W3C home > Mailing lists > Public > public-html@w3.org > September 2009

Re: what is dt?

From: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Date: Fri, 18 Sep 2009 16:59:36 +0200
Message-ID: <4AB3A058.9080505@xn--mlform-iua.no>
To: "Tab Atkins Jr." <jackalmage@gmail.com>
CC: Shelley Powers <shelleyp@burningbird.net>, Maciej Stachowiak <mjs@apple.com>, Smylers@stripey.com, public-html@w3.org
Tab Atkins Jr. On 09-09-18 16.06:

> On Thu, Sep 17, 2009 at 9:56 PM, Shelley Powers
> <shelleyp@burningbird.net> wrote:
>> Leif Halvard Silli wrote:
>>> I have no idea at all why the draft suddenly says that we can use <p> for
>>> dialog. I took part in the the debate and saw no compelling arguments.
>>> Perhaps it was Microsoft's un-support that made the deal?
>>> [snip]
>> Yet a decision is made that has generated out and out rejection from every
>> web page author who has happened to see it. Vehemently rejected.
> No it has not.  Not even close.  A handful of people got themselves in
> a twist because they saw <b> being recommended for something.

Since you demand such accuracy in language: No. This doesn't 
reflect anything I have seen on this list. I brought arguments 
against use of <b> - Smylers hasn't refuted them yet. But I have 
nothing against <b> whenever <b> is a synonym for <span> - 
whenever <span> is the right element.

> Overall, though, the current recommendation is nothing more than an
> outlining of *current practice*. 

There are different practices, currently. The chat section of the 
Microformats web site was brought into the discussion, and that 
page shows many current solutions, including the use of <dl>.

> We argued, *on this list*, that
> <dialog> wasn't sufficient for marking up dialog, and that people were
> already marking up dialog happily in various other manners (lists,
> paragraphs, and sometimes tables), and that these alternate methods
> *did* work because we had full flexibility.

When you say "lists", then I wonder if you count in <dl>?

You do have full flexibility in a <dl> list. However, it is 
necessary that you adapt to the key/value logic, of course.

>  Ian just chose a path
> that seemed the most reasonable out of all of those (and I agree that
> it's pretty reasonable and will probably change my chat room to
> generate markup along those lines).

The draft currently says:

"The dl element is inappropriate for marking up dialogue. Examples 
of how to mark up dialogue are shown below."

I will not respect that viewpoint - in fact, I'll probably file a 
bug about it. Not least because there were no compelling arguments 
in this debate that showed any fundamental problems with using 
<dl>. I think I provided several arguments to the contrary.

If anything, the draft should show how one /can/ mark up dialog 
with <dl>.

> Please don't mistake mailing list heat for internet-wide rejection.  A
> lot of us around here *like* arguing, and we did so well enough to
> convince the editor that his idea wasn't worth the effort.

So what /was/ the compelling argument then?

I tried to separate two issues:

a) <dialog> was OK, but the limitations on how <dt> and <dd> 
should be used compared to <dl> was not beneficial.
b) simply using <dl> could be good enough - with the option for 
using a role attribute to separate semantics.

Instead, something else happened. Away with <dialog> and away with 
<dl> as dialog container.
leif halvard silli
Received on Friday, 18 September 2009 15:00:17 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:51 UTC