W3C home > Mailing lists > Public > public-webapps@w3.org > April to June 2009

Re: [webidl] XMLHttpRequest interface object

From: Ian Hickson <ian@hixie.ch>
Date: Thu, 30 Apr 2009 16:58:18 +0000 (UTC)
To: Anne van Kesteren <annevk@opera.com>
Cc: WebApps WG <public-webapps@w3.org>
Message-ID: <Pine.LNX.4.62.0904301650150.7824@hixie.dreamhostps.com>
On Thu, 30 Apr 2009, Anne van Kesteren wrote:
>
> For environments where the global object is represented by the Window 
> object XMLHttpRequest needs to be associated with a Document object for 
> determining the origin and base URL. The problem is that the 
> XMLHttpRequest interface object can be copied around, e.g.
> 
>   var x = frames[0].XMLHttpRequest
> 
> in that scenario x needs to be associated with Document object 
> associated with frames[0]. (If frames[0] is later navigated the Document 
> object would change, so it is important that this happens at this 
> point.)

There is a 1:1 mapping from global object to Document object now that we 
have WindowProxy objects.

So you can probably just say something like:

   "...the Document object associated with the Window object for which the 
   XMLHttpRequest interface object was created..."


> To be clear, assuming a and b are distinct global objects, in
> 
>   a.x = a.XMLHttpRequest
>   b.x = a.x
> 
> the XMLHttpRequest interface object in b.x has a pointer to the Document 
> object associated with a. When the XMLHttpRequest constructor is invoked 
> this pointer needs to be somehow copied to the constructed object.

I think that is covered by the incantation above.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Thursday, 30 April 2009 16:58:57 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 18:49:31 GMT