Re: ol, ul, nl, dl, oh my! (was Re: [XHTML 2] removal of navigation list element)

On Wed, 1 Jun 2005, Shane McCarron wrote:

> While it is incompatible, it is not *arbitrarily* incompatible, if you
> see what I mean.  For example, we could have easily gotten rid of the
> "a" and "img" elements, but we have not because people like and use
> them.

Sorry, but that sounds ridiculous. If you have decided to make a change
that prevents _all_ current browsers from rendering XHTML 2 documents, you
are paying quite a price, so why make irrelevant changes like the
introduction of the cryptically named <l> to replace <br> which people
like and use very much, yet leave badly named and badly designed elements
alone? It's foolish enough to have to explain that <a> is <a> because
someone had an anchor in his mind, but it's much worse to have to explain
that the brand new XHML 2 _still_ has such an odd element. If
compatibility with existing HTML is not required, a link should be <link>
of course. (Hiding linking into attributes alone violates fundamental
principles of good semantic markup. But I digress.)

> These types of lists are in common use, and have different
> default presentation.

The <font> tag is also in common use.

> Since CSS support is not required, we cannot
> eliminate one type of list and say "use CSS to style these items how you
> want".  It wouldn't be portable.

We can, just as we can say "use CSS to suggest fonts". It isn't portable
either.

We can and should say "if numbering is essential, use numbering in the
content". Just as we can and should say about numbering of headings

> >It's still a _list_. Not a collection or set of items.
> >
> True.  Whereas the paragraphs in a document are not a list.  Reordering
> those arbitrarily would violate the "principle of least surprise".

Paragraphs appear in a specific order in HTML source. They do not
constitute an ordered sequence in any other meaning, though common sense
says that the rendering order should normally be the same as the source
order. But a <ul> element is even stronger ordered in the sense that it is
defined as a _list_.

> Reordering the items in a UL so that they are shortest first or
> something would be a nice transformation on the data, and as an author I
> personally wouldn't really mind.

I'm sure many authors would. The choice of the order in which you make
your points can be a difficult and important issue, even if you end up
with making your points bulleted.

-- 
Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/

Received on Thursday, 2 June 2005 07:42:04 UTC