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

RE: rationalize presentation [was: Use consistent presentation]

From: Geoff Deering <gdeering@acslink.net.au>
Date: Fri, 11 Jan 2002 07:34:24 +1100
To: <gian@stanleymilford.com.au>, <w3c-wai-gl@w3.org>
Message-ID: <NBBBJPNFCLNLAADCLFJBOECPDEAA.gdeering@acslink.net.au>
Yes, most of the best tools do not support structured markup and CSS design
very well, and even though I think a lot of these designers are trapped in a
certain pixel based mentality, you can't blame them when their tools, which
are focused on that approach, and until we get really good integrated tools
that support this, it will be difficult to change this.

Most of the "Content Management Systems" (CMS) store all the site navigation
in databases or XML repositories.  They store them as text labels and URLs.
The designers put styles to them, and the sites structure and navigation is
generated from those repositories.  It is too difficult to use image-based
navigation when building sites with these systems.  You can automate
building images from text, but the quality is poor, and you can't anti alias
them.  In the case of CMSs that implement personalisation, the navigation is
dynamically data driven.

Amazon is a case in point.  But it is also its own animal.  It has a graphic
top menu bar, and much of the other navigation in boxes is text based and
data driven from personalised information (as is also parts of the top bar).
But you need $ to try and copy this model successfully.  They cache their
images in server memory for rapid delivery and have a separate server to
serve them (g-images.amazon.com) and they have built and compiled most of
the CGI logic into their own version of the Apache web server in the C
language to get their site working at maximum speed.  Few can copy this
model because of the specialisation/$. But it is still by and large a low
graphics site.

The trouble is, users are used to this type of performance and delivery on
the good sites.  They want it on theirs too.  Clients want their sites to be
as good and as fast at amazons.  One of the best ways to mimic faster
delivery and still give an elegant design is through CSS.  I guess there are
trade offs everywhere.

I just don't feel designers are really trying to use CSS to its max, and you
have to know the bugs.  It's too long and painful I guess.  But I hope and
believe it will be worth the effort to move towards this model.

Geoff Deering

-----Original Message-----
From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org]On Behalf
Of gian@stanleymilford.com.au
Sent: Thursday, 10 January 2002 1:10 PM
To: w3c-wai-gl@w3.org
Subject: RE: rationalize presentation [was: Use consistent presentation]

I think you will find that graphics designers still don't believe that
CSS etc could provide the kind of artistic expression they are used to,
and even if it does, it will take a lot to convince them. I understand
that a lot of pages are data-driven, but this doesn't mean that their
navigation is data-driven- one would hope, actually, that the navigation
remains static to ensure rational [consistent] presentation!

-----Original Message-----
From: gdeering [mailto:gdeering@acslink.net.au]
Sent: Thursday, 10 January 2002 10:33 AM
To: Gian Sampson-Wild; w3c-wai-gl
Subject: RE: rationalize presentation [was: Use consistent presentation]

I take your point, and I agree designers have an unwillingness to
their pixel based design tools for a more liquid based design approach,
I don't agree with you that markup has not progressed to the point of
placing graphical navigation.

Most analysis of current web logs shows a very very high percentage of
capable browsers (eg
So designers should know that most people are seeing their design as
designed it, or close to it.  It is also becoming uneconomical to design
navigation using pixel based designs as more and more sites require data
driven generation of pages.  All major News sites, and any site that
produces a large amount of information uses this approach.

Designers can still control the fonts, colors, text and hover effects,
most importantly it scales better. There are so many benefits from at
abandoning graphic labels in navigation.  I know designers are
but it I feel it is more from habit and having a fixed pixel based 800 x
mentality, that being restricted by the markup.  Also pixel based
navigation can look terrible and difficult to read under certain

Re: Checkpoint 3.1 WCAG1.

If you have specific instances, can you please refer to them?

My feeling is that at this point, preparing WCAG2, if we DO feel that
based graphics navigation is still a legitimate choice, then WCAG2 needs
reorientation.  There seems to be a great change from WCAG1 to WCAG2.
set out the 3 levels of Priority, and if you have a site designed
to any flavour, you can still go over it and adopt all the Priority 1
recommendations to repair it to make it more accessible without
fundamentally causing a redesign.  Unfortunately a lot of companies
(like my
last place of work) just use Bobby to get a Priority 1 error free
It does not matter that Bobby did not pick up that there were no
tags for each <SCRIPT> tag (6.3).

WCAG2 seems to have moved beyond that approach.  It seems to be saying,
this place and time: this is the way to address markup in order to
accessibility (I am making my own assumption here, please correct me).
If a
designer really looks at this thoroughly, they should not feel that
freedom and creativity are being taken away, instead, it is our job to
teach and communicate how they can be just as creative, and more so,
also making their design more accessible and usable.

I do feel that the delivery technology is up to this, but the basic
tools definitely are not.  In good web teams, it is often best for the
designer to produce the design in their pixel based tools, then a
HTML/CSS/WAI specialist take that design and render it in markup.  1)
Because the tools just aren't good enough, and may not be for a long
and 2) Most designers cannot come to terms with a liquid design

Geoff Deering

My two cents on this issue is that appropriate markup for navigation
(HTML/XML/CSS/XSL) has not progressed to the point of being able to
replace graphical navigation. Until that is the case designers will want
to use graphical navigation ahead of any markup.


-----Original Message-----
From: gdeering [mailto:gdeering@acslink.net.au]
Sent: Tuesday, 8 January 2002 5:53 PM
To: asgilman; w3c-wai-gl
Subject: RE: rationalize presentation [was: Use consistent presentation]

I feel this is an important distinction.  WCAG2 seems to have advanced
a way from WCAG1 in the way it now focuses accessibility issues for
producers.  To me, WCAG2 does not seem like a clarification of WCAG1,
is what was originally intended, but a redefining and refinement of core
issues of Web Content Accessibility.

It seems a powerful statement that WCAG2 starts with Guideline 1;
the focus back on the users needs and preferences.  This was one of the
objectives of CSS; the power of design, so that the author based design
could easily be overridden or substituted by the users own CSS.  This
to be lost on most of the web designers.  If this is lost on most web
designers, it is difficult enough for them to come to terms with a model
where the user is actually in control of the form and how they
the content.

It is very hard as a designer to feel that your site could or should be
"skinned" (have another CSS wrapper put around it, other than your own).
This is an ideal, but very hard to achieve or deliver for the simple
of using abstract class definitions that a base User CSS will not
Still basic declarations like font properties, etc, should be able to be

What strikes me most about WCAG2 as compared with WCAG1 is the
back on marking up content to facilitate full accessibility to all
With such focus in WCAG2 on user control, there seems to be an implicit
statement that presentation markup follows strict guidelines and syntax
(correct use of CSS / XSL), and these are the type of guidelines that
"Markup your document's presentation style in such a way that users can
enforce their own style and rules on your documents and sites".

If this is the intention, what are the cases for using navigation bars
comprised of graphic images?  It seems in WCAG2 we have gone beyond
saying now that there is an appropriate markup for navigation
(HTML/XML/CSS/XSL), graphic text images are obsolete.

Is this the implication?  Or am I missing something?

Geoff Deering

-----Original Message-----
From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org]On
Of Al Gilman
Sent: Saturday, 5 January 2002 3:15 AM
To: w3c-wai-gl@w3.org
Subject: rationalize presentation [was: Use consistent presentation]

Well, we are building a problem for ourselves as soon as we use
'consistent' as
the headline.  This sends too superficial a message.  It creates an
that consistency of appearance is the success criterion, whereas this is
the point.  What we are after is a consistent relationship between
in appearance and distinctions in the rhetoric of the site content.

The success criterion here is that most presentation differences within
site can be explained by a systematic binding [pattern of relationship]
presentation effects to one consistent set of organizing principles
whole site.  The organizing principles cover concepts not only of parts
whole but also kinds of stuff found here.  This includes both
differences.  It is possible to quantify this criterion with entropy

What the content provider should accomplish is a balance of a dominant
rationality - functional application of effects - and a light ripple of
- un-rationalized variation - to combat highway hypnosis.

If we just say "rationalize your presentation" and people do this to the
the result _will_ be boring.  A modicum of reasonableness is required in
how we
try to get people to curb their wildest flights of fancy and employ
in presentation.

The rational core (of presentation decisions) follows the XAG
- have a model
- bind the model to interface specifics
- stick to this plan
- share the reasoning for those who have to re-bind

Changes in presentation are rocks that turn over.  People expect to find
information under those rocks.  When there is no sense to the change,
distinctions are treated as senseless.  The consistent styling of
to indicate the consistent functional role that they play across
pages will be defeated if one does not at the same time hold to a dull
changes in presentation that are _not_ explainable by site-organizing

What Gian has had trouble selling to his customers is good design.
presentation for exceptional circumstances.  It is all backed up by a
well-developed design concept or model.  But the neanderthals with the
don't have a basis for judging what is a bona-fide exceptional
unfortunately.  So they cling to shallow rules that actually work
application of the better, deeper, rule.

Not a problem with a simple answer, because in many situations we need
on the funding authorities to sit on the heads of the designers and make
exercise some discipline in addressing access issues.


At 07:07 PM 2002-01-03 , gian@stanleymilford.com.au wrote:
>I think we had a very productive meeting this morning.
>On the subject of consistency of presentation, I think what we are
>trying to say (while still following our own guideline 14.1!) is that
>the presentation is similar in design throughout the entire site, and
>over a finite period of time (as in, one does not expect the
>presentation to continually change on a week-to-week basis).  I did go
>talk to one of our graphics designers and he did not like the word
>'predictable'!  So perhaps we can consider other words.  A search in
>thesaurus for predictable found:
>- obvious
>- unsuprising
>- expected
>- anticipated
>- evident
>- observable
>A search in the thesaurus for consistent found:
>- reliable
>- constant
>- uniform
>And secondly, on the actual amount of consistency, I have found myself
>arguing against the guidelines when discussing this checkpoint with
>project managers. For a large site I usually advocate the use of
>secondary navigation appearing when on a particular page (for an
dhs.vic.gov.au/ds/disabilitysite.nsf/pages/prov vs
b.dhs.vic.gov.au/ds/disabilitysite.nsf/pages/prog or
ound.com.au/dev/ultra.html).  However when I
>suggest this I get checkpoint 13.4 thrown at me and told that it can't
>be done (usually I then ask why they have employed me if they aren't
>going to listen to me- as you can tell this is a sore point!).  So I
>believe we should somehow incorporate these navigational mechanisms
>the guidelines, or at the very least, ensure that they aren't outlawed
>by the guidelines.  The other thing to consider is the use of
>trails (see:
>w~ or
>Breadcrumb trails are a very good way of indicating where in the site
>you are, especially for those sites that are very large.
>Happy New Year to everyone!
>Gian Sampson-Wild
>Member: Web Content Accessibility Group Working Group
>W3C Web Accessibility Initiative
>Stanley & Milford
>A Software Communication Group Company
>Level 16
>644 Chapel Street
>South Yarra VIC 3141
>Tel. 613 9826 5829
>Fax. 613 9826 8336
>Mob. 0404 498 030
>Email gian@stanleymilford.com.au
>This message contains privileged and confidential information intended
>only for the use of the addressee named above. If you are not the
>intended recipient of this message you must not disseminate, copy or
>take any action in reliance on it. If you have received this message in
>error, please notify Software Communication Group immediately. Any
>expressed in this message are those of the individual sender except
>where the sender specifically states them to be the views of Software
>Communication Group.
Received on Thursday, 10 January 2002 15:34:41 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:40 UTC