- From: Jukka K. Korpela <jukka.k.korpela@kolumbus.fi>
- Date: Mon, 11 Nov 2013 15:34:00 +0200
- To: public-html@w3.org
2013-11-11 15:07, Léonie Watson wrote: > > Jukka J. Korpela wrote: > > “I wonder what the specific issue is there. Which software is this about? > > What is the difference, from the user perspective, between > > <nav><ul><li><a href=...>...</a>...</ul></nav> and a <nav> containing > > just <a> elements with some separators, like “→”, between them? How much > > does this matter, and why?” > > The difference is that the nav element doesn’t convey (and isn’t > intended to convey) information about a collection. The list provides > the semantic information that tells any/all screen readers that there > is a collection of items displayed on the page. > A <ul> element contains a sequence of <li> elements. It can be called a collection, but so what? Calling it a sequence or list or collection does not really change anything. > So from the screen reader user’s point of view: The nav element > indicates that the region of the page relates to navigation and > provides a landmark for efficiently navigating to it, where as the > list indicates there is a collection containing a definite number of > items. > I fail to see what impact this would have on software and users. The number of items is no more (and no less) definite than in the other case. (And a browser cannot tell the number without parsing and counting the <li> elements, just as it can count <a> elements.) A speech browser can read <nav><ul><li>foo<li>bar<li>zap</ul></nav> (I’m omitting link markup here for simplicity, as it is the same in both cases) as “bullet foo, bullet bar, bullet bar”, and a simple <nav>foo → bar → zap</nav> as “foo, arrow, bar, arrow, zap”. What is the difference? A speech browser might have a reading mode where list items are read under user control, so that a user action is required to proceed to next item. While this can be useful for <ul> elements with lengthy items, would it really be useful for a breadcrumb? Since the entries are links, isn’t it enough to separate them with separator characters? (And modern software could probably handle them without separators, too; separators, like an arrow or greater than sign or other symbol, are really more for the eye than for the ear, especially when a speech browser anyway precedes each link with a word like “link”.) There can be other specialties in handling <ul> (and <ol>) in speech browsers, affecting the issue, but I think they should be discussed in concrete terms and not with abstract arguments about semantics. And for any specialty, we should ask whether it is just an oddity in some software, relevant to its users, but something that should not dictate what HTML specifications and recommendations say. -- Yucca, http://www.cs.tut.fi/~jkorpela/
Received on Monday, 11 November 2013 13:34:26 UTC