Speaking of hacks, here is an example of what I was talking about. It is from the <www.digitalmedievalist.org> website's screen css http://www.digitalmedievalist.org/dmlscreen.css.
b) Screen Geography */
div { /* controls width and location of all blocks except div.navigation and children */ margin-left: 20%; max-width: 35em; top: 0; }
div.navigation { margin-left: 0; position: absolute; /* position: absolute is for MSIE only; the style is removed below */ top: 15px; left: 4px; width: 18%; }
/* MSIE workaround */ /* This instruction hides "position:fixed" from MSIE5.5~6.0 */ body>div.navigation { position: fixed; }
This is a harmless enough hack, I suspect (it would be better if the second instruction referred to "body div.navigation" to keep the two instructions exactly parallel). Currently it works; if MSIE later implements the > child rule, then it won't work for MSIE anymore, and users of MSIE will end up with scrolling navigation. I am beginning to wonder if it is not better, however, to design a sheet to the standard, and then allow users of MSIE to choose a simpler sheet if some effect doesn't work well in their browser. Comments? -dan