Re: compliance and layout tables revisited

Hi Denise,

Thanks to Joe and other sources, I'm aware of those NN4-friendly CSS layout
templates however the site I was trying to validate was already done in
tables because of checkpoint 5.3 which states, "Do not use tables for layout
unless the table makes sense when linearized". I had already put alot of
effort into that. I don't want to go back and re-do the site now. I'll know
for the next site I do.

> I initially found the same problem with border="0" used as part of the img
tag
> being reported as invalid by the HTML validator. I changed the doc type to
> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> and the
> validator no longer reports this as a problem. I agree with Joe that use
of CSS
> to control the appearance of images with respect to borders is preferable.
> However, many browsers do not recognise the class and still show the
border. So
> I decided to use the border="0" for the sake of browser compatibility.

I had been using the transitional doctype you mentioned all along. My
problem is not with an image tag but with an input tag (having a graphic as
a button in a form). border="0" doesn't validate for the input tag. So, I
can't claim compliance unless I allow the border to show in some browsers.
Or, I do claim compliance knowing that my site is accessible and hope no one
checks. Those are the options. Sucks to be me, I guess.

Thanks for bringing all this out in the open. Good luck with your site.

Michelle Podd, Web Designer


----- Original Message -----
From: "Denise Wood" <Denise.Wood@unisa.edu.au>
To: "'Michelle Podd '" <mpodd@iqnetcom.com>; "'WAI (E-mail) '"
<w3c-wai-ig@w3.org>
Sent: Tuesday, May 07, 2002 2:09 PM
Subject: RE: compliance and layout tables revisited


> Michelle
>
> You raise an interesting point about the Priority 2 validation issues
relating
> to use of tables for layout. I too had trouble reconciling those points
but in
> the end decided that I would aim to use style sheets for layout in any
case as
> it is just so much better to be able to separate layout from content. You
are
> right, Joe did respond to this discussion (and I note his new post on this
> topic has just arrived in my inbox).If you recall however, Joe also
included
> suggestions of html pages (templates) that effectively use CSS2 style
sheets
> for layout and look fine in NS4 and other lower/broken browsers. That is
the
> approach I took to be comfortable that the site does achieve triple A
> compliance.
>
> I initially found the same problem with border="0" used as part of the img
tag
> being reported as invalid by the HTML validator. I changed the doc type to
> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> and the
> validator no longer reports this as a problem. I agree with Joe that use
of CSS
> to control the appearance of images with respect to borders is preferable.
> However, many browsers do not recognise the class and still show the
border. So
> I decided to use the border="0" for the sake of browser compatibility.
>
> Denise
>
> -----Original Message-----
> From: Michelle Podd
> To: WAI (E-mail)
> Sent: 5/8/2002 1:56 AM
> Subject: compliance and layout tables revisited
>
> I've been reading the "compliance and html validation" thread with
> interest as I've had similar questions.
>
> The people involved in the discussion seemed to have settled on the fact
> that if you use tables for layout, your page cannot meet Priority 2 WAI
> standards. Is that correct?
>
> This is a snippet from the 1.0 Guidelines: <snippet>
>
> 11 style_alignmentLayout, positioning, layering, and alignment
>
> Checkpoints in this section:
>
> * 3.3 <http://www.w3.org/TR/WCAG10-TECHS/#tech-style-sheets>  Use
> style sheets to control layout and presentation. [Priority 2]
>
> * 5.3
> <http://www.w3.org/TR/WCAG10-TECHS/#tech-avoid-table-for-layout>  Do not
> use tables for layout unless the table makes sense when linearized.
> Otherwise, if the table does not make sense, provide an alternative
> equivalent (which may be a linearized version). [Priority 2] .
>
> Layout, positioning, layering, and alignment should be done through
> style sheets (notably by using CSS floats and absolute positioning): ...
>
> </snippet>
>
> It seems as if checkpoint 5.3 is saying you can use layout tables. For
> instance, if I use layout tables for a form where I have 2 columns (the
> left holds the field name and the right holds the form control), that
> would make sense when linearized.  If I have a nav bar that goes
> horizontally across the top of my web page made up of a table with
> several images in it, that would make sense when linearized.
>
> Are these two points contradictory? Is it true that if I don't only use
> CSS to layout my page that I can't say my page conforms to Priority 2
> requirements?
>
> Denise also brought up another question related to html validation that
> I'd like to clarify. My particular problem is that I am using a
> graphical image as a button in a form. Here is the code:
>
> <input TYPE="Image" SRC="images/buttons/update.gif" border="0"
> VALUE="Update Basket" ALT="if you changed a quantity, Update Basket">
>
> border="0" doesn't validate for the html 4.01 transitional doctype (the
> most lenient) on the input tag yet if I take it out, a border shows up
> around the image in some browsers.
>
> Joe's answer to that was (and thanks Joe, as you were the only one to
> provide an answer):
>
> "Your only choice is to use a browser fork and serve different HTML
> to different browsers, only some of which will validate."
>
> That seems very extreme to get rid of an image border. My question is,
> if there really is no other way to remove the border and still validate,
> how far do we, as web developers have to go to be able to say our sites
> meet certain accessibility standards? Does the border element in an
> input tag make my site (or that particular part of the page)
> inaccessible? No, it doesn't. I'd love to have all the little icons on
> my site that tell people we've made efforts to provide an accessible
> website however I can't see serving up different pages for different
> browsers for things that don't make any difference to the accessibility
> of my page anyway. Any thoughts or other solutions to the validation
> problem?
>
> partially asking and partially venting,
>
> Michelle Podd, Web Designer
>
>
>
>
>
>
>
>

Received on Tuesday, 7 May 2002 15:38:00 UTC