W3C home > Mailing lists > Public > public-html@w3.org > March 2014

Re: 4.13.1 Bread crumb navigation - use of right angle brackets

From: Bruno Racineux <bruno@hexanet.net>
Date: Sat, 01 Mar 2014 17:24:55 -0800
To: "Jens O. Meiert" <jens@meiert.com>
CC: Steve Faulkner <faulkner.steve@gmail.com>, HTMLWG WG <public-html@w3.org>
Message-ID: <CF362929.82686%bruno@hexanet.net>
On 2/28/14 8:49 AM, "Jens O. Meiert" <jens@meiert.com> wrote:


>> > http://webaim.org/projects/screenreadersurvey5/
>>
>> Which of the Breadcrumb examples on this page (link opens a new
>> windows/tab) do you prefer?
>> Response# of Respondents% of Respondents
>> #1 - Unordered List 156 11.9%
>> #2 - Ordered List 364 27.8%
>> #3 - Nested List 96 7.3%
>> #4 - Text 268 20.4%
>> I Don't Know 427 32.6%
>
>I question the value of these results because they¹re purely
>preference-based and non-qualitative. Unless I miss something then
>there¹s no reason provided as to why‹there¹s no indication that an AT
>user ran into issues with one or more of these options, or that
>someone had an objection that is new to our discussion here.
>
>I repeat the position that I expressed earlier [1] and recommend to
>keep it simple with a ³text² solution, but am otherwise slowly moving
>away from this debate. :)

It doesn't matter what W3C recommends if Google or Bing suggest different
patterns. Right now Google uses the right angle bracket in their doc.
While Bing suggest either right angle bracket or double right angle
bracket.

Why don't you guys at W3C consult with search engine folks. Getting on the
same page would be very nice, instead of having 3 different syntax
recommendations.

Not to appear dismissive, but ultimately we have to go with what search
engines are actually parsing properly and telling us to use regardless of
what W3C recommends, if we want to be SEO friendly.
http://microformats.org/wiki/breadcrumbs-formats

> >> I found no indication of accessibility users complaining about
> breadcrumbs.
> > Do you have any indications of such problems?

In regards to the survey:

Perhaps one of the follow up question on that survey should have been or
should be: Do you feel any of these solutions are appropriate for
breadcrumbs?

The majority of 'I Don't Know' may potentially suggest that neither is
adequate, which is what usually blog posts on the subject end up
conveying. As a dev I find neither very satisfying.

 On Sun, Jan 27, 2013 at 7:48 PM, Steve Faulkner wrote:

> Upon looking into the issue further what I did find was that providing
>  a label providing context is accessible best practice [1]

What's lacking to me is the notion of tree level as opposed to ordered
numbers. The breadcrumb often feels like a 'tree' to me as opposed to a
navigation. Unless we are talking of that kind of breadcrumbs:
http://codecanyon.net/item/xbreadcrumbs-dynamic-expandable-navigation/54592
38
Not sure how AT users may feel about such an expandable breadcrumb ...

> Suggest the current advice/examples be augmented to include a text cue
> such as "You are here" at the start of the trail.
>
> <nav>
> <p> You are here:
> <a href="/">Main</a> >
> <a href="/products/">Products</a> >
> <a href="/products/dishwashers/">Dishwashers</a> >
> <a>Second hand</a>
> </p>
> </nav>

I am thinking that an ARIA 'current-page' state might be more helpful than
a "you are here" in complementing breadcrumbs as well as other types of
pagination or navigations.

Question like this come up from developers for very legitimate reasons.
http://stackoverflow.com/questions/19604677/ (and badly accepted answer.)

I have no info on how many AT users end up clicking on the *current* page
in a 
navigation, but since we have no such notion beside the css :visited
selector, even non AT users can very easily pointlessly reload the same
page. Hence a 'current-page' ARIA state seems an adequate suggestion.

As devs, we can of course remove the link to a 'current page/category' but
that 
implies that we would be removing it from the 'skimming though links'
process which doesn't seem appropriate for AT users, nor for search
engines.
We can always treat that problem independently with javascript for non-AT
users if they 
click-on or touch it.
Received on Sunday, 2 March 2014 01:25:24 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:16:38 UTC