Bert Mahoney wrote:
> First Merry Christmas and Happy Hanukkah to list members.
>
> I have stumbled upon a hover bug in IE PC that doesn't show in FF IE or PC.
>
> I'm using an IE PC hack to use fixed positioning and I am using the
> Gilder-Pixy method for using rollover images. When you hover over the link
> the scroll bar moves in about 50px from the right every time.
>
> I've looked at the source and can only figure that FF isn't honoring the
> height of 101% on the #bin-left div.
>
> If anyone on the list can offer other advice I would be grateful.
>
> I'm indebted to Ingo Chao who made the following simplified test case.
>
> Here is the link.
>
> http://www.satzansatz.de/cssd/tmp/compaction.html
>
I must admit I don't have much experience with the position:fixed
emulation methods for IE/Win, so I hope others who use it regularly
can give a better answer.
Looking again at the method, the principle is to have a scrollable
pane, and any position:absolute which relates to the html element
will look like positioned fixed with respect to the scrollable pane.
In standards mode, this can be accomplished by giving overflow:auto
to body and overflow:hidden to html.
* html { overflow:hidden;
width:100%;
height:100%;
}
* html body { overflow:auto;
width:100%;
height:100%;
}
* html #bin_right {position:absolute}
#bin_right {position:fixed}
But this triggers the mentioned bug when redraw events are induced
like in your case where the image replacement requires
background-positioning on a:hover.
Other methods do not use body as scrollable pane, but an inner
wrapper as the first child of body instead:
* html,
* html body{
width:100%;overflow:hidden;
height:100%;
}
* html #wrapper {
overflow: auto;
width: 100%;
height: 100%;
}
* html #bin_right {position:absolute}
#bin_right {position:fixed}
This stops the compacting effect of the bug, see
http://www.satzansatz.de/cssd/tmp/bm/
The general drawback is that any absolute positioned element appears
fixed, and by horizontally resizing the window, elements overlap the
vertical scrollbar, so I'd prefer a degradation instead of emulation
for IE.
Ingo
--
http://www.satzansatz.de/css.html
______________________________________________________________________
css-discuss [EMAIL PROTECTED]
http://www.css-discuss.org/mailman/listinfo/css-d
List wiki/FAQ -- http://css-discuss.incutio.com/
Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
