W3C home > Mailing lists > Public > www-style@w3.org > February 2013

Re: position fixed, z-index and Google Chrome.

From: James Robinson <jamesr@google.com>
Date: Tue, 5 Feb 2013 10:30:54 -0800
Message-ID: <CAD73mdJt4CenOQTJPT-997GNjMC7j9RTbknyzaH3RkJMnX0esg@mail.gmail.com>
To: lists@m8y.org
Cc: www-style@w3.org
On Tue, Feb 5, 2013 at 10:24 AM, <lists@m8y.org> wrote:

> On Tue, 5 Feb 2013, Simon Fraser wrote:
>
>  On Feb 5, 2013, at 10:48 AM, lists@m8y.org wrote:
>>
>>  On Tue, 5 Feb 2013, Simon Fraser wrote:
>>>
>>>  On Feb 5, 2013, at 9:04 AM, lists@m8y.org wrote:
>>>>
>>>>  So.  http://m8y.org/tmp/chrome_wtf.**xhtml<http://m8y.org/tmp/chrome_wtf.xhtml>
>>>>> Is a mockup of a problem I encountered recently.
>>>>>
>>>>> I was told to bring this up here.
>>>>> How does Chrome expect us to workaround this behaviour?  Or are we
>>>>> just not supposed to implement layouts like this anymore?
>>>>>
>>>>
>>>> What is the exact problem that you're seeing? How does Chrome's
>>>> behavior differ from other browsers?
>>>>
>>>
>>> Appears to be related to:
>>> http://updates.html5rocks.com/**2012/09/Stacking-Changes-**
>>> Coming-to-position-fixed-**elements<http://updates.html5rocks.com/2012/09/Stacking-Changes-Coming-to-position-fixed-elements>
>>>
>>> The page has instructions to reproduce on it, did you try them? I can
>>> take screenshots if that helps.
>>>
>>
>> Indeed. WebKit creates stacking context for position:fixed, in order to
>> optimize scrolling. This is not strictly compatible with the spec, but was
>> thought to have minimal real-world impact.
>>
>
> Real world is a relative thing I guess.  When IE controlled the web, one
> could say IE6 not supporting position:fixed at all had little real-world
> impact 'cause we had no choice.  Certainly, I don't see a way to do the
> layout on that page anymore.
> It seems to me if one is deliberately breaking the spec such that position
> no longer works right, it'd be nice to offer -webkit-position so we could
> target the broken implementation w/o javascript.
> Man. Maybe there needs to be a generic spec for the equivalent of IE
> condcoms :)  Then I could just target the UA.
>
> I can't tell what sort of layout you are attempting to achieve here, but
in general the solution for issues like this where you want to order
descendants of a position:fixed element is to apply the z-index to the
position:fixed element itself.

- James
Received on Tuesday, 5 February 2013 18:31:22 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:21:05 GMT