[whatwg] markup as authored in practice

[Anne van Kesteren]

On Sat, 02 Dec 2006 23:41:34 +0100, Robert Sayre <sayrer at gmail.com> wrote:
>>>> There's probably no way you can serialize that document.
>>>
>>> Hmm, Sam's example displayed correctly in Safari, Firefox, Opera, and
>>> recent WebKit nightlies.
>>
>> Yes. Rendering it is different from serializing it though. I agree that  
>> it has to work as it does.
>
> What is the benefit of refusing to specify a serialization?

I didn't refuse anything.


>> I'm not sure I see the relation to HTML5.
>
> It's not conclusive, but the fact that
>
> <http://intertwingly.net/stories/2006/12/02/whatwg.logo>
>
> rendered correctly in WebKit nightlies while
>
> <http://www.intertwingly.net/blog/2006/12/01/The-White-Pebble>
>
> regressed (in several different ways depending on the revision) is a
> sign that the two ways of serving (X)HTML have more in common than the
> HTML5 specification claims.

Not at all. The HTML5 specification merely defines a way of mapping an  
input string to a DOM. That you can append other types of nodes to this  
DOM, such as SVG or MathML elements, isn't really a strange thing. It's  
not like we discovered something new here or anything.


> I don't understand why it's useful to pretend those pages live on
> separate planets because they  have different MIME types.

The way the page is initially processed is different. One is parsed using  
an HTML parser and the other using an XML parser. After that you deal with  
a DOM which is sort of similar except for certain APIs. element.tagName  
has to return the uppercase tag name for instance.


-- 
Anne van Kesteren
<http://annevankesteren.nl/>
<http://www.opera.com/>