WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Accessibility of responsive, repositioned sidebars?


From: Robert Fentress
Date: Oct 31, 2016 11:26AM


I certainly *can* change the visual order using flexbox or CSS, but my
question is if I *should*, when that means the tab order no longer matches
the visual order.

For instance, assume the following for a desktop view: There is a row
containing a full-width banner at the top of the page followed by a row
containing a horizontal main navigation bar that stretches across the full
width of the page, followed by a row with two columns, where the left
column is the main page content, including some links embedded in a
paragraph, and the right column is a vertical secondary nav.

If I tab past the last menu item in the main horizontal navigation at the
top of the page, I, personally, would expect the next tab stop to be the
first link in the main content area (left column), not the first link in
the secondary navigation menu (right column). This would be the case
whether that secondary navigation was made to appear there using
float:right with CSS or by using the order property in the flexbox model.
If this is, indeed, problematic, is it better, then, to use Javascript to
rearrange the DOM, so that the tab order always matches the visible order?
Though this seems less elegant, if one used matchMedia, it would be less

Jonathan, I'm interpreting what you've said as falling on the "It's
probably not that big a deal either way" side of things in the scenario
described above. That sound right?

Adding a further wrinkle, sometimes there are layouts where a secondary
navigation menu and other supplementary content both appear together in a
right-side column on large viewports, but, on small viewports, the
secondary menu breaks above the main content, while the other supplementary
content breaks below. Ugh!

On Wed, Oct 26, 2016 at 11:54 AM, Asa Baylus < <EMAIL REMOVED> > wrote:

> Hi Robert,
> I just went through this same issue. It's trivial to solve if you're using
> flexbox. With flexbox you can change the order without JS
> Browser support is pretty decent these days
> Just a thought and your browser support matrix may not allow for it
> Asa
> Sent from my iPhone
> > On Oct 26, 2016, at 8:45 AM, Robert Fentress < <EMAIL REMOVED> > wrote:
> >
> > Bummer. Nobody has an opinion, or was I just not clear in my explanation
> > of the issue?
> >
> >> On Tue, Oct 18, 2016 at 10:21 AM, Robert Fentress < <EMAIL REMOVED> > wrote:
> >>
> >> Hello, all.
> >>
> >> Anybody got opinions on the accessibility of responsive designs where a
> >> secondary menu appears as a column to the right of the main content on
> >> desktop breakpoints, but, on mobile, the menu appears above the main
> >> content, perhaps in a disclosure widget of some sort? In both cases,
> the
> >> DOM order has the menu before the main content, with the visual changes
> >> being accomplished using CSS.
> >>
> >> From a design perspective, the fact that the main content dominates
> >> visually on desktop is probably good and making the menu easier to get
> to
> >> by preceding the content on mobile makes sense (though it gets
> complicated
> >> when you consider ease of access and hand position). My question,
> though
> >> is how bad it is that the DOM order now does not match the tab order for
> >> keyboarders? Certainly, this is not an uncommon pattern on the web
> these
> >> days. Perhaps, it would not terribly violate user expectations.
> >>
> >> Beyond the tab order, simply having the visual order not match the DOM
> >> order appears to be a no-no for various reasons, based on "C27: Making
> >> the DOM order match the visual order"
> >> <https://www.w3.org/TR/WCAG20-TECHS/C27.html>. However, this is but
> one
> >> sufficient condition for meeting the related success criterion, 1.3.2
> >> Meaningful Sequence
> >> <https://www.w3.org/WAI/WCAG20/quickref/#qr-content-
> structure-separation-sequence>.
> >> C27 says:
> >>
> >> "If this is a sufficient technique for a success criterion, failing this
> >> test procedure does not necessarily mean that the success criterion has
> not
> >> been satisfied in some other way, only that this technique has not been
> >> successfully implemented and can not be used to claim conformance."
> >>
> >>
> >> So, it seems ambiguous to me.
> >>
> >> Would it be better to actually move the content with JavaScript at
> smaller
> >> breakpoints so the DOM order matches the visual order? That strikes me
> as
> >> inelegant (though less so as matchMedia
> >> <http://caniuse.com/#feat=matchmedia>; gains ground) and more likely to
> >> break. It may also confuse users who move between devices.
> >>
> >> What do you folks think about the relative advantages and disadvantages
> of
> >> various approaches?
> >>
> >> Thanks.
> >>
> >> Best,
> >> Rob
> >>
> > > > > > > > >
> > > > >