The background image is placed one-third of the way across the page, and halfway down it, as depicted in Figure 6-42.

Table 6-2 gives a breakdown of keyword and percentage equivalencies.
The names of these properties make them fairly self-explanatory. Here we have one possible solution for the example given in the previous section:
top: 10%; bottom: 20%; left: 50%; right: 10%; min-width: 20em; min-height: 30em;
Of course, this still isn't a very good solution, since it forces the element to be at least 20em wide by<TD><A HREF="read.html">My Writing</A></TD><TD><A HREF="fun.html">Fun Stuff!</A></TD><TD><A HREF="links.html">Other Links</A></TD><TD><A HREF="write.html">Contact Me</A></TD></TR></TABLE>
However, on each page, we want the cell containing the current pageto be highlighted in some fashion. This is really easy. All we haveto do is add a class to each table cell, like this:
properties add up to the required total. It's a semantic dodge, but it's valid behavior.Let's consider another example, illustrated in Figure 8-20, where the left margin is set to be negative:
DIV {width: 400px; border: 1px solid black;}
P.wide {margin-left: -50px; width: auto; margin-right: 10px;
border: 3px solid gray;}

In this case, not only does the paragraph spill beyond the borders of
In order to understand how a useragent determines the heaviness, or weight, of a given font variant,not to mention how weight is inherited, it's easiest to startby talking about the keywords 100 through900. These numberkeywords were defined to map to a relatively common feature of fontdesign in which a font is given nine levels of weight. OpenType, forexample, employs a numeric scale with nine values. If a font hasthese levels built in, then the numbers are mapped directly to thepredefined levels, with 100 as the lightest