Re: HTML5 script start tag should select appropriate content model according to src

David Woolley:
> Tina Holmboe wrote:
>
>>   There is not now, nor has there EVER been, a legitimate use case  
>> for
>>   the I or B elements in HTML.
>
> I think that depends on what you see the purpose of HTML to be.  I  
> think you and I, and I believe TBL, have a higher view in that it  
> is to help people to communicate, but I don't think that that is  
> what the WHATWG's constituency wants.  They do want a  
> presentational language, and they are in the vast majority.

There is no reason we cannot have both in one language. Actually we  
do have it already, if we consider both |div|+|span|+|class| and | 
table|+|font|+|img| documents to be of one kind and files with more  
exotic (read: semantic) elements to be of the other kind. Then there  
are applications (|form| etc.), too.

I very much like semantic markup, but I also welcome the unity of  
languages for commercial (advertising and basic application  
interfaces) and scientific (hypertext) purposes, although there is in  
practice some wisdom required on behalf of user agents to decide how  
serious the theoretical semantic meanings of the elements used are to  
be taken in a given document.

> Unfortunately, the average HTML author sees semantic markup as  
> ivory towerism.

We cannot force mainstream authors to use or even understand semantic  
markup. They are too many, they have been there too long, and things  
worked too well. It is of no practical use to moan about this.

We can try to guesstimate their minds. We may allow them to use  
meaningless presentational code that is rendered without extra costs  
within the same user agents as semantic markup is. Everything else is  
lost terrain.

Received on Thursday, 3 May 2007 20:27:47 UTC