W3C home > Mailing lists > Public > www-style@w3.org > July 2010

Re: Property Selector

From: Boris Zbarsky <bzbarsky@MIT.EDU>
Date: Mon, 19 Jul 2010 22:27:54 -0400
Message-ID: <4C4509AA.8020800@mit.edu>
To: Garrett Smith <dhtmlkitchen@gmail.com>
CC: www-style <www-style@w3.org>
On 7/19/10 10:07 PM, Garrett Smith wrote:
> On 7/19/10, Boris Zbarsky<bzbarsky@mit.edu>  wrote:
>> On 7/19/10 8:12 PM, Garrett Smith wrote:
>>> On 7/19/10, Boris Zbarsky<bzbarsky@mit.edu>   wrote:
>>>> On 7/19/10 3:39 PM, Garrett Smith wrote:
>>>>> On 7/19/10, Boris Zbarsky<bzbarsky@mit.edu>    wrote:
>>>>>> On 7/19/10 2:50 PM, Garrett Smith wrote:
>>>>>>> So the problem is where the property is a getter and that getter has
>>>>>>> side effects and the side effects include modifying the DOM.
>>>>>>
>>>>>
>>>>> [...]
>>>>>
>>>>>> In Gecko, sure.  For example offsetWidth.
>>>>>>
>>>>>
>>>>> What is the side effect in the DOM from getting offsetWidth?
>>>>
>>>> In Gecko, due to the existence of XBL, arbitrary script can run under
>>>> the getter (since getting offsetWidth requires flushing out pending
>>>> style and layout changes and flushing out pending style changes can
>>>> attach XBL bindings and run their constructors).
>>>>
>>>
>>> Wouldn't pending layout changes need to be flushed prior to the query?
>>
>> I'm not sure I follow the question....  What query?
>
> Any NodeSelector.query.

No.  Why would those need anything flushed?

Note that the NodeSelector case is NOT the common case for selector 
matching, by the way.  The common case is style sheets.

-Boris
Received on Tuesday, 20 July 2010 02:28:27 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:29 GMT