Re: Cleaning House

On May 3, 2007, at 23:35, Philip & Le Khanh wrote:

> I believe we should provide the functionality to allow an
> author to add his/her own preferred semantic markup, since the
> one things on which I am confident we /all/ agree is that a finite
> set of elements can never carry all the possible semantic nuances
> that an author might wish to convey.

Focusing on authors wanting to convey things is a common fallacy in  
semantic markup advocacy.

Semantic markup is about communication between parties. The purpose  
of Web specs is to establish a common understanding between all  
parties so that a priori bilateral agreements between particular  
authors and markup consumers are not needed. If you pull "semantic"  
markup from your sleeve, the consumer has no way of knowing what you  
mean without a prior agreement.

Suppose I send your UA this:
   <kappale>This is <korostettu>important</korostettu>!</kappale>
how is your UA going to implement behavior appropriate for the  
semantics of the elements without either
a) a spec for this markup language and a human implementing behavior  
for it in your UA
b) artificial intelligence that reads Finnish

Perhaps the problem is that the semantics are expressed in Finnish,  
but so far UAs don't have English AI, either.

Henri Sivonen

Received on Friday, 4 May 2007 15:09:50 UTC