Tee G. Peng wrote:

Georg, can you kindly take a look on IE6? The horizontal menu doesn't load smoothly, when the page is fully loaded, the header's part reloads, I suspect it has to do with the clear both class yet I
 can't figure  a fix for IE (tried all tricks from hasLayout)

The page loads very much slower in IE6 than in other browser, presumably
because IE6 needs so many fixes. Can't see any reloads.

The use of 'hasLayout' related tricks should be kept at a minimum -
especially at early dev-stage, as that "bug/fix" can ruin any layout
more than it helps, and make all attempts to fix things in IE6 more or
less impossible.
Adding too many 'hasLayout' triggers is like bracing a building to the
point where one has no space left to move around on inside it. Not much
of a building then, and IE/win (any version) won't cooperate well either.

Once loaded, the right part is dropped below the left part, see...


...resulting in overlapping. It will jump back into place with smaller
than normal font-resizing, but won't stay there with normal or larger

I think it's a case of "auto-expansion" - IE6 doesn't respect declared
dimensions and can't handle overflow properly. The only cure against
that bug is to hide the horizontal overflow with 'overflow-x: hidden' on
problematic containers, which _may_ hide a bit too much at times.

Generally it looks like you have mixed 'em', 'px' and '%' for
positioning and dimensions in ways that doesn't add up under anything
but ideal conditions, and the weaknesses show in any browser. Resizing
window and adding a bit of font-resizing create overlapping and
overshooting in all browsers.
No "elastic" or "conditional elastic" method works well for layouts with
that many fixed dimensions - images. Although it is possible to make
something like that work in better browsers, IE6 will put up such a
fight that it simply isn't worth the hassle, IMO.

A fixed-width approach will _just work_ in all browsers, and make it
much easier to get IE6 to behave like a browser.


List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm

Reply via email to