no description. - Added
- Step 5 user interaction should be 'no' instead of 'yes'. -Not changed, explanation below
- The comment on step 5 raises the question, if in this SC we should check for functionality of the elements ("Otherwise, or if the element provides any functionality..."). In AngularJS for example, it is possible to add events to any element, such as div or span.
- Step 5: T1 should be presented to the user. - Done
- "Element containing background images positioned and in same order as defined by CSS" I'm not sure I understand what's going on here -Changed
- Description: Added check for description provided in text nearby.
- Step 4: removed "eyecatching" from help-text
- Step 4: changed to context-sensitive
- Step 5: Introduced variable T1 used in user input question
- Step 4 and 5: aligned item presented to the user
- Helptext sums up use cases we should be able to automatically determine, namely:
- And can css background images be used as spacers? - - changed "spacers" in step 4 to "separator"
- Fix numbering of result identifiers: pass1, pass2, ...
- I don't think we should be referring to the DOM. If this is somehow important (and I think it is) we should present the text available in the nearby DOM structure to the user instead.
- note that since css3 multiple are possible: http://www.css3.info/preview/multiple-backgrounds/
- I think it's worth including a step 5 to test if the text alternative is somewhere adjecent to the image.
- additional test: test if the image is covered with text above the image (a text box covering the area from left to right) -> we can assume the image is decorative
- Step 1 checks the background-repeat property. Do we also need to check the background property?
- test if a image is used to convey information by color or shape, can be dealt with SC1-4-1 and SC1-3-3
- should we take care about iconfonts?
- Selector: Include "list-style" shorthand for completeness.
- Step 2: I think we need an explanation for "surrounding text".
- The :before and :after pseudo-selectors are a big subject all on their own, and one that is relevant to other criteria as well. It's part of a bigger question of what to do with shadow DOM
- I would prefer if we could capture this test in a single question, has this been tried yet?
- Step 2: What is 'surrounding text'? Is a user supposed to locate this text on the page? What about hidden text? Should we instead pull the surrounding text from the page and ask the user to compare that, so that we can include any AT-only text?
- it's now manual testing only, I would like to see some automatic tests at the start. Two suggestions:
- we can assume that as soon as background-repeat is used, it's decorative
- we can assume that as soon the image is less than 10 pixels width or height, it's decorative