2009-10-03 21:47, Tab Atkins Jr. skrev:
Well, no amount of proof would do so; only a convincing enough
argument. I, personally, do not feel that<dt>'s semantics change
between<dl> and<details>. Nor do I feel they have different syntax
at all -<dl> and<details> do have slightly different syntaxes, but
it's very minor and pretty much bound up in the fact that<dl> has
multiple name/value pairs while<details> has only one, so<details>
doesn't *have* to worry about ordering in the same way that<dl> does.
>
> etc
In what way is the SYNTAX different? We seem to agree on this:
First and foremost, in <dl> the order is all important. Here it would
not matter.
In <dl> one may have several <dd> for each <dt> (or several <dt>'s in a
row), here one may not.
You call this "minor", I say confusing. But we have in fact created a
new syntax - why is that better than creating new elements?
In what way is the SEMANTICS different?
> So, in my mind, <dt>/<dd> do *not* hold some special meaning that
> locks them into only ever being used in <dl>. <dt> is a heading
> element, nothing more, effectively equivalent to <h1>*.
Well, that is not what the SPEC says is it?
> I mean, would you complain about using <title> or <caption> or <label>
> or <legend>... in <details>?
Yes, I would.
I am arguing in favor of introducing a new element, which would be the
zero cost solution, since <details> is new anyway.
+ No hacks besides those that we already use to get details working as
such in legacy browsers.
+ When implementing details the browser vendors will not have a harder
time using a new element than they would using dt/dd.
+ We would keep the several meanings per element count down, which from
a teachability POV is more important than keeping the total number of
elements down.
And from that POV nuances are often harder to pick up than anything else.
--
Keryx Web (Lars Gunther)
http://keryx.se/
http://twitter.com/itpastorn/
http://itpastorn.blogspot.com/