W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > January to March 2002

RE: CSS versus tables

From: Scott Luebking <phoenixl@sonic.net>
Date: Tue, 1 Jan 2002 09:20:41 -0800
Message-Id: <200201011720.g01HKfiG016500@newbolt.sonic.net>
To: charles@w3.org, phoenixl@sonic.net
Cc: andrew.mcfarland@unite.net, w3c-wai-gl@w3.org
Hi,

Just wondering  -  is this true for most rectilinear layouts of
cells or does it refer to certain types of layout, e.g. the common
format of a main area bordered by smaller areas above, below, etc?
I think for this common type of layout CSS is easier (ignoring
the browser problems).  If the layout gets more complex and information
needs to have certain related alignments to other information for visual
appeal, tables can be pretty fast.

Scott

> I'm not Andrew, but here are my answers. The easiest thing for me to do rough
> work is use a tool that has a nice interface - I point out the regions and
> layout features I want, and the tool does the coding. In that case it doesn't
> matter as a user which way the code is done, but as a tool developer it is in
> fact easier to use CSS.
> 
> If I am hand coding it is easier to use CSS than tables for the code.
> 
> The problem is that there are some browsers which not only do not implement
> CSS, but actually implement it in a way which means that CSS causes problems
> - this does not happen with the pre-CSS browsers, nor with those whose
> implementation is quite good, but with the early, buggy, test implementations
> that were necessary to work out whether the spec worked, and what was
> possible and useful, but which should have been advertised as such so people
> didn't consider them production-standard and insist on using them as if they
> were.
> 
> Tihs, combined with the fact that tables were implemented earlier than CSS
> and some rudimentary support is available for them in the same broken
> browsers that fail if I use good CSS, means that I have to work out whether
> to make pages that are broken for new browsers, very old browsers, and ones
> where the user has customised it in unpredictable ways, but work for oldish
> browsers (2-5 years old, but covering a significant number of users), or
> whether to make content that works for very old and very new browsers, and
> that any browsers I have never heard of but which conform to the
> specification will know how to handle, at the price of wierd presentation in
> some common but poor browsers.
> 
> Sorry for the last sentence. If someone can rewrite that in english people
> speak I would be happy.
> 
> cheers
> 
> Chaals
Received on Tuesday, 1 January 2002 12:20:44 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:18 GMT