W3C home > Mailing lists > Public > www-html@w3.org > March 2001

RE: XHTML 2.0 -- A Chance to Improve Document Structure?

From: Klaas De Waele <klaas@gracegraphics.be>
Date: Fri, 30 Mar 2001 09:26:19 +0200
Message-ID: <F16ABE403FD3D311A61500608CC0D1670C1AAA@pdc>
To: "'Daniel Hiester'" <alatus@earthlink.net>, "'www-html'" <www-html@w3.org>


Problem with this CSS positioning using no tables... I've done some html
4.01 strict encoding yesterday, based purely upon CSS 2.0 positioning, and
I've noticed when having only an <a> tag in a DIV, you get IE 5.0 and higher
to correctly position using CSS (what I did was select the text of the a-tag
then print screen and count pixels in PhotoShop) but Mozilla/NS6.0 is three
pixels off.  So I used a table, which I gave a fixed height (through CSS)
high enough and positioning was the same in both systems then.

Some other thing: when including only an image in a DIV tag, the DIV would
get a size that is about 4 pixels higher than the image.  This is especially
bugging me when using a background (image or colour) in the DIV, since it
will repeat/continue for those few pixels.  Can be worked around through
repeatx, but then again, we shouldn't need workarounds.  In these cases,
tables were necessary to have the exact same behavious in the different
browsers.

I have also made a page using 4.01 transitional that has minimal deviation
throughout all main browsers, using a combination of tables and CSS2.0.
Works (almost) perfectly in NS4.7 and higher throughout IE4.0 and higher.
That's as long as you're not using javascript.  Main problem there is the
frameborders etc. in NS 4.7 which can't be less than 1, and the framesizes
which have to be a multitude of 8.

I can provide source files for those interested.


- Kayjey -



-----Oorspronkelijk bericht-----
Van: www-html-request@w3.org [mailto:www-html-request@w3.org]Namens
Daniel Hiester
Verzonden: vrijdag 30 maart 2001 3:01
Aan: www-html
Onderwerp: Re: XHTML 2.0 -- A Chance to Improve Document Structure?


"I would agree with Daniel on this, however, that tables are likely not to
go
away any time soon, and furthermore I'd argue that they probably shouldn't.
A table structure is an easy construct to make using XSLT, whereas
positioning via CSS can be a real nightmare to handle well in that milleau."

I'm afraid I don't understand xslt at all...
But it is far from impossible to do a decent layout without extensive use of
tables. For those who are interested, a fairly good example of what can be
done with CSS positioning can be seen here:
http://www.interplay.com/giants
They use some tables for small things, but the major parts of the page are
done with css positioning. It really works very well, and offers better
opportunities than HTML tables ever could.
Unfortunately, because it is javascripted to be NS4-friendly, the site does
not display properly in Mozilla / NS6. I've done similar pages which render
in IE5.5 and Mozilla without the use of any javascripting at all, although
they aren't anywhere near as well thought-out.

This is off-topic for the html list, I know... but the point is to say that
HTML tables *should* be discouraged, *if* it were a realistic expectation.

I don't know what xslt tables are... however, since it's off-topic for the
html list, if anyone wants to explain it to me, it would probably best to do
so off-list.

Daniel
Received on Friday, 30 March 2001 02:32:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:15:45 GMT