Where and what is the ICB? - Different views of browsers.
Sent by Philippe Wittenbergh on 16 April 2005 07:07
On 16 Apr 2005, at 7:40 am, Adam Kuehn wrote:
>> There are problems [overflow: auto; versus overflow: visible;]
>> regarding with HTML-element and positioned elements. I wanted
>> to know it exactly and thus, I wrote a test side.
>> < http://www.kriton.de/TEST/ICB.HTML >
> It is possible you are seeing some artifacts. Your test case is not
> valid HTML. You failed to close the head element, and did not make a
> body element at all. Some of the differences you are seeing may be
> related to the different ways the browsers recover from the invalid
> code. Try validating, and then see if the problems persist.
Correction, it is a perfectly valid html401 strict document. Remember,
<body>is implied for text/html 401 docs. (but this not a discussion for
this list, he hasten to say).
File is at <http://dev.l-c-n.com/_temp/ICB.html>
> You are correct, though, that some differences are to be expected,
> since the initial containing block is up to the browser-makers. This
> area is pretty vague, and I'm afraid I don't have a decent summary for
FF 1.0+ nightly build renders your test perfectly. Both Safari (1.2.4
and 1.3) and Opera 8b have some problems with the styling of the body
(even if you explicitly add it to the markup). They move the overflow
to the viewport. UA's have some freedom to do this according to the CSS
2.1 rec. There are some discussion about the validity of this, however.
But other wise, all 3 browser render the test fairly identically.
Add margin:0 to your h1
All three browsers use <html> as the ICB.
FF1.02 positions the overflow (scrollbar) on the viewport. This has
been tightened with the nightly builds.
now live : <http://emps.l-c-n.com/>
code | design | web projects : <http://www.l-c-n.com/>
IE5 Mac bugs and oddities : <http://www.l-c-n.com/IE5tests/>
List wiki/FAQ -- http://css-discuss.incutio.com/
Supported by evolt.org -- http://www.evolt.org/help_support_evolt/