W3C home > Mailing lists > Public > www-dom@w3.org > July to September 2006

Re: [dom3core] Cross-document appending of nodes

From: Anne van Kesteren <annevk@opera.com>
Date: Tue, 29 Aug 2006 17:36:41 +0200
To: "Boris Zbarsky" <bzbarsky@mit.edu>
Cc: "DOM mailing list" <www-dom@w3.org>
Message-ID: <op.te1xnfpo64w2qv@id-c0020.oslo.opera.com>

On Thu, 24 Aug 2006 18:01:07 +0200, Boris Zbarsky <bzbarsky@mit.edu> wrote:
> Anne van Kesteren wrote:
>> FYI: For compatibility reasons with other browsers and websites (which  
>> rely on their behavior) Opera now does an implicit call to adoptNode  
>> when appendChild, insertBefore or replaceChild are invoked.
>
> Intetesting.  Can you possibly elucidate on why this was needed?   
> IE/Windows does not exhibit this behavior, and we've been trying to  
> eliminate it in Gecko (because it causes security issues in many cases);  
> I'm hoping that this behavior will be gone in Firefox 3.
>
> Are there specific sites that do browser-sniffing and serve different  
> content to IE and other browsers, with the "other browsers" content  
> broken?  If so, wouldn't it make more sense for Opera, Mozilla, and  
> Apple to evangelize them than to start reverse-engineering each others  
> bugs here?

As I understand it we've tried to do that but failed. However, I suppose  
Firefox doing the right thing would give some sites the incentive to fix  
their bugs. When that happens, I suppose we revert our behavior as well.


(FWIW, we also reverse-engineered several extensions to the Range object  
and other things Mozilla introduced at some point because sites are using  
those. For us it's not just Internet Explorer that's causing issues :-(.)


-- 
Anne van Kesteren
<http://annevankesteren.nl/>
<http://www.opera.com/>
Received on Tuesday, 29 August 2006 15:37:06 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 22 June 2012 06:13:58 GMT