And how could it be that it was changing back and forth between the wanted rendering and the faulty one, seemingly random?
Or is it random?
No! It's actually painfully consistent. Well, when the bug is actually there, at least. It seems to come and go. Right now, I am not allowed to disclose any details about the project, so quick descriptions will have to do for now.
OK, so what I've got is a positioned (explicit) block level element with it's width also explicitly set to 100%. This block level element sits inside of another floating block level element with its position set to relative.
So, what's the dealio?
The issue, of which I was introducing above, is differences in the rendering. In exactly the same browser in every single way. Same OS, same computer.
IE6 renders the web page in two different ways – I can make IE6 render it the first way just by pushing the refresh button or hitting F5. The second rendering version I trigger by marking the URL and hitting Enter, or alternatively pushing the Go button to the right of the URL. The innermost block level element's width will expand out to its parent element (wanted behaviour), or to a parent element further up (faulty behaviour).
Now, this is beyond me. Why are the results different from one another, just because I loaded the page different ways? I've tried hard refreshing about three billion times, so this should not be caused by som weird, annoying cache issue.
PS: If you're visiting my blog using IE6, you'll see I really don't care all that much about optimizing for it. This is, as stated above somewhere in the text, a site I'm coding at work for quite a big customer. They care.
Legg igjen en kommentar