Patrick H. Lauke wrote:
Geoff Deering wrote:
How do you know what device configuration is receiving your design?
Because if you do not, and cannot be absolutely sure your design is
not clashing with this principle, you cannot *ensure* you have
succeeded.
But that is true of pretty much any element and situation where you're
applying style. And that's why we have separation of content and
presentation: if the presentation does present a problem to the user,
the user agent should provide the user with a simple way of
overriding, or completely ignoring, the styles suggested by the designer.
Firstly, this still does not address the initial problem.
Secondly, by this recommendation you are actually addressing the flip
side of the problem I am trying to address.
The case you are addressing here is
1) A recommendation of how to deal with styles that may conflict with a
form element that is in an activated state.
2) What I was addressing was dealing with styles that may conflict with
a form element that is in a non activated state.
Either way, that these recommendation could be feasible in practice, is
for the functions within the user agent being able to detect at least
two conditions;
for your condition;
if(FormElementWebDesignerStyle != FormElementClientDefaultValue) &
(FormElement == active) then do {apply correct display of state};
and for my condition;
if(FormElementWebDesignerStyle != FormElementClientDefaultValue) &
(FormElement != active) then do {apply correct display of state};
I can't see how this type of functionality will ever be added to a user
agent because it goes against the fundamental interface principles of
OSs. The WAI/UAAG would come under scrutiny if they did this. And with
all the bugs and unimplemented recommendations in user agent
development, I can't see this ever seeing the light of day.
I'm really not up to date with CSS behaviour at all, but if someone can
show me a real world example of this recommendation and how it could be
applied, without programming the user agent functionality, and therefore
changing the UAAG specs, I'd be happy to see it.
You could probably do it with client side scripting, but then that
breaks when it's turned off.
Unless you know for sure how users have configured their interface
you are swimming in the sea of uncertainty. [...] If accuracy of
communicating [...] is a primary principle to your design philosophy,
> you cannot be
sure you have not interfered with this process
With my intentional omissions (not trying to misquote you, just making
it generic), I'd say this applies to any styling, not just the
specific case of form elements.
If you are suggesting this is to be handled by the user agent, how are
you going to implement it? I'd like to know your suggestions on this,
and the functional logic.
I know many people feel that about the WAI GL, but I have never felt
that. People complain about the WAI police and the lengthy drawn out
debate that goes on there, but I mostly see a lot of people concerned
*not* to write recommendations that restrict design.
Fair enough...I'm probably thinking of some of the more radical (and
possibly most vocal and entrenched) elements here.
Yes, I know, but the way they are still worked with and appreciated for
their input is inspiring. I used to play the devils advocate there
because in the long run it helps to apply as much scrutiny and analysis
as possible for the most clear and comprehensive outcome. It's
incredibly hard work to develop accurate standards that are also free of
unnecessary restrictions. It requires that type of process.
-----------
Geoff Deering
******************************************************
The discussion list for http://webstandardsgroup.org/
See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list & getting help
******************************************************