E-mail List Archives

Re: Block Links & Other Validation Issues in XHTML 1.0 Transitional


From: Patrick H. Lauke
Date: Feb 25, 2016 5:17AM

On 25/02/2016 11:55, Dennis Deacon wrote:
> We use block links throughout our website for advertisements, offers, etc.
> I've recently learned that block links are officially only supported
> starting with HTML5. So while the web browser handles them fine
> (seemingly), and only irritate screen reader users (unfortunately), this
> technically fails validation. How does this relate to the "R" in WCAG's
> POUR principles (Robust)?
> I see mention that under parsing, web pages should validate, with the HTML
> conforming to specifications. Block links under a XHTML doctype would fail
> this. Does this mean these pages fail compliance (especially if mandated)?
> Additionally, if a page has validation errors, would this be an
> accessibility compliance concern?

Note that validation is not mandated, but simply a sufficient technique
for 4.1.1 - meaning that even if a markup validator throws up errors,
that does not necessarily mean that it's a FAIL of 4.1.1.

The main concern of 4.1.1 is to ensure that user agents consistently
interpret the markup and generate the same DOM structure from it. As
most modern browsers interpret broken markup the same(ish) way (as the
markup parsing algorithm has been more thoroughly defined in HTML5 and
implemented in browsers), the chance of broken markup messing up in
specific user agents (and then leading to accessibility problems) is
lower nowadays. Nonetheless, the main things that should still be
avoided (even though these should still actually result in the same DOM)
are mainly making sure you have correct opening/closing tags, you don't
misnest things (like <p><strong>...</p></strong> or similar), don't have
duplicate ID attributes, don't have duplicate attributes with
conflicting values.

But, picking up from Steve's comment: why are you using and XHTML
doctype? Simply replacing that with the minimal HTML5 doctype of
<!DOCTYPE html> should not have any effect on how your content is
actually interpreted/displayed in browsers, and will have the added
benefit of automatically running validation against the more modern spec
(meaning fewer errors will be flagged up).

Patrick H. Lauke

www.splintered.co.uk | https://github.com/patrickhlauke
http://flickr.com/photos/redux/ | http://redux.deviantart.com
twitter: @patrick_h_lauke | skype: patrick_h_lauke