ZG>Aren't tables a trick as well?
HS> Not really. I tend to regard tables as a gridded layout.
The interesting thing about this point is that it identifies the HTML
table element as an element that describes presentation, versus an
element that describes content. This is where Zoe and I and the
"semantics crowd" see table layouts as tricks, where we are using
HTML elements designed for defining tabular data to instead define a
HS> tricks I mean things like putting a table around a table for
HS> the purpose of putting a border around the entire content. My
HS> main problem with tricks is that they are hard to explain,
HS> and therefore hard to maintain.
I agree with this -- if something is not clear, then it will be
difficult to maintain.
Much of the difficulty I think comes from the transition from table
layouts to CSS-P. As it stands, CSS does grid layouts with difficulty.
In order to do grid layouts well, it requires a pretty strong
understanding of the underlying theory behind CSS.
The flip side of this is that once the understanding is present, it
becomes fairly easy to use it all over the place. The concepts used in
the version I made with positioning are fairly standard CSS-P --
padding to offset space taken by absolute elements, font size
management, etc. -- but make no sense within a table perspective.
Admittedly, table layouts and image slicing techniques were considered
tricks at one time, also. The concept that a 100%-wide table cell is
actually not 100% wide is now common knowledge, but theoretically
sounds kind of odd.
ZG> They've already done that. The display: table-cell and related
ZG> properties do this. It's just that IE doesn't support them.
HS> I tried this, and it didn't help, though I might not have done
HS> so exhaustively.
HS> As to IE support, I guess I don't care much, since this is a
HS> vanity page.
Cool -- if that's the case, then display: table-cell will do the trick.
(Sorry, I couldn't resist. ;)
I've updated my page to use table-cell, to permit borders. The nice
thing about browsers that support table layout is that it then applies
most of the standard table rules to those non-table elements.
The main challenge with CSS-P is that it gives you much greater freedom
to use HTML as it was intended -- for giving meaning to your content --
but at the cost of changing perceptions about HTML's use as a layout
 Don't know if anyone's used this term before, but it seems to
 http://home.earthlink.net/~landisdesign/date-alignment.html (in
case it got lost from the last part of the thread)
List wiki/FAQ -- http://css-discuss.incutio.com/
Supported by evolt.org -- http://www.evolt.org/help_support_evolt/