On Apr 04, 2006, at 22:15, Michael Champion wrote:
> > From: robin.berjon@e...
>
> > That's a red herring, you can trivially switch on  
> arguments.length. I
> > think that whatever constraints that brought them to do this were
> > different. I'm not the OMG IDL specialist but could the overloading
> > issues come from there?
>
> I believe that was the strongest argument against method  
> overloading in the DOM.  As I've argued many times on this list,  
> it's easy to flip the bozo bit on the DOM (and I personally think  
> it's time to consider end of life planning for the poor ol' dog).  
> But to understand where it came from one needs to remember those  
> days of yore when CORBA was considered the wave of the future, and  
> its constraints -- e.g. no overloading, the assumption of language  
> neutrality, etc. -- were baked into DOM.

The Web API WG is considering dropping compliance to OMG IDL (as can  
be seen in the freshly released http://www.w3.org/TR/XMLHttpRequest/)  
what would folks here think of that? This drop of compliance would  
likely not concern DOM 3 Events and DOM 3 XPath which are also in our  
care.

There are concerns that some implementations (including browsers) are  
relying on compliant IDL, and that Java bindings (yes, people  
apparently want to use these APIs there, why is open to your  
speculations) are often generated from them.

Any input from this gang of heretical DOM-haters?

-- 
Robin Berjon
    Senior Research Scientist
    Expway, http://expway.com/