Mark Lundquist wrote: > [...] > Something along those lines?
Guess one _can_ describe IE's behavior that way, but your understanding of this 'auto-expansion' bug is still too weak to leave it at that. Both the bug and the fix are complex - and neither makes no sense outside IE6, so the following description will be just as complex and "sense-less" :-) 1: IE6 does *not* treat 'width' as 'min-width' in any real sense, so 'width' can't be used as a substitute for 'min-width' in that bugger. Part of IE6' problem is that it doesn't treat 'overflow: visible' (default or declared) at all. IE6 can't overflow properly. IE6 treats _all_ elements as if they were _partly_ under HTML table rules, where 'width' *is* and should be treated as 'min-width' and where 'overflow' is ignored. 2: IE6 *does* treat 'width' under conditions like yours as: "you can't be serious - the content doesn't fit so I'm going to ignore you". You'll see that - completely unlike 'min-width' in a good browser - the otherwise normal-width content inside an 'auto-expanded' container doesn't fill the extra width provided by the bug, so it's only the 'auto-expanded' container itself that has become too wide. So, we have basically these options: a} make IE6 respect given 'width' by adding 'overflow: hidden' or 'overflow-x: hidden' on the container. This will make IE6 respect 'width', but the overflowing parts of "too wide" children will of course be hidden and we will have to make them visible again. b) make the "too wide" child be narrow enough not to create a "pushing-effect". This works fine, but restricts what we can put into that child and how we want it to appear. c) make the "too wide" child _appear to be_ narrow enough not to create a "pushing-effect". All browsers are calculating the horizontal width-space needed for or occupied by an element based on where its outer margins are. Thus, if we "pull back" the 'backside margin', the element will take up less space. ---- I used a variant of c), but it's one that only works because IE6 is buggy and other browsers respect 'width' on the container regardless of what I do to its children. I could therefore "pull back" the right margin on the relative-positioned #footer and get the effect I wanted in IE6. I mentioned that the whole thing "makes no sense", didn't I :-) You can read more about IE6' 'auto-expansion' bug here... <http://www.positioniseverything.net/explorer/expandingboxbug.html> ...and then you can try to figure out how wide IE6 thinks the sloping line of "dead herrings" is in this page... <http://www.gunlaug.no/contents/molly_1_06.html> The essential CSS parts of that sloped line is in the page-head, and you should make the browser-window /really/ narrow to see the slope extend beyond both sides of the center-column, and even beyond the entire page. regards Georg -- http://www.gunlaug.no ______________________________________________________________________ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7 information -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
