Re: what is dt?

Shelley Powers On 09-09-17 19.32:

> Leif Halvard Silli wrote:
>> Smylers On 09-09-17 18.18:
>>> Shelley Powers writes:
>>>> Smylers wrote:
>>>> Except now, when seemingly dt and dd have become HTML5 silly putty:
>>>> good for a hundred and one uses.
>>
>>> So it's actually <dl>, <details>, and <figure> which have different
>>> uses.  <dt> and <dd> are 'building blocks' which each of those happen to
>>> contain.
>>
>> Agree in principle about 'building blocks'. <li> is also the building 
>> block of both <ul> and <ol>.
> 
> But the li element is used exactly the same in ul and ol. The same does 
> _not_ apply to dt/dd in Figure, Details, and dl.

  [...]

> Look at the syntax for dt now:
> 
> Before dd or dt elements inside dl elements.
> In a figure element containing no other dt element children.
> As the first child of a details element.


In <table>, the <caption> has to be the first element. I see no 
problem with saying that, in a <figure> - or as I prefer <dl 
figure > - the <dt> _must_ come before the <dd>. That is what is 
said about <detail>. And <details> and <dl> are fairly similar in 
that respect, except that <details> only allow one dt/dd group.

If we said it like this, then the placement of the <dt> would have 
to be styled via the CSS caption side property.

In HTML 4, the <tfoot> has to come before <tbody>. Not so in HTML 
5. So there is a tendency to dislike requirements to have elements 
at a certain place.

 
> One could look at how dt is used within dl, and extrapolate from that 
> usage, how the dt element could be used in Figure. Except for one thing: 
> it would be wrong. In fact, extremely wrong. There are significant 
> differences to how dt/dd are syntactically used in all three of the 
> elements. I'm not talking about minor differences in semantics and 
> syntax, I'm talking about completely and absolutely different.

This is - or was - the same for <dialog> as well. That is why I 
proposed that there should be the exact same rules for dt/dd in 
dialog as in dl.

> What happened is people searched for a set of constraints based on 
> legacy browser physical behavior and just grabbed whatever made it 
> through the filter. The confusion caused to existing web developers and 
> designers, as well as new web page authors in the future was, from what 
> I can tell, never once factored into the decision.


I agree that it sometimes seems a bit to accidental what solutions 
one ends up with. For my own taste, a real <dl> - <dl figure> etc 
- therefore seems better.
-- 
leif halvard silli

Received on Thursday, 17 September 2009 17:54:38 UTC