I was under the impression that "if the content requires more height than is specified for the [block's] attribute, the content requirements override the attribute value" (Danny Goodman)
This impression is incorrect. This _is_ how IE/Windows acts, but it is _NOT_ what the CSS specification says. (And I've never seen Danny Goodman's name involved in any CSS discussion; I'd love to know who he is.)
I cannot find a logical reason for this mish-mosh behavior. How exactly does the W3C spec allow/cause this to happen?
You mean "how does the W3C spec allow a website author to shoot himself in the foot?" By being a tool (that can be misused) rather than a toy. Do you wonder why sites can set background and foreground both to the same color? ;)
