[whatwg] register*Handler and Web Intents

Thu Aug 9 00:11:02 PDT 2012

I also agree with Henri and James. I would be opposed to implementing the feature in WebKit the way it is currently proposed. The aesthetic benefit is not great enough to be worth the breakage. Consider in particular that  the following proposed markup:

  <intent
    action="edit"    
    type="image/png"  
    scheme="mailto"   
    href=""           
    title="Foo"       
    disposition="">
</intent>

could just as easily be:

  <link intent
    action="edit"    
    type="image/png"  
    scheme="mailto"   
    href=""           
    title="Foo"       
    disposition="">

or:

  <meta intent
    action="edit"    
    type="image/png"  
    scheme="mailto"   
    href=""           
    title="Foo"       
    disposition="">

Both of which are no more verbose and read about as well.

 - Maciej

On Aug 6, 2012, at 8:11 AM, Henri Sivonen <hsivonen at iki.fi> wrote:

> On Fri, Aug 3, 2012 at 12:00 PM, James Graham <jgraham at opera.com> wrote:
>> I agree with Henri that it is
>> extremely worrying to allow aesthetic concerns to trump backward
>> compatibility here.
> 
> Letting aesthetic concerns trump backward compat is indeed troubling.
> It's also troubling that this even needs to be debated, considering
> that we're supposed to have a common understanding of the design
> principles and the design principles pretty clearly uphold backward
> compatibility over aesthetics.
> 
>> I would also advise strongly against using position in DOM to detect intents
>> support; if you insist on adding a new void element I will strongly
>> recommend that we add it to the parser asap to try and mitigate the above
>> breakage, irrespective of whether our plans for the rest of the intent
>> mechanism.
> 
> I think the compat story for new void elements is so bad that we
> shouldn't add new void elements. (<source> gets away with being a void
> element, because the damage is limited by the </video> or </audio> end
> tag that comes soon enough after <source>.) I think we also shouldn't
> add new elements that don't imply <body> when appearing in "in head".
> 
> It's great that browsers have converged on the parsing algorithm.
> Let's not break what we've achieved to cater to aesthetics.
> 
> -- 
> Henri Sivonen
> hsivonen at iki.fi
> http://hsivonen.iki.fi/