Sent by Bruno Fassino on 3 September 2004 07:07
Ingo Chao wrote:
> yes, in quirks mode, the margins stay correct. Its the
> standard render mode which behaves quirky.
:-))
> The bug is the wrong calculation of the %-margin before jumping.
[...]
> 1000px x 10% = 116px: error: +16px;
> 500px x 10% = 116px : error: +66px;
> 300px x 10% = 116px : error: +86px;
I'm not sure, but at least on my PC, IE6 seems to follow this "logic":
The wrongly computed margins are obtained applying the percentage not to the
container width, but to the body width.
Try this: resize the window, reload and measure the applied wrong margins
(before the jumping):
If the window width is 1000px, the 10% margins are all 100px; if the window
is 500px the margins are 50px, and so on...
Is this reproduceable for you as well?
Having seen this, I tried wrapping each container in another one, with the
same width. This seems to convince IE6 to use the width of these containers
in computing the %-margin. Even assuming that this workaround works
consistently (I'm absolutely not sure), it is not so nice to have to use two
nested wrappers...
(I also tried adding a javascript which, after page loading, forces any
element to have 'layout'. This seems to force a sort of re-rendering and
move the shifting boxes in the correct positions, but I need to check
better.)
> Someone might tell me the name of the bug
I think you can choose a name for it!
Bruno
______________________________________________________________________
css-discuss [EMAIL-REMOVED]]
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/
