WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Z-Order and Tag Order Need to Match?


From: Duff Johnson
Date: May 27, 2020 8:26PM

> Quote: " One of the assistive technologies that relies on the Content order rather than the tags is the PDF reader in Read&Write from TextHelp, which is a popular tool used in education settings." /EndQuote
> Exactly. It's used from primary ed to post-graduate.
> When I've had the chance to talk with these technology manufacturers (not specifically Read and Write, but others), they've said they really don't have the R&D budget to retool their software/hardware to meet the PDF/UA-1 standard. They cater to a small, select market and massive R&D changes would price their product out of that market. Both the companies and the end users would suffer.

As stated; it's a business decision. Tagged PDF is not news; it's been around for 20 years.

And more importantly, if a viewer doesn't use tagged PDF then it cannot convey (for example) "info and relationships" between elements of content in the document, but can only guess at the document's structure. No-one would ever consider this as acceptable in a web browser; why it is acceptable in a PDF viewer?

> A colleague wondered if we have built a system that, although intended to include people, ends up excluding those who can't afford the better, and more expensive technologies that do the job right.

PDF is an ISO standardized open technology; Tagged PDF has been around for 20 years. Apple's productivity software makes tagged PDF by default - you can't turn it off. Microsoft's Office 365 makes tagged PDF as as a matter of routine, and has done so for years. Open Office has made tagged PDF for over 10 years and that software is FREE! When do the excuses run out?

> A form of economic discrimination, in a way. Definitely not what was intended.

That's what some people said about WCAG 2.0; "it's so much work to support it". Yep, but if you want to call your website "accessible" there are some things you have to be able to do. Representing the semantics of content is one of them. In PDF, semantics are encoded via tags. No tags, no semantics…. no universal accessibility