Re: 3.8. Sections

Hi Sander,

On Aug 8, 2007, at 9:38 AM, Sander Tekelenburg wrote:

>
> At 15:59 +0200 UTC, on 2007-08-08, Anne van Kesteren wrote:
>
>> On Wed, 08 Aug 2007 15:50:22 +0200, Sander Tekelenburg  
>> <st@isoc.nl> wrote:
>
> [...]
>
>>> How does "parent sectioning element" leave room to think
>>> it might apply to a non-sectioning parent element?
>>
>> Because it refers to the parent in a tree.
>
> No, it doesn't. It specifically refers to the parent *sectioning*  
> element. It
> says so.
>
>> And parent (same for ancestor)
>> has a well defined meaning in a tree. [...]
>
> All I'm stating is that (I think) a simpler terminology would be  
> preferable.
> I'm *asking* for an explanation why that simpler terminology  
> doesn't cover
> what is being expressed. But your response is a statement. Without an
> explanation that's not very helpful.
>
> AFAIK "parent" == "nearest ancestor". If not, an explanation or a  
> pointer to
> one would be useful.

The confusion here, I think, is that "parent sectioning element of A"  
can refer to A that must then have a document (element) tree parent  
that is a sectioning element. You see this at times when the spec  
describes a TBODY and says the "TBODY element's parent TABLE  
element".  This sound very close to the way you're suggesting we use  
parent sectioning element, but means a completely different thing..  
To use it the way you're suggesting, I think we would first need to  
introduce the concept of a sectioning tree and differentiate that  
from the document (element) tree. I'm not saying that wouldn't be  
possible or even desirable, but it is not necessarily a  
simplification. FWIW, we already have a distinction between the DOM  
node tree and the document element tree.

Take care,
Rob

Received on Wednesday, 8 August 2007 18:29:51 UTC