Re: Non-text content loophole

Thanks Patrick, I think what confuses me is that there's a failure for CSS 'background-image' (F3) and CSS ::before ::after (F87) so I don't really see why there couldn't be a similar one for 'inserting non-decorative content by using css background-color'.

I see the argument Kevin that it's the inverse in a sense - you could argue that the input field is the control which is there programmatically and that the visual mechanism is the alternative for that. However, I would argue that the visual mechanism is acting as an image conveying information, and that although that image is functionally redundant to screen reader users, it coveys information to other users and is more than just a decorative image and therefore needs some sort of semantic mark-up so that AT knows it is an image and non-decorative.

It seems to me that when you take the visual dimension into account, how something is 'programmatically determined' also includes CSS (and sometimes it includes the CSS anyway, for example, if display:none; is used). So if something is rendered invisible in the CSS, I would make the case that it is 'programmatically determined' as such. Therefore, it is technically not programmatically determinable in a way that visual technologies can recognise as important content. Likewise for the background-color <div> checkmark. So I would say there is no programmatically determined visual mechanism, but that's just how I look at it.

Anyway, I appreciate your insights! I understand at the moment this is not covered by WCAG. It is frustrating because to someone like my nanna who is very low vision and heavily reliant on inverted colour schemes, this obviously renders some forms entirely unusable, so it seems like letting her down to class the site as WCAG compliant, but there we go!

Thanks

Sarah

Sent from Outlook for iOS<https://aka.ms/o0ukef>