- From: Jonathan Avila <jon.avila@ssbbartgroup.com>
- Date: Tue, 15 Jun 2010 09:11:10 -0400
- To: ng@tjkdesign.com, w3c-wai-ig@w3.org
Thierry Wrote: "A better approach would be to ... avoid losing these images when users switch to high contrast styles sheets." Actually, converting images of text to text would be good for many low vision users. The WCAG 2 Level AA success criteria indicates that images of text should not be used unless essential or the technology does not support. Your point is well taken for images that are not text. However, I do have several thoughts. Since only one image is used, automated testing tools will need to pull the viewport coordinates from the code to determine which area of the larger image is applicable to the current element. For instance, an automated testing tool displays the whole image when needed. In the example from http://tjkdesign.com/articles/how-to_use_sprites_with_my_Image_Replacement _technique.asp text is used since this is a heading where text fits. However, many sites that use this technique will likely use icons where there is not sufficient room to place on-screen text without messing up the layout. In those cases alt text would need to be applied to the image and I'm sure some developers won't do this. When an automated tool detects the missing alt text it will display the entire image for the user to review to determine what the appropriate alt text is. The example also uses a lot of fixed pixel coordinates. I am unclear if relative sizing can be used effectively with this solution. Jonathan
Received on Tuesday, 15 June 2010 13:11:45 UTC