Re: Coding practices of large sites

Jonas Smithson <smithsonNOSPAM (AT) REMOVETHISboardermail (DOT) com> writes:

1. If, as the regulars in this newsgroup believe, CSS is clearly
superior to table code for defining page layout, why are *none* of
these sites relying on CSS for that purpose?

It wasn't more than a couple of years ago that common browsers weren't
really capable of coping with CSS layouts. Once Mozilla/Opera 5/IE6
came out it became much easier.

Table layout is not going to kill a website, and most of them have
bigger problems to fix anyway (count the missing alt attributes on a
lot of big commercial sites).

Probably they have bigger priorities than a template rewrite.

2. Can anybody provide links to any large commercial or institutional
sites (whose subject matter is not "coding practices") that *do* use
CSS to define their page geometry, so that I can look at their code?

"Kris" <kristiaan (AT) xs4all (DOT) nl.omitthis> wrote

} http://www.uoguelph.ca/~stuartr/
} http://www.interactionbydesign.com/ (yeah, well, it is different!)
} http://www.opera.com/
} http://www.marsorange.com/ (again, the different-category)
} http://www.frightenstein.com/
} http://www.minid.net/
} http://www.asteric.co.uk/
} http://mbusa.com/brand/index.jsp (Mercedes Benz USA)
} http://www.onepointzero.com/

There's a few others as well.

And we use stylesheet layout: http://www.dur.ac.uk/its/

--
Chris

Chris Morris wrote:

Jonas Smithson <smithsonNOSPAM (AT) REMOVETHISboardermail (DOT) com> writes:
1. If, as the regulars in this newsgroup believe, CSS is clearly
superior to table code for defining page layout, why are *none* of
these sites relying on CSS for that purpose?

It wasn't more than a couple of years ago that common browsers weren't
really capable of coping with CSS layouts. Once Mozilla/Opera 5/IE6
came out it became much easier.
[snip]

Was that really the problem? For example, my CSS layouts work in IE 5, and I
think so do those of most other authors. And those other browsers haven't
really gone away, have they? Except, perhaps, for NN4, people still tend to
cater for older browsers. (And IE 6 introduced some weird float-bugs of its
own!)

I think what is much more relevant is that a lot more is now known about how
to do CSS layouts. Not long ago, a flexible 3-column layout was "the holy
grail". Now it is pretty routine. There has been a steady accumulation of
knowledge of bugs & non-compliances, and hacks & workarounds to cater for
specific problems. I've just been looking at one of the CSSs at W3C, and it
runs through a whole set of hacks for Konqueror 3.1, Safari 1.0b, IE 6 &
others, Opera 6, etc. (See CSS for the W3C page below). And it is also easier,
perhaps, to run a range of different browsers on your system, for example
multiple versions of IE, for testing purposes.
http://www.w3.org/Style/Examples/007/figures.html

In addition, I think people are using different design concepts, that are more
sympathetic to CSS techniques. For example, minimising images slices in the
document/foreground, while making more use of backgrounds via CSS. Sometimes
this makes a visible difference, (which some people criticise), while in other
cases it is "just" a different, (and often better), way of achieving the same
visual design.

--
Barry Pearson
http://www.Barry.Pearson.name/photography/
http://www.BirdsAndAnimals.info/
http://www.ChildSupportAnalysis.co.uk/

"Barry Pearson" <news (AT) childsupportanalysis (DOT) co.uk> writes:

Chris Morris wrote:
Jonas Smithson <smithsonNOSPAM (AT) REMOVETHISboardermail (DOT) com> writes:
1. If, as the regulars in this newsgroup believe, CSS is clearly
superior to table code for defining page layout, why are *none* of
these sites relying on CSS for that purpose?

It wasn't more than a couple of years ago that common browsers weren't
really capable of coping with CSS layouts. Once Mozilla/Opera 5/IE6
came out it became much easier.
[snip]

Was that really the problem? For example, my CSS layouts work in IE 5, and I
think so do those of most other authors. And those other browsers haven't
really gone away, have they? Except, perhaps, for NN4, people still tend to
cater for older browsers. (And IE 6 introduced some weird float-bugs of its
own!)

Hmm, it's NN4 and IE4 that I'm really referring to. NN6 for that
matter had some messed up stylesheet support (and indeed fairly broken
table support), but hardly anyone used it.

--
Chris

Barry Pearson wrote:

.... And it is also easier, perhaps, to run a range of
different browsers on your system, for example
multiple versions of IE, for testing purposes....

Sorry if this is slightly off-topic, but you've touched on an issue
that has been driving me slightly batty...

Testing has been a definite issue for me in constructing CSS,
especially CSS-P. Currently I routinely test in three browsers: Win
IE6, Win NS6, and Mac IE5; and sometimes also in Mac NS7. (I've given
up on NS4, may it rot in hell.) I'd like to also test in Windows IE5 or
maybe IE5.5 (which is more common?) but a guy who knows more about
Windows than I do told me to forget it. He said it's either impossible
or very difficult to get more than one version of IE running on the
same Windows box, even if you don't try to run them at the same time.
(I've had no problem running more than one version of Netscape, even
simultaneously, on both my PC and my Mac -- I suppose because it's not
as tightly bound to the OS as WinIE is.)

My PC uses Windows 2000 Pro, BTW. Can you give me any information, or
point me somewhere, that would help me run more than one test version
of IE on a PC? That would greatly increase my incentive to use more CSS
-- right now I can switch the box model of IE6 to "standards mode"
(more or less) using a DTD, but I've read that there's no way to get
WinIE5 to follow spec on box widths using a DTD switch. I've read of
some hacks/workarounds for the box model in IE5 but I'm reluctant to
trust them unless I can test them in my actual layouts.

Thanks much!

Jonas Smithson wrote:

Barry Pearson wrote:

.... And it is also easier, perhaps, to run a range of
different browsers on your system, for example
multiple versions of IE, for testing purposes....
[snip]
My PC uses Windows 2000 Pro, BTW. Can you give me any information, or
point me somewhere, that would help me run more than one test version
of IE on a PC? That would greatly increase my incentive to use more
CSS -- right now I can switch the box model of IE6 to "standards mode"
(more or less) using a DTD, but I've read that there's no way to get
WinIE5 to follow spec on box widths using a DTD switch. I've read of
some hacks/workarounds for the box model in IE5 but I'm reluctant to
trust them unless I can test them in my actual layouts.

I'm not sure exactly where I got IE 5 from, but it may have been here:
http://www.zeit.ca/mozie/
Or, more likely, here:
http://browsers.evolt.org/?ie/32bit/5.01_SP2

I just unzipped it into a folder, and it ran. It didn't reconfigure the
system, it just used the existing IE 6 options. I use just IE 5 (in addition
to IE 6), but others use a wider variety. I've been using it successfully
since last December. It is very illuminating.

--
Barry Pearson
http://www.Barry.Pearson.name/photography/
http://www.BirdsAndAnimals.info/
http://www.ChildSupportAnalysis.co.uk/

Jonas Smithson wrote:

Barry Pearson wrote:

.... And it is also easier, perhaps, to run a range of
different browsers on your system, for example
multiple versions of IE, for testing purposes....

My PC uses Windows 2000 Pro, BTW. Can you give me any information, or
point me somewhere, that would help me run more than one test version
of IE on a PC?

Joe Maddalone is your man. Check out the following article

http://www.insert-title.com/web_desi...s/dev/multi_IE

It's really pretty easy . . .

Of course, I couldn't remember the web address offhand, so a quick
Google for <multiple ie one machine> pointed me in the right direction .
.. .

P