[thelist] CSS, Netscape, .class oddity?

[Peter-Paul Koch]

>>having a lot of styles on the page (doesn't matter what and where).
>
>borders maybe... on images.  I don't know where else it is dangerous.
>"Having a lot of styles..."  What is that supposed to mean?

Just a great whacking lot of style sheets for 14 different classes all with
borders and backgrounds and paddings and margins etc.

As to borders: I had terrible trouble with a 'border: none' declaration on
a P (to get it to make a nice block of the bgcolor). NN4's crashed on my
site constantly; when I removed the border it stopped.

>Some things conflict, some don't.  You are generally okay with CSS1. 
>The only time I ever heard about CSS1 crashing Netscape was that 
>thing with Zeldman's pages... and I don't recall what the end 
>diagnosis of that was.

NN4 can make a mess of even CSS1, though if you use only a couple of styles
it should be safe. My personal guess as to the reason of Zeldman's pages
crashing is also a border declaration.

>It is never a good idea not to close tags when you're using CSS.

In Netscape, no, it's not a good idea. IE doesn't care.

>  >My recommendation is to remove all of your CSS properties for your table
>My recommendation is always run it thru the W3C validator.  The 
>validator showed that his CSS was fine.  

Valid CSS does NOT mean that Netscape won't crash on it or have other
problems. For instance,

position: relative;
margin-top: -20px;

is supposed to be perfectly all-right, but when you a relatively positioned
P or DIV partly overlaps the previous one (as the margin-top does in this
example) the links inside the P suddenly don't work any more. No validator
will ever catch this problem that isn't even supposed to be a problem.

IE4 has a similar problem that's caused by quite different declarations in
the style sheet (not yet sure which ones, but I had a 'height: 90%' that
closed down all links). This, too, was caused by CSS1 that was valid and
OK, but just didn't work correctly.

ppk