W3C home > Mailing lists > Public > www-style@w3.org > August 2002

RE: CSS3[the box model]: Overflow values

From: Adam van den Hoven <list@adamvandenhoven.com>
Date: Tue, 27 Aug 2002 16:47:53 -0700
To: <www-style@w3.org>
Message-ID: <2CFBE0D9CD992F41BE13069173C5C1C9019925@c2kxch.cucbc.com>

I have one very good reason to be afraid of table elements. In my
company, our web app is used for a number (hundreds) of different
clients, all Canadian credit unions who what to apply their own look and
feel to the application. If I use tables for layout, I'm restricted to
one particular layout. If I suddenly get client who wants to customize
the look and feel in such a way that the navigation goes from a
"Masthead image a the very top, Main navigation across the top, context
navigation down the left" to an Apple-esque "Graphical tabs along the
top for main navigation (put the logo in the first tab), context
navigation below", I have to produce a completely new set of jsps to do
it. Even with intelligent use of includes, there are is a lot of JSP
work to do. 

I'll have to do a completely new accessibility audit. And since a lot of
functionality is inline with the code, my QA guy will insist I do a
complete regression test of the application. 

Multiply that time and effort by 100 and the money will probably come
out of my salary. 

It gets even worse when they tell me that they want the form fields to
be on the line below the label instead of having them all line up at
200px from the left edge. I will need to duplicate a lot of work. 

Besides, I now have nice, clean, semantically meaningful HTML that is
completely meaningful when you look at it in lynx (or more importantly
with a screen reader). And I can achieve 80+% of my design goals. Why,
oh why would I clutter it up with tables which add no meaning, in and of

All I'm looking for is the tools that allow me to do reasonable things
without having to penalize myself for insisting on using only clean,
semantically meaningful HTML. 

To my way of thinking, the argument that you can use tables to do it
now, so extending the box model is unnecessary is equivalent to saying
that the default rendering of BLOCKQUOTE is to offset the left edge of
its contents so we shouldn't bother with margin left. 

I may be in the process of becoming a markup zealot. The process may be


-----Original Message-----
From: www-style-request@w3.org [mailto:www-style-request@w3.org] On
Behalf Of Coises
Sent: Tuesday, August 27, 2002 3:06 PM
To: www-style@w3.org
Subject: Re: CSS3[the box model]: Overflow values

[Tue, 27 Aug 2002 13:03:12 -0700 (PDT)] Bill Daly:

My immediate thought is that Adam's "problem" is not so much evidence
that the behavior of floats should be extended --- though I do believe
it should, for other reasons --- as that fear of the <TABLE> element has
gotten entirely out of hand.
Randall Joseph Fellmy aka Randy@Coises.com
Received on Tuesday, 27 August 2002 19:58:01 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:03 UTC