W3C home > Mailing lists > Public > www-html@w3.org > May 2005

Re: Comments on the XHTML 2.0 WD

From: Laurens Holst <lholst@students.cs.uu.nl>
Date: Sat, 28 May 2005 22:22:00 +0200
Message-ID: <4298D2E8.4050602@students.cs.uu.nl>
To: Maxwell Terpstra <terpstra@myrealbox.com>
Cc: ACJ <ego@acjs.net>, www-html@w3.org

Maxwell Terpstra wrote:
>> I understand the difference in type of lists, and that is exactly why
>> I've recently started pondering (I was a fan when I first read about it)
>> whether navigation should be one of them—it seems to me "navigation"
>> says something about its contents on a different level.
> I agree.  The current list types have important structural differences - 
> <dl> presents a list of data pairs, <ol> presents a list whose ordering 
> is important, and <ul> presents a "grab-bag" list.  <nl> is just the 
> same as <ul> in this respect, differring only in it's role in the 
> document.  Now that there exists an attribute to express this "role," 
> there seems to be no reason to create an <nl> element.

I agree as well.

I understand the purpose of creating the <nl> list is to catch a case 
that is currently commonly done with a scripting solution:

"Part of the design of XHTML 2 has been to observe existing use of HTML 
and identify what is perceived as missing, for instance by use of 
scripting to achieve ends not supported directly in HTML. One obvious 
component of very many HTML pages is the 'navigation list', consisting 
of a collection of links to other parts of the site, presented 
vertically, horizontally, or as a drop-down menu."

I think the thought is nice, but in practice web site authors will want 
to control the orientation of such a list, and have their own styling 
applied to it anyway. Hence, a default representation of such a thing 
(which for the sake of discoverability might justify a separate element) 
would not add anything of real value to the website author.

I think that we can nowadays depend on the page author knowing that a 
navigation list is just a list (after all, that is I think after the 
misuse of the <table> element the second-most stressed example when 
explaining the authoring of semantic documents), and let the role of the 
list be defined with the role="" attribute.

Which I happen to like a lot :). The alternative is defining separate 
elements like the WHATWG does, such as <article>, <aside>, <nav>, etc., 
which I feel not comfortable with. Which is exactly what <nl> does as 
well right now.

However, with the role of a section defined as an attribute, it will be 
somewhat difficult to style without shorthand in CSS that e.g. the class 
attribute offers (which I used as ‘role’ before). I think a solution 
needs to be found for that.

Be it either by writing role as ‘class’, which kind of means the same 
thing and in XHTML 1.0 documents I was (most of the time) using it for 
that purpose anyway. Or by adding a new shorthand for the role attribute 
in CSS (which would be something the CSS WG should do), or mapping the 
role attribute to the CSS ‘.’ class selector (which the HTML WG can do, 
but might be a little confusing). I think I personally favour the last 

As a secondary issue, I also find the requirement of placing a <label> 
tag in navigation lists unnecessary - this element may be useful, but 
should be optional. The contents of the containing <li> of subnavigation 
lists can fulfill the same role.

I hope the W3C will address these thoughts.


Ushiko-san! Kimi wa doushite, Ushiko-san!!
Laurens Holst, student, university of Utrecht, the Netherlands.
Website: www.grauw.nl. Backbase employee; www.backbase.com.
Received on Saturday, 28 May 2005 20:22:01 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:06:10 UTC