[whatwg] [br] element should not be a line break

Aryeh Gregor Simetrical+w3c at gmail.com
Fri Aug 6 13:29:06 PDT 2010


On Thu, Aug 5, 2010 at 1:48 PM, Ryosuke Niwa <ryosuke.niwa at gmail.com> wrote:
> That's totally incorrect in HTML5 as Thomas has pointed out.

As I pointed out, it's only theoretically incorrect.  <b> still means
"something that's conventionally boldface", and <i> still means
"something that's conventionally italic".

> Let me ask you
> a question.  What do you suppose non-visual user agent should do when they
> encounter br?  Simply ignore them because it only signifies a line break?
>  Or read out that there's a line break?  Neither seems user friendly to me.
>  If anything, a momentary pause will be appropriate because what's what we
> usually do when reading a book and a line break appears.  This clearly isn't
> *line break*.

No, but it's a stand-in for a class of semantics that can only fairly
be summarized as "the places where you would always use a line break
in print".  There is no single behavior that screen readers could use
to correctly present <br>, but the same is true for any number of
other cases.  How to pronounce the word "minute" depends on context
too, because the sequence of letters M-I-N-U-T-E can signify multiple
concepts that happen to be represented the same way textually, but
vary when spoken.

There is no realistic way to avoid this kind of thing.  Even if you
eliminate it on the markup level, it remains on the level of text, so
you haven't actually made the problem go away.  Instead, we rely on
the fact that a listener can usually extract the meaning pretty well
even if some of the fine distinctions are lost, and focus
accessibility efforts on avoiding only drastic misrepresentations
(like missing content images).

This discussion would not even be occurring if not for incidental
choices in the underlying technology.  If HTML respected Unicode line
breaks, no one would propose that Unicode line breaks must be axed in
favor of a semantic solution.  Insisting that every single HTML
element must be fully semantic and media-independent, while ignoring
the fact that web pages are written in text and that is
*intrinsically* not media-independent, does not make any sense.



More information about the whatwg mailing list