[whatwg] Tag Soup: Blocks-in-inlines

Wed Jan 25 03:19:16 PST 2006

Anne van Kesteren wrote:
> Quoting Lachlan Hunt <lachlan.hunt at lachy.id.au>:
>> 1.
>> <em><p>X</em>Y</p>
>>
>> BODY
>>   + P
>>     + EM
>>       + #text: X
>>     + #text: Y
>>
>> The theory is that any inline elements
> 
> This gives problems for new elements I assume... We already have a 
> problem with
> <header><h1>test</h1></header>...

I don't see how this affects new elements, it should only affect known 
inline elements.

>> 2.
>> <em><p>XY</p></em>
>>
>> BODY
>>   + P
>>     + EM
>>       + #text: X
>>       + #text: Y
> 
> And this likely breaks existing content. Perhaps not for EM, but 
> certainly for
> other inline elements, like <span>.

I'm not saying it won't break anything, but every single change we make 
to the parsing could possibly break any number of the billions of pages 
on the web in any number of browsers.  However, the chances are that 
such pages are already broken is several browsers already (probably 
built for IE only, who's quirks we are definitely not keeping), so I 
don't see this as a huge problem.

There's nothing wrong with saner parsing at the expense of a few broken 
pages which I'm sure will still remain readable (even if they don't look 
perfect) and/or be easily fixed by their authors.  Trying to remain 100% 
compatible with 100% of the web is physically impossible.

However, span does show some interesting behaviour which should be made 
more consistent with other inline elements.

<!DOCTYPE html><span><p>X</span>Y</p>

Firefox:
HTML
   + HEAD
   + BODY
     + SPAN
       + P
         + #text: X
     + #text: Y

Opera 9/Win:
HTML
   + BODY
     + SPAN
       +P
         +#text: X
         +#text: Y

IE6:
HTML
   + HEAD
     + TITLE
   + BODY
     + SPAN
       + P
         + #text: X
         + #text: Y
       + #text: Y (Highlighted in red in the DOM view)

-- 
Lachlan Hunt
http://lachy.id.au/